From chris.gonnerman at newcenturycomputers.net  Sat Sep 28 02:15:18 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sat, 28 Sep 2002 01:15:18 -0500
Subject: zlib needs the python15.dll on python22 ???
References: <an3g6c$j5t$00$1@news.t-online.com>
Message-ID: <000d01c266b6$6c6785c0$ba01010a@local>

----- Original Message ----- 
From: "Ralf Claus" <102030405 at gmx.net>


> Hello NG,
> by importing the zlib module on python 2.2 (windows)
> i get a popup that means "python15.dll not found in path 
> //winnt/system32/......and so on...
> 
> On that computer, there was never installed python 1.5 !?
> 
> What is wrong ?
> 
> Thanks a lot
> Ralf

You have an old version of the zlib module on your system.

Good luck finding it...

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net
 




From monnetb at yahoo.com  Tue Sep 10 09:20:15 2002
From: monnetb at yahoo.com (bruno Monnet)
Date: 10 Sep 2002 06:20:15 -0700
Subject: embedded Python sample
Message-ID: <2a0f4cc4.0209100520.1d3a723b@posting.google.com>

Hi,

I'm trying to call a python script from my c++ code. All I need is to
invoke the script and check the return status.

does anyone has a simple example ? All I found until now is very
complex and ugly code :-((

Regards,
Bruno


From aahz at pythoncraft.com  Sun Sep 29 13:19:52 2002
From: aahz at pythoncraft.com (Aahz)
Date: 29 Sep 2002 13:19:52 -0400
Subject: Newbie References Question -> Exemple
References: <3d91c72f$0$8558$79c14f64@nan-newsreader-02.noos.net> <3d91d2cf$0$1484$79c14f64@nan-newsreader-01.noos.net>
Message-ID: <an7cno$on7$1@panix1.panix.com>

In article <3d91d2cf$0$1484$79c14f64 at nan-newsreader-01.noos.net>,
Guy Rabiller <grabiller at 3dvf.net> wrote:
>
>What I want to do is simple, and I suppose that this is my 'way of thinking'
>regarding Python that is wrong.

Yup.

>I have a list of points, and a list of polygons that should contain
>'references' to points.
>
>So for me:
>-> points list:
>p = []
>p.append([0,0])
>p.append([0,1])
>p.append([1,1])
>p.append([1,0])
>-> polygon:
>quad = []
>quad.append(p[0])
>quad.append(p[1])
>quad.append(p[2])
>quad.append(p[3])

Make the actual points class instances.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From anthony at computronix.com  Fri Sep 13 14:07:19 2002
From: anthony at computronix.com (Anthony Tuininga)
Date: 13 Sep 2002 12:07:19 -0600
Subject: calling Oracle functions
In-Reply-To: <3d821f71$1@news.mt.net.mk>
References: <3d81f811$1@news.mt.net.mk>
	<mailman.1031929598.17983.python-list@python.org> <3d8210a3@news.mt.net.mk>
	<mailman.1031935337.20279.python-list@python.org> 
	<3d821f71$1@news.mt.net.mk>
Message-ID: <1031940440.472.97.camel@chl0151.edmonton.computronix.com>

On Fri, 2002-09-13 at 11:25, ?????? ?. wrote:
> 
> >> Now a simpler question:
> >> 	is it possible to specify the data source name directly not through
> >> 	the alias defined in tnsnames.ora. In perl DBD I can specify a
> >> 	connection string like cs = "dbi:Oracle:host=1.10.100.200;sid=TEST"?
> 
> > cx_Oracle is passing the string directly to Oracle, so whatever it
> > accepts, cx_Oracle will accept. Specifically, you can take the entire
> > string that is in your tnsnames.ora file and pass it directly as in
> 
> > v_TNS =
> > (DESCRIPTION=(ADDRESS_LIST=(ADDRESS=(PROTOCOL=TCP)(HOST=opus)(PORT=1521)))(CONNECT_DATA=(SID=DEV)))
> 
> > v_Connection = cx_Oracle.connect("user", "pw", v_TNS)
> 
> > and that works just fine. Does that answer your question?
> 
> Yes. Thanks. I would only suggest that you can add a function to
> cx_Oracle to make that big ugly string from 3 parameters.
> something like:
> 
> def makeDSN(host,port,sid):
> 	return 	"""(DESCRIPTION= (ADDRESS_LIST= (ADDRESS= (PROTOCOL=TCP)
> 	(HOST=%s) (PORT=%s))) (CONNECT_DATA=(SID=%s)))""" %
> 	(host,port,sid)

Sounds reasonable. I'll put it on my list of things to do.

> 
> 
> 
> 
> 
> -- 
> ??????
> 
>            Intel: where Quality is job number 0.9998782345!
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
-- 
Anthony Tuininga
anthony at computronix.com
 
Computronix
Distinctive Software. Real People.
Suite 200, 10216 - 124 Street NW
Edmonton, AB, Canada  T5N 4A3
Phone:	(780) 454-3700
Fax:	(780) 454-3838
http://www.computronix.com




From sismex01 at hebmex.com  Thu Sep 26 10:10:19 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Thu, 26 Sep 2002 09:10:19 -0500
Subject: Python threading?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062330260465A@mail.hebmex.com>

> 
> I wonder if generators can come to a rescue here?
> Still trying to get them into my head...
> 
> Thomas
> 

I was thinking exactly the same thing, generators as
cooperative multithreading.

I wonder if a "task", or a "service", in an async server
could simply initialize and then "yield" when it's ready
to be paused, and between yield's do some processing.

The core server would simpy call the service factory when
it's requested, receive a generator back, add it to it's
pool of active services, and iteratively call every
generators ".next()" method, so everybody gets a bit of
attention and love each cycle.

You could even use priorities, something simple, where high-
priority services would receive a bit more attention every
complete cycle.

I dunno, but it's an idea who's time has time, I think.

:-)

-gus


Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From marklists at mceahern.com  Tue Sep 17 10:28:15 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 17 Sep 2002 09:28:15 -0500
Subject: newbie-question: private attributes
In-Reply-To: <3D873A42.1070400@snafu.de>
Message-ID: <JHEOKEOOLIGLDHCMAHMOAEDGDPAA.marklists@mceahern.com>

[T. Kaufmann]
> What is the difference between the two attributes in this class?
>
> class Test:
>
>      def __init__(self):
>
>          self.__var1 = 1 # I think this is a private attribute
>          self._var2  = 2 # ... and this? Maybe protected?

Short answer:

  Both of them are ways of indicating a variable is private.  Both of them
are private by convention only.

  _var requires no hoops to circumvent.
  __var variables are mangled, but you can access them by jumping through
simple hoops.

Consider this:

#!/usr/bin/env python

class Test:

     def __init__(self):

         self.__var1 = 1 # I think this is a private attribute
         self._var2  = 2 # ... and this? Maybe protected?

     def show(self):
         print self.__class__.__name__
         for k, v in self.__dict__.items():
             print "\t%s = %s" % (k, v)

     def dosomethingwithvar1(self):
         self.__var1 += 1

class SubclassOfTest(Test):

    def dosomethingwithvar1(self):
        try:
            self.__var1 += 1
        except AttributeError:
            print "expected"
        self._Test__var1 += 1

t = Test()

print t._var2
try:
    print t.__var1
except AttributeError:
    print "expected error"
print t._Test__var1

t.show()
t.dosomethingwithvar1()
t.show()

s = SubclassOfTest()
print s._var2
try:
    print s.__var1
except AttributeError:
    print "expected error"
print s._Test__var1

s.show()
s.dosomethingwithvar1()
s.show()

-




From sholden at holdenweb.com  Mon Sep 16 07:47:06 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 16 Sep 2002 07:47:06 -0400
Subject: [ANN] PySQLite 0.3.0 released
References: <mailman.1031924237.12603.python-list@python.org> <a05h9.6196$Lg2.1046166@news2.news.adelphia.net>
Message-ID: <Wyjh9.1922$Cf.1588@fe03>

"Bob X" <bobx at linuxmail.org> wrote ...
> sismex01 at hebmex.com wrote:
> <snip>
> >
> >
> > Yeesh! You had to quote the whole message just to
> > add two words?
> >
> > :-(
> >
> > -gus
>
> My bust on that one. I usually do a <snip> like I did above. Sorry bout
dat.
>
> On that note...your two paragraph sig is okay right?  ; )
>

No, his two-paragraph "sig" is a PITA, but probably it's not aa sig but
something that his company's mailer sticks on the end of all message quite
gratuitously, just so some corporate lawyer can sleep better (fancying his
ass is covered).

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From jim.vickroy at noaa.gov  Tue Sep  3 16:32:26 2002
From: jim.vickroy at noaa.gov (j vickroy)
Date: Tue, 3 Sep 2002 14:32:26 -0600
Subject: Tk BLT / Pmw 1.1/ Python 2.2.1/ Ms Windows
Message-ID: <al368q$io2$1@mwrns.noaa.gov>

Hello,

So far, I have not been able to get the Tk BLT extension package to work
with Python 2.2.1 and Pmw 1.1.

I have tried placing the BLT24.dll various places, but the error message,
when running a sample script, is always:

TclError: invalid command name "::blt::graph"

Any suggestions would be appreciated.

Thanks.




From herald at ns1.nabitel.com  Mon Sep 30 00:42:37 2002
From: herald at ns1.nabitel.com (herald at ns1.nabitel.com)
Date: Mon, 30 Sep 2002 13:42:37 +0900
Subject: (ad)Strong WebRobot/eMailId Collector: Free Download !
Message-ID: <NS1DQpOII2XYRExc4hX00002e2a@ns1.ns1.nabitel.com>


Sorry for interrupting you - click refuse
<mailto:mailer at ns1.nabitel.com>  for no more mail... 	
? 	
- Welcome to NabiTel's <http://www.nabitel.com/English.asp>  software
products and portal services - 	
Software Products 	
 <http://www.Nabitel.com/English.asp> 

Web Robot: also called web spider or web crawler, collects useful web
page informations by navigating world wide web sites. 

Download free trial version now ! <http://www.nabitel.com/English.asp> 

 <http://www.Nabitel.com/English.asp>  	eMail ID Collector: Collects
email ids publicly opened on various web pages, with good intention. 

Download free trial version now ! <http://www.nabitel.com/English.asp> 

Portal Services 	
 <http://www.nabitel.com/English.asp>  	Web Portal: Do you have your own
home page and want to broadcast it all over the world ? Register your
home page to NabiTel Portal Now !! (nabi=a butterfly)

Register your home page now, it's free !
<http://www.nabitel.com/English.asp> 

 <http://www.AllThatCars.com/English.asp>  	Automobiles: Do you want
to sell or buy automobiles ? Cars, trucks, limos, airplanes, ships,....
All That Cars are here !

Register your vehicles now, it's free !
<http://www.AllThatCars.com/English.asp> 

 <http://www.AllThatComputers.Com>  	Computers: Do you want to sell
or buy computers ? PCs, printers, scanners, servers, mainframes, ....
All That Computers are here ! 

Register your computers now, it's free !
<http://www.AllThatComputers.com/English.asp> 

 <http://www.AllThatFoods.Com/English.asp>  	
Food & Restaurants: Are you seeking for a nice place to eat ? Or do you
run a restaurant ? Foods of the world, restaurants of the world, ....
All That Foods are here ! 

Register your restaurant now, it's free !
<http://www.AllThatFoods.com/English.asp> 

Have a nice day.  Thank you. 	
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020930/cdf2c3af/attachment.html>

From magnus at thinkware.se  Sun Sep 29 18:58:33 2002
From: magnus at thinkware.se (=?ISO-8859-1?Q?Magnus_Lyck=E5?=)
Date: Mon, 30 Sep 2002 00:58:33 +0200
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05>
Message-ID: <3D978599.1040206@thinkware.se>

Steve Holden wrote:
> Of course, none of this is necessarily an argument that f-p is the *best*
> representation for money, but I'm always amused to see people arguing
> against it on general grounds when it's perfectly acceptable for many such
> purposes *when correctly used*.

My opinion is that Python has a lot of the features
required to become "the next COBOL".

Yes, I admit it sounds horrible, I hate COBOL as much
as anybody ;) but try to understand what I mean.

Python has the potential to be used to quickly and in
a practical way build millions of applications and tools
by people who aren't always from a CS, math or engineering
background.

Python also has the potential to become the next BASIC--a
simple language for education, autodidacts and power-users.

It can also become the next Pascal, but I don't think we
need to change any language features for that.

But to suit beginners and...well, amateurs...we need to
follow the principle of least surprise.

It IS surprising for a lot of people that
 >>> 4/5
0.80000000000000004

You know that Steve, you see it here on c.l.py now and then.

It IS surprising for some people that for some numeric
values of 'w':
 >>> 1+w+w+w+w == w+w+w+w+1
0

And
 >>> 4/5
0

was certainly surprising for many, and that was
changed. The other big surprise was obviously

 >>> w = 5
 >>> print W
Traceback (most recent call last):
   File "<interactive input>", line 1, in ?
NameError: name 'W' is not defined

but that I think, is not so difficult to explain:
We don't consider w and W to be the same thing.
It's a design decision, and it could have been the
other way. Guido choose this way, just like he choose
= for assignment and == for equality, instead of
:= and = as in Pascal. End of story.

But we can't explain why floating point arithmetic
works the way it does without going into details the
groups I refer to feel that they shouldn't have to
understand.

It's a bit like demanding that a car driver understands
the physics and chemistry of combustion engines. You
can't sell a lot of cars if you need a degree in science
to drive them...

Or, to put it the other way around, you can sell a lot
more cars if people can drive them without a B.Sc, and
you can "sell" Python to a lot more people if they don't
need a B.Sc. to understand it...

On the other hand, I feel that when you make the tool
so simple that it can be handled by any fool, any fool
will be put in charge of the tool, and regardless of the
quality of the tool, they will make a mess because they
lack the required thinking ability.

In other words, we who actually understand computers and
problem solving will have to clean things up. But
a) I'd rather fix real logical problems in the business
    sphere than correct trivial programming blunders, and
b) We shouldn't make life more difficult than needed for
    those who are clever enough, but just don't have the
    expertize in our field. Finally,
c) we don't think that pen and paper ought to be difficult
    to use, so that people need to learn a lot of things
    unrelated to the ideas they want to put in writing.

Maybe this not such a big issue as I sometimes think it
is, but I do feel that Python is way ahead of the
competetition from a pedagogic perspective in most cases,
but this is one where it's...average.

Perhaps the right trade-off is still to educate the
programmers, and to build tools or routines to prevent
problems. But I think the issue is worth discussing.

It seems to me that a really smooth handling of dates,
times and decimal numbers would make Python even more
useful for most people who don't program today, but
could benefit from it if they did...

-- 
Magnus Lycka, Thinkware AB
Alvans vag 99, SE-907 50 UMEA, SWEDEN
phone: int+46 70 582 80 65, fax: int+46 70 612 80 65
http://www.thinkware.se/  mailto:magnus at thinkware.se



From chris.gonnerman at newcenturycomputers.net  Mon Sep 30 22:12:00 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Mon, 30 Sep 2002 21:12:00 -0500
Subject: Decimal arithmetic, with example code
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05> <an7bvd$burft$1@ID-125932.news.dfncis.de> <an7p7t$cpg$0@216.39.172.122> <mailman.1033338759.6775.python-list@python.org> <an8knv$mhr$0@216.39.172.122>
Message-ID: <001601c268ef$f79980e0$ba01010a@local>

Ok, here it is.  I have written some Python code which
displays the problems with float (floating binary)
arithmetic applied to monetary amounts.  Run this, and
many the places where the simple percentage times
amount calculation in floating binary falls down will
become obvious.

You need the fixedpoint.py module installed to run this.

    http://fixedpoint.sourceforge.net

(code attached)

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: fp-test.py
URL: <http://mail.python.org/pipermail/python-list/attachments/20020930/9f24f73d/attachment.ksh>

From tdelaney at avaya.com  Wed Sep 11 23:52:46 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 12 Sep 2002 13:52:46 +1000
Subject: "str.contains(part)" or alternatives?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A592@natasha.auslabs.avaya.com>

> From: Joseph A. Knapka [mailto:jknapka at earthlink.net]
>
> So in 2.3, can I do this and get the expected result?
> 
> >>> "" in "abc"
> 1

Yes - this has been hotly debated and defined as true.

> >>> for subst in "abc":
> ...   print subst
> ...
> <empty string>
> a
> ab
> abc
> b
> bc
> c

No. Iterating over a string is defined to return each single character in
turn. Remember:

    for subst in "abc":

is equivalent to

    for subst in iter("abc"):

and iter("abc") returns "a", "b", "c"

whilst

    if subst in "abc":

is equivalent to

    if "abc".__contains__(subst):

(falling back to if subst in iter("abc") if "abc" does not define
__contains__)

and str.__contains__ performs a substring match.

Tim Delaney



From peter at engcorp.com  Tue Sep 17 08:32:06 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 17 Sep 2002 08:32:06 -0400
Subject: faster way for adding many strings
References: <am6tgt$lqm$01$1@news.t-online.com>
Message-ID: <3d8720b8$1@news.sentex.net>

Ulli Stein wrote:
> Hi,
> 
> What is the faster way: appending many strings to a list and then joining 
> them, or writing to 
> StringIO and then getvalue()?

Alex' detailed response notwithstanding, there's another usual
answer to your question:

"Premature optimization is the root of all evil in programming"
(D. Knuth)

It means you usually shouldn't care which is faster until (a) your
software is working fully, and (b) you have profiled the code
and determined that in fact you have an unacceptable bottleneck
which must be optimized to meet your requirements.

Don't get me wrong, there _are_ times to optimize.  You just
shouldn't see them very often... and life is simpler that way. :)

-Peter



From kroc at blueyonder.co.uk  Tue Sep 24 04:18:45 2002
From: kroc at blueyonder.co.uk (Keith)
Date: Tue, 24 Sep 2002 08:18:45 GMT
Subject: Elementary confusion
References: <mailman.1032817169.21706.python-list@python.org>
Message-ID: <874rcf242i.fsf@smtp.blueyonder.co.uk>

>>>>> "Mark" == Mark McEahern <marklists at mceahern.com> writes:


    Mark> This is a FAQ.  Please read the FAQ.  This has to do with
    Mark> floating point math:

    Mark>   http://www.python.org/doc/FAQ.html#4.98

OK - read it. Thanks.

 I am reassured that it is not me, my machine, or a bug. However it
has rapidly become apparent that there are all sorts of pitfall
waiting me                              :-(

I guess thats where the fun is!         :-)

-- 
Keith.


From cbrown at metservice.com  Thu Sep  5 16:38:11 2002
From: cbrown at metservice.com (Colin Brown)
Date: Fri, 6 Sep 2002 08:38:11 +1200
Subject: Project idea
Message-ID: <3d77c0b4$1@news.nz.asiaonline.net>

Hi

If you are interested in a Windows platform project for automated testing
then read on.

There is an excellent freeware package "AutoIt" available from
http://www.hiddensoft.com/AutoIt/
that provides a DLL module that could be readily interfaced to Python.

"AutoIt" is a simple tool that can simulate key presses, mouse movements and
window commands (maximize, minimize, wait for, etc.) in order to automate
any windows based task (or even windowed DOS tasks).

Combining these features with say the ability to produce checksums of
snapshotted screen frames for
test comparisons would provide the basis for a nice GUI regression test rig.

Unfortunately I do not have the time to follow up on this myself so I though
I would pass it on.

Colin Brown
PyNZ





From joconnor at cybermesa.com  Fri Sep  6 23:11:13 2002
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sat, 07 Sep 2002 03:11:13 GMT
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <3D7958DB.50903@nospam.free.fr> <k08e9.25533$w51.7064239@news2.news.adelphia.net> <mailman.1031344988.4517.python-list@python.org>
Message-ID: <3d796e2f.44509032@news.cybermesa.com>

On Fri, 06 Sep 2002 15:41:52 -0500, "Bradley D. Larson"
<blarson at crary.com> wrote:

>That is what is nice about Vi   paren,bracket matching!

Vi, Emacs, half a hundred other code editors

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


"God himself plays on the bass strings first, when he tunes the soul"


From see_reply_address at something.invalid  Tue Sep 24 22:41:13 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Wed, 25 Sep 2002 14:41:13 +1200
Subject: Python Compiling
References: <amfveg$5g64k$1@ID-53055.news.dfncis.de> <j4lm5w9zft.fsf@informatik.hu-berlin.de> <tyf7khczziq.fsf@pcitapi22.cern.ch> <j4r8fkg7s8.fsf@informatik.hu-berlin.de>
Message-ID: <3D912249.6040409@something.invalid>

> Jacek Generowicz <jacek.generowicz at cern.ch> writes:
> 
> Results from one language don't transfer to another. I don't know
> (really) why Lisp compilers work that well, and what restrictions on
> the Lisp language had to be made to make them work well.


I think Lisp compilers work by a combination of help
from optional declarations, and heroic amounts of
type inference.

Also, Python is even more dynamic than Lisp in some
ways -- namespaces can have names added and removed
at run time, built-in operators can be overloaded,
etc. You can hardly rely on *anything* staying the
same from one moment to the next in Python!

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From rolf.kalbermatter at citeng.com  Mon Sep  2 10:41:59 2002
From: rolf.kalbermatter at citeng.com (Rolf Kalbermatter)
Date: Mon, 2 Sep 2002 16:41:59 +0200
Subject: version independant object usage counting
In-Reply-To: <j465xoldg2.fsf@informatik.hu-berlin.de>
Message-ID: <EKEJLGEFHHFGJOBEGJECMECBCEAA.rolf.kalbermatter@citeng.com>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Martin v. Lowis
> Sent: Monday, September 02, 2002 3:38 PM
> To: python-list at python.org
> Subject: Re: version independant object usage counting
>
>
> "Rolf Kalbermatter" <rolf.kalbermatter at citeng.com> writes:
>
> > However although this loader works quite fine now, I found out that
> > I gained little with it because the current Python architecture is
> > not really designed to link dynamically to it.
>
> Yes, this is the case - mainly because the API changes in slight ways,
> and because therefore, extension modules explicitly link to
> pythonxy.dll.

I realize that extension modules might get other problems with API changes
and that this is more likely to be a problem than the layout of the object
interface.

> > The problem arises in fact with the refcounting of Python
> > objects. The macros Py_INCREF and Py_DECREF access directly
> > structure members in the object layout and this layout is different
> > for debugging and non debugging versions and potentially different
> > between Python versions as well.
>
> That is only a minor problem: those macros are correct across all
> released Python versions, back to 1.4 or earlier.

But there is a serious difference between the DEBUG and NON-DEBUG
PyObject_HEAD structure layout which I would like to not have to worry
about if possible. As I'm really embedding Python (in this case into
LabVIEW), by trying to access the whole Python core as version independant
as possible I want to make it unimportant if the Python DLL is compiled
as DEBUG or non DEBUG or actually what version it really is.

> > Is there a specific reason that these two API functions are not
> > available and one is forced to use the according macros to do proper
> > refcounting, or did I miss those functions completely?
>
> Nobody has proposed to introduce them, because there is no need to do
> so, and because other ABI changes break the ABI more often - INCREF
> and DECREF have never done so, and it is unlikely that they ever will.
>
> > Any thought about this?
>
> Feel free to submit a patch to SourceForge. However, you might want to
> consider the following issues:

Does anyone see any problems with this code?

RCS file: /cvsroot/python/python/dist/src/Include/object.h,v
retrieving revision 2.111
diff -c -2 -r2.111 object.h
*** object.h	8 Aug 2002 20:55:20 -0000	2.111
--- object.h	2 Sep 2002 13:44:46 -0000
***************
*** 347,350 ****
--- 347,352 ----

  /* Generic operations on objects */
+ PyAPI_FUNC(int) PyObject_IncRef(PyObject *);
+ PyAPI_FUNC(int) PyObject_DecRef(PyObject *);
  PyAPI_FUNC(int) PyObject_Print(PyObject *, FILE *, int);
  PyAPI_FUNC(void) _PyObject_Dump(PyObject *);


diff -c -2 -r2.192 object.c
*** objects/object.c	24 Aug 2002 06:31:34 -0000	2.192
--- objects/object.c	2 Sep 2002 13:51:37 -0000
***************
*** 166,169 ****
--- 166,191 ----

  int
+ PyObject_IncRef(PyObject *op)
+ {
+ 	if (op) {
+ 		Py_INCREF(op);
+ 		return op->ob_refcnt;
+ 	}
+ 	return 0;
+ }
+
+ int
+ PyObject_DecRef(PyObject *op)
+ {
+ 	if (op) {
+ 		int refcnt = op->ob_refcnt;
+ 		Py_DECREF(op);
+ 		if (refcnt > 1)
+ 			return op->ob_refcnt;
+ 	}
+ 	return 0;
+ }
+
+ int
  PyObject_Print(PyObject *op, FILE *fp, int flags)
  {

This interface explicitedly would not require anymore to import _Py_Dealloc
and _Py_RefTotal which are only implemented and therefore exported from
DEBUG
core DLLs. It would still be anybodies own decision if to use the existing
macros or the according accessor functions instead.

> - various places assume presence of the ob_type field:
>   * the Py*_Check functions (e.g. PyList_Check) directly access the
>     ob_type field.

Hmm. I see your point in this. However I check this by having a private
PyObject_MyCheckType defined which accesses the ob_type field over the
PyObject_Type function and then checking for specific type through
PyObject_MyCheckType(op, &Py_<Type>).

>   * PyObject_INIT relies on initializing the object members.
> - PyType_HasFeature relies on the position of the tp_flags field

I'm not using these function at all until now. May have to do so when
I start to add an extension module however?

> - various accessor macros rely on the internal structure, e.g.
>   * PyTuple_GET_ITEM
>   * PyString_AS_STRING
>   * ...

I try to avoid them and use the according function interface instead such as
PyString_AsString() which was successful until now.

Rolf Kalbermatter
CIT Engineering Nederland BV    tel: +31 (070) 415 9190
Treubstraat 7H                           fax: +31 (070) 415 9191
2288 EG Rijswijk	http://www.citengineering.com
Netherlands		mailto:rolf.kalbermatter at citeng.com





From maurice at akst.tn.tudelft.nl  Fri Sep 13 08:48:14 2002
From: maurice at akst.tn.tudelft.nl (Maurice van de Rijzen)
Date: Fri, 13 Sep 2002 14:48:14 +0200
Subject: Evaluating strin composed of multiple parts?
Message-ID: <alsmqf$dke$1@news.tudelft.nl>

Dear all,
I'm using numpy.
In numpy one can take a part of a multi array just as in matlab.

  	partArray=totalArray[1:10,1;10,:,:,:

Now I've encounterd the following problem.
I'm using the followoing code:

class SubDataMatrix_5D:
	def SetInput(self,DataInput):
		self.DataInput=DataInput
	def SetDimensions(self,dim1,dim2,dim3,dim4,dim5):
		self.Dim1=dim1
		self.Dim2=dim2
		self.Dim3=dim3
		self.Dim4=dim4
		self.Dim5=dim5
	def CreateSubMatrix(self):
self.Sub=self.DataInput[eval(self.Dim1),eval...
		(self.Dim2),eval(self.Dim3),eval(self.Dim4),...
		eval(self.Dim5)]
	def GetOutput(self):
		return self.SubMatrix

temp2=SubDataMatrix_5D()
temp2.SetInput(temp.GetOutput())
temp2.SetDimensions("1:10","1:10","1:10","1:2","1:2")
temp2.CreateSubMatrix()


Doing this gives the following error:

self.SubMatrix=self.DataInput[eval(self.Dim1),eval(self.Dim2),eval(self.Dim3

),eval(self.Dim4),eval(self.Dim5)]
   File "<string>", line 1
     1:10
      ^

I know there goed something wrong with the strings but I have no clue of
how to solve it.
Preferrebly I would like to set the dimensions as follows:


temp2.SetDimensions(1:10,1:10,1:10,1:2,1:2)

Can someone please help me?


Maurice



From igetenoughspamalreadythanksjlr at doctor.com  Mon Sep  2 07:23:18 2002
From: igetenoughspamalreadythanksjlr at doctor.com (John La Rooy)
Date: Mon, 02 Sep 2002 21:23:18 +1000
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <mailman.1030907703.26266.python-list@python.org> <akuifb$h0e$2@solaris.cc.vt.edu> <1030962064.628232@ampungk.ozonline.com.au> <al0djh$a33$1@solaris.cc.vt.edu>
Message-ID: <1031008966.343893@ampungk.ozonline.com.au>

Carl Banks wrote:

>>It would be more likely to look like this (I haven't put too much 
>>thought into this)
> 
> 
> No kidding.
> 
> 
> 
>>"anything,anything,anything,same_as_3rd,same_as_2nd,same_as_1st"
>>or would you like to suggest something else?
> 
> 
> How about:
> 
> pattern = Group(Any()) + Group(Any()) + Group(Any()) \
>           + GroupRef(3) + GroupRef(2) + GroupRef(1)
> 
Err symantically that's exactly the same as the re and my suggestion
only the syntax is different. It's still nothing like saying

pattern = "6 character palindrome"

John



From marklists at mceahern.com  Fri Sep 20 12:32:19 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 20 Sep 2002 11:32:19 -0500
Subject: isFloat: Without Exception-Handling
In-Reply-To: <JHEOKEOOLIGLDHCMAHMOKECJEAAA.marklists@mceahern.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOMEDAEAAA.marklists@mceahern.com>

For what it's worth, here's a slightly tweaked version of the test harness.
It shuffles the function list before testing and does multiple test runs and
then averages the results.  I'm posting this mainly so that I can archive
this on this list without having to bother figuring out where to put it on
my own hard disk.  <wink>

isFloatExcept comes out pretty consistently almost twice as fast as the
RE-based approach.  Now, wouldn't I feel real dumb to find Python comes with
a builtin isfloat?

If not, are there other alternatives for doing this?

Easy come, easy go.

Cheers,

// m

#!/usr/bin/env python

import re
import random
import operator
from time import clock

# Compile up front.
floatPattern = re.compile(r"^[-+]?(\d+\.?\d*|\.\d+)([eE][-+]?\d+)?$")

def isFloatExcept(s):
    try:
        float(s)
        return True
    except (ValueError, TypeError), e:
        return False

def isFloatRE(s):
    floatRE = r"^[-+]?(\d+\.?\d*|\.\d+)([eE][-+]?\d+)?$"
    return re.match(floatRE, str(s)) is not None

def isFloatRECompiled(s):
    return floatPattern.match(str(s)) is not None

def time_it(fn, test_data, n):
    time_in = clock()
    for x in xrange(n):
        for (s, expected) in test_data:
            actual = fn(s)
            assert actual == expected
    time_out = clock()
    return time_out - time_in

def show(fn, r, fastest):
    avg = average(r)
    ratio = (avg / fastest) * 100
    print "%-20s  %1.3f  %2.f" % (fn.func_name, average(r), ratio)

def average(seq):
    return reduce(operator.add, seq) / len(seq)

def test():
    test_data = [(1.2, True),
                 (1, True),
                 (-1, True),
                 (1e+10, True),
                 ("1e-10", True),
                 ("banana", False),
                 ("5j", False),
                 (4+0j, False),
                 (isFloatExcept, False),
                 (__import__, False),
                 (clock, False)]
    test_runs = 10
    n = 100
    results = {}
    functions = [isFloatExcept, isFloatRE, isFloatRECompiled]
    for t in xrange(test_runs):
        random.shuffle(functions)
        for fn in functions:
            r = time_it(fn, test_data, n)
            results.setdefault(fn, [])
            results[fn].append(r)

    keys = results.keys()
    keys.sort(lambda x, y: cmp(average(results[x]), average(results[y])))
    fastest = average(results[keys[0]])
    print "%-20s  %s  %s" % ("name", "clock", "cmp")
    for fn in keys:
        show(fn, results[fn], fastest)

if __name__ == "__main__":
    test()
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: junk.py
URL: <http://mail.python.org/pipermail/python-list/attachments/20020920/9617ca91/attachment.ksh>

From remailer at xganon.com  Fri Sep  6 03:31:08 2002
From: remailer at xganon.com (xganon)
Date: Fri, 6 Sep 2002 02:31:08 -0500
Subject: Timothy Rue
Message-ID: <898662b7d85d2c966c0bd45d563d5d2d@xganon.com>

Does Timothy Rue still Troll here? 
If not, how did you get rid of him?
We need him out of our group now,
maybe we could get him into a Perl
group for a while and let them deal
with him?

(:



---
This post was anonymized at http://www.xganon.com
---



From dsavitsk at e-coli.net  Wed Sep 18 00:50:34 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Wed, 18 Sep 2002 04:50:34 GMT
Subject: COM object registration
References: <afph9.9468$yt3.4669302@newssrv26.news.prodigy.com> <T_Rh9.17227$Ee4.51953@news-server.bigpond.net.au>
Message-ID: <uCTh9.10343$yt3.4866032@newssrv26.news.prodigy.com>

"Mark Hammond" <mhammond at skippinet.com.au> wrote in message
news:T_Rh9.17227$Ee4.51953 at news-server.bigpond.net.au...
> dsavitsk wrote:
> > if the module below is run from the command line, it registers and works
> > fine.  if, however, it is imported and the register() method is executed
it
> > does not work (giving a 'localserver' error).
>
> What doesn't work?  Registration or subsequent use of the object?  Can
> you be more specific about the error you got, or else attach code I can
> run without modification to see the error.
>
> Thanks,
>
> Mark.
>

here are 2 modules. call the first anything, I called the second
REGISTER_MODULE2.py for the example. (I tend to do anything COM related in
caps). Place them both on the path on one computer, and place the second on
a second computer.

the problem is illustrated below.

# -------- 1st ------
"module to register Python COM components"
import pythoncom
class regmodule:
    _reg_clsid_ = '{7E63BF81-E120-4CD9-A9AE-13B5B6BB4ADB}'
    _reg_clsctx_ = pythoncom.CLSCTX_LOCAL_SERVER
    _public_methods_ = ['REGISTERM']
    _reg_progid_ = 'this_is.MODULE_REGISTER'

    def REGISTERM(self, mod):
        exec('import ' + str(mod) + ' as mr')
        mr.register()
if __name__ == '__main__':
    import win32com.server.register
    win32com.server.register.UseCommandLine(regmodule)
# -------------------
# -------- 2nd ------
import pythoncom
class regmodule2:
    _reg_clsid_ = '{2B1120DD-83F8-4850-91F6-B4F9601C16B7}'
    _reg_clsctx_ = pythoncom.CLSCTX_LOCAL_SERVER
    _public_methods_ = ['GET_VALUE']
    _reg_progid_ = 'ecp.MODULE_REGISTER2'
    def __init__(self):
        self.x = 0
    def GET_VALUE(self):
        self.x += 1
        return self.x # it needs to do something :-)
def register():
    import win32com.server.register
    win32com.server.register.UseCommandLine(regmodule2)
if __name__ == '__main__': register()
# -------------------

1. register the first on computer 1
2. register the second on computer 2
3. do the following on computer 1
# instance the server
>>> f = win32com.client.Dispatch('this_is.MODULE_REGISTER')
# use it to locally register other server
>>> f.REGISTERM('MODULE_REGISTER2')
# instance the second on server
# my server is named camus
>>> a = win32com.client.DispatchEx('ecp.MODULE_REGISTER2', 'CAMUS')
# works fine
>>> a.GET_VALUE()
1
>>> a.GET_VALUE()
2
>>> del a
# try instance it locally
>>> a = win32com.client.Dispatch('ecp.MODULE_REGISTER2')
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
  File "C:\Python\win32com\client\__init__.py", line 92, in Dispatch
    dispatch, userName =
dynamic._GetGoodDispatchAndUserName(dispatch,userName,clsctx)
  File "C:\Python\win32com\client\dynamic.py", line 81, in
_GetGoodDispatchAndUserName
    return (_GetGoodDispatch(IDispatch, clsctx), userName)
  File "C:\Python\win32com\client\dynamic.py", line 72, in _GetGoodDispatch
    IDispatch = pythoncom.CoCreateInstance(IDispatch, None, clsctx,
pythoncom.IID_IDispatch)
com_error: (-2147467259, 'Unspecified error', None, None)

if you go back and regularly register the second module locally (double
clicking or what not) the local instancing will work fine. In some sense,
this might be useful to only allow remote use, but i suppose this can be
done by distributing a shell of the module as well.

Thanks,

-doug





From pedronis at bluewin.ch  Tue Sep 17 15:02:51 2002
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Tue, 17 Sep 2002 21:02:51 +0200
Subject: Calling Python from C# - please help
References: <6de16ac9.0209130508.5d1b56c6@posting.google.com> <j4elbyt3yw.fsf@informatik.hu-berlin.de> <ur8fyxau0.fsf@online.no> <0iCg9.2350$Ee4.6516@news-server.bigpond.net.au> <3D82FA5B.3050003@mxm.dk> <uptvgvn1h.fsf@online.no> <3%Dh9.15217$Ee4.45118@news-server.bigpond.net.au>
Message-ID: <3d877db8_5@news.bluewin.ch>

Mark Hammond <mhammond at skippinet.com.au> wrote in message
3%Dh9.15217$Ee4.45118 at news-server.bigpond.net.au...
> Syver Enstad wrote:
...
> >
> > Now that J# (Java .NET) has been released, I was wondering about the
> > possibility of compiling Jython with the MS J# compiler to create a
> > Jython .NET. Has anyone considered the possibility?
>
> I'd love to see some Jython technology used for .NET.  At the time
> Python.NET started, there was very strict NDA, so discussing the project
> with jython-dev, or even python-dev, was taboo.  Greg Stein and I made a
> pragmatic decision to go with the p2c tool as a framework over jython as
> we alrady had experience with p2c and didnt have the time-lines to
> understand jython before embarking on our pilot code.

How were you going to support eval and exec?

For the moment the unload granularity of .NET is AppDomains only and only
upon explicit request, and types/modules/assemblies are not garbage
collected but simply live as much as the AppDomain,

that's OK for static languages, or environments with a clear run/design time
distinction (one can setup an AppDomain and tear it down for each
run-cycle), or for running scripts in isolation but not if one wants IL/JIT
speed for a enviroment that allows for redefinitions and eval, ...

In Jython we create Java bytecode and dynamically load classes for all the
code, also eval code because in Java classes are elegible for garbage
collection.

In .NET it seems one needs a pure interpreter that does not compile to IL
for exec and eval support, and that seems what JScript does for jsc compiled
code containing eval, that means the script is compiled but the evaluated
code is only interpreted,

otherwise a long running process using eval could go out of memory.

That's something people wanting to write a feature complete impl of Pyhton
for .NET should consider or wait MS to implement a more
dynamic-environment-friendly unload policy.

This is also an issue versus runtime subclassing of framework classes,
Jython creates distinct concrete Java subclasses each time a Java class is
subclassed, under .NET one should probably limits this  ...

regards.




From gerhard.haering at opus-gmbh.net  Fri Sep 20 07:32:20 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 20 Sep 2002 11:32:20 GMT
Subject: M$SQL connections from linux
References: <1032528860.599940@news.etel.ru>
Message-ID: <slrnaom1sc.2e0.gerhard.haering@haering.opus-gmbh.net>

Vasiliy A. Chernoivan wrote:
> Hello all
> Is there some module for python which performs the task?
> Is it possible at allto access data under M$SQL from linux machine?

What /I/'d do first is ask Google, like with:

http://groups.google.de/groups?q=group%3Acomp.lang.python+microsoft+sql+server+linux&ie=UTF-8&oe=UTF-8&hl=en&btnG=Google+Search

I remember reading here that Dave Cole's Sybase module can used to access SQL
Server in combination with the FreeTSD library.

-- Gerhard


From wlfraed at ix.netcom.com  Wed Sep 25 23:19:37 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Wed, 25 Sep 2002 20:19:37 -0700
Subject: transform strings list in a int lists or other type???
References: <mailman.1032996693.1693.python-list@python.org>
Message-ID: <9cutma.ov3.ln@ix.netcom.com>

jubafre at brturbo.com fed this fish to the penguins on Wednesday 25 
September 2002 04:29 pm:

        Haven't you asked a version of this /each/ day for the last few days?

>                   Can i tranform a list of strings in a list of int,
>                   or other type in python???
> and also write in a file? I think the open(path,'wb') in __builtin__
> dont?t write in a file other types, just strings, rigth???
> 

        ALL I/O in Python is effectively a stream of bytes. The "b" option on 
the open ONLY controls the interpretation of line-endings. IE, on 
Windows reading a file with "b" means you see both the <cr> and the 
<lf> character; without the "b" (normal mode) the <cr><lf> is converted 
to just a "newline" (<lf>) for internal use.

        To write a "binary" file requires you to either manipulate /each/ byte 
of your data, OR to use the struct module to pack your data into a 
string (full of unprintable characters) which can be written.

import struct
ots = struct.pack(">2L2H12s30pB", 0x10346f89, 563993, 0x0d0a, 32767,
        "Some Text", "More text with a count byte", ord("\r")

print ots

Lots of unprintables in that... Cut&Pasted from the terminal window:
4o?       ore text with a count byte

print repr(ots)

'\x104o\x89\x00\x08\x9b\x19\r\n\x7f\xffSome Text\x00\x00\x00\x1bMore 
text with a count byte\x00\x00\r'

        I used the ">" in the pack format string to ensure that the byte 
sequence for output is "big-end first" so the output representation is 
the same as the input numbers -- default on my machine is "little-end 
first".

Compare:
0x10346f89      to the output... You have to take into account that 
printable characters show up as single byte characters rather than an 
encoded \xnn set.

0x10   34   6f   89
\x10    4    o \x89

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From k.robert at gmx.de  Thu Sep 26 16:39:36 2002
From: k.robert at gmx.de (Robert)
Date: Thu, 26 Sep 2002 22:39:36 +0200
Subject: http file upload from Python (multipart/formdata) how-to?
Message-ID: <3d937091$0$181$9b622d9e@news.freenet.de>

want to upload file data (multipart formdata) - same as upload from
web-browser <input type=file ..>

urllib.urlencode  + urlopen takes only simple name/value tuples

how do I render multipart formdata ?

Robert









From syver-en+usenet at online.no  Mon Sep  9 17:22:52 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Mon, 09 Sep 2002 21:22:52 GMT
Subject: win32com.client.gencache and py2exe
References: <uit1ftba1.fsf@online.no> <alih9e$1qjt12$1@ID-59885.news.dfncis.de>
Message-ID: <uptvmrh7u.fsf@online.no>

"Thomas Heller" <theller at python.net> writes:

> It's already fixed in CVS. It should be sufficient if you replace
> the existing build_exe.py by the new one which you can grab
> from the CVS repository:
> http://cvs.sourceforge.net/cgi-in/viewcvs.cgi/py2exe/py2exe/py2exe/build_exe.py

Cool, I'll check it out.

-- 

Vennlig hilsen 

Syver Enstad


From sholden at holdenweb.com  Fri Sep 13 07:43:31 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 13 Sep 2002 07:43:31 -0400
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid> <alln71$1s2s$1@nntp6.u.washington.edu> <b16e4ef7.0209110616.48dfe04c@posting.google.com> <3D7FFC43.10002@something.invalid> <b16e4ef7.0209120504.62d3c5bd@posting.google.com> <alqiue$1ejo$1@nntp6.u.washington.edu> <alqt2q$q8u$1@reader1.tiscali.nl>
Message-ID: <%ckg9.3117$WV.1764@fe03>

"GerritM" <gmuller at worldonline.nl> wrote ...
> "Russell E. Owen" <owen at nospam.invalid> schreef ...
> <...snip...>
> > True. It would help if the command callback sent an arguement that
> > identified the widget (making it look a bit more like an event binding).
> > In any event, it doesn't and command is unfortunately still the best way
> > to handle button widgets.
>
> In a recent thread it was pointed out that you can also bind to s function
> of a specific object instance. Out of the top of my head:
>
> Class A:
>     def printarg( self, arg ):
>         print arg
>
> anA = a()
> functionBoundToInstance = anA.printarg
>
> functionBoundToInstance("Hello Python")
>
> Hello Python
>
> this would work also for widgets, command = aWidget.activate
>
It also works for widget subclasses, which means that you have "self"
available to access various attributes of the widget. But it still doesn't
get you the event data, which only get explicitly passed when you establish
a binding for an event rather than using the command argument.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From astrophels at yahoo.com  Fri Sep  6 12:48:40 2002
From: astrophels at yahoo.com (Micah Mayo)
Date: Fri, 06 Sep 2002 16:48:40 GMT
Subject: Getting a proccesses ID
References: <Qu3e9.188987$_91.214523@rwcrnsc51.ops.asp.att.net> <mailman.1031326272.10078.python-list@python.org>
Message-ID: <I%4e9.190045$_91.214443@rwcrnsc51.ops.asp.att.net>

Sean 'Shaleh' Perry wrote:
> On Friday 06 September 2002 08:05, Micah Mayo wrote:
> 
>>Hi,
>>
>>I'm writing a little tool for SysAdmins on FreeBSD - I'm trying to make
>>it as portable as possible(as far as the unices go), and have come upon
>>a problem - one of the things this tool does is restarts various deamons
>>on the system, in order to do that I have to send a killsignal to the
>>pid. The problem is getting the process id - right now I've got a hack
>>using ps, grep, and awk, which I doubt will translate well to other
>>flavors of *nix because they all seem to have their own method of
>>delivering info via ps. Can anyone tell me a cross platform way of
>>grabbing a particular processes ID
>>
>>Here's what I'm doing now, by the way:
>>
>>pid = os.popen("ps -x | grep sshd | grep -v grep | awk '{ print $1 }'")
>>pid = pid.atoi()
> 
> 
>>>>import os
>>>>os.getpid()
>>>
> 29421
> 
> (-:
> 
Ah, but os.getpid get's my pid, I need to get another programs, and in 
fact, several other program's pid. :)

Thnx,
Micah



From marklists at mceahern.com  Wed Sep 25 10:50:38 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 25 Sep 2002 09:50:38 -0500
Subject: List/Tuple bug or feature?
In-Reply-To: <87ofamgnbh.fsf@nospam.eml.cc>
Message-ID: <JHEOKEOOLIGLDHCMAHMOAEGIEBAA.marklists@mceahern.com>

[Mahesh Padmanabhan]
> It is simply a list of constants so I could not generate it
> automatically. I then use this list to filter out one element of the
> list using the builtin filter function. I don't think I can break down
> the list so that I can view missing commas easier because all
> constants seem to logically group together. Something like this:
>
> constantList = [1,2,3,4,5]
>
> filteredList = filter(some_function, list)

How does the above example resemble your original problem, though?

You said it's a list of constants.  Are they stored in a file?  In what
format?  Do you see where my questions are leading?  I'm basically trying to
find out how you happened to get:

  ['constant1', 'constant2' 'constant3', 'constant4', ..., 'constantn']
                           ^
                            \-- missing comma

Cheers,

// mark




From whisper at oz.net  Wed Sep 25 02:51:00 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 24 Sep 2002 23:51:00 -0700
Subject: PythonWin CVS problem
In-Reply-To: <L4ak9.478$vJ5.43427085@newssvr21.news.prodigy.com>
Message-ID: <GCEDKONBLEFPPADDJCOEIENOFCAA.whisper@oz.net>

I'm using the same repository I've been using for the past 6 months or more:
	:pserver:anoncvs at cvs.pythonpros.com:/home/cvsroot

Just tried it again and got the same: "permission denied".

David LeBlanc
Seattle, WA USA 

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Dilton McGowan II
> Sent: Tuesday, September 24, 2002 20:30
> To: python-list at python.org
> Subject: Re: PythonWin CVS problem
> 
> 
> David LeBlanc wrote:
> 
> > Second day it's inaccessable with "permission denied"
> > 
> > David LeBlanc
> > Seattle, WA USA
> 
> Are you using cvs.pythonpros.com? It worked for me yesterday and 
> again just 
> now. They seem to have the latest source. Just a friendly suggestion.
> -- 
> http://mail.python.org/mailman/listinfo/python-list



From kkuroda at crl.ucsd.edu  Tue Sep  3 12:31:37 2002
From: kkuroda at crl.ucsd.edu (Kow Kuroda)
Date: Tue, 3 Sep 2002 09:31:37 -0700
Subject: RedHat 7.3 and Python2.2.1
In-Reply-To: <3d74a6e8@news.sentex.net>
Message-ID: <9DF2DF67-BF5A-11D6-8D0D-00050287C0A6@crl.ucsd.edu>

On Tuesday, September 3, 2002, at 05:11 , Peter Hansen wrote:

> Kow Kuroda wrote:
>> As a last thought, doesn't Red Hat have different versions of the 7.3 
>> distribution?
>
> Maybe you installed a beta?

I don't know because I just used the installer available at an institute I'
m working, and I'm back to RHL 7.2 now.

> I'm pretty sure there's only one release version to date.  I could be 
> wrong.  The one I installed I took from
> their web site just two weeks ago.  uname says it's version:
>
>   2.4.18-3 #1 Thu Apr 18 07:37:53 EDT 2002 i686
>
> and /etc/redhat-release says "release 7.3 (Valhalla)"

I'm not on 7.3 now, so I can't tell exactly which version/release it was.
Nevertheless, it was surely release 7.3 Valhala, though the kernel version 
was different, I suspect.

>
> Oh well, there's no telling with software. :)

Very true ;-)

kOw




From joconnor at cybermesa.com  Tue Sep  3 11:53:42 2002
From: joconnor at cybermesa.com (Fearless Freep)
Date: 3 Sep 2002 08:53:42 -0700
Subject: Why it is so dramatical?
References: <mailman.1030957743.4021.python-list@python.org> <mailman.1030962662.4274.python-list@python.org> <20020902.104707.1789376348.1554@cybermesa.com> <mailman.1031039705.9065.python-list@python.org>
Message-ID: <7d3dc526.0209030753.27bf08a8@posting.google.com>

"Bo M. Maryniuck" <b.maryniuk at forbis.lt> wrote in message news:<mailman.1031039705.9065.python-list at python.org>...
> On Monday 02 September 2002 05:47, Jay O'Connor wrote:
> > As a result Perl is much faster at some kinds of operations than
> > Python, and vice-versa.
> 
> Yeah, actually, but I just would like to find some optimal Python solutio
> n :)

Well..is it more efficient to treat it as a list of characters?

>>> x = " " * 10
>>> x = list (x)
>>> y = 'test str'
>>> x[0:len(y)] = list (y)
>>> z = 'another test string'
>>> x.extend(list(z))
>>> s =''.join (x)
>>> print s

'test str  another test string'

Treating a string as a list of character elements would allow you to
append to it and change the internals.   May be an improvement in
performance.

This assumes you are having a noticeable performance problem with
concatenating strings.  Beware of premature optimization.  Just
because an fairly simple operation in one language is slower than the
equivalent in another language doesn't mean that the program as a
whole will be slower enough to matter

Take care,
Jay


From fromclay at maxitec.co.za  Tue Sep  3 10:36:40 2002
From: fromclay at maxitec.co.za (Morne)
Date: Tue, 3 Sep 2002 16:36:40 +0200
Subject: heeelp
Message-ID: <001f01c25357$53cb7ca0$2500a8c0@maxitec.co.za>

Could someone please help me with this?
Im really new to any kind of programming 
 
If posible would you be
able to answer in detail.
Thanx 
morne

is it wrong of me to send the ppp.log file to just anyone?

HERE IS WHAT I HAVE:

import re
import string

secondsMatcher=re.compile("Connect time: (\d+) secs")

def getStuff(file):
    closeline = 0
    errorcount = 0
    secondscount = 0
    connectcount = 0
    data_outcount = 0
    data_incount = 0
    ppplist = file.readlines()

    for line in ppplist:

      if string.find(line,"LayerUp") >= 0:
          print "found connect"
          connectline = string.split(line)
          connectcount = connectcount + 1

      elif string.find(line,"ERROR") >= 0:
          print "found error"
          errorcount = errorcount + 1

      elif string.find(line,"Connect time") >= 0:
          print "found close"
          connectSecs=secondsMatcher.search(line).group(1)
          print "Connected for",connectSecs
          closeline = string.split(line)
          secondscount = secondscount + string.atoi(connectSecs)

def main():
    f=open("C:/my documents/ppp.log","r")
    getStuff(f)
    return None

if __name__=="__main__":
    main()

THIS IS WHAT I NEED THE PROGRAM TO DO:

I want the program to read the attached ppp.log file and give me the
following type of report:


PPP Log file summary for 5 Jul 2002
-----------------------------------
Jul  5 18:15:01 Connected to internet (IP 155.239.150.146)
Jul  5 18:18:08 Closed connection. ( Connect time: 197 secs: 5091 octets in,
1864 octets out )
-----------------------------------
Successful connections: 5
Failed connections: 2
Total time online: 456 seconds
Total data in: 66576 octets
Total Data out: 66349 octets














-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020903/5d02f49d/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ppp.log
Type: application/octet-stream
Size: 33059 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020903/5d02f49d/attachment.obj>

From sismex01 at hebmex.com  Wed Sep  4 09:36:57 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Wed, 4 Sep 2002 08:36:57 -0500 
Subject: Daemonize?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233022B66EC@mail.hebmex.com>

Sorry to ask here, but I don't have any "unixy" python
documentation at hand.

Normally, I put an app into "daemon" mode, by first
forking, then reopening stdin/out/err to /dev/null,
and then forking again. Works good enough for my
purposes.

But, am I right? Is this the correct way to do this?
Also, I can't find any "reopen" in python to apply
to sys.stdin/out/err.

Any help will be most appreciated.

Thanks.

-gus
--

Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From marklists at mceahern.com  Wed Sep 18 07:16:43 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 18 Sep 2002 06:16:43 -0500
Subject: Parsing Logs...Please Advise
In-Reply-To: <CXSh9.504829$UU1.83831@sccrnsc03>
Message-ID: <JHEOKEOOLIGLDHCMAHMOGEHKDPAA.marklists@mceahern.com>

[sensei zeki]
> I am looking for a good tool to parse system logs.

You provided a sample log, which is great.  How about fleshing out what you
mean by 'parsing'?  I.e., what output/results do you want?  Or are you
hoping we fill in that blank for you?  <wink>

Cheers,

// m

-




From TOGLIMIcuni at programmazione.it  Fri Sep 20 10:19:01 2002
From: TOGLIMIcuni at programmazione.it (Antonio Cuni)
Date: Fri, 20 Sep 2002 14:19:01 +0000
Subject: Copy with __slots__
References: <mailman.1032419786.7187.python-list@python.org> <amd082$r5j$1@newsreader.mailgate.org> <cUoi9.134531$pX1.4784251@news2.tin.it>
Message-ID: <amfj0b$lb2$1@newsreader.mailgate.org>

Alex Martelli wrote:

> This doesn't deal with inherited __slots__. 
[cut]

you're right, I couldn't think of it :-(

> Some care
> is needed, though (I can think of several possible anomalies).

what kind of anomalies?

Anto
-- 
"Computer science is not about computers any more than astronomy
is about telescopes." -- EW Dijkstra


From tdelaney at avaya.com  Wed Sep  4 19:16:12 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 5 Sep 2002 09:16:12 +1000 
Subject: How to force built-in commands over imported.
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A574@natasha.auslabs.avaya.com>

> From: Simon Brunning [mailto:SBrunning at trisystems.co.uk]
>  
> You are, I assume, importing 'os' as follows:
> 
> from os import *
> 
> *Don't* *do* *this*. Use:
> 
> import os
> 
> instead. There *are* reasons for using the former method, but 
> they are few
> and far between. The latter form should always be the one you 
> use unless you
> have a good reason.

Indeed. The only time I have *ever* found to do this (I don't use TKinter ;)
is in the following scenario:

	# UnitTests.py

	import unittest

	from tests1 import *
	from tests2 import *

	if __name__ == '__main__':
		 unittest.main()

	# tests1.py

	import unittest

	class Test1 (unittest.TestCase):

		 def test1 (self):
		pass

	if __name__ == '__main__':
		 unittest.main()

	# tests2.py

	import unittest

	class Test2 (unittest.TestCase):

		 def test2 (self):
		pass

	if __name__ == '__main__':
		 unittest.main()

which allows me to perform each set of tests independently, and run them all
through a single command. Even so, it should probably specify the actual
names of the test cases in case there is a conflict somewhere, but that
would fail to pick up any new test cases.

Tim Delaney



From a at b.c  Sun Sep 29 02:29:12 2002
From: a at b.c (Michael Bakkemo)
Date: 29 Sep 2002 06:29:12 GMT
Subject: Why doesn't this work?
Message-ID: <an66jo$cb6$0@216.39.145.102>

Why doesn't the rectangle start somewhere in the middle of the canvas (where
one would expect (40,40) to be, instead of all the way over on the left?

---

import Tkinter
import Pmw
import tkFont


class MovingRect:
    def __init__(self, master, sc, x, y,width, height):
        self.master = master
        self.sc = sc
        self.x = x
        self.y = y
        self.width = width
        self.height = height

        self.rectID = self.sc.create_rectangle(self.x, self.y,
self.x+self.width, self.y+self.height,fill="black", outline="black")

        self.poll()

    def poll(self):
        sc.move(self.rectID, 4, 0)

        self.timerID = self.master.after(500, self.poll)


    def width(self, width):
        self.width = width

    def height(self, height):
        self.height = height


    def Destroy(self):
        self.master.after_cancel(self.timerID)


def bail():
    menu.quit()
    rect.Destroy()


root = Tkinter.Tk()

# create a menu
menu = Tkinter.Menu(root)
root.config(menu=menu)

Quitmenu = Tkinter.Menu(menu)
menu.add_cascade(label="Quit", menu=Quitmenu)
Quitmenu.add_command(label="Quit", command=bail)

sc = Pmw.ScrolledCanvas(root, borderframe=1, hscrollmode='dynamic',
vscrollmode='dynamic')

sc.xview("moveto", 0)
sc.yview("moveto", 0)

sc.resizescrollregion()
sc.pack(fill='both',expand=1)

rect = MovingRect(root, sc, 40,40, 10, 10)
root.mainloop()





From greek_bill at yahoo.com  Wed Sep 18 16:58:13 2002
From: greek_bill at yahoo.com (Morfeas)
Date: Wed, 18 Sep 2002 21:58:13 +0100
Subject: SWIG and C++
References: <Xo3i9.3567$8t.155902@newsfep1-win.server.ntli.net>
Message-ID: <oM5i9.844$Ew4.62552@newsfep2-gui>

Morfeas wrote:

> I don't know if this is the right place to post this, but I couldn't find
> anywhere better.
> 
...
> 
> Thanks!
> Morfeas
> 

It's ok I got the answer now! I was reading the documentation for an earlier 
version of SWIG and I didn't name the shared library correctly (missed the 
underscore!)

Morfeas
-- 
www.freecfm.com/m/morfeas


From aleax at aleax.it  Tue Sep 24 03:33:00 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 24 Sep 2002 07:33:00 GMT
Subject: Python threading?
References: <f9Pi9.10275$Lg2.2191294@news2.news.adelphia.net> <amogmi$r1l$1@panix1.panix.com> <T5Rj9.13058$6g7.39977@news-server.bigpond.net.au>
Message-ID: <MyUj9.134183$ub2.2937335@news1.tin.it>

Mark Hammond wrote:
        ...
> I would have thought that a brave call.  If you are performing lots of
> IO and need many many connections, you are almost certainly better off
> using an "asynchronous" approach, much like asyncore.  With a large

Yes BUT -- as of course you know, select, on Windows, can only deal with
sockets.  Other kinds of asynchronous I/O require a completely different
architecture on Windows wrt Unixoid systems -- file I/O, specifically,
has the best performance with Windows' on asynchronous-IO-with-callbacks-
on-completion, an arrangement very reminiscent of old VMS systems.  Even
for sockets, I benchmarked back in NT 3.something times that "native"
(async-with-messages) winsock was vastly more scalable than "BSD emulating
sockets and select" (not sure if that's still true these days).

twisted.internet does have multiple implementations of the Reactor
design patterns, including among others both a portable one based
on select, and a Windows-specific one with a specialized message loop
at its core (and others yet, such as ones working with GUI frameworks'
events, poll in lieu of select, etc) -- haven't examined implementation
in detail, but that does seem like the only sane approach if you need
servers that are BOTH cross-platform AND highly scalable.

> number of threads (where "large" can be as low as single figures) the OS
> overhead of switching will be greater than your internal overhead of a
> more complicated design.

I guess so, depending of course on the OS -- threading is more and
more likely to be well-optimized these days, but still the context
switch has to cost something compared to a "cooperative" asynchronous
architecture you can implement within one thread (or a very few
threads of your own -- twisted lets you do that, too).


> A "one thread per connection" model is doomed from the start -

Hear, hear.

> therefore, in most applications, I would consider 5 threads a large
> number - particularly when all are actively doing something.  OTOH, if
> you are willing to accept that your first version will accept a max of
> ~20 connections, the brain-power saved by using threads beats the
> additional OS overhead by a mile ;)

I think it only LOOKS simple if you don't look into it deeply enough.
If the threads have almost no need to cooperate or share resources,
then you may well be right.  If the cooperation needs are limited
enough that you can smoothly deal with them with Python "Queue"
instances, you still have a chance to be right.  But generally, I
find that Ousterhout was right -- I'm sure you're familiar with his
talk on "Why threads are a bad idea for most purposes" (I guess one
can still find the slides on the net)... his point is more about
"brain-power saved" (or consumed by debugging in a just-too-hard
environment, with transient errors, race conditions, etc, etc) by
using event-driven (i.e., async) approaches rather than threading,
rather than about performance and scalability.

Python may not yet support event-driven programming quite as smoothly
and seamlessly as Ousterhout's Tcl, but, thanks to such developments
as twisted, I think we're drawing progressively closer.  With a solid
enough underlying framework to dispatch events for you, I think event
driven programming (perhaps with a few simple specialized threads off
to one side, interfacing to some slow external resource, and getting
work requests off a Queue, posting results back to a Queue...) can
be made conceptually simpler as well as more scalable than threading.
After all, in event-driven / async programming, ONE thing is happening
at a time -- you only "switch" at known points, and between two such
points you need not worry about what is or isn't atomic, locking, &c.
It's also easier to debug, I think...


Alex
 


From Janssen at rz.uni-frankfurt.de  Tue Sep 17 04:37:37 2002
From: Janssen at rz.uni-frankfurt.de (Michael Janssen)
Date: Tue, 17 Sep 2002 10:37:37 +0200
Subject: pickle magically not working
References: <3d865dac@nntp.server.uni-frankfurt.de> <mailman.1032224577.11771.python-list@python.org> <m3it15cgpv.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3d86ea57@nntp.server.uni-frankfurt.de>

Martin v. Loewis wrote:

> jepler at unpythonic.net writes:
> 
>> If I had to guess, I'd say the "fix" was in Pickle itself, you might
>> turn
>> something up by investigating Python's CVS tree.  Or somebody else
>> might be able to immediately indicate the relevant code...
> 
> It's in structseq.c:structseq_reduce (rev. 1.4 and 1.5), also fixed in
> Python 2.2.1.
> 
> Regards,
> Martin

thanks much,

now i can comment out my workaround and be sure it won't happen with 
any other object's.

Michael

PS: happy to see that every "black magic" is downtrackable :-)



From martin at v.loewis.de  Fri Sep 20 03:37:45 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 20 Sep 2002 09:37:45 +0200
Subject: Suggesting a filename on a HTTP-request by sending someting in the header of the reply
References: <Pusi9.20169$0p1.363761@news2.ulv.nextra.no>
Message-ID: <m33cs5un5i.fsf@mira.informatik.hu-berlin.de>

"Thomas Weholt" <2002 at weholt.org> writes:

> It seem to be the "dl"-part of the url above. I've tried to read the
> RFC 2616-document and sent "Content-Disposition"-headers etc. but
> with no luck.

You had no luck reading a document ?-)

What browser are you using? Many Mozilla releases ignore the
Content-Disposition field. Try downloading a patch from SourceForge
with your browser, and see whether it uses the SF file name, or
"download.php". SF works correctly in this respect, so if it won't
work for you, it's a browser bug.

Regards,
Martin



From syver-en+usenet at online.no  Wed Sep 25 10:57:18 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Wed, 25 Sep 2002 14:57:18 GMT
Subject: help
References: <amqclg$6uu$1@server.ltid.inpe.br>
Message-ID: <ur8fi5d8k.fsf@online.no>

"Juliana Braga" <juliana at pdi.inpe.br> writes:

> I need help.
> For example:
> x = ?name1?
> How can I get the value of variable x (in this case the value of
> variable x
> 
> is ?name1?). After , I?d like to do this:
> value of variable x = ?teste?.

I am not sure I understand what you mean, but here's a shot:

x = 'name1'
saved = x
x = 'teste'
print x
teste
print saved
name1

Hope this helps.



-- 

Vennlig hilsen 

Syver Enstad


From jonathan at onegoodidea.com  Mon Sep 30 03:32:49 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Mon, 30 Sep 2002 08:32:49 +0100
Subject: [Mac] Editor on Macintosh?
References: <1fj7za2.1erbqps1uvzpj8N%news@agapow.net> <B9BC89EF.12189%jonathan@onegoodidea.com> <290920020905027300%pecora@anvil.nrl.navy.mil>
Message-ID: <B9BDBCB1.12206%jonathan@onegoodidea.com>

On 29/9/2002 14:05, in article 290920020905027300%pecora at anvil.nrl.navy.mil,
"Louis M. Pecora" <pecora at anvil.nrl.navy.mil> wrote:

> From the looks of it this is OS X only, right?  -- Lou

The shell stuff is OS X only (though you could always get MPW for the
authentic experience under 9). Otherwise, BBEdit is a Carbon app that runs
fine on 9.

Jonathan



From aleax at aleax.it  Wed Sep 18 11:26:31 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 18 Sep 2002 17:26:31 +0200
Subject: monetary applications [was: Python GUI app to impress the bos s?]
In-Reply-To: <31575A892FF6D1118F5800600846864DCBD817@intrepid>
References: <31575A892FF6D1118F5800600846864DCBD817@intrepid>
Message-ID: <E17rgjC-0002ci-00@mail.python.org>

On Wednesday 18 September 2002 04:34 pm, Simon Brunning wrote:
> > From:	Alex Martelli [SMTP:aleax at aleax.it]
> >
> > > Perhaps the lack of a FixedPoint type?
> >
> > http://starship.python.net/crew/aahz/FixedPoint.py
>
> A candidate for the standard library, perhaps? I use it all the time, and

Definitely (Aahz's Decimal may be a better alternative), but it needs a
paladin to convince the BDFL to include it in the core.  Right now, it
seems to me that FixedPoint and Decimal are not even among the
most popular extensions -- Numeric, mxDateTime and other Lemburg
extensions, and many others, seem to be vastly more popular.

> if nothing else, it would be nice for DB-API modules to know that it was
> there, so that they could return instances of it for NUMERIC and DECIMAL
> values.

To play devil's advocate: DB-API compliant modules have long used
mxDateTime happily, for date-time values, even though mxDateTime
was not (and likely will never be) in the core.  Why is FixedPoint (or
Decimal or some future Currency -- whatever) different from mxDateTime
in this regard?  How come DB-API compliant modules seem not to
even have felt any need to fish around for a type with which to
represent NUMERIC and DECIMAL, as they did for dates and times?

I don't know the answers, but if "help out the DB-API" is among the
arguments you plan to muster to convince the BDFL, you might
want to be prepared in case these questions get raised.  No, sorry,
I won't volunteer as a paladin for this PEP -- I'll gladly help out with
code and even drafting the PEP if need be, but this isn't really my
own crusade...


Alex



From rajarshi at presidency.com  Mon Sep 16 10:05:30 2002
From: rajarshi at presidency.com (Rajarshi Guha)
Date: Mon, 16 Sep 2002 10:05:30 -0400
Subject: getting username from uid
Message-ID: <pan.2002.09.16.10.05.30.511334.24338@presidency.com>

Hi,
  when I do a stat on a file I can get the UID of the owner. IS there
anyway to get from UID to username (or directly to username for that
matter)?


From aleax at aleax.it  Thu Sep 26 09:08:36 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 26 Sep 2002 13:08:36 GMT
Subject: object changing itself to another object
References: <amu0lv$6ac$1@otis.netspace.net.au>
Message-ID: <oFDk9.145692$ub2.3210552@news1.tin.it>

Glen Murphy wrote:
        ...
> that input goes. In the highly simplified example below, I would like the
> output to be "Obj1","Obj2","Obj1" - but not suprisingly, I just get Obj1,
> Obj1, Obj1
> 
> 
> ### code begin  ###
> 
> class Obj1:
>     def key1(self):
>         # do obj1 specific stuff
>         self = Obj2()
>         print "Obj1"
> 
> class Obj2:
>     def key1(self):
>         # do obj1 specific stuff
>         self = Obj1()
>         print "Obj2"
> 
> a = Obj1()
> 
> # simulate user keypresses
> a.key1()
> a.key1()
> a.key1()

Then change the classes to:

class Obj1:
    def key1(self):
        # do obj1 specific stuff
        self.__class__ = Obj2
        print "Obj1"

class Obj2:
    def key1(self):
        # do obj1 specific stuff
        self.__class__ = Obj1
        print "Obj2"



> I know I could achieve the result I want by doing horribly complicated
> trees of ifs and such, but I'd like my code to be nice and expandable (in
> this example, to easily add new Objs). I've looked through the Python FAQ,
> Google Groups and Various O'Reilly books, but I don't really know the
> terminology for what I'm looking for, so I haven't found anything so far,
> so does anyone have any pointers?

You have not checked O'Reilly's Python Cookbook, I think, because
this particular idea of changing an object's __class__ is exemplified
there.  http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/68429
in the online version, but I think you'll find the discussion more
complete, &c, in the printed version.


Alex



From gerhard.haering at opus-gmbh.net  Thu Sep  5 07:36:42 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 5 Sep 2002 11:36:42 GMT
Subject: newbie problem: use socke lib to retrieve one web page:
References: <czAd9.6803$yt3.3242044@newssrv26.news.prodigy.com>
Message-ID: <slrnanegb5.268.gerhard.haering@haering.opus-gmbh.net>

koko wrote:
> I write this to retrieve one web page using socket lib.
> 
> # retrieve webpage
> import socket
> HOST='www.uic.edu' # remote host
> PORT=80     # remort port
> s=socket.socket(socket.AF_INET, socket.SOCK_STREAM)
> m="""GET /home/events.shtml HTTP/1.0
> From: hh at uic.edu
> User-Agent: test/1.0
> """
> [...]

You're missing the empty line that terminates a HTTP 1.0 or later HTTP
request.

-- Gerhard


From jk3380 at naida.org  Sun Sep 22 12:02:46 2002
From: jk3380 at naida.org (Pino Gargiulo)
Date: Sun, 22 Sep 2002 18:02:46 +0200
Subject: Smalltalk vs. Python and COM was Re: Byte Magazine and Smalltalk
References: <MXFh9.23885$jF4.2037258@twister.southeast.rr.com> <3d87a686@tobjects.> <rd8fougqbgdivimn7vuest3ugoj57lvrr6@4ax.com> <Xns928CA588EF17Dtrippyparcplacecom@64.164.98.7> <3d87dd7c$1@tobjects.> <g9Vh9.313655$kp.1041470@rwcrnsc52.ops.asp.att.net> <Xns928D58DF09B6trippyparcplacecom@64.164.98.6> <20020918.115128.1605894428.1527@cybermesa.com>
Message-ID: <amkpk0$or2$1@fe2.cs.interbusiness.it>

"Jay O'Connor" <joconnor at cybermesa.com> wrote in message >

... snip ...

> I use Python in a lot of places where others would probably use Perl.
> Pythons regex support is not a strong as Perl's, but its support for data
> abstraction(and hence, more complicated problems)  is much more powerful.

This brings up a question I was about to ask. I'm getting involved in a
small project
that requires a couple of COM servers, whatevere these are exactly... the
nice thing
is that any developmente language can be chose. My backgroung is Unix &
CORBA,
dabble a bit with VWNC (no serious work) and this is my second time I write
something on m$ platform.

For this project I was considering Smalltalk and Python. The first so I
could be
some real stuff done and Python because apparently it permits quite easy
access
the the mysterious (to me) m$ APIs. I really would want to stay away from
VC++,
and I'm looking for some quite turnaround environment since I'll be learning
COM on the job.

Anybody care to give some advice? I played around with DST but never looked
much at the existance of a COM wrapper/interface.

TIA

Pino




From Nicola.Musatti at ObjectWay.it  Fri Sep 20 10:36:00 2002
From: Nicola.Musatti at ObjectWay.it (Nicola Musatti)
Date: Fri, 20 Sep 2002 16:36:00 +0200
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com> <amet8v$e50$1@scotsman.ed.ac.uk> <L6Gi9.119475$ub2.2533804@news1.tin.it>
Message-ID: <3D8B3250.BA964BCF@ObjectWay.it>


Alex Martelli wrote:
[...]
> I think you need substantial capacity and appetite for abstract thought
> to really _enjoy_ SICP -- it doesn't spoon-feed you, nor does it use
> baby-talk, lots of somewhat-repetitive examples, and other standard
> didactical techniques.  But if the book's style does match your own
> character and inclination, you're likely to fall in love with it.

Or, to put it in another way, gurus like this book. ;-)

Cheers,
Nicola Musatti


From altis at semi-retired.com  Fri Sep 13 20:10:00 2002
From: altis at semi-retired.com (Kevin Altis)
Date: Fri, 13 Sep 2002 17:10:00 -0700
Subject: Python GUI/graphics package recommendation?
References: <Xns9288738D9E9EEnospamanytime@216.148.227.77>
Message-ID: <T6vg9.137$NS3.67606@news.uswest.net>

I don't think it is very nice to make your aunt do your programming for you
<wink>, but she can see if the instructions below are easy enough to give
PythonCard a try.

http://pythoncard.sourceforge.net/installation.html

There are some walkthroughs to help you along once you have everything up
and running.

ka

"David Ehrens" <no.spam at any.time> wrote in message
news:Xns9288738D9E9EEnospamanytime at 216.148.227.77...
> I am using Python 2.2 on Win32 for an AI class and would like to
> be able to display a graphical representation of an agent-based
> game I am developing. I have seen a number of GUI and graphics
> packages for Python, and would appreciate some recommendations
> since I really don't have the time to install, learn, and try all
> of them.
>
> My main criteria are: (1) my great aunt should be able to install
> it and immediately use it; (2) it should work with Win32; (3) it
> should do menus and graphics.
>
> If it happens to be the #1 tool of its kind, that would be nice
> too.
>
> Thanks, David




From jpm at papercut.org  Mon Sep 30 17:51:16 2002
From: jpm at papercut.org (Joao Prado Maia)
Date: Mon, 30 Sep 2002 17:51:16 -0400 (EDT)
Subject: Question about __getattribute__/__getattr__
In-Reply-To: <20020930234554.Y30315@prim.han.de>
Message-ID: <Pine.LNX.4.44.0209301746400.12980-100000@mercury.sabren.com>

On Mon, 30 Sep 2002, holger krekel wrote:

> The important lines from the above link are:
> 
> def __call__(self, *args):
>         if not self.memo.has_key(args):
>             self.memo[args] = self.fn(*args)
>         return self.memo[args]
> 
> and 'args' is a tuple of arguments.  
> 'args[0]' would give you the first argument, for example.  
> 

Correct. However, the same is not valid for __getattr__. The first and 
only parameter of this magic method is the attribute/method name that you 
are trying to access. It doesn't give you the list of arguments passed to 
a method.

So on my example:

"""
class Real_Class:
    def real_method(self, name):
        print name

class Cache:
    r = None

    def __init__(self):
        self.r = Real_Class()

    def __getattr__(self, attr):
        cached_value = self.get_cache(attr)
        if cached_value == None:
            return getattr(self.r, attr)
        else:
            return cached_value

    def get_cache(self, attr):
        return None

if __name__ == '__main__':
    t = Cache()
    t.real_method('spam')
"""

'attr' is the actual method name I was trying to call, using the 
't.real_method('spam')' line in there.

My question is, how do I get the actual argument list from inside 
__getattr__, so I can create an unique method/argument list signature for 
the cached files ? :)

I'm guessing there must be some deep magic to do this, but I couldn't find 
this anywhere.

Cheers,
Joao




From irmen at NOSPAMREMOVETHISxs4all.nl  Mon Sep 23 13:38:58 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Mon, 23 Sep 2002 19:38:58 +0200
Subject: UpToDate SOAP-module ??
References: <ATAj9.21315$0p1.376966@news2.ulv.nextra.no>
Message-ID: <amnjji$8l0$1@news1.xs4all.nl>

Thomas Weholt wrote:
> I've looked around for a SOAP-module to use in a WebService-based project,
> but the SOAP-implementations in python seem to be a bit ... unstable ??
[...]
> Any opinions, hints or ideas are welcome. Thoughts on ease of use, list of
> features, comparisons etc. are also interesting.

I've had good results with the updated SOAP.py found in the PyGOOGLE library,
http://diveintomark.org/projects/#pygoogle

But I find XML-RPC much easier to use.
It has a lot less features than SOAP has, but in this particular
case that is a *good* thing. Plus the xmlrpclib module that
comes installed is very good.

Irmen de Jong



From owen at nospam.invalid  Fri Sep 13 15:55:29 2002
From: owen at nospam.invalid (Russell E. Owen)
Date: Fri, 13 Sep 2002 12:55:29 -0700
Subject: Tkinter listbox clicking problem
References: <sGog9.307183$aA.53945@sccrnsc02>
Message-ID: <altfrk$1eva$1@nntp6.u.washington.edu>

In article <sGog9.307183$aA.53945 at sccrnsc02>,
 "Matthew Bonig" <mbonig at hotmail.com> wrote:

>Right now I have it bound so when a Tkinter listbox gets a <Button-1>
>(lb.bind("<Button-1>", someFunc)) it calls function someFunc. Now,
>someFunc() simply prints the curselection(). The problem is it's lagged...
>if I first select list item 1 (index 0), and then 2, and then 3 (array
>indexes 0,1,2) it prints out themout incorrectly.
>(I have someFunc actually do some try, except stuff so I don't get uncaught
>exceptions) Instead of printing 0,1,2 it prints "nothing selected", 0, 1.
>The function get's called before the listbox updates it's selection markers!
>
>Does anybody know how I could force listbox to update it's selection markers
>so instead of printing "nothing selected, 0,1  it prints 0,1,2 like it
>should?

I'm hoping somebody has better ideas, but here are some to try:

Call update_idletasks before you ask for the selection. I have no idea 
if this will work, but it might.

An ugly solution that would probably work:
- extract the y coordinate from the event callback
- call nearest(y_coord) to get the item closest to the mouse click
- explicitly select that index item. You could try skipping this step, 
but you risk a possible error between what you think they clicked on and 
what is hightlighted. Unfortunately, this step will may be a real pain 
for a list that supports multiple selected items.

Poll for selection changes -- thus it may take a bit of time before your 
code notices a change, but it will always be right once it gets there.

-- Russell


From maxm at mxm.dk  Tue Sep 17 03:48:10 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 17 Sep 2002 09:48:10 +0200
Subject: Windows AutoRunning Python Hello -- feedback requested
References: <Hmuh9.70$Ha6.5066815@newssvr14.news.prodigy.com>
Message-ID: <3D86DE3A.2040903@mxm.dk>

Dennis Reinhardt wrote:


> 1) Is this the right group to discuss/get feedback?


Yes

> 2) Is this functionality useful enough to others to pursue?


It is probably to "in-between" to be of interrest to a lot of people. To 
high level to be a short learning example, and to low level to be of 
real use.

It will probably be interresting to only a small subset of developers. 
But here is the place to look for them ;-)


regards Max M



From gerhard.haering at opus-gmbh.net  Mon Sep  2 07:55:45 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 2 Sep 2002 11:55:45 GMT
Subject: Does IDLE handle unicode?
References: <QdIc9.2412$Ak.1132837@kent.svc.tds.net>
Message-ID: <slrnan6kcv.eg.gerhard.haering@haering.opus-gmbh.net>

Edward K. Ream wrote:
> IDLE took an exception writing a .py file containing the copyright
> character.  This happened a while ago; IIRC, I lost the file.
> 
> Is this a known problem?

Yes.

> I see nothing about this in the FAQ.

http://www.faqts.com/knowledge_base/view.phtml/aid/11712 has a workaround.
 
-- Gerhard 


From tjabo.kloppenburg at unix-ag.org  Tue Sep 10 14:28:23 2002
From: tjabo.kloppenburg at unix-ag.org (Tjabo Kloppenburg)
Date: Tue, 10 Sep 2002 20:28:23 +0200
Subject: Trouble upgrading to 2.2.1
In-Reply-To: <slrnans81k.jc.gerhard.haering@lilith.my-fqdn.de>
References: <aXof9.29813$w51.7932884@news2.news.adelphia.net>
	<slrnans81k.jc.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <20020910202823.2bff6572.tjabo.kloppenburg@unix-ag.org>

On 10 Sep 2002 16:39:19 GMT
Gerhard H?ring <gerhard.haering at gmx.de> wrote:

GH> > libcrypto.so.2
GH> > libsel.so.2
GH> 
GH> That should probably be libssl...

on a suse system, you find a file ARCHIVES.GZ on the first cdrom. copy it to the hard disk, and do a "zgrep libcrypto ARCHIVES.GZ" -- you'll see the packages with files with this name...

I allways have/had:) a small script called "wofile (wherefile)" on my suse linux systems to find the right packages -- yast1/2 was not my friend... :)

--
Tjabo Kloppenburg
  --=- In 104.16 Tagen ist Weihnachten! -=--

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020910/80b45abc/attachment.sig>

From bridgman at wyeth.gsfc.nasa.gov  Mon Sep 23 16:22:24 2002
From: bridgman at wyeth.gsfc.nasa.gov (Tom Bridgman)
Date: Mon, 23 Sep 2002 16:22:24 -0400
Subject: Saving & Restoring Python objects in XML?
Message-ID: <3D8F7800.F757A72D@wyeth.gsfc.nasa.gov>

Hello,

I'm setting up some codes where I want to save the state of the code in
an XML file at some time and reload it later to continue the processing.

It's relatively easy to save out the innards of Python objects into XML
regardless of how the objects are nested.  But what about loading this
state information back in from the XML file?  

Is there a way to write the parser handlers and/or other components (I'm
using PyXML 0.7 and higher) such that they can assemble the required
objects as they read in the file?  I'm wondering if there is an elegant
solution that might not be immediately obvious from the 'brute force'
method I keep coming up with.  I'd prefer to do this using SAX (since
I've worked with it some), but using DOM is not out of the question if
it solves the problem in a straightforward way.

Thanks,
Tom
--
Dr. William T."Tom" Bridgman           Scientific Visualization Studio
Global Science & Technology, Inc.      NASA/Goddard Space Flight Center
Email: bridgman at wyeth.gsfc.nasa.gov    Code 935
Phone: 301-286-1346                    Greenbelt, MD 20771
FAX:   TBD                             http://svs.gsfc.nasa.gov/


From soodvarun at janu.info  Sat Sep 28 05:02:00 2002
From: soodvarun at janu.info (Varun Sood)
Date: 28 Sep 2002 02:02:00 -0700
Subject: problem with GUI editor
Message-ID: <c2b9ddcd.0209280102.2f532c9d@posting.google.com>

hi there !!!!
i am very new to python and to this world of developers I need help !!

1) TUTORIALS which could be downloaded.

2) PYTHON's GUI (SHELL) editor is not good. As it does not shows SECONDARY
PROMPT and it do not run "if....elif" statement as it gives errors
on 
it. "INDENTATION NOT SET" but the same works properly in python's
COMMAND 
LINE shell. Wts the problem?? 
I am using python 2.2

varun


From pinard at iro.umontreal.ca  Tue Sep  3 21:11:06 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Tue, 03 Sep 2002 21:11:06 -0400
Subject: winsound.Beep() frequencies
In-Reply-To: <al3lmt$uq6$0@216.39.172.122> (bokr@oz.net's message of "4 Sep
 2002 00:55:57 GMT")
References: <3ab719f7.0209031229.33667461@posting.google.com>
	<al3lmt$uq6$0@216.39.172.122>
Message-ID: <oq1y8asgnp.fsf@titan.progiciels-bpi.ca>

[Bengt Richter]

> But I believe the best sounding scale will not have 12 equal ratios of
> 2**(1/12).

How to tune has been the source of hot debates for centuries.  When the equal
tuning established itself and won, the debates progressively faded out, and we
now take that a dull equal tuning for granted.  Most people never heard a good
tuning, and would feel astonished or uncomfortable at first, hearing something
like a perfect third.  It take some time to rediscover the pleasure of such
things, after what, todays's ubiquitous equal tuning appears even more dull...

For beeping a computer however, it might not matter so much! :-)

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



From imbosol at vt.edu  Sun Sep  1 19:17:19 2002
From: imbosol at vt.edu (Carl Banks)
Date: Sun, 1 Sep 2002 19:17:19 -0400
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de>
Message-ID: <aku8j5$3i2$2@solaris.cc.vt.edu>

Gerson Kurz wrote:
> So basically, couldn't one come up with
> a *human readable* syntax for re, and compile that instead? Python
> prides itself on its clean syntax, and human readability, an bang -
> import re, get perl-ish code instantly! 


You might want to look at the Plex package.  It defines patterns by
constructing data structures.  Something like this:

    symbol = Range("A-Za-z") + Any(Range("0-9A-Za-z") | Char("_"))

However, three points:

First, this will certainly be slower than regular expressions, since
there are many Python calls needed to build the structure. (Of course,
after you've compiled it, it can be as fast as regexps.)

Second, even if you use re module, it is still nowhere near Perl-ish
ugliness.  You still have Python's clean syntax outside of the
pattern.

Third, readability is not a unilateral good thing; conciseness is also
important, and sometimes opposed to readability.  Sacrificing a little
readability to get a lot of conciseness is usually a good thing.  I
think, as long as the regexp is not too obnoxious, it is probably
better to keep it concise.  (Of course, this depends a lot on what
you're doing and how flexible you need to be.)



-- 
CARL BANKS
http://www.aerojockey.com


From ezcoder at mailcity.com  Fri Sep 13 23:20:52 2002
From: ezcoder at mailcity.com (Glenn Scott)
Date: 13 Sep 2002 20:20:52 -0700
Subject: Question regarding mysqldb module and variable table names
Message-ID: <45ac92e5.0209131920.23b1c22a@posting.google.com>

Hi all,

I'm currently using the MySQLdb module to write data to a mySQL
database.

I've run into a problem while changing my code that I was hoping
someone might have some suggestions on.

Originally I had code that looked like this:

c.execute("""INSERT INTO mytable \
        SET playerkey=%s, link=%s, snippet=%s, time_created=%s, \
        last_seen=%s, relevancy=%s, doc_title=%s """, \
         (playerKey, urlLink, snippetText,
self.getCurrentTime(),self.getCurrentTime(),\
         noOfMentions, re.escape(str(self.doc_title))) )

I liked using the function of this type rather than that using the "%"
string formatter since it automatically fixed problems with odd
characters in my data.  Worked great!

I want to change it so that the "mytable" value in this example is a
variable rather than a constant.  So the code would look like this:

 c.execute("""INSERT INTO %s \
        SET playerkey=%s, link=%s, snippet=%s, time_created=%s, \
        last_seen=%s, relevancy=%s, doc_title=%s """, \
         (infoTable, playerKey, urlLink, snippetText,
self.getCurrentTime(), self.getCurrentTime(),\
         noOfMentions, re.escape(str(self.doc_title))) )

But this doesn't work...for whatever reason, using "%s" in an
assignment SET variable works, but as a table name it doesn't.

I suppose I can use the "%" string formatter at the end of the
triplq-quotes to fix this problem, but I'll have to jump through other
hoops to deal with oddness with my data that is taken care of for me
by using the "," alternative.

Any ideas as to how I might approach this?

TIA


From heikowu at ceosg.de  Sat Sep  7 17:25:10 2002
From: heikowu at ceosg.de (Heiko Wundram)
Date: 07 Sep 2002 23:25:10 +0200
Subject: Killing/Terminating a running thread
In-Reply-To: <Xns9282E6FBFE1C2cliechtigmxnet@62.2.16.82>
References: <mailman.1031419029.11871.python-list@python.org> 
	<Xns9282E6FBFE1C2cliechtigmxnet@62.2.16.82>
Message-ID: <1031433910.17902.107.camel@d168.stw.stud.uni-saarland.de>

Am Sam, 2002-09-07 um 22.42 schrieb Chris Liechti:
> sound like the news thread "Interrupting Python" could be interresting for 
> you...

I've looked at that, but the main problem is that this thread deals with
interrupting fullblown processes on the fly (from within the code), not
with interrupting a process by an exterior signal.

> the clean solution is to test for a flag and terminate if the flag is set.
> 
> e.g. you could provide a function:
> 
> def interrupted():
>     	return global_flag_if_user_wants_to_abort
> 
> and in you function in the queue:
> 
> def worker():
>     	while not interrupted():
>     	    	do_stuff()

This is what I do to shutdown the worker cleanly. The problem is:
do_stuff() doesn't return until it's complete, and I have no control
whatsoever about the internal workings of do_stuff().

What I actually wish to do is while the worker still is in do_stuff(),
to kill the worker as a whole, to discard all its data, when a timeout
on returning a value from do_stuff() occurs. It's not about clean
termination of a thread, but about killing the thread in case it takes
too long (in case do_stuff() takes to long).

Thanks for the advice anyway! :) I guess I at least did the general
interrupt on the worker right then... :)

Yours,

	Heiko Wundram
	Netzwart Wohnheim-D
	Zimmer 2206 - Universit?t 18 - 66123 Saarbr?cken

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 196 bytes
Desc: Dies ist ein digital signierter Nachrichtenteil
URL: <http://mail.python.org/pipermail/python-list/attachments/20020907/b2535ae5/attachment.sig>

From Padraig at Linux.ie  Fri Sep 13 12:33:23 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Fri, 13 Sep 2002 16:33:23 GMT
Subject: question about dictionary type..
References: <alt22f$pf8$1@newsreader.mailgate.org>
Message-ID: <3D82132C.1020203@Linux.ie>

eugene kim wrote:
> hi..
> 
> i'd like to have a dictionary which looks like
> table = { category1 : { category2 : { category3 : 'garbage value' } } }
> 
> category1 has many category2
> category2 has many category3
> 
> i expected
> table[ 'computer' ][ 'language' ][ 'python' ] = 0
> to create {computer : { language : { python :0 }}}
> table[ 'computer' ][ 'language' ][ 'java' ] = 0
> to becomes { 'computer' : { 'language' : { 'python' :0 , 'java' :0 }}}
> 
> but python doesn't allow me to do this..
> can anyone help me?
> 

It only creates 1 level at a time
because ypur not acutally assigning
to table[ 'computer' ] or table[ 'computer' ][ 'language' ],
you're only referencing them.

I guess you have to do something like:
table.setdefault('computer',{}).setdefault('language',{})['python']=0

P?draig.



From pyth at devel.trillke.net  Wed Sep 11 16:56:31 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 11 Sep 2002 22:56:31 +0200
Subject: Any C++/Python sig members here?
In-Reply-To: <l1Kf9.1470$Lg2.157618@news2.news.adelphia.net>; from Oschler@earthlink.net on Wed, Sep 11, 2002 at 04:18:25PM +0000
References: <l1Kf9.1470$Lg2.157618@news2.news.adelphia.net>
Message-ID: <20020911225631.T3893@prim.han.de>

Robert Oschler wrote:
> I joined the C++ sig for Python, to talk with others about BoostPython.  I
> joined more than a day ago and got my subscription confirmation request from
> the sig server.  I have not seen any messages yet, is it a very low traffic
> sig or something?

I don't think so.  but to be sure you can look at the archives. the link
should have been sent with your welcome-message.

    holger



From aleax at aleax.it  Wed Sep 25 17:15:46 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 21:15:46 GMT
Subject: help with print function in extension
References: <tA8k9.270440$AR1.11448440@bin2.nnrp.aus1.giganews.com> <4jfk9.163306$pX1.5842487@news2.tin.it> <Cwpk9.283583$5r1.12035771@bin5.nnrp.aus1.giganews.com>
Message-ID: <6Ipk9.143767$ub2.3151188@news1.tin.it>

Eric Hagemann wrote:
        ...
> Alex, Thanks -- missed that section in the manual.  I blindly assumed that
> I could use the repr() as the print func......

You can!  Just leave the tp_print slot to 0 and all should be fine.


> ok so for the next trick question. I implemented a function that has the
> proto
> int print_func(PyObject *obj, FILE *fp,int flags)
> 
> when I try to print to fp -- as in --> fprintf(fp,"<fmt>", , )... I crash
> the program.

I'll have to admit that I never filled tp_print myself (oh the shame),
always relying on other slots, so I can't help you here...


Alex



From cben at tx.technion.ac.il  Tue Sep 10 15:35:50 2002
From: cben at tx.technion.ac.il (Beni Cherniavksy)
Date: Tue, 10 Sep 2002 22:35:50 +0300
Subject: "str"(expr)"str" interpolation syntax
Message-ID: <Pine.GSO.4.33.0209102221070.10002-100000@techst02.technion.ac.il>

Hi all.

I was searching for a new interpolation syntax that would use quotes
for starting/ending the interpolated part (I don't feel comfortable
putting code inside unclosed quotes...)

After contemplating some horrors like $"""str"expr"str""" or
$("str" expr "str"), I noticed that currently both  "(  and  )"  are
illegal (first because strings are not callable, second is a syntax
error).  So I propose that a parenthesized expresion near a string would
imply concatenation.  In other words,
	"foo"(x * 3)"bar"
would be equivallent to
	"foo" + str(x * 3) + "bar".
This works of course with only one side, e.g. "foo"(x) or (y)"bar" but in
longer cases it would be good style to add empty strings on either end,
e.g. ""(y)"bar".  And of course it all works with single quotes too.  One
of the benefits over current interpolation PEPs is that you can change the
quote type in the middle...

It is slightly more verbose (especially when interpolating into tripple
quote strings: """foo"""(x)"""bar""")  but it is more readable and
understandable (just extends the current concatenation of constant
strings), allows clear nesting of strings into the expr:
	"foo"('%3d' % x)"bar"
and most importantly, you never need to remember arbitrary rules where the
interpolation ends...

What do you think?

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>

Awake! Awake! Fear, Fire, Foes! Awake! Fire, Foes! Awake! [Tolkien, LOTR]
Fight software patents:  http://lpf.ai.mit.edu/
BTW, does somebody know what's the situation in Israel
concerning software patents?




From greek_bill at yahoo.com  Sat Sep 21 19:55:13 2002
From: greek_bill at yahoo.com (Morfeas)
Date: Sun, 22 Sep 2002 00:55:13 +0100
Subject: [newbie] Calling Python from C++
References: <8L5i9.841$Ew4.62552@newsfep2-gui> <3D8B3EA0.6050500@thinkware.se> <amffjq$ofg$1@bob.news.rcn.net>
Message-ID: <uE7j9.6097$DR.337426@newsfep2-win.server.ntli.net>

David Abrahams wrote:

> While I appreciate the plug and I do think Boost.Python is a great
> solution for C++, SWIG has evolved since I first did my analysis and it
> understands a lot more of C++ than it used to, though it is still missing
> the ability to instantiate templates. The Boost comparisons page is
> woefully outdated and I've been waiting for David Beazley to write a new
> blurb for SWIG.
> 

I concur. I've managed to do what I want in the end, using SWIG. Altough, I 
must admit I'm not using it (SWIG) the way it was designed to be used! If I 
understand correctly, SWIG (and probably a lot of other similar tools) has 
been designed to make your C (or C++) code visible to Python (and other 
interpreted languages) code. i.e. it helps you write extensions to Python.

What I was trying to do (and finally succeeded) was :
1. to extend an embedded Python interpreter, and
2. allow the python code to interact with existing instances of C++ objects

It took a bit of messing about and hacking at code, but it worked (hey, 
what's another couple of 16 hours days in front of the computer? :)

Morfeas
-- 
www.freecfm.com/m/morfeas


From aleax at aleax.it  Tue Sep 24 11:19:31 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 24 Sep 2002 15:19:31 GMT
Subject: Getting started
References: <uojinmcfapjf45@corp.supernews.com> <mailman.1032567390.29374.python-list@python.org> <0%Aj9.129252$ub2.2813083@news1.tin.it> <877khb7884.fsf@pcdh91.cern.ch>
Message-ID: <7o%j9.159168$pX1.5714084@news2.tin.it>

Tibor Simko wrote:

> Hello
> 
> On Mon, 23 Sep 2002, Alex Martelli wrote:
>> as far as I know, the CAML dialect of ML (also in the O'Caml
>> extension, last I looked) is one of the very few programming
>> languages that doesn't have a polymorphic + operator -- you have to
>> use + for integers only, +. for floating-point.
> 
> There is a good reason why OCaml does not overload the + operator:

There is a reason -- whether it's a GOOD one is more debatable:-)

> type inferring.  In OCaml you do not need to declare variables
> explicitly in the program because OCaml infers types for you.
> However, this does not match well with operator overloading, since the
> compiler would not understand whether "a + b" stands for float or
> integer.  Unlike SML, OCaml decided to avoid this kind of type
> inferring problem by not overloading the + operator.

Compare this "solution" to Haskell's typeclasses to see why I
love Haskell so:-).  While SML doesn't seem to suffer unduly
from "overloading" (as you say) `+` , Haskell generalizes this
to a kind of generic-programming ability -- no overloading,
just the right amount of generality.  And without any damage
at all to strong compile-time type safety, nor to type
inferencing (or inferring), of course.


Alex



From skip at pobox.com  Wed Sep 18 17:19:00 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 18 Sep 2002 16:19:00 -0500
Subject: linux python ideas
In-Reply-To: <200209181347.39562.shalehperry@attbi.com>
References: <3D88E578.6010001@diespammerdieuselesspython.com>
        <200209181347.39562.shalehperry@attbi.com>
Message-ID: <15752.60868.893189.395705@12-248-11-90.client.attbi.com>

    >> I'll be giving a presentation on Python to my local LUG
    >> (http://lugoj.org) in a few weeks, and would like to point out at
    >> least a few things that would be of particular interest to linux
    >> users.

    Sean> obvious ideas are things like zope and mailman as well as cheetah.

Also rdiff-backup: <http://rdiff-backup.stanford.edu/>

Skip



From fb at frank-buss.de  Thu Sep  5 07:45:52 2002
From: fb at frank-buss.de (Frank Buss)
Date: Thu, 5 Sep 2002 11:45:52 +0000 (UTC)
Subject: How do i print to a printer using python?
References: <6afb9d2c.0209041044.43c5f5f1@posting.google.com> <al63r8$3bo$1@news7.svr.pol.co.uk>
Message-ID: <al7g5g$a29$1@newsreader2.netcologne.de>

"G. Willoughby" <never at mind.info> wrote:

> I would really like to know how to do this too, basically i want to
> print to a standard bubblejet print (non postscript) from a Tkinter
> Text widget, any ideas?

Perhaps you can draw a PDF file with Reportlab (http://www.reportlab.com/) 
and printing it. Because you are using Outlook Express, I assume you need 
it for Windows. An example:

from reportlab.pdfgen import canvas
from reportlab.lib.pagesizes import A4
c = canvas.Canvas('c:\\tmp\\test.pdf', pagesize=A4, bottomup=0)
c.setFont('Helvetica', 14)
c.drawString(10, 20, 'Hello World!')
c.save()

If you want direct printing, without user interaction and you know the 
path of the Acrobat Reader, you can use this line:

system('"C:\\Programme\\Adobe\\Acrobat 5.0\\Reader\\AcroRd32.exe" /p /h c:
\\tmp\\test.pdf')

Perhaps it's easier with the ShellExecute command, because then you don't 
need to know the path of AcroRd32.exe, but then you need this, if you 
don't have ActiveState Python installed:
http://starship.python.net/crew/mhammond/win32/Downloads.html

If you just want to show the user the page and let the user decide to 
print or close the document, this lines are sufficient:

from os import startfile
startfile('c:\\tmp\\test.pdf')

-- 
Frank Bu?, fb at frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de


From blalor at insight.com  Fri Sep  6 11:39:58 2002
From: blalor at insight.com (Brian Lalor)
Date: Fri, 06 Sep 2002 15:39:58 GMT
Subject: help needed creating xml document with xml.dom
Message-ID: <m34rd33za5.fsf@localhost.localdomain>

Good morning, all.  Several times, now, I've used xml.dom.minidom to create
raw XML documents, but I don't think I've ever really done it the "right"
way.  I blame that on my seeming inability to find documentation on the
"right" way to do it in Python, or in any language, for that matter.  Pointers
to said documentation are welcomed. :-)

That out of the way, here's my current problem.  I'm trying to write a script
that will spin through a set of images taken with my digital camera, pull out
the EXIF data and save it to an XML document and also extract the JPEG
thumbnail from those images.  Gene Cash's excellent EXIF.py[1] does the dirty
work with getting the EXIF data.  I also want to create an XHTML document that
is the index of all of these thumbnails, and that's where things started going
downhill.  A "proper" XHTML document requires a well-formed XML document along
with a doctype definition(?) below the <?xml ?> declaration and an xmlns
attribute to the <html/> tag.  Using minidom, I cannot seem to get either the
namespace or doctype to print out.

Here's the relevant code snippet:

    import xml.dom
    
    # is this the right way, instead of directly importing xml.dom.minidom?
    DOM = xml.dom.getDOMImplementation()
    
    # now create the doctype and the document element
    doctype = DOM.createDocumentType('html',
    								 '-//W3C//DTD XHTML 1.0 Strict//EN',
    								 'http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd')
    
    index_page = DOM.createDocument(xml.dom.XHTML_NAMESPACE, 'html', doctype)

    index_page.documentElement.setAttribute('xml:lang', 'en')
    index_page.documentElement.setAttribute('lang', 'en')


So, now if I do a index_page.toprettyxml(), I only get
    <?xml version="1.0" ?>
    <html xml:lang="en" lang="en"/>

... but I *want*
    <?xml version="1.0" encoding="UTF-8"?>
    <!DOCTYPE html
            PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
            "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
    
    <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en"/>

Is this a bug in minidom, or (more likely) am I doing something wrong?  I'm
using Python 2.2.1 but I have PyXML 0.7 installed, so that's where minidom.py
is coming from.

Thanks,
B

[1] http://home.cfl.rr.com/genecash/digital_camera.html

--
      Brian Lalor                 |    http://introducingthelalors.org/
  blalor at ithacabands.org (email)  |  blalor at jabber.ithacabands.org (jabber)
                       N33?27.369' W111?56.304' (Earth)



From pedronis at bluewin.ch  Sat Sep  7 07:32:20 2002
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Sat, 7 Sep 2002 13:32:20 +0200
Subject: [curiosity/info] Parrot long-term goals/prospects
Message-ID: <3d79e4c5_4@news.bluewin.ch>

http://archive.develooper.com/perl6-internals%40perl.org/msg12437.html


[ For people prone to just skim the pointed message, it is worth to repeat
that it is talking about things that should become real in the "distant"
future, for some value of "distant".This link is posted just as
curiosity/information, no negative/positive opinion from my part on Parrot
should be inferred from me posting this, I'm not involved with Parrot, apart
from skimming their mailing list from time to time. ]

regards.




From sholden at holdenweb.com  Sat Sep  7 17:30:49 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 7 Sep 2002 17:30:49 -0400
Subject: about parse a link
References: <cVUd9.6962$yt3.3362358@newssrv26.news.prodigy.com> <HXXd9.9644$HY3.2204710@newsc.telia.net>
Message-ID: <Mfue9.18474$l_4.7243@atlpnn01.usenetserver.com>

"Fredrik Lundh" <fredrik at pythonware.com> wrote in message
news:HXXd9.9644$HY3.2204710 at newsc.telia.net...
> "koko" wrote:
>
> > if I have extracted the links on the page:
> > e.g:
> >
> > http://www.uic.edu/index.htm
> > on this page: there are
> > a.htm
> > b.htm
> > c.htm
> > http://www.uic.edu/home/e.htm
> >
> > how can I log the a.htm, b.htm, c.htm with the full web address?
>
>     base = "http://www.uic.edu/index.htm"
>
>     url_list = [
>         "a.htm",
>         "b.htm",
>         "c.htm",
>         "http://www.uic.edu/home/e.htm"
>     ]
>
>     import urlparse
>
>     for url in url_list:
>         print urlparse.urljoin(base, url)
>
> prints
>
>     http://www.uic.edu/a.htm
>     http://www.uic.edu/b.htm
>     http://www.uic.edu/c.htm
>     http://www.uic.edu/home/e.htm
>
> </F>
>
> <!-- (the eff-bot guide to) the python standard library:
> http://www.pythonware.com/people/fredrik/librarybook.htm
> -->
>

In case you haven't done much of this kind of stuff, note tha you should use
the URL of the page you are inspecting as the base, unless it contains a
<BASE...> tag, in which case you need to use the URL from that.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                        pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From sgclift at earthlink.net  Sun Sep  1 22:47:25 2002
From: sgclift at earthlink.net (Steve Clift)
Date: Sun, 01 Sep 2002 19:47:25 -0700
Subject: urllib bug?
Message-ID: <3.0.6.32.20020901194725.007d4cd0@mail.earthlink.net>

Is this *really* a urllib bug, or am I missing something blindingly obvious?

  Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
  Type "copyright", "credits" or "license" for more information.
  IDLE 0.8 -- press F1 for help
  >>> import urllib
  >>> f = urllib.urlopen('http://www.python.org')
  Traceback (most recent call last):
    File "<pyshell#4>", line 1, in ?
      urllib.urlopen('http://www.python.org')
    File "C:\PROGRAM FILES\PYTHON22\lib\urllib.py", line 73, in urlopen
      return _urlopener.open(url)
    File "C:\PROGRAM FILES\PYTHON22\lib\urllib.py", line 178, in open
      return getattr(self, name)(url)
    File "C:\PROGRAM FILES\PYTHON22\lib\urllib.py", line 276, in open_http
      if not host: raise IOError, ('http error', 'no host given')
  IOError: [Errno http error] no host given

Fails the same way with 2.1. Works OK with 1.5.2 and (IIRC) 2.0.

-Steve



From joconnor at cybermesa.com  Sat Sep  7 12:43:05 2002
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sat, 07 Sep 2002 16:43:05 GMT
Subject: None or 0
References: <k7ip4-0p3.ln1@junker.stroeder.com> <jwbaxter-E9C054.09080807092002@corp.supernews.com>
Message-ID: <3d7a2b6b.92959728@news.cybermesa.com>

On Sat, 07 Sep 2002 09:08:08 -0700, John Baxter <jwbaxter at spamcop.net>
wrote:

>In article <k7ip4-0p3.ln1 at junker.stroeder.com>,
> Michael Stroder <michael at stroeder.com> wrote:
>
>>  >>> repr(0 or None)
>> 'None'
>>  >>> repr(None or 0)
>> '0'
>>  >>> repr('' or None)
>> 'None'
>>  >>> repr(None or '')
>> "''"
>>  >>>
>> 
>> Is it guaranteed to work like this or should that be avoided?
>
>It seems very convenient to implement or (and and) by returning the 
>first object which decides the answer.

I noticed that happening.  It seems sensible for short-circuiting
that in an "or", you just return th second term , regardless of what
it is


>My guess is that enough people are counting on the behavior that even if 
>the underlying hardware or implementation language were to shift enough 
>to make this implementation quite inconvenient, it would nonetheless 
>survive.
>
>Having said that, I tend not to like counting on it, because it 
>stretches the extra-Python definition of the operators quite a bit.  
>Particularly on the "and" side:
>
>>>> print 5 and 1
>1
>
>Go into (English-speaking) classrooms, and you're going to hear chants 
>of "5 and 1 is 6" not "5 and 1 is 1".

Because in that contexts "and" means "plus" and in this context "and"
means "logical and"

Take care,

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


"God himself plays on the bass strings first, when he tunes the soul"


From anton at vredegoor.doge.nl  Sat Sep 14 09:22:50 2002
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Sat, 14 Sep 2002 15:22:50 +0200
Subject: How to represent sets
Message-ID: <alvd8p$1sq$1@news.hccnet.nl>

Hello All,

In the past there have been some discussions on various platforms about
whether sets should be implemented using long integers (possibly leading to
longs becoming iterable arrays of booleans), or by using dictionaries. The
matter seems to be mostly settled and there is a sets.py script in the CVS
directory at sourceforge which uses dictionaries.

I would like to reopen this discussion.

Starting this discussion again when the CVS script is already very mature
seems a bit inappropriate, but I believe there is a way to make sets faster. I
only recently realized this so I could not act earlier.

To give some material for the discussion I am providing a beta version of a
possible implementation af a set class using long integers to represent sets.
It is basically the same script I presented a year ago but I have made several
improvements to this script so that it is now a bit faster, and I hope the
docstring is a bit better now. Please keep in mind this is just a test script,
for a safer and more complete implementation the script in CVS is preferable.

http://home.hccnet.nl/a.vredegoor/universe/universe.py

or the html version:

http://home.hccnet.nl/a.vredegoor/universe/universe.html

Anton.



From bdesth at nospam.free.fr  Fri Sep  6 21:39:39 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Fri, 06 Sep 2002 21:39:39 -0400
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com>
Message-ID: <3D7958DB.50903@nospam.free.fr>

Rod Stephenson wrote:
> Slashdot has a list of questions posed to Larry Wall (perl). When
> asked his thoughts on other scripting languages, he makes the
> following observation about python
> 
> "Python is cool to look at small bits of, but I think the "outline"
> syntax breaks down with larger chunks of code. I'm with Aristotle on
> the structure of discourse--a story should have a beginning, and
> middle, and an end. So should blocks"
> 
> I'm not quite sure what he's trying to get at here - I guess that for
> a long heavily indented chunk of code, you could lose track of the
> overall structure, but I don't write code this way.
> 
> Any comments?

def fun(args):
#BEGIN
     if ceci or cela:
     #BEGIN
         do_this(with_that)
         [snip a lot of code]
     #END
#END


So what the problem with blocks ?

laotseu



From stephen.boulet at motorola.com  Fri Sep 27 15:23:35 2002
From: stephen.boulet at motorola.com (Stephen Boulet)
Date: Fri, 27 Sep 2002 14:23:35 -0500
Subject: Better terminal for windows?
Message-ID: <3D94B037.2020202@motorola.com>

I'm using python 2.2.1 on winnt.

If I want to run python interactively (and perhaps using ipython), is 
there something better than the "cmd" terminal?

I'm used to bash on linux at home, but I'd like to avoid installing 
cigwin for now. Thanks.

-- Stephen



From abuse at nospamtron.riot.com.au  Thu Sep 26 09:43:00 2002
From: abuse at nospamtron.riot.com.au (Glen Murphy)
Date: Thu, 26 Sep 2002 13:43:00 GMT
Subject: object changing itself to another object
References: <amu0lv$6ac$1@otis.netspace.net.au> <oFDk9.145692$ub2.3210552@news1.tin.it>
Message-ID: <E9Ek9.23498$6g7.74155@news-server.bigpond.net.au>

> You have not checked O'Reilly's Python Cookbook, I think, because
> this particular idea of changing an object's __class__ is exemplified
> there.  http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/68429
> in the online version, but I think you'll find the discussion more
> complete, &c, in the printed version.

Thanks Alex, I have actually subscribed to the book on Safari, but it's
kinda hard to flick through there :D

And thank you to everyone who submitted answers - very much appreciated,

~ Glen




From Steven_Shaw at adc.com  Tue Sep 10 01:19:26 2002
From: Steven_Shaw at adc.com (Steven_Shaw at adc.com)
Date: Tue, 10 Sep 2002 15:19:26 +1000
Subject: dynamism
Message-ID: <4A256C30.001DA17F.00@ssxrgw01.aus.ssd.adc.com>


>> I've been away from Python for awhile and I'm wondering whether Python got
>> less "dynamic" while I was away.
>>
>> I thought you could just extend an object's attributes at
>> run-time like this:
>>
>> >>> p = object()
>> >>> p.x = 1
>> Traceback (most recent call last):
>>   File "<stdin>", line 1, in ?
>> AttributeError: 'object' object has no attribute 'x'
>
>object is the base type of new-style classes.  Try something like this
>instead:
>
>>>> class foo:pass
>...
>>>> f = foo()
>>>> f.x = 1
>>>> f.x
>1

Thanks!

>
>Or 'class foo(object):pass' works too.
>
>// m





From chris.gonnerman at newcenturycomputers.net  Sat Sep 21 12:31:50 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sat, 21 Sep 2002 11:31:50 -0500
Subject: Local Python Server Installation
References: <3d8b47c6.277669@news.jaring.my> <slrnaomn6m.39s.gerhard.haering@haering.opus-gmbh.net> <3d8c993c.2190426@news.jaring.my>
Message-ID: <003f01c2618c$67e3e720$0101010a@local>

----- Original Message ----- 
From: "Kenneth Gomez" <kennethg at pd.jaring.my>


> I tried your method. But instead of getting a cgi listing of 
> my environmental listings, all I saw on my browser was the 
> two input lines ..i.e. import cgi and cgi.test()

Your cgi is being displayed rather than executed.  The 
configuration options for Apache are so complex, that entire 
books have been written about configuring it... and Gerhard's
suggestion only works on Windows platforms (AFAIK).  You didn't 
name your platform, did you?  Or did I just miss it?

You need to verify that Apache 1) recognizes .py files as CGI 
scripts, and 2) allows you to run a CGI in the folder you're 
putting it in.  You probably should put your script in the
cgi-bin directory to start with.  If you are on a Unix/Linux
system, you need to make sure that the script is marked 
executable.

Really, I think you need to spend some time with the Apache 
docs.  It isn't a Python problem per se.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From martin at v.loewis.de  Mon Sep 16 01:43:05 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Sep 2002 07:43:05 +0200
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <mailman.1032152058.11043.python-list@python.org>
Message-ID: <m3admizdzq.fsf@mira.informatik.hu-berlin.de>

Gerhard H?ring <gerhard.haering at gmx.de> writes:

> You should perhaps reconsider if and why you need to hide your source
> code.

You should also recognize that, if your computer can execute and
perform a certain algorithm, users can always observe simulate the
computer operations, thus breaking any protection scheme.

You'll have to give your customers a sealed computer and no permission
to install additional software, or communicate with the internet, if
you want to reliably prevent them from analysing your program.

Regards,
Martin


From walterm at parque.homelinux.net  Thu Sep 19 20:01:48 2002
From: walterm at parque.homelinux.net (Walter Moreira)
Date: Thu, 19 Sep 2002 21:01:48 -0300
Subject: RPy for Windows (unstable version)
Message-ID: <20020920000148.GA5899@parque.homelinux.net>

Hi. I have uploaded a Windows version of RPy to SourceForge. It is very
unstable (I'm not sure whether it even works on a Windows other than
mine :-) Please, send me your comments or problems. The next release
(mid or end of october) will be a real one. Download at

    http://sourceforge.net/projects/rpy

Two important problems with this release:
1. It doesn't locate the R dll's. You may need to copy R.dll to some
   location. Mine is 'Program Files\r\rw1404\bin\R.dll'
2. It only works from the DOS prompt, because the I/O is directed to
   console.

Regards:
Walter


1.  WHAT IS RPy?

RPy is a very simple, yet robust, Python interface to the R Programming
Language (http://www.r-project.org).  It can manage all kinds of R
objects and can execute arbitrary R functions (including the graphic
functions).  All the errors from the R language are converted to Python
exceptions.  Any module that later were installed on the R system, can
easily be used from within Python, without introducing any changes. 



From brian at sweetapp.com  Thu Sep 26 17:38:14 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Thu, 26 Sep 2002 14:38:14 -0700
Subject: Who knows somefunction?
In-Reply-To: <3D937B08.4090704@aon.at>
Message-ID: <00d701c265a5$053d0710$df7e4e18@brianspiv1700>

> Is there a function <somefunction> in Python
> for which
> 
> something is <somefunction>(id(something))
> 
> returns invariably true?

Nope. You might want to consider creating your own dictionary to map ids
to the objects that you are interested in.

Cheers,
Brian




From thorsten at thorstenkampe.de  Fri Sep 13 16:05:45 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Fri, 13 Sep 2002 22:05:45 +0200
Subject: Coding Style: Quotes
References: <bt1g9.17429$Pf7.590136@news1.west.cox.net>
Message-ID: <altgev$10u79$1@ID-77524.news.dfncis.de>

* John Waycott
> I'm curious if others have adopted any standards for choice of single vs.
> double quotation marks to delimit strings. A look through the standard
> library reveals the choice of one over the other is rather arbitrary. I
> suspect it really makes no difference, but the question has come up during a
> code review.

Python itself seems to "prefer" single quotes slightly:
1: >>> "String"
1:     'String'
2: >>>

Thorsten


From martin at v.loewis.de  Wed Sep 18 17:24:26 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 18 Sep 2002 23:24:26 +0200
Subject: Remarks to Python 2.3
References: <20020918222421.328088eb.use-net@schabi.de> <amannn$p4b$1@panix1.panix.com>
Message-ID: <m3u1knj8j9.fsf@mira.informatik.hu-berlin.de>

aahz at pythoncraft.com (Aahz) writes:

> Normally, I tell people posting to python-dev that they're in the wrong
> place and should go to c.l.py.  However, this time, I think the reverse
> suggestion is in order: you should post your questions to python-dev.

Or, better yet, submit bug reports.

Regards,
Martin


From sross at connectmail.carleton.ca  Wed Sep 25 23:18:59 2002
From: sross at connectmail.carleton.ca (Sean Ross)
Date: Wed, 25 Sep 2002 23:18:59 -0400
Subject: Removing all occurences of a character from a string.
References: <amr8nh$rd8$1@driftwood.ccs.carleton.ca><DL9k9.201596$216.7001230@bin4.nnrp.aus1.giganews.com><amrav1$4rk$1@driftwood.ccs.carleton.ca> <mailman.1032965672.24886.python-list@python.org>
Message-ID: <3%uk9.7131$l44.1367805@news20.bellglobal.com>

True.
In fact, I will be using a dictionary as you say. However, before I could
use the dictionary in the manner you've indicated it was first necessary, or
at least preferred by me, to remove the single quotes around the attribute
name and attribute values. Thanks to the suggestions here, that is
accomplished. And, yes, the list of lists has since been converted into a
dictionary with attribute name keys and list of attribute values. Although,
I'm still working on my solution to the larger problem (making a decision
tree), and I may choose to change the dictionary keys to indexers into my
data set. Anyway, that's neither here nor there. For now, let me thank you
all again for your time,
Thanks,
Sean

"John Hunter" <jdhunter at ace.bsd.uchicago.edu> wrote in message
news:mailman.1032965672.24886.python-list at python.org...
>
> From the example you gave, I was wondering if a dictionary might be a
> more suitable data structure:
>
> attr = { "gill-attachment" : ["a", "d", "f", "n"],
>          "gill-spacing"    : ["c", "d", "w"],
>          "gill-size"       : ["b", "n"]
>          }
>
> print attr["gill-size"]
>
> John Hunter
>




From bloke at ii.net  Tue Sep 17 21:17:01 2002
From: bloke at ii.net (Rob Hall)
Date: Wed, 18 Sep 2002 09:17:01 +0800
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <Xns928CED692541cliechtigmxnet@62.2.16.82> <6_Ph9.479711$q53.16315681@twister.austin.rr.com>
Message-ID: <3d87d3f9$0$28706@echo-01.iinet.net.au>


> Well, our shop is win32 only, but on a personal level I'd like to be more
> flexible (I mostly use Linux at home).  Isn't Tk still considered the
> "official" GUI lib?
>
> For a newbie, what's easier, Tk or wxWindows?

Yes, tkinter is the official one, but many believe it should be wxPython
(www.wxpython.org) .  wxPython is an extremely good x-platform
implementation.  It is fast, with a native look and feel.

wxpython also has some good ide packages.  Boa is pretty good.
PythonCardPrototype is also very good.

If you use rational rose, I've started work on a python implementation.  It
will also work with visual modeller.


Rob




From gmuller at worldonline.nl  Thu Sep 12 16:20:37 2002
From: gmuller at worldonline.nl (GerritM)
Date: Thu, 12 Sep 2002 22:20:37 +0200
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid> <alln71$1s2s$1@nntp6.u.washington.edu> <b16e4ef7.0209110616.48dfe04c@posting.google.com> <3D7FFC43.10002@something.invalid> <b16e4ef7.0209120504.62d3c5bd@posting.google.com> <alqiue$1ejo$1@nntp6.u.washington.edu>
Message-ID: <alqt2q$q8u$1@reader1.tiscali.nl>

"Russell E. Owen" <owen at nospam.invalid> schreef in bericht
news:alqiue$1ejo$1 at nntp6.u.washington.edu...
<...snip...>
> True. It would help if the command callback sent an arguement that
> identified the widget (making it look a bit more like an event binding).
> In any event, it doesn't and command is unfortunately still the best way
> to handle button widgets.

In a recent thread it was pointed out that you can also bind to s function
of a specific object instance. Out of the top of my head:

Class A:
    def printarg( self, arg ):
        print arg

anA = a()
functionBoundToInstance = anA.printarg

functionBoundToInstance("Hello Python")

Hello Python

this would work also for widgets, command = aWidget.activate

Sorry didnot try it out yet :-)

regards Gerrit
--
www.extra.research.philips.com/natlab/sysarch/





From Oschler at earthlink.net  Fri Sep  6 15:15:37 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Fri, 06 Sep 2002 19:15:37 GMT
Subject: Interapp communication under Python[Linux]?
References: <mailman.1031334429.8860.python-list@python.org>
Message-ID: <t97e9.25447$w51.7054198@news2.news.adelphia.net>

<brueckd at tbye.com> wrote in message
news:mailman.1031334429.8860.python-list at python.org...
> On Fri, 6 Sep 2002, Robert Oschler wrote:
>
> It _sounds_ like you want the separate pieces to act as if they were all
> one giant process, so maybe really linking them together is the way to go.
> Personally I really like a socket-based approach (roll your own protocol
> or use XML-RPC/SOAP/etc).
>
> Putting such questions aside, I'd first evaluate your solutions in terms
> of difficulty in Java and C++ because all of the ones you listed are
> pretty easy in Python.
>
> -Dave

Dave,

Currently, most of the modules external to Python will either do a large
intensive database lookup, or a heavy esoteric calculation.  There will be
frequent calls to them, but based on what I just said you can see that the
percentage ratio of the time it takes to receive the request and then return
an answer to Python, versus the lengthy operation itself, will be quite
large.  (So pretty much any of the communication strategies I outlined won't
be signicant when it comes to the overhead of the call itself, regardless of
the protocol used.)  Yes I too really like the socket approach, although
setting up the data object transmission format, especially the quality
control and validation parts of it, and creating a robust async receipt
mechanism on the client side, is a headache.

thx





From roffe at aqualene.uio.no  Sun Sep  1 03:12:36 2002
From: roffe at aqualene.uio.no (Rolf Marvin Bøe Lindgren)
Date: Sun, 01 Sep 2002 09:12:36 +0200
Subject: [Q] win32com/Excel
Message-ID: <lbzsn0uw5cr.fsf@aqualene.uio.no>

I would like to emulate this Visual Basic statement from Python, using
Com:

        Range(Range("A1"), ActiveCell.SpecialCells(xlLastCell)).Select

this statement selects the active worksheet (which is, incidentally,
not similar to UsedRange).  

any clues?  

thank you,

-- 
Rolf Lindgren                                            http://www.roffe.com/
roffe at tag.uio.no


From loewis at informatik.hu-berlin.de  Fri Sep 13 06:12:27 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 13 Sep 2002 12:12:27 +0200
Subject: no RHL 7.1 RPMs for python2.1
References: <mailman.1031910461.5602.python-list@python.org>
Message-ID: <j4sn0etcz8.fsf@informatik.hu-berlin.de>

"Ben Gerblich" <benji_75 at hotmail.com> writes:

> Why now does the RPMs for Python2 not appear on python.org?

Because they are contributed, instead of being released by Python
Labs.

> Is there a known problem?

No.

Regards,
Martin


From peter at engcorp.com  Wed Sep 18 18:56:10 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 18 Sep 2002 18:56:10 -0400
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <mailman.1032368428.1671.python-list@python.org>
Message-ID: <3d89047d$1@news.sentex.net>

Bradley D. Larson wrote:
> I might have missed it... (I read most of the posts on this subject).
> Why not just use a significant info registration key. The registration
> key would be sensative to date, registered name and/or some other
> piece of information that would be required to get the software to work.
> 
> At least then  (if the person buying the pirated software) would have to:
> 
>     1. Set their computer date BACK to the DATE of the original reg key use.
>     2. Enter the name of the original owner.
>     3. Enter the other information (ie. version, address -- if they move they get a
> new regkey).
> 
> This would do several things.
> 
>     1. Become pretty obvious to the purchaser that the software is pirated.
>     2. The pirater and/or purchaser would have to know all items relating to
>             the reg key.

You overlook the very common situation of pirates seeing this as a
challenge and breaking the program enough to install a fake registration
key that has user name "Johnny H4X0R" etc, and posting it to a dozen
warez sites.  Then all the other pirates download it and us it without
any original registry key involved.

Given the speed with which every program with such protections makes
it onto the warez sites with broken protection, you have to wonder why
people still think this is of much use.

-Peter



From fredrik at pythonware.com  Thu Sep  5 13:27:13 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 05 Sep 2002 17:27:13 GMT
Subject: Max line coords canvas.create_line() can display?
References: <fa06e058.0208291055.4b0dd338@posting.google.com> <5pid9.5885$e5.1033551@newsb.telia.net> <117880a1.0209042325.68193ca7@posting.google.com>
Message-ID: <RtMd9.9505$HY3.2176148@newsc.telia.net>

J?rgen Hansen wrote:

> the small program below fails to draw the line, when it is called with
> an integer larger than or equal to 16384. I only tested the script on
> a Windows 98 machine with Python 2.1.3.

this works just fine under Windows 2K.

a little digging reveals that this is a limitation of the Windows 9X
GDI implementation:

http://msdn.microsoft.com/library/default.asp?url=/library/en-us/gdi/linecurv_10bp.asp

    Windows 95/98/Me: Polyline cannot draw more than a certain
    number of points. The limit depends on the line width (that is,
    the width of the pen selected into the DC), as shown in the
    following table.

    Line width -- Maximum points allowed
    line width is 1 -- 16K
    line width > 1 (that is, wideline) and device supports wideline -- 16K
    line width > 1 but device does not support wideline -- approximately
        1360 (that is, approximately 16K / 12)

    Any extra points are ignored. To draw a line with more points,
    divide the data into groups that have less than the maximum
    number of points and call the function for each group of points.
    Remember to connect the line segments.

(looks like a 64k limit somewhere in the device driver interface)

</F>




From richiereynolds at GETRIDOFTHISntlworld.com  Sun Sep 29 18:10:48 2002
From: richiereynolds at GETRIDOFTHISntlworld.com (Richard Reynolds)
Date: Sun, 29 Sep 2002 23:10:48 +0100
Subject: idea for a much needed application
References: <576c1752.0209291326.56dcdf31@posting.google.com>
Message-ID: <JTKl9.1293$_X3.109133@newsfep1-win.server.ntli.net>

check out www.zoneedit.com

"mike henley" <rsrchstr at hotmail.com> wrote in message
news:576c1752.0209291326.56dcdf31 at posting.google.com...
> I remember a while ago reading about tim burners-lee and how his
> vision of the WorldWideWeb as a medium where all can freely and easily
> publish information has not materialized properly, and i felt it so
> strongly over the past week when trying to look at options for hosting
> a website. Hosts are either expensive or limit you in terms of what
> languages, modules, applications, and OSs you can use. They also limit
> you in terms of space, bandwidth used per period, number of executions
> of CGIs... etc. Some of those who give good deals, seem difficult to
> trust with either your credit card number or your content.
>
> I wish i can just use my own computer, with whatever technologies I
> choose, keeping my content that i build up over time safe here on my
> own machine, but unless you have a broadband connection this is not
> possible. It may be possible using a dynamic DNS provider, but then
> again either they add cost, end up with an ugly URL, and, most
> importantly, there's no way for others to know when you get online. So
> here comes the idea that occured to me, and i post it here to ask if
> you know of an application that already provides this functionality,
> then please tell me what it is, or if you think it is worth doing on
> an open source basis, or even proprietary, in which case i invite
> anyone to do so, or if you think it's a completely stupid and
> impossible idea, then perhaps you can please tell me why.
>
> You probably guessed if you read as far as the second paragraph that
> what i'm thinking about is an application functionality that enables
> instant messaging to provide the dynamic DNS functionality. So instead
> of giving others a URL I can just give them an IM username they can
> add to their client, and when i go online, through my humble dial-up
> connection that changes IP whenever i log in, they'll be aware of my
> presence, and be able to connect to a server i host on my machine.
>
> It sounds like a simple thing; IM presence, automatic IP query of my
> username, click on my username and or use a contextual menu option or
> whatever and a browser launches to go to the site i'm hosting on my
> machine. Perhaps it can use existing open-source platforms, such as
> jabber or gnutella, or as an add-on to existing open-source
> applications/clients. Basically what i'm suggesting is a little tool
> to enable a server such as apache to join the whole IM/p2p thing and
> have "presence". The more automated it is, and the easier it is, the
> better. I think such a functionality will be especially useful for
> certain uses, in my case it will be hosting a wiki, as i leave my
> machine online a lot through my dial-up, can't predict the number of
> executions of CGI or the amount of space needed, wish to be free to
> use whatever wiki application is best for my needs and in whatever
> language regardless of what a host will support, and do not wish to
> pay a monthly fee, extra charges, setup fee or a domain registration.
> Perhaps there can be other areas where it is useful, and i'm sure
> there are other areas where it will not be suitable. Perhaps there can
> also be a simple way to enable a page hosted in such a way to be
> submitted to search engines, i haven't thought about that yet, but
> tell me if u see any.
>
> What do you think? already done, worth doing, stupid, or impossible?
> (p.s. please also let me know if there are other newsgroup to which
> posting this will be a good idea)




From quiteblack at yahoo.com  Thu Sep 19 06:11:58 2002
From: quiteblack at yahoo.com (black)
Date: Thu, 19 Sep 2002 03:11:58 -0700 (PDT)
Subject: surface of pygame
Message-ID: <20020919101158.58549.qmail@web21305.mail.yahoo.com>


Hi,all~

I touched something about pygame today and confused with surface type. I didnt remember surface data type when learning python, and it displays strange to me, for example, I create a new surface by loading a picture as follows:

pygame.image.load("myPic.bmp") 

it works fine but I coulnt see any surface appears on the screen.
any ideas ?





---------------------------------
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020919/a28c1985/attachment.html>

From maxm at mxm.dk  Mon Sep  9 10:22:13 2002
From: maxm at mxm.dk (Max M)
Date: Mon, 09 Sep 2002 16:22:13 +0200
Subject: Zope Status?
References: <Xns9281221A07167bosahvnetscapenet@66.185.95.104> <cSOe9.3024$io.1100@nwrddc03.gnilink.net> <3D7C4833.2080308@mxm.dk> <c_1f9.2$on1.1@nwrddc01.gnilink.net>
Message-ID: <3D7CAE95.7090202@mxm.dk>

Jeff Sasmor wrote:

 From what I've seen, Zope 3 is more regular and better organized
than Zope 2, but it _is_ different.

Anyway, that's all I meant. Nothing bad, just inevitable. Anyway,
it's just my opinion, so it doesn't matter all that much!



I havn't poked round that much in Z3 either, but from the example 
product in the "Products3" repository in cvs I could actually go in and 
read and understand what was going in from the source only. In current 
Zope I have never been able to do that ;-)

It seems almost as simple as Python itself.



regards Max M

 ><evolve>
   >    >



From member at dbforums.com  Mon Sep 16 21:06:56 2002
From: member at dbforums.com (hpyhpy)
Date: Tue, 17 Sep 2002 01:06:56 +0000
Subject: A question about C/API of python.
References: <1818979.1032155554@dbforums.com> <mMhh9.5390$V7.1451374@twister.socal.rr.com>
Message-ID: <1822704.1032224816@dbforums.com>

I try and thanks.

--
Posted via http://dbforums.com



From brueckd at tbye.com  Fri Sep  6 14:53:05 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Fri, 6 Sep 2002 11:53:05 -0700 (PDT)
Subject: Interapp communication under Python[Linux]?
In-Reply-To: <St5e9.25080$w51.7041171@news2.news.adelphia.net>
Message-ID: <Pine.LNX.4.44.0209061146160.1256-100000@ns0.tbye.com>

On Fri, 6 Sep 2002, Robert Oschler wrote:

> My Python app will talk to 3 or 4 modules written in Java and C/C++.
> Fortunately I have the source for the Java and C/C++ modules and I have
> experience in both languages.  Currently the external modules have no
> provision for being accessed outside of the process they are running as.  So
> I'm wondering what the best protocol would be used to "glue it all
> together".  There a dizzying array of possibilities:
> 
> 1) Have everybody talk to each other via SOAP (web services).
> 2) Have everybody talk to each other via XML-RPC (Don't know much about this
> method).
> 3) Use Boost's SWIG interface for C++ to link the C/C++ modules to Python.
> How to link to Java in the same manner?
> 4) Use "sockets" and my own custom protocol for everybody.
> 5) Code in Jython instead and use various methods above to talk to C/C++
> modules from Jython app.
> 6) CORBA? (yikes!)
> 
> So as I said, a lot of ways to get things done.  I'd like to hear from those
> of you that have actually done a lot of cross-app communication with Python
> on what your experiences have been, and the pros and cons you encountered.

It depends on the nature of the communication between the various parts. 
Is the communication lots of frequent messages? Are the messages large or 
small? Does communication mostly flow in one direction?

It _sounds_ like you want the separate pieces to act as if they were all 
one giant process, so maybe really linking them together is the way to go. 
Personally I really like a socket-based approach (roll your own protocol 
or use XML-RPC/SOAP/etc).

Putting such questions aside, I'd first evaluate your solutions in terms 
of difficulty in Java and C++ because all of the ones you listed are 
pretty easy in Python.

-Dave




From jjl at pobox.com  Sat Sep  7 09:04:43 2002
From: jjl at pobox.com (John J. Lee)
Date: Sat, 7 Sep 2002 14:04:43 +0100
Subject: PyQT picture display
In-Reply-To: <albp27030dq@enews4.newsguy.com>
References: <albp27030dq@enews4.newsguy.com>
Message-ID: <Pine.SOL.4.44.0209071403200.22048-100000@mimosa.csv.warwick.ac.uk>

On Fri, 6 Sep 2002, Jim wrote:

> I've a widget with three buttons and a label. I want to display a JPEG
> picture when either of the first two buttons is clicked. When I run the
> code below, the widget displays on screen but when either button is clicked,
> the picture does not appear( the text in the label disappears as it should

Haven't read your code, but have you checked that JPEG support is compiled
into your Qt library?  It's optional.


John



From HRG at popmail.com  Thu Sep 19 00:54:51 2002
From: HRG at popmail.com (HRG)
Date: Thu, 19 Sep 2002 04:54:51 GMT
Subject: cephes.pyd for python2.2?
Message-ID: <Xns928DF2CE3405DHRGpopmailcom@204.189.71.76>

Does anybody have cephes.pyd compiled for python 2.2?
Thanks,
Horatio


From jb at cascade-sys.com  Fri Sep 27 05:05:37 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Fri, 27 Sep 2002 02:05:37 -0700
Subject: Can't suspend a thread?
References: <QZ4j9.10630$Lg2.2393215@news2.news.adelphia.net> <amognf$r6g$1@panix1.panix.com> <juj2puko63pmq5k4j4tm2eonoadvpup50d@4ax.com> <uIpk9.11787$Lg2.3232725@news2.news.adelphia.net> <rqk7pu8d8nl3op3pln8bpqtrrhontqg07t@4ax.com>
Message-ID: <3D941F61.3090108@cascade-sys.com>

Tim Roberts wrote:

>Hmm, you're quite right.  I didn't realize you could suspend another thread
>in Win32.  The documentation does caution against using it in anything
>other than a debugger, but it certainly exists.
>
Yeah, it's possible but I question the wisdom of doing it.  Starting and 
stopping threads altogether or using synchronization primitives seems 
the better way to go in every case I can think of.

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From rsrchstr at msn.com  Mon Sep 30 15:35:08 2002
From: rsrchstr at msn.com (mike henley)
Date: Mon, 30 Sep 2002 20:35:08 +0100
Subject: idea for a much needed application
References: <576c1752.0209291326.56dcdf31@posting.google.com> <pasepuc0r73mh5vlcsu1l7erl4e2b17qvi@4ax.com> <an831g$bfrqn$1@ID-157863.news.dfncis.de> <7l0gpugn6lehtmj46a1e5a0stg2cfmreqa@4ax.com>
Message-ID: <ana91j$c47s0$1@ID-157863.news.dfncis.de>

no comment.

"TGOS" <tgos at invalid.invalid> wrote in message
news:7l0gpugn6lehtmj46a1e5a0stg2cfmreqa at 4ax.com...
>
> On Mon, 30 Sep 2002 00:40:24 +0100 "mike henley" <rsrchstr at msn.com>
> wrote in comp.lang.java.programmer:
>
> >> To publish information, all you need is writing it into a HTML file and
> >> a server that hands out the HTML file when getting asked for it.
> >> You don't need any special languages, modules or applications on the
> >> server to publish your information, other than the ones already present
> >> on a web-server.
> >
> > "other than the ones already present on a web-server", exactly! i want
to be
> > ably to use whatever language, modules or applications i choose, and
over
> > the year use others, freely.
>
> You just put your HTML file online, how this is served to users
> requesting it is non of your business. Your webhost guarantees that it
> is served. It doesn't matter if the WWW server is Apache or Zeus.
>
> > if i choose perl, php, or even python, ruby or tcl, then i want to be
> > able to do that.
>
> Can use all these on my webhost. However, you need NONE of these to
> publish information and that's all you want to do according to your own
> words.
>
> > if i want to use mysql or something else that i prefer, then i want
> > to have the freedom to do that.
>
> Can use that, too. But again, not necessary for your described task.
>
>  <snip>
>
> > Agree, if that's all you want, but beyond keeping some text online, it's
> > difficult.
>
> Keeping some text online (and/or some pictures) means publishing of
> information on the WWW.
>
>  <snip>
>
> > my post is not about tim-burners lee or what he said.
>
> To quote yourself:
>
> >>> I remember a while ago reading about tim burners-lee and how his
> >>> vision of the WorldWideWeb as a medium where all can freely and
> >>> easily publish information has not materialized properly
>
> This vision is already reality, make a HTML file, say what you have to
> say and put it on some server. Information... published. His vision did
> not include the usage of any script or database, as these are not
> necessary to publish information.
>
>  <snip>
>
> > my post is also not about the economics or politics of hosting;
>
> Of course it is. You demand to get something for free, others (webhosts)
> must pay for. How shall this work?
>
> Despite that, you already pay for your Internet connection, don't you?
> So when you publish something from your own PC, it's not for free
> either: You pay the fees for your dial-up account! And if anyone would
> publish from home through a dial-up account, these fees would raise
> noticeably very soon.
>
> > this sounds much like the open source vs proprietary software arguments.
>
> OpenSource is "free" software, but "free" like in "free speech"; not
> "free" like in "free beer". Nowhere is written that a service must be
> offering OpenSource software to you for free (without any costs
> involved). They can charge you whatever it pleases them. OpenSource just
> means that they will provide the source code for the software as well
> (the source code is not kept a trade secret). And often it means that
> you may use this source code in your own software, as long as it will be
> OpenSource again (GPL, but GPL != OpenSource; not all OpenSource is GPL
> software).
>
>  <snip>
>
> >> IRC, ICQ, AIM... the list is endless.
> >
> > Yes, that is what my post is about, combining instant messaging presence
> > with dynamic DNS.
>
> Then you must create your own shiny, new instance message service, as
> the already existing providers like ICQ (belongs now to AOHell, doesn't
> it?) and AIM (always belonged to AOHell) don't allow you play with
> *their network*. They write the server software, they write the client
> software, they alone decide about the network protocol (at least they
> did for years, now they pretend to seek for open standards, all bullshit
> IMHO).
>
> So what you are looking for is not a new software, but a new SERVICE.
> How will this service be financed if it shall offer everything for free?
> AOHell can finance their free services, they make enough money from 12
> millions moron stupid enough to sign-up for the joke they call "Internet
> Access".
>
>  <snip>
>
> >> Allow users to see your IP address when you are connected to an IRC
> >> network or to ICQ for example, then they can enter your IP address into
> >> their browser address bar and will get a webpage if you run a webserver
> >> and are not behind a firewall that blocks port 80.
> >>
> >
> > Perhaps it can be possible to do a little application to automate this.
>
> Perhaps users that are too lazy to copy/paste your IP address into their
> browser should go to hell?
>
> --
> TGOS




From cliechti at gmx.net  Fri Sep 27 13:30:38 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 27 Sep 2002 19:30:38 +0200
Subject: file open verification
References: <9eabe547.0209270710.17ca112d@posting.google.com>
Message-ID: <Xns9296C686777B1cliechtigmxnet@62.2.16.82>

python473 at yahoo.com (John Howard) wrote in 
news:9eabe547.0209270710.17ca112d at posting.google.com:

> Two questions - somewhat related:
> 
> Situation - I have several files on an apache server that are being
> accessed by
> python programs that are receiving data from html forms.
> 
> q1: How can the status of a file be checked?  If two forms have been
> submitted
> and the data is being added to a common file, how do I know that data
> from both forms are being written correctly to the common file?

don't ever write simultaneous from two processes to one file. your choices 
are:
- use "locks", so that only writes at a time
- one server writes the file, logs are directed to that server (like syslog  
on linux)
- use a database
- ...
all these options ensure that one writes after the other. avoiding problems 
is better than fixing.

> q2: Related to above - really more of a cgi question - does the apache
> server keep all these file accesses separate some way?

don't know apache. if you manage to limit it on one thread, only one can 
access the file. but iin that case the performace of the server is poooooor 
;-)

chris
-- 
Chris <cliechti at gmx.net>



From aleax at aleax.it  Wed Sep 25 05:47:08 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 09:47:08 GMT
Subject: how to check if a function signature "matches" without calling the function?
References: <a76c1561.0209241305.454f7ae6@posting.google.com>
Message-ID: <wCfk9.140788$ub2.3074390@news1.tin.it>

Mike McCandless wrote:

> Does Python 2.2.1 make it possible for me to see some details about
> the expected arguments that a function takes, without actually calling
> the function?

Yes.  inspect.getargspec lets you learn all you need:

import inspect
argnames, has_varargs, has_kdws, default_values = inspect.getargspec(func)


> Specifically, if I have a function or method object, and I have a
> tuple + dict of arguments I'd like to apply it on, is there a function
> I can call that would check if the tuple + dict can be properly
> "coerced" to the formal parameters?

You still have some work to do for your checking, but it doesn't
seem to be much.  You need to check you supply each mandatory
argument exactly once, and each optional argument 0 or 1 time;
you can have other positional args only if has_varargs, and
other named args only if has_kwds.  E.g. for clarity:


class NoWay(Exception): pass

def times_supplied(argname, argposition, atuple, adict):
    return (argname in adict) + (argposition < len(atuple))

def check_args(argnames, default_values, atuple, adict):

    number_mandatory = len(argnames) - len(default_values)

    for pos in range(len(argnames)):
        name = argnames[pos]
        n = times_supplied(name, pos, atuple, adict)
        if n<1 and pos<number_mandatory:
            raise NoWay, "Mandatory argument %s (%s) missing" % (name, pos)
        elif n>1:
            raise NoWay, "Argument %s (%s) given twice" % (name, pos)

def check_funcargs(func, atuple, adict):
    import inspect
    names, has_varargs, has_kdws, defaults = inspect.getargspec(func)
    check_args(names, defaults, atuple, adict)
    if not has_varargs and len(atuple)>len(names):
        raise NoWay, "Supplied %d positional args, max is %d" % (
            len(atuple), len(names) )
    if not has_kwds:
        for name in adict:
            if name not in names:
                raise NoWay, "Unknown argument %s" % name


You can express this more concisely &c, of course, or use other
means than raising a dedicated exception to diagnose problems
(I only diagnose the first problem if any, that can be tweaked
too, of course).


Alex



From peter at engcorp.com  Tue Sep  3 23:44:28 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 03 Sep 2002 23:44:28 -0400
Subject: default tuple unpacking?
References: <slrnanaknb.1gh.huaiyu@gauss.almadan.ibm.com>
Message-ID: <3d758180$1@news.sentex.net>

Huaiyu Zhu wrote:
> It is often very convenient to have default arguments such as
> 
>     def func(a, b, *c): ...
> 
> Is there a way to use this in statements too?  Examples:
> 
>     a, (b, *c), *d = e

Presumably d would end up as () if there was nothing to put there,
just as in the default argument syntax for functions?

>     for a, *b in c: ...
>     return a, *b

These are all very readable to me, having learned to grok
the basic "def func(a, b, *c):" syntax.  I'd be in favour
of seeing this added, as the first idiom above is something
I've needed fairly often.

Is there an analog for the **x syntax?  I hope nobody
suggests that having

 >>> d = { 'a' : 5, 'b' : 7 }
 >>> **d

should be equivalent to

 >>> a = 5
 >>> b = 7

:-)

-Peter



From Michael.Sparks at rd.bbc.co.uk  Fri Sep 13 06:55:09 2002
From: Michael.Sparks at rd.bbc.co.uk (Michael Sparks)
Date: Fri, 13 Sep 2002 11:55:09 +0100
Subject: wrapping yield ?
References: <alqehl$9qv$1@nntp0.reith.bbc.co.uk> <Xns92885F082A21Cduncanrcpcouk@127.0.0.1> <alsdpt$2nq$1@nntp0.reith.bbc.co.uk> <Xns9288759449799duncanrcpcouk@127.0.0.1>
Message-ID: <alsg06$3e2$1@nntp0.reith.bbc.co.uk>

"Duncan Booth" <duncan at NOSPAMrcp.co.uk> wrote in message
news:Xns9288759449799duncanrcpcouk at 127.0.0.1...
> The syntax wouldn't look quite so bad if your function wasn't returning a
> value.

Problem is, it would in most situations, also I'm trying to keep the number
of generators to a minimum whilst a lower overhead than things like threads
and processes if I had several 10s of thousands going, keeping throughput
up may become an issue. (Hence the reason for wanting the functions called
part of the same generator rather than a separate, extra one)

> There definitely isn't any way to put a yield in a function and not have
it
> turn into a generator. When you yield from a generator, only a single
frame
> is saved, not a stack, so you can only yield from the outer loop.

Ah... that makes sense then...Pity - after all, "all you'd need" (looking at
it
simplistically) is a marker in the stack frame to tell the system how much
of the
stack to save. Amiga E had something very similar (from a use perspective)
for
determining how far back to throw exceptions without requiring the
equivalent
of try/(except|catch) blocks.

Thanks!


Michael.




From wlfraed at ix.netcom.com  Sun Sep 22 16:58:07 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Sun, 22 Sep 2002 13:58:07 -0700
Subject: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amabpb$5rm$1@news.nems.noaa.gov>
Message-ID: <0talma.rq3.ln@ix.netcom.com>

ChrisBarker fed this fish to the penguins on Wednesday 18 September 
2002 10:01 am:


> 
> just like VB.
>
        Just located one of my references (O'Reilly: VB & VBA).

        Based upon the description, VB /currency/ data type is a 64-bit 
(binary) integer with implicit scaling for FOUR decimal places -- so 
one STILL has to be concerned with rounding to get to whole cents.

        The other "applicable" data type is /decimal/, which appears to be a 
28-digit packed BCD format. (Though where VB stores the sign and 
decimal point info I don't know -- book says the type takes 14-bytes).



--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From loewis at informatik.hu-berlin.de  Tue Sep  3 12:28:26 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 03 Sep 2002 18:28:26 +0200
Subject: xml question
References: <3D74CB6B.7697013E@millfilm.co.uk>
Message-ID: <j4znuzf36d.fsf@informatik.hu-berlin.de>

Eric Texier <erict at millfilm.co.uk> writes:

> From Disk: (unicode)
> N1 [(None, u'fileName'), (None, u'OutputType'), (None, u'name'), (None,
> u'UId'), (None, u'version'), (None, u'node')]
> From Memory:
> N2 ['node', 'version', 'UId', 'fileName', 'OutputType', 'name']
> 
> What will be the simplest thing to do in this case?

It appears that when parsing from disk, you have used a
namespace-aware mode, and when building the document from memory, you
use a namespace-unaware mode.

I recommend to use the same mode of operation for both cases. To get
namespace-unaware parsing from disk, assuming you use minidom.parse,
you should create a SAX parser that has the namespaces feature turned
off.

HTH,
Martin



From rajarshi at presidency.com  Tue Sep 10 09:56:53 2002
From: rajarshi at presidency.com (Rajarshi Guha)
Date: Tue, 10 Sep 2002 09:56:53 -0400
Subject: os.tempnam() warning
Message-ID: <pan.2002.09.10.09.56.50.41586.29853@presidency.com>

Hi,
  whenever I use the function os.tempnam() I get

/usr/local/adapt/bin/striphin.py:14: RuntimeWarning: tempnam is a potential security risk to your program
  outfilename =  os.tempnam()

The C library advises against using tempnam() and suggests using
mkstemp(). Is there any Python equivilant?

(I use Python 2.2)

Thanks,
Rajarshi Guha


From maxm at mxm.dk  Wed Sep 18 15:31:30 2002
From: maxm at mxm.dk (Max M)
Date: Wed, 18 Sep 2002 21:31:30 +0200
Subject: cgi binary upload
References: <d5t2JzAIBNi9EwtI@jessikat.demon.co.uk>
Message-ID: <3D88D492.1080909@mxm.dk>

Robin Becker wrote:

> I know there's a lot of skilled people using python so I thought I'd ask
> how one can get an HTML page to do a proper binary upload using the 
> <input type=file> form tag. We want to do this for PDF documents, but it
> seems as though IE/Mozilla/Netscape clients insist on putting 
> 
> Content-Type: data/pdf
> 
> or similar headers into the multi part file section and the files have
> their line endings corrupted.


That should not be a problem. I have not heard of the cgi module having 
those kind of problems at least.


> On the receiving side using cgi.py we are certainly seeing some
> filetypes OK eg zip files, but surely we should be able to fix this more
> reliably.


cgi usually is reliable.

> I should note that experimentally renaming the file as
> xxx.xpdf made no difference as the mime type was being correctly
> determined by the browser. 


I don't think your problem is with Python. You are probably having a 
problem somewhere else in the chain.

At least we should see some code, or you need to be more specific.

regards Max M




From jdavis at empires.org  Fri Sep 27 23:34:06 2002
From: jdavis at empires.org (Jeff Davis)
Date: Sat, 28 Sep 2002 03:34:06 GMT
Subject: Python Process that can survive
References: <571739c3.0209270629.10575f47@posting.google.com>
Message-ID: <Oq9l9.68435$U7.25221497@twister.socal.rr.com>

$ python -c "
> import time
> time.sleep(100) " &

That seems to get me a background process. Can you describe your problem in 
more detail? perhaps show what you did with perl?

Regards,
        Jeff Davis


Christopher Peery wrote:

> Alright... I love python but I think I may have found a definite
>  problem. I can't seem to run a python process, kick it to the
>  back-ground, and then kill the terminal that started it. With most
>  other things, the python process would keep running in the
>  back-ground regardless. For me, the process always dies with the
>  terminal.
> 
>  So my question: is there a way to separate a python script from the
>  underlying terminal so that this will not happen.
> 
>  I'm hoping for some scheme that's supported by the language. I've
>  found a few hacks on-line where the python process is daemonized
>  but this requires shutting down all the standard io streams.
>  There's got to be a better way especially since it can be done in
> Perl very easily with no special tricks.
> 
> By the way, I'm doing this on a linux box.
> 
> Chris



From h_schneider at marketmix.com  Fri Sep 27 03:12:55 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Fri, 27 Sep 2002 09:12:55 +0200
Subject: _winreg.DeleteValue - Permission denied ...
Message-ID: <an108i$ddv$03$1@news.t-online.com>

Hi,

go a prob here with _winreg.

When I try to delete a registry key, I get a WindowsError (Errno5) /
Permission denied, even thought I am logged in as admin.
Deleting the same key with regedit works.

The function looks like this:

    def del_cfg_path(self):
        key =
_winreg.OpenKey(_winreg.HKEY_LOCAL_MACHINE,'SOFTWARE\TEST\SUBTEST)
        _winreg.DeleteValue(key,"TESTVAL")
        _winreg.CloseKey(key)

Any ideas ?

Thanks a lot,
Harald





From peter at engcorp.com  Tue Sep  3 08:17:01 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 03 Sep 2002 08:17:01 -0400
Subject: Python to surpass C performance by 2030
References: <kuptw4o65q.fsf@lasipalatsi.fi> <ad052e5c.0209022303.32319d0b@posting.google.com>
Message-ID: <3d74a81f$1@news.sentex.net>

Dan Bishop wrote:
> Erno Kuusela <erno-news at erno.iki.fi> wrote in message news:<kuptw4o65q.fsf at lasipalatsi.fi>...
> 
>>just for fun i ran pystone under python versions since 1.5.2 to
>>current cvs. here are the results:
>>
>>[benchmark results snipped]
>>
>>since 1.5.2 is about 4 years old, the improvement per year is about
>>13.2% on average. if we assume that 1.5.2 is 50 times slower than
>>C (probably conservative?), then at 13.2% per year since 1998 it
>>will add up to 53x improvement in 32 years.
>>
>>i wonder if this would be a good application for the time machine...
>>
>>learn python today, and in 28 years you will have a job writing
>>the speed-critical parts of c programs in python :)
> 
> 
> Before that could ever happen, the Python interpreter would have to be
> written in something other than C.

Clearly if it will be 53x faster in 32 years even if written in C,
then it should be rewritten in Python for another 53x speedup!

Imagine the possibilities.  You could even implement another virtual
machine underneath the current one, to increase speed even more,
provided *that* one was executing Python bytecode as well.  Scary
really.  I understand this plan was discussed at a recent PSU mee



From erikprice at mac.com  Fri Sep  6 17:07:20 2002
From: erikprice at mac.com (Erik Price)
Date: Fri, 6 Sep 2002 17:07:20 -0400
Subject: successor to htmllib
Message-ID: <A1A08646-C1DC-11D6-9B02-00039351FE6A@mac.com>

I noticed that the htmllib module is really best suited to HTML 2.0 
documents.  I was wondering if there was a newer (4+) HTML or even 
XHTML parsing module in development right now.  (By XHTML parsing lib I 
mean perhaps an XML parser that is specifically written for the XHTML 
1.0 DTD.  I realize that any XML parser -should- work.)



Erik





--
Erik Price

email: erikprice at mac.com
jabber: erikprice at jabber.org




From JeffH at ActiveState.com  Wed Sep 11 12:57:51 2002
From: JeffH at ActiveState.com (Jeffrey Hobbs)
Date: Wed, 11 Sep 2002 16:57:51 GMT
Subject: Tkinter, updateStringProc for type font
References: <1907a5e4.0209040455.64ae73b1@posting.google.com>
Message-ID: <3D7F772C.80205@ActiveState.com>

Pier Paolo Glave wrote:
> I'm facing a problem with my Python program, which uses Tkinter to
> interface Tcl/Tk.
> I sometimes get this error message from Tcl: 
> "UpdateStringProc should not be invoked for type font".
> As a consequence, the application crashes with SIGABRT (6).
> 
> I'm using: Python 2.2, Tcl/Tk 8.3.3-69.
> 
> I've seen in the tcl sources that this happens when a function named 
> Tcl_GetString() is called for a "type" which has the "byte" field set
> to NULL, and no "updateStringProc" function pointer: this seems to
> happen for Tk type "font", sometimes in my application.
> Anyway, I'm not able to understand more.
> 
> Has anyone encountered this problem before?

Do you have any other C extensions in your application?  I've not seen this 
occur myself, except in bugs where other C extensions were abusing Tcl objects 
internally.

-- 
     Jeff Hobbs                     The Tcl Guy
     Senior Developer               http://www.ActiveState.com/
         Tcl Support and Productivity Solutions
  Join us in Sept. for Tcl'2002: http://www.tcl.tk/community/tcl2002/



From gleki at gol.ge  Thu Sep  5 05:39:00 2002
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Thu, 05 Sep 2002 11:39:00 +0200
Subject: SWIG 1.3.14::: import error
Message-ID: <3D772634.575A0AEB@gol.ge>

Hi all!

I well may have been missing something in new features of SWIG.

Any way, I have not found this in manual.

When I try to import a newly wrapped module, I get this error:
 >>>import PyArray
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ImportError: dynamic module does not define init function (initPyArray)


***


I guess _PyArray is linked somehow with libPyArray.so? What and where
did I miss simething?

I thought I did everything as I should:
1. I made PyArray.i
2. swig -c++ -python PyArray.i
3. g++ -c  PyArray_wrap.cxx -I/usr/local/include/Python2.0
4. g++ -shared Array.o PyArray_wrap.o -o PyArraymodule.so

Thank you in advance.

Regards, Giorgi







From fperez528 at yahoo.com  Mon Sep 23 13:12:07 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 23 Sep 2002 11:12:07 -0600
Subject: numpy array assignment
References: <m2bs6on46w.fsf@mother.paradise.lost> <mailman.1032794130.2250.python-list@python.org>
Message-ID: <amni18$44s$1@peabody.colorado.edu>

John Hunter wrote:

>>>>>> "John" == John Hunter <jdhunter at ace.bsd.uchicago.edu> writes:
> 
>     John> I have a 2 dimensional array X which is NxM and a 1
>     John> dimensional array v which is Nx1.  I want to assign v to the
>     John> k-th column of X
> 
> This is what I've come up with so far -- if there is a better/faster
> way, please let me know....
> 
> import Numeric
> 
> def putcol(k, Z, v):
>     (n,m) = Z.shape
>     Numeric.put(Z, k+m*Numeric.arange(n), v)
> 
> def putrow(k, Z, v):
>     (n,m) = Z.shape
>     Numeric.put(Z, k*m+Numeric.arange(m), v)
> 
> 
> Z = Numeric.zeros( (3,5), Numeric.Float  )
> vc = Numeric.ones( (3,1), Numeric.Float )
> vr = Numeric.ones( (1,5), Numeric.Float )
> 
> putcol(3, Z, vc)
> putrow(1, Z, vr)
> print Z

How about:

In [38]: z=Numeric.zeros( (3,5), Numeric.Float  )

In [39]: z[1] = vr[0]

In [40]: z[:,3]=vc[:,0]

In [41]: z
Out[41]:
array([[ 0.,  0.,  0.,  1.,  0.],
       [ 1.,  1.,  1.,  1.,  1.],
       [ 0.,  0.,  0.,  1.,  0.]])

Cheers,

f.


From gerhard.haering at gmx.de  Mon Sep 16 00:44:44 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Mon, 16 Sep 2002 06:44:44 +0200
Subject: pypgsql -- not preserving transactions ?
In-Reply-To: <am3id4$1epu$1@nntp6.u.washington.edu>
References: <am3id4$1epu$1@nntp6.u.washington.edu>
Message-ID: <20020916044444.GA1566@lilith.ghaering.test>

First, I'd recommend to ask questions like this on the pypgsql-user
mailing list. You'll find the link on  http://pypgsql.sf.net/

* Frank Miles <fpm at u.washington.edu> [2002-09-16 03:15 +0000]:
> I've discovered that the 2.0 version of pyPgSQL packaged with Debian/woody
> doesn't preserve transactions (at least with Postgres).  For example, if
> a transaction consists of a series of INSERTs, and an error occurs partway
> through the series, a COMMIT at the end of the series will result in the
> INSERTs up to the erroneous one being thrown out (rollback), but unfortunately
> those following the error being committed.

I can't reproduce that behaviour here. What do you do to ignore the
PgSQL.OperationalError? Or is it a different exception that gets thrown?

> In looking at the imported code, there are comments about trying to
> recover from a defective state, with the class variable
> 'inTransaction' being cleared.
> 
> Does any newer version fix this nonsense?

This area of code hasn't changed up to 2.2, IIRC. Could you provide a
test case with a minimal schema, minimal code and the output of your
code. To see which SQL commands pyPgSQL sends, you can simply access
pyconn.conn.toggleShowQuery, where pyconn is your DB-API connection
object. This output would be useful, too. Also, which PostgreSQL server
version are you using?

-- Gerhard



From mcherm at destiny.com  Tue Sep 10 12:26:51 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Tue, 10 Sep 2002 12:26:51 -0400
Subject: Pickle problem
References: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220D10@hendrix.empolisuk.com>
Message-ID: <all6g8$t21$1@main.gmane.org>

>>Are you trying to store the MatchObject returned by a call to 
>>a regular 
>>expression object's match() method?  If so, it looks like you're 
>>forgetting the parens at the end of "re.compile(y).match".

> No, I haven't forgotten the parentheses. I'd like to pickle the match method 
 > of the compiled regular expression.

Three different objects to consider.

  (1) "the compiled regular expression object":  re.compile(y)
  (2) "the match method":                        re.compile(y).match
  (3) "the MatchObject, ie the results":         re.compile(y).match()


Objects (1) and (3) can be pickled, I don't think (2) can. Here's a 
quick summary of what they're there for.

   Using (3) will allow you to examine the results of the match.
   Using (2) will allow you to re-run the match on a different string.
   Using (1) will allow you to perform a match, search, findall, or
      other functions on any string.

Since you can't pickle (2), I suggest storing (3) unless you need to 
re-run it on a different string, then store (1).

-- Michael Chermside

PS: If what you are trying to do is save in the pickle the exact way 
that the "match" method works because you are afraid that the 
maintainers of Python may come out with a new version of Python in which 
match works differently and is no longer backward compatible, and you 
want your code to still work the same way in this hypothetical new 
version of Python, then you really *DO* need to store (2). 
Unfortunately, it's still not possible (AFAIK). Fortunately, the 
maintainers Python wouldn't do that to you.






From ifls at rz.tu-clausthal.de  Mon Sep 30 02:08:27 2002
From: ifls at rz.tu-clausthal.de (Lutz Schroeer)
Date: 30 Sep 2002 06:08:27 GMT
Subject: Getting last char in string.
References: <amue3q$h59$1@lust.ihug.co.nz>
Message-ID: <an8por$23p3$1@ariadne.rz.tu-clausthal.de>

CheapSkate <gua81 at xxxyahoo.com> wrote:
: my Q now is how to get the last char???

>>> s = "Hello!"
>>> print x[-1:]
!

greets
Lutz


From benji_75 at hotmail.com  Thu Sep 19 01:09:04 2002
From: benji_75 at hotmail.com (Ben Gerblich)
Date: Thu, 19 Sep 2002 14:39:04 +0930
Subject: FIXED. try import except want lineno from Exception
Message-ID: <F131qqgNjWOiy54DgI60000042f@hotmail.com>

In the file "progfile.py", replace between the print start/end with:
<a more glorified version>
--<snip>--
self.userDefs[searchPath][file[:-3]] = {}
try:
    self.userDefs[searchPath][file[:-3]]['class'] = __import__(file[:-3])
    code = "object = self.userDefs[searchPath][file[:-3]]['class'].%s()" % 
file[:-3]
    exec code
    self.userDefs[searchPath][file[:-3]]['obj'] = object
except Exception, e:
    errortext = e
    for badfile, badlineno, badfunction, badtext in traceback.extract_tb( 
sys.exc_info()[2]):
        pass

    # ask the user to 1.'skip' this MDF continue loading other MDF's
    #              or 2. quit the application
    message = "The Module Definition File has an error.\n\n"
    message = message + "Filename: '%s'\n" % fullfilename
    message = message + "Line number: %s\n" % badlineno
    message = message + "Source Code: %s\n" % badtext
    message = message + "Function: %s\n" % badfunction
    message = message + "Error message: %s\n" % e
    message = message + "\n\nSkip this module, but load others?\n"
    message = message + "\nOr CANCEL the application?"

    response = wxMessageBox(message , "Confirm", wxOK | wxCANCEL| 
wxICON_EXCLAMATION, self)
    if response == wxOK:
        print '   ** Did not load this Module Definition File **'
        del self.userDefs[searchPath][file[:-3]]
    elif response == wxCANCEL:
        sys.exit(1)


>From: "Ben Gerblich" <benji_75 at hotmail.com>
>To: python-list at python.org
>Subject: try import except want lineno from Exception
>Date: Thu, 19 Sep 2002 13:53:39 +0930
>
>I am importing module files, and want to catch an error (if any) and
>print the filename and line number of the error.
>The "Exception ,e" does not return, or does not have the attirbutes listed 
>above. It only returns e="line() takes exactly 2 arguments (3 given)"
>
>I think I want to use SyntaxError as it has the attributes: filename, 
>lineno, offset and text.
>I have found no info on traceback.
>With the code below, how do I find the **line_number**?
>Thanks.
>--<snip>--
>[13:45:40]~/py/dev/import2:cat runme.py
>from progfile import *
>
>[13:45:44]~/py/dev/import2:cat progfile.py
>class Foo:
>    def __init__(self):
>        self.lines = []
>
>    def line(self,x):
>        self.lines.append( x )
>
>print 'start'
>try:
>    klass = __import__('module')
>    code = "object = klass.%s()" % 'Bar'
>    exec code
>except TypeError, e:
>    print 'TypeError\n%s' % e
>except Exception, e:
>    print 'Exception\n%s' % e
>print 'end'
>
>[13:45:51]~/py/dev/import2:cat module.py
>from progfile import Foo
>
>class Bar(Foo):
>    def __init__(self):
>        Foo.__init__(self)
>        print 'line-1'; self.line(20)
>        print 'line-2'; self.line(40,60) # <--- an error!!!
>        print 'line-3'; self.line(80)
>
>[13:45:57]~/py/dev/import2:python2 runme.py
>start
>line-1
>line-2
>TypeError
>line() takes exactly 2 arguments (3 given)
>end
>
>[13:46:07]~/py/dev/import2:
>
>
>_________________________________________________________________
>Join the world?s largest e-mail service with MSN Hotmail. 
>http://www.hotmail.com
>
>
>--
>http://mail.python.org/mailman/listinfo/python-list




_________________________________________________________________
MSN Photos is the easiest way to share and print your photos: 
http://photos.msn.com/support/worldwide.aspx




From mstenner at phy.duke.edu  Wed Sep 18 17:27:13 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Wed, 18 Sep 2002 17:27:13 -0400
Subject: monetary applications (et al)
In-Reply-To: <E17rm1M-0002iZ-00@mail.python.org>; from tg5027@citlink.net on Wed, Sep 18, 2002 at 05:35:32PM -0400
References: <E17rm1M-0002iZ-00@mail.python.org>
Message-ID: <20020918172713.A10897@phy.duke.edu>

On Wed, Sep 18, 2002 at 05:35:32PM -0400, terry wrote:
> I do admit to being OO inhibited, so maybe you could take it from this 
> short piece code and execution results, and, provide a brief 
> description of what you would put into this money class that would 
> eliminate this error message, while maintaining the code simplicity.

While you're at it, please write me 4-liner that predicts the stock
market's behavior.  :)

Take a look at this:
http://python.org/doc/2.2.1/ref/numeric-types.html

for how to emulate numeric types.

					-Michael

> -------------------------------------------------
> class money:
> 	def __init__(self,name):
> 		self.name = name
> 
> unitprice = money(100)
> qty = 2
> amount = qty * unitprice
> 
> terry at terry:~> python money.py
> Traceback (most recent call last):
>   File "money.py", line 7, in ?
>     amount = qty * unitprice
> TypeError: unsupported operand type(s) for *: 'int' and 'instance'
> -------------------------------------------------
> 
> Support for mixed number types in expressions involving this money 
> class must come from somewhere - where?  And, it must be capable of 
> determing and setting amount to be of class money to be consistent with 
> the rest of Python's complexity hierarchy for the handling of numbers.
> 
> The rest of the problem is that the above is a trivial calculation.  
> What happens to your recommendation when faced with an additional 
> number type?
> 
> taxrate = 0.06
> amount = qty * unitprice + (qty*unitprice)*taxrate
> 
> As I see it, if Python had an intrinsic number type of 'money', the 
> logic for determing the results of the above calculation would be 
> easily determined and consistent with expectations, as in VB.
> 
> Thanks for your time,
> 
> terry
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From prouleau at impathnetworks.com  Thu Sep 12 13:28:57 2002
From: prouleau at impathnetworks.com (Pierre Rouleau)
Date: Thu, 12 Sep 2002 13:28:57 -0400
Subject: Is there a portable way to copy files in Python?
References: <mailman.1031842155.17931.python-list@python.org>
Message-ID: <3D80CED9.6050703@impathnetworks.com>

Oops... next time i'll lokk a bit more in the lib documentation...

Thanks!

Mark McEahern wrote:
> [Pierre Rouleau]
> 
>>Subject: Is there a portable way to copy files in Python?
> 
> 
> Have you tried shutil?
> 
> // m
> -
> 
> 



From marklists at mceahern.com  Tue Sep 17 12:52:38 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 17 Sep 2002 11:52:38 -0500
Subject: handling exceptions
In-Reply-To: <fa576652.0209170827.4ba29201@posting.google.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOMEECDPAA.marklists@mceahern.com>

> The problem is: how can I access the message string of the exception?

try:
  raise RuntimeError("Gee, I wonder what this is?", "Oh, and how about this
too?")
except RuntimeError, e:
  print e

// m

-




From donn at u.washington.edu  Fri Sep  6 12:19:52 2002
From: donn at u.washington.edu (Donn Cave)
Date: 6 Sep 2002 16:19:52 GMT
Subject: Timothy Rue
References: <898662b7d85d2c966c0bd45d563d5d2d@xganon.com>
Message-ID: <alakj8$20qo$1@nntp6.u.washington.edu>

Quoth xganon <remailer at xganon.com>:
| Does Timothy Rue still Troll here? 
| If not, how did you get rid of him?
| We need him out of our group now,
| maybe we could get him into a Perl
| group for a while and let them deal
| with him?

If the problem is only that he posts to your group, then your
group is way ahead of comp.lang.python.  When he was here, a
number of regulars utterly lost touch with whatever common sense
they may ordinarily possess, and it went on for what seemed like
months.  If you don't want to engage in his discussion, then don't.
I think you'll find that it works as well as anything.

	Donn Cave, donn at u.washington.edu


From donal.k.fellows at man.ac.uk  Fri Sep  6 06:38:12 2002
From: donal.k.fellows at man.ac.uk (Donal K. Fellows)
Date: Fri, 06 Sep 2002 11:38:12 +0100
Subject: Could Python supplant Java?
References: <Xns926F8B90671A0AusYourStandingInIt@130.133.1.4> <slrnam3o3f.qo.gerhard.haering@haering.opus-gmbh.net> <8d3f4438.0208202348.578ff9c8@posting.google.com> <H16r29.KxE@bath.ac.uk> <ak025t$p8q$2@srv38.cas.org> <um80n06kf34081@news.supernews.com> <akv697$qrr$1@wagner.wagner.home>
Message-ID: <3D788594.79DE84C6@man.ac.uk>

Victor Wagner wrote:
> Dan Johnson wrote:
[I think this attribution is correct]
>> What is so developer-hostile about Windows?
> 
> "Friendly" interface. To be a "friendly" is to consider oneself equal
> with his friends. If you want your friend to do something for you,
> you ASK him, not ORDER him.
[...]

"Friendly" interfaces are interfaces that let users do things in the order they
want to, and they drive everything from the users' viewpoint.  Leaving that
critical field blank because the person who knows the information is off on
vacation?  That's a friendly thing to do.  Explaining that you can't save the
information because some fool that shares the computer with you has decided to
share MPEGs from the machine?  That's a friendly thing to do.  Both are hard to
program.

Let's face it, even with the help of a top-notch GUI toolkit, writing a good GUI
is hard.  And playing well in the Windows world demands a GUI (by culture, even
if for no better reason.)  That's why developing for Windows is hard.  (A vast,
baroque API with only mediocre documentation doesn't help, but that's not a
problem limited to Windows.)

Give me a nice clean protocol (where any deviation allows me to say "go away, I
don't want to talk to you") any day.  So much easier to implement!

Donal.
-- 
Donal K. Fellows   http://www.cs.man.ac.uk/~fellowsd/   donal.fellows at man.ac.uk
-- Of course, she was referring not to the current but to a past chair, unless
   Babbage is very, very old, *and* tied to a chair in Hawking's basement.
   Which, admittedly, would be pretty cool.                   -- Chris Hammock


From aleax at aleax.it  Thu Sep 19 11:40:08 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 19 Sep 2002 15:40:08 GMT
Subject: strong/weak - dynamic/static [Was: Getting started]
References: <mailman.1032445617.19510.python-list@python.org> <87ofaut2xt.fsf@pokey.henrik-motakef.de>
Message-ID: <sdmi9.114692$ub2.2440760@news1.tin.it>

Henrik Motakef wrote:
        ...
>> Python 2.2.1 (#1, Jun 25 2002, 10:55:46)
        ...
> There's a difference between a type and a class, I guess...

In the release of Python you use, the default metaclass for
your classes is still such as to force a difference, yes.

Just use one of the many ways to make your classes new-style
(e.g., a global variable __metaclass__=type will do) if you
are bothered by this distinction (gradually going away, but
with care for backwards compatibility, of course).


Alex




From tjreedy at udel.edu  Fri Sep 27 08:48:55 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 27 Sep 2002 12:48:55 GMT
Subject: a python puzzle
References: <Pine.LNX.4.44.0209251736470.16806-100000@hydrogen.sabren.com> <mailman.1032991473.7292.python-list@python.org> <an1i31$1r3$1@news.netpower.no>
Message-ID: <XsYk9.318161$5r1.14075173@bin5.nnrp.aus1.giganews.com>

"David Brown" <david at no.westcontrol.spam.com> wrote in message
news:an1i31$1r3$1 at news.netpower.no...
> I'm not too interested in the crypogram puzzle, but your solution
here has
> just given me a solution to something that has been bugging me for
some
> time - finding an easy way to convert a list of characters into a
string:
>
> >>> a = "Hello"
> >>> b = [x for x in a]

or
>>> list(a)
['H', 'e', 'l', 'l', 'o']
which is probably faster

> >>> "".join(b)
> 'Hello'




From jknapka at earthlink.net  Sat Sep  7 01:53:55 2002
From: jknapka at earthlink.net (Joseph A. Knapka)
Date: Sat, 07 Sep 2002 05:53:55 GMT
Subject: Python class ( Prolog ) ???
References: <20020905234600.11264.60328.Mailman@mail.python.org> <3D7914FD.5D713A48@anansispaceworks.com> <mailman.1031353570.17391.python-list@python.org>
Message-ID: <3D799469.EB2E47DC@earthlink.net>

Terry Hancock wrote:
> 
> Well, I haven't seen any posts go by on the python-logic
> SIG, so I thought I'd carry over this topic from the
> python list as a conversation starter:
> 
> While we're on the subject -- I've taken a couple of
> tries at Prolog in the past, but didn't really get
> anywhere with it.  I understand the concept of
> expressing logic in code, but got rather bogged down
> when it came to any practical examples.

When done properly, Prolog is beautiful, easy to understand,
easy to maintain. It is not well-applicable to a wide
range of problems, as is Python; largely because
procedural algorithms, which are the meat-and-potatoes
of mainstream IT, are rather painful to express in
Prolog. But if you happen to have a problem whose
structure is easily expressed in first-order predicate
logic, you can often simply write down that problem
description in your editor and run it as Prolog code.
Many problems to which we typically apply procedural
algorithms are easily expressed in FOPL; we don't
do it simply because efficient procedural algorithms
have been around for far longer than practical
declarative-logic languages.

> Python, with it's object/procedural focus seems like
> it would complement Prolog well (or vice-versa). Can
> anyone fluent in both compare what sort of applications
> you might apply a prolog-python solution to?  Where
> would you draw the line between the two? What would
> using prolog help me with? What's a good problem to
> solve as a learning project?

I would apply Prolog to a problem that is naturally
expressed as a search of a solution space for
solutions that satisfy some set of constraints. Of
course, any problem -can- be expressed in such
fashion; the question is, is it a -natural-
description. For trivial examples of the kind
of problem I mean, see any book of logic
puzzles of the "The man who wears blue lives in
the house closest to the river; the man with the
parrot lives next to the man who wears green"
variety. In this sort of problem, it's not obvious
how I would efficiently find an arrangement of
houses and owners that satisfies all the given
constraints, short of exhaustive generate-and-test;
but I can easily write Prolog rules expressing the
constraints themselves, and let the interpreter
perform the search for me. If I express the
constraints properly, the result will be as
fast as anything I could code up by hand in
Python or perhaps even C.

Prolog is (IMO) not particularly suitable for
applications that involve lots of user interaction
(though this may be because the Prolog UI toolsets
I've used have generally been thin wrappers around
underlying procedural toolkits - declarative UI
design could be quite powerful, as in XUL, etc).
The UI is probably best expressed using something
like Python. It's also not a good choice for
a problem that is mainly a long grindy numeric
computation - there are better languages to
express such things. When well-known, fast procedural
algorithms are applicable, apply them. Prolog
is good when you know a lot about the syntax of the
solution space, so to speak, but not a great deal
about its semantics -- "I ain't never seen one
before, but I'll know it when I see it!"

Incidentally, there's a Prolog interpreter buried
within the WinNT kernel; it handles configuration
of network devices at boot time (among possibly
other things).

> Any more book or website
> recommendations for a learner?
> 
> Conversely is there anyone who thinks Prolog is
> a waste of time for a Python programmer, or that
> it would be better to use something else in such
> a role?

I think that any programmer would benefit from learning
Prolog well (and Lisp, and Haskell, and assembler, and C).
It will probably be quite a drastic change if all you
know is OO and structured programming, and when you
"get it" you will know things about programming that
you didn't before. (Disclaimer: Prolog was the second
language I learned really well, after BASIC :-) so my
experience with it is not typical. When I learned C
later on, *that* made my brain hurt.)

Cheers,

-- Joe
  "I'd rather chew my leg off than maintain Java code, which
   sucks, 'cause I have a lot of Java code to maintain and
   the leg surgery is starting to get expensive." - Me


From 53ab2750 at meowing.net  Sat Sep  7 11:42:36 2002
From: 53ab2750 at meowing.net (itsy bitsy meowbot)
Date: Sat, 07 Sep 2002 15:42:36 GMT
Subject: None or 0
References: <k7ip4-0p3.ln1@junker.stroeder.com>
Message-ID: <M7pe9.1061$8u.746@nwrddc04.gnilink.net>

Michael Str?der wrote:
> >>> repr(0 or None)
> 'None'
> >>> repr(None or 0)
> '0'
> >>> repr('' or None)
> 'None'
> >>> repr(None or '')
> "''"
> >>>
> 
> Is it guaranteed to work like this or should that be avoided?

That's what the language reference specifies, in Doc/ref/lambda.html --
but each of the 73 Python license layers disclaims responsibiilty for
anything, so Pythonlabs are still free to change it if they're feeling
particularly mean and nasty, so don't tick them off.


From whisper at oz.net  Tue Sep 17 22:08:12 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 17 Sep 2002 19:08:12 -0700
Subject: Python GUI app to impress the boss?
In-Reply-To: <E17rTMz-0006XA-00@mail.python.org>
Message-ID: <GCEDKONBLEFPPADDJCOEAEKBFAAA.whisper@oz.net>

<snip>
> Is there any way to get the Python world to allow for the processing of
> monetary applications without some hokey 60's constructs that involve
> in-line conversions or anticipations of calculation characteristics?
>
> Regards,
>
> terry

What prevents the creation of a decimal money type? I agree, it would be
nice if python could do something akin to calculator aritmetic with types
like scientific (covered) engineering (improper scientific with the exponent
as a power of 3 and the (oops, forgot the term - mantissa?) more then a
single digit as in 12.2e3), fixed precision (1..n to the right of the
decimal). The forthcoming fractions are going to be interesting... (chinese
interpretation).

Be nice too if math could do rads, grads and decimal degrees.

Just my $0.02 worth.

Dave LeBlanc
Seattle, WA USA




From duncan-news at grisby.org  Wed Sep 11 05:16:24 2002
From: duncan-news at grisby.org (Duncan Grisby)
Date: Wed, 11 Sep 2002 09:16:24 GMT
Subject: Returning lists from python COM servers
References: <3d7731f5$1@pull.gecm.com> <3d7738f5$1@pull.gecm.com> <qFce9.1228$_l1.2030@news-server.bigpond.net.au> <3d7c87ff@pull.gecm.com>
Message-ID: <IRDf9.158$hZ1.1067@news13-win.server.ntlworld.com>

In article <3d7c87ff at pull.gecm.com>,
 Olly Smith <oliver.smith at removethisbit.baesystems.com> wrote:

>I'll remember to only chuck lists across the COM boundary.  Just out of
>interest, are there any other distributed object architectures that are
>better supported by Python? (corba, soap, ......)

CORBA is very well supported by Python. Check out omniORBpy or Fnorb:

  http://omniorb.sourceforge.net/
  http://www.fnorb.org/

I wrote omniORBpy, so that is the best one, of course :-)

SOAP is a bit of a mess (not just in Python -- everywhere) since the
specification has been changing too much, and nobody seems to agree on
its interpretation. It's also not strictly speaking a distributed
object architecture, since it doesn't have objects in the same way COM
and CORBA do.

Finally, you might want to consider Dopy and Pyro, which are Python
specific remote object systems. Personally I don't like being locked
into a single language (even if it's a language as nice as Python),
but they might work for you.

Cheers,

Duncan.

-- 
 -- Duncan Grisby         --
  -- duncan at grisby.org     --
   -- http://www.grisby.org --


From tjreedy at udel.edu  Mon Sep 23 21:04:46 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 24 Sep 2002 01:04:46 GMT
Subject: list problem
References: <mailman.1032818129.8126.python-list@python.org>
Message-ID: <OSOj9.250161$5r1.10250490@bin5.nnrp.aus1.giganews.com>

<jubafre at brturbo.com> wrote in message
news:mailman.1032818129.8126.python-list at python.org...
> i have:
> x=[['0020', '0000x9'], ['0030', '0000xa'], ['00B4', '0000x8'],
['0030', '0000xb'], ['00F0'], ['0001'], ['0003'], ['0005']]
>
> and i want:
> y=['0020', '0000x9', '0030', '0000xa', '00B4', '0000x8', '0030',
'0000xb','00F0', '0001',
> '0003', '0005']
>
> how to do this?
>>> [b for a in  x for b in a]
['0020', '0000x9', '0030', '0000xa', '00B4', '0000x8', '0030',
'0000xb', '00F0',
 '0001', '0003', '0005']

For deeper flattening, try google: Python list flatten

Terry J. Reedy




From wlfraed at ix.netcom.com  Tue Sep  3 03:40:08 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Tue, 03 Sep 2002 00:40:08 -0700
Subject: syntax error in Pyton shell
References: <e895ad50.0209022304.44c32987@posting.google.com>
Message-ID: <p0p1la.th5.ln@ix.netcom.com>

lion fed this fish to the penguins on Tuesday 03 September 2002 12:04 
am:

> I'm learning exception handling, I type these code in Python shell:
> 
>>>> try: f=open("nofile")
> ... except IOError: pass
> ... print "The exception is just passed!"
>   File "<stdin>", line 3
>     print "The exception is just passed!"
>         ^
> SyntaxError: invalid syntax
> 
> But these code run without any error as a program(or script). Could
> anyone tell me why?

        Off hand... Your indentation is invalid... You need to return to the 
>>> level for the print

-- 
--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From mgerrans at mindspring.com  Thu Sep 26 01:52:17 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Wed, 25 Sep 2002 22:52:17 -0700
Subject: "which" command?
References: <mailman.1032998974.31888.python-list@python.org>
Message-ID: <amu7ba$rjt$1@slb6.atl.mindspring.net>

"Brian Quinlan" <brian at sweetapp.com> wrote in message
news:mailman.1032998974.31888.python-list at python.org...
> The actual software is available here:
> http://starship.python.net/~tmick/

This one is nice, especially on the Windows platform where it looks at both
the path and pathext.

I wrote my own which.py a while ago to support these, but that was before I
heard of this one (i.e. just now).   I also made mine find all occurrances
of matching programs in the path, not just the first one.




From skip at pobox.com  Sun Sep  8 16:27:56 2002
From: skip at pobox.com (Skip Montanaro)
Date: Sun, 8 Sep 2002 15:27:56 -0500
Subject: transform a list of lists in a lit of strtings???
In-Reply-To: <algb9t$1q0jg0$1@ID-71831.news.dfncis.de>
References: <mailman.1031507832.24527.python-list@python.org>
        <algah9$1p0kp3$1@ID-71831.news.dfncis.de>
        <algb9t$1q0jg0$1@ID-71831.news.dfncis.de>
Message-ID: <15739.45772.400520.334648@12-248-11-90.client.attbi.com>

    James> labels2 = [i for item in labels1 for i in item]

    James> (I find this oddly hard to read.)

I rarely use list comprehensions for anything complex, but for this sort of
thing I tend indent it:

    labels2 = [i for item in labels1
                     for i in item]

Which is to say, if it gets too complex, I more-or-less just fall back to
the statement form:

    label2 = []
    for item in labels1:
        for i in item:
            labels2.append(i)

:-)

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From phncontact.libero.it at   Tue Sep  3 11:49:44 2002
From: phncontact.libero.it at  (Ben Jamin)
Date: Tue, 03 Sep 2002 15:49:44 GMT
Subject: Servizi Internet
References: <mailman.1031056925.7383.python-list@python.org>
Message-ID: <3bl9nu8924gihfj2nlpog874vfm73qqbh9@4ax.com>

excuse me but aren't commercial promotion and such things
forbidden on newsgoups?


From max at alcyone.com  Thu Sep  5 03:01:42 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 05 Sep 2002 00:01:42 -0700
Subject: Graham's spam filter
References: <mailman.1029992314.27908.python-list@python.org> <3D647B4B.28D9BF04@alcyone.com> <mailman.1030050253.11330.python-list@python.org> <3D655AE6.A097AA25@alcyone.com> <ak44j1$1f2ft4$2@ID-125932.news.dfncis.de> <c2ef5940.0209041912.16e102fb@posting.google.com>
Message-ID: <3D770156.2E55864C@alcyone.com>

Aaron Swartz wrote:

> I've been using bogofilter[1], Eric Raymond's Graham-derived spam
> filter which threw away base64-encoded data and 90% of all spam that
> got past the filter was base64-encoded. Therefore, I think that base64
> content really needs to be decoded. I wrote a base64-decoding filter
> in Python for it and the problem has gone away.

Indeed.  I've been finding very much the same thing with my rule-based
filter; about 90% of the spam that's getting through is base64 encoded. 
I haven't yet taken the next step of automatically decoding the base64
text parts (and then just processing that), but as you have discovered
it is an obvious solution to the obvious problem.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ There is nothing so subject to the inconstancy of fortune as war.
\__/ Miguel de Cervantes
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From sludtke at bcm.tmc.edu  Mon Sep 30 15:41:23 2002
From: sludtke at bcm.tmc.edu (Steve Ludtke)
Date: Mon, 30 Sep 2002 14:41:23 -0500
Subject: XMLRPC calls don't work in threads, or rather, they don't play
 well  with others ! (help)
References: <3D97D190.9030003@bcm.tmc.edu>
Message-ID: <3D98A8E3.7020006@bcm.tmc.edu>

Ok. Sorry to waste everyone's time. This seems to be a machine-specific 
problem. I put together a test case to demonstrate it, and it worked (on 
a different machine). On the original machine (running and identical 
OS/python version) the problem still persists, so I'll have to figure 
out what's wrong with that specific machine.  Very odd. Thanks everyone.



From bdesth at nospam.free.fr  Sun Sep  8 23:11:48 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Sun, 08 Sep 2002 23:11:48 -0400
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com> <4f49d232.0209050418.7668c21@posting.google.com> <al7ub8$crj$1@panix2.panix.com> <3D77FB95.3070300@nospam.free.fr> <al8amr$d3q$1@panix2.panix.com>
Message-ID: <3D7C1174.5060905@nospam.free.fr>

Roy Smith wrote:
> I wrote:
> 
>>>If you're learning languages to study computer
>>>science, lisp is looking a lot better.
>>
> 
> laotseu  <bdesth at nospam.free.fr> wrote:
> 
>>I think once you know them both, you know *how* to program, whatever
>>the language you have to use.
> 
> 
> I believe we actually both said the same thing, just worded
> differently :-)

Sorry, my misunderstanding...

laotseu



From sismex01 at hebmex.com  Thu Sep 12 18:31:57 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Thu, 12 Sep 2002 17:31:57 -0500
Subject: [ANN] PySQLite 0.3.0 released
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233024251DB@mail.hebmex.com>

Oops, you are SO right. :-)

-gus
--

> 
> >> for row in cursor:
> >>        print "%14s, %15s, %19s, %8s, %25s" % tuple(row)
> >> ?
> >> 
> >> Terry J. Reedy
> >> 
> 
> > You can already do it, using iter():
> 
> > for row in iter(cursor, None):
> > 	print ....
> 
> you mean:
> 
> for row in iter(cursor.fetchone, None):
> 	print ...
> 	
> 
> 
> -- 
> ??????
> 
> 	In a world without walls and fences,
> 	do we care about Windows and Gates ?
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 

Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compania. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electronico enviado para o desde esta direccion sera
procesado por el sistema de correo corporativo de HEB. Tal correo
electronico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el proposito de monitorear que se cumplan
las normas de seguridad de la empresa.



From erict at millfilm.co.uk  Tue Sep  3 10:47:07 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Tue, 03 Sep 2002 15:47:07 +0100
Subject: xml question
Message-ID: <3D74CB6B.7697013E@millfilm.co.uk>

I am trying to compare 2 node from 2 different xml dom.
On it just read from disk and the other one has been
generated from memory: I am finding myself having to compare
the same thing in 2 differents state:
>From Disk: (unicode)
N1 [(None, u'fileName'), (None, u'OutputType'), (None, u'name'), (None,
u'UId'), (None, u'version'), (None, u'node')]
>From Memory:
N2 ['node', 'version', 'UId', 'fileName', 'OutputType', 'name']

What will be the simplest thing to do in this case?
Thanks,
Eric







From gerhard.haering at gmx.de  Sun Sep  1 17:44:38 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sun, 1 Sep 2002 23:44:38 +0200
Subject: Add methods to a class at runtime?
In-Reply-To: <wmvc9.21757$FJ1.3181934@e3500-atl1.usenetserver.com>
References: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com> <slrnan4dl3.18h.gerhard.haering@lilith.my-fqdn.de> <wmvc9.21757$FJ1.3181934@e3500-atl1.usenetserver.com>
Message-ID: <20020901214438.GA4349@lilith.ghaering.test>

* Robert Oschler <Oschler at earthlink.net> [2002-09-01 17:19 -0400]:
> Ok bear with me guys (Gerhard and Peter).  I'm not lazy, the reason I
> don't test certain things I should is because Python is capable of
> such amazing introspective dynamism that, having a _very_ long
> experience bank in strongly typed "static" languages like C++ and
> Pascal, I find I don't even consider yet even trying certain
> possibilies that Python is capable of.
> Some interesting tidbits I found while trying out your suggestions, as
> you will see in the two ".py" files that follow, I discovered that to
> create a method you can dynamically add to a class _instance_, you
> need to omit the 'self' parameter in the method declaration.

Yep, I found that out today too, when I wrote my last reply. Though I've
never used that in real life.

> [...] I'm running 2.1.1, anybody know if 2.2.2 is going "disable"
> either of these abilities?

It will be strictly (even binary) compatible with Python 2.2.{0|1}, so
the answer is that these features will stay.

> (class or class instance dynamic method addition).  I saw a note about
> such activities requiring a "dynamic" keyword, which is fine.

I never saw such a note. Maybe you were reading a message from the
(dormant) types-sig, where ways to introduce optional (!) stronger
typing into Python are discussed. I never saw Guido make any comments
about removing dynamic features from Python, though. With /one/
exception, which you didn't mention, yet: assigning to the __class__
attribute:

>>> class Foo:
...   def action(self): print "foo"
... 
>>> class Bar:
...   def action(self): print "bar"
... 
>>> foo = Foo()
>>> foo.__class__
<class __main__.Foo at 0x80d39ac>
>>> foo.action()
foo
>>> foo.__class__ = Bar
>>> foo.action()
bar
>>> foo.__class__
<class __main__.Bar at 0x80e75c4>
>>> 

AFAIK Guido isn't too fond of this feature. I guess it might go away
some day.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))



From jfaris at sbcglobal.net  Mon Sep  9 19:08:50 2002
From: jfaris at sbcglobal.net (Jordan Faris)
Date: Mon, 9 Sep 2002 16:08:50 -0700
Subject: Diving into the Python pool....
Message-ID: <005c01c25855$dc5752d0$e680cf3f@leader1>

Hi...after strolling through these threads for some while now, and
comparing the generally benign tone here with the rather elitist edge of
other programming communities, I think I'd like to get my feet firmly onto
Terra Pythonia.
I'm a little reluctant to plow in headfirst, however, mainly because I had
my heart set on becoming a Java programmer, but I think I need to be more
realistic about programming fundamentals, since I am doing this completely
self-study.
Python is, if I understand correctly, an object-oriented language with a
much simpler syntax than Java, yet compatible in applicability, especially
web-related, is this correct?
My primary focus is to create streamlined, rapid-deployment web applications
(PHP and JSP portal sites, etc), but from what I've seen, Python can be used
or adapted to achieve comparable results; am I on the right track, or should
I continue to nurture the Java bean?
Any frank, level-headed input would be much appreciated.
Thanks......
Jordan






-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020909/57b5ec09/attachment.html>

From dwelch91 at nospam.attbi.com  Sat Sep 21 13:08:17 2002
From: dwelch91 at nospam.attbi.com (djw)
Date: Sat, 21 Sep 2002 17:08:17 GMT
Subject: PyQt setup on Linux Mandrake 8.2
References: <3D8B4898.20808@nospam.attbi.com>	<ddc19db7.0209210259.7beb52d4@posting.google.com> <3d8c9b92$0$182$9b622d9e@news.freenet.de>
Message-ID: <3D8CA781.3010000@nospam.attbi.com>

<snip>
> Unfortunately, Mandrake is somewhat more perl than python biased.
> 

Do you mean there is more support/newer versions of Perl or is it more 
of a philosophical thing: There are many ways vs. there is one way?

<snip>
> If you know a bit about RPM, perhaps you, dwelch, can manage
> to update these packages (rpm -Uhv xx.src.rpm, a few changes ..., 
> rpm -ba SPECS/sip.spec).
> Otherwise you could contact me. In about a week I've got time
> to do this.

Uhhh... considering my extreme novice status on Linux and especially 
RPMs (as witnessed by my problems), I don't think I am the right person 
for this task.

<snip>
> The RPM  stuff (or DEB) has got real advantages compared to 
> "setup.exe". 
> And if the offered versions are outdated, fetch the
> new source, learn from the (e.g.) sip.spec file somewhat about
> the compilation, and make a new package!
> 

Well, here I have to disagree (at least from a user's perspective). 
Having come from over 10yrs of Windows development experience (including 
writing Windows installers from scratch), I find RPM package management 
(can't say about DEB format, never used it) to be very complex, 
difficult, and it rarely works for me.

Case in point is two systems I tried to install at work... one Mandrake 
8.x and one Windows 2000. The packages I was installing were Python 
2.2.1 with a bunch of extensions like MySQLdb and mxDateTime, MySQL 4.0, 
Jabber x.xx (can't remember version) and an FTP daemon. The Windows box 
was up and running in less than an hour after running a handful of 
setup.exe-type programs, and the Linux box still wasn't fully installed 
after two days of trying. I completely gave up trying to upgrade a copy 
of MySQL 3.23 to 4.0 after every way I tried failed. In the end, the 
system became corrupted that I had to re-install Linux (probably a real 
Linux person would laugh at this, but after a while /usr/bin got so 
filled with junk that I didn't know what I had anymore and wanted to 
start over "clean").

Not to say that this sort of thing doesn't happen in Windows-land, but 
setup.exe works on 99% of the packages I have tried 99% of the time. 
Don't get me wrong, I really like Linux, but this is one area where I 
believe that Linux still has a ways to go to "catch up" in the 
ease-of-use department.

Don




From wlfraed at ix.netcom.com  Fri Sep 20 18:18:47 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Fri, 20 Sep 2002 15:18:47 -0700
Subject: Python Compiling
References: <amfveg$5g64k$1@ID-53055.news.dfncis.de>
Message-ID: <8s6gma.ti3.ln@ix.netcom.com>

Terje Johan Abrahamsen fed this fish to the penguins on Friday 20 
September 2002 01:12 pm:

> But, I guess my question boils down to, why is not Python a compiled
> language?
>
        Firstly, Python started life as a scripting language -- like REXX, 
DCL, PERL... As such, it was meant to avoid such intervening confusions 
as compile/link phases.

        Besides, there is a virtual machine intermediate -- modules that are 
mentioned in an "import" statement are compiled to the VM byte-codes 
(creating a <module>.pyc file) so subsequent invocations save time.

> Is it not possible to write a complete compiler like the C compiler
> for Python? Or is it just that nobody has done it? Or must the
> language be constructed differently?
>
        Do you know of a C compiler that builds executables that can read a C 
expression as input and compile/execute that expression? Most scripting 
languages can do that, as the interpreter is already present. 


-- 
--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From christophe.delord at free.fr  Tue Sep  3 16:09:44 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Tue, 3 Sep 2002 22:09:44 +0200
Subject: efficient list merging
References: <ced73313.0209031012.13624f3@posting.google.com>
	<mailman.1031080984.2275.python-list@python.org>
Message-ID: <20020903220944.475d7222.christophe.delord@free.fr>

On Tue, 03 Sep 2002 15:19:44 -0400
pinard at iro.umontreal.ca (Fran?ois Pinard) wrote:

> [Peter Saffrey]
> 
> > I have two lists of lists of strings which I would like to merge
> > efficiently without repeats.  [...] Or is there a better way?
> 
> I am not sure about the relative speed of everything, but I would be
> tempted to try:
> 
>    dict(zip(list1 + list2, [None] * (len(list1) + len(list2)))).keys()

This may be written using list comprehension:

>>> list1 = list('abcdebxyz')
>>> list2 = list('pqqr')
>>> dict([(k,1) for k in list1+list2]).keys()
['a', 'c', 'b', 'e', 'd', 'q', 'p', 'r', 'y', 'x', 'z']





-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From roy at panix.com  Mon Sep 30 07:32:59 2002
From: roy at panix.com (Roy Smith)
Date: Mon, 30 Sep 2002 07:32:59 -0400
Subject: is cPickle really this slow?
References: <slrnapfs9u.8re.dag@omally.animagicnet.no>
Message-ID: <roy-FD4DA7.07325930092002@reader1.panix.com>

Dag <dag at animagicnet.no> wrote:
> So I pickle my dictionary, which ends up
> being a 3.2MB file and try to run my script again, this time getting 
> the pickled dictonary instead of building it at runtime.  This time
> however the script takes over 35 seconds to run, close to three times
> as long.

Did you use the optional third argument which tells pickle to write the 
file out in binary?  My experience is that the difference in performance 
between reading binary and ascii pickles will be substantial.


From marklists at mceahern.com  Tue Sep 10 08:27:56 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 10 Sep 2002 07:27:56 -0500
Subject: dynamism
In-Reply-To: <Xns92855D48384F9duncanrcpcouk@127.0.0.1>
Message-ID: <JHEOKEOOLIGLDHCMAHMOCEJNDNAA.marklists@mceahern.com>

[Duncan Booth]
> New style classes (which are those with 'object' or other builtin classes
> as a base class) do not all allow you to add arbitrary attributes:

Au contraire:

$ python
Python 2.2.1 (#1, Jun 25 2002, 10:55:46)
[GCC 2.95.3-5 (cygwin special)] on cygwin
Type "help", "copyright", "credits" or "license" for more information.
>>> class foo(object):pass
...
>>> f = foo()
>>> f.x = 1
>>> f.x
1
>>>

Cheers,

// m

-




From senux at senux.com  Wed Sep 18 04:16:32 2002
From: senux at senux.com (Brian Lee)
Date: Wed, 18 Sep 2002 17:16:32 +0900
Subject: python does not start
Message-ID: <20020918081632.GA4156@mercury.senux.com>

$ ./python
ld.so.1: ./python: fatal: libdb-3.2.so: open failed: No such file or directory
Killed

On SunOS 5.7, python which is compiled on other sun machine does not start.
How can I fix this?

-- 
Brian Lee <senux at senux.com>



From gerhard.haering at gmx.de  Tue Sep 24 13:46:30 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Tue, 24 Sep 2002 19:46:30 +0200
Subject: Python and cgi and <a> problem!
In-Reply-To: <amp4se$ik4$1@lust.ihug.co.nz>
References: <ammn6e$lj7$1@lust.ihug.co.nz> <Xns9292EA042C259cliechtigmxnet@62.2.16.82> <slrnaov5q3.nrp.jon+usenet@snowy.squish.net> <Xns929330F5A61B0cliechtigmxnet@62.2.16.82> <slrnaovmel.dma.jon+usenet@snowy.squish.net> <amp4se$ik4$1@lust.ihug.co.nz>
Message-ID: <20020924174629.GA787@lilith.ghaering.test>

* CheapSkate <gua81 at XXXyahoo.com> [2002-09-24 17:40 +1000]:
> thanks dudes,
> everything works fine,
> im just doing some simple things, nothing that will break the code.

I'd recommend to use urllib.quote() anyway. It's better to do the right
thing in the first place, as software often lives longer and in stranger
places than you'd have expected it to.

-- Gerhard



From thorsten at thorstenkampe.de  Tue Sep 17 14:01:40 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Tue, 17 Sep 2002 20:01:40 +0200
Subject: IDLE
References: <wpbf9.264320$aA.45691@sccrnsc02> <3D7D4F84.318CDB0F@bellatlantic.net>
 <v4Pg9.428631$me6.49396@sccrnsc01> <bs70f7ks.fsf@online.no> <am1efh$1ohti$1@ID-77524.news.dfncis.de>
 <EOjh9.2033$Cf.81@fe03> <am557m$2nldt$2@ID-77524.news.dfncis.de> <mailman.1032205103.25956.python-list@python.org>
Message-ID: <am7qoi$3emja$1@ID-77524.news.dfncis.de>

* Cliff Wells
> On Mon, 2002-09-16 at 10:43, Thorsten Kampe wrote:
>> "Hardware: 386 enhanced, 75MHz, 32MB RAM."
>
> This can't be correct. [...]
>
> The actual CPU is either an AMD/Cyrix 486 clone or an early Pentium.

You know hardware better than me, for sure. But even with a 486/586 
running with 75 MHz he surely will have problems installing and 
running Python 2.2.1: the python shell for example is about 4 MB 
running idle, Pythonwin is about 9 MB.

Thorsten 


From b.maryniuk at forbis.lt  Mon Sep  2 04:08:00 2002
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Mon, 2 Sep 2002 10:08:00 +0200
Subject: Python GUI
In-Reply-To: <akv3b8$7dv$1@usenet.Stanford.EDU>
References: <3d8d3c0.0208262339.e7bc02c@posting.google.com> <r794nugam0su9metpaas0nobn4jsm84j3j@4ax.com> <akv3b8$7dv$1@usenet.Stanford.EDU>
Message-ID: <200209021008.00596.b.maryniuk@forbis.lt>

On Monday 02 September 2002 09:16, Andy Salnikov wrote:
> Problem with X11 is that there is no "OS style manager" (whatever
> this could mean) for X11, or there is no _single_ style manager.

No. Problem with the license. Use wxWindows instead.

-- 
Regards, Bogdan

main(k){float i,j,r,x,y=-16;while(puts(""),y++<15)for(x
=0;x++<84;putchar(" .:-;!/>)|&IH%*#"[k&15]))for(i=k=r=0;
j=r*r-i*i-2+x/25,i=2*r*i+y/10,j*j+i*i<11&&k++<111;r=j);}
/* Mandelbrot in ASCII. */




From teaandbikkie at aol.com  Wed Sep 11 21:01:59 2002
From: teaandbikkie at aol.com (TeaAndBikkie)
Date: 12 Sep 2002 01:01:59 GMT
Subject: file upload using msvcrt
References: <20020911204906.00669.00000156@mb-fh.aol.com>
Message-ID: <20020911210159.00669.00000162@mb-fh.aol.com>

>data = fileinfo.file.read( )
>if not data: break

I just thought of another possible problem here, read() is supposed to read in
the entire file, and maybe 8k is all data can take :)

So you could change it to read(1024) for 1k chunks.

Also, maybe read() recognises an EOF character in the word doc? Hopefully being
binary read this should not happen...

Kind regards,


From pedronis at bluewin.ch  Wed Sep  4 12:31:40 2002
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Wed, 4 Sep 2002 18:31:40 +0200
Subject: jython memory usage
References: <hVm5ydAY2hd9Eww$@jessikat.demon.co.uk>
Message-ID: <3d76366a$1_5@news.bluewin.ch>

From: Robin Becker <robin at jessikat.fsnet.co.uk>
> A colleague has been running tests of reportlab under jython-2.1 several
> of the tests use prodigious amounts of memory (one even failed). We're
> using win2k+j2sdk1.4.0_01 as the jython engine.
>
> Anyone know what sorts of things cause x10 memory usage? I find one
> particular test needs more that 1Gb and then falls over for lack of
> memory.

That means you are running jyhton through:

java -mx1024m  -cp ...jython.jar org.python.util.jython ...

and it uses all that memory and would need more?

regards.




From dwelch91 at nospam.attbi.com  Fri Sep 20 01:46:31 2002
From: dwelch91 at nospam.attbi.com (dwelch)
Date: Fri, 20 Sep 2002 05:46:31 GMT
Subject: newbie seeks advise.
References: <77c84db2.0209192137.51167ebb@posting.google.com>
Message-ID: <3D89F605.4030303@nospam.attbi.com>

akirasugiura wrote:
> Hi,
> 
> I've been learning python for about a month. I had no prior knowledge
> of programming.
> 
> My question is, I have a tuple which is, data = (69, 40, 74, 41, 56,
> 57, 38, 58, 26, 55). How I can make a function that picks up a largest
> number
>  from this "data" tuple. If "data" was a  'List' I would be able to
> use "sort" but since sort doesn't  work with tuple. I am stuck here.
> 
> Thanks in advance
> 
> Sincerely,
> 

Simple way, probably not optimal:

 >>> max_element = max(list(data))
 >>> max_element
74



From bokr at oz.net  Mon Sep  9 21:10:08 2002
From: bokr at oz.net (Bengt Richter)
Date: 10 Sep 2002 01:10:08 GMT
Subject: Windows/DOS: double clicking a .py file
References: <mailman.1031613313.24086.python-list@python.org>
Message-ID: <aljgpg$1vs$0@216.39.172.122>

On Mon, 9 Sep 2002 16:17:08 -0700, "David LeBlanc" <whisper at oz.net> wrote:

>cmd.exe will accept a flag telling it to stay open after something is run.
>You can create a shortcut along the lines of:
>
>cmd /K python myscript.py <args>
You may want to specify full paths, depending on where things are etc.

You can also do _several_ things in the cmd /k line by using "&". E.g., the following

%SystemRoot%\system32\cmd.exe /x /k prompt [$T$H$H$H$H$H$H] $P$G & title MyPy & D:\Python22\python.exe

will
  - start a new cmd window
  - set the prompt for it so it will show current time and directory, e.g.,
    [17:19] C:\pywk>
  - set the title of the window to MyPy
  - start interactive python
the shortcut also lets you set the working directory. If you have a place you like
to test code, note that by setting the working directory there, everything will be
visible to import, aside from any other startup initialization. Thus you can separate
things by people or project or whatever. When you exit Python, you'll be at a cmd
prompt as specified, and in the working directory specified. Handy.

Don't forget that you can also change the icon associated with whatever you run via
the shortcut. I use the swiss army knife for many of my various "DOS" console shorcuts.
>
>You could also add this to the cmd shortcut itself if you have one, but that
>generally turns out to be a major pain since all dos-window based apps will
>then leave their windows open.
You mean the file extension association for .bat and .cmd? Yes, that doesn't work out
as a good general thing.
>
>A nice aspect of using a custom cmd shortcut is that you can set colors and
>fonts and window size just for this app which might help distinguish it if
>you have a lot of dos windows open. You can set the screen buffer height to
>a high number so you can scroll really far back (not a good idea to make the
>screen buffer width greater then the window width though).
>
A nice thing is that you can change things just for your current session,
using the properties selection from the system (left click title bar icon for it) menu.

>you can type 'help cmd' in a dos window for more info on cmd switches.
>
>HTH,
>
>David LeBlanc
>Seattle, WA USA
>
>> -----Original Message-----
[...] ;-)

Regards,
Bengt Richter


From phr-n2002b at NOSPAMnightsong.com  Sat Sep 14 02:20:47 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 13 Sep 2002 23:20:47 -0700
Subject: EDI parsing
References: <746127ef.0209110128.53665a23@posting.google.com> <mailman.1031755813.4692.python-list@python.org>
Message-ID: <7xd6rh6qio.fsf@ruckus.brouhaha.com>

"Emile van Sebille" <emile at fenx.com> writes:
> > 2) Is there a (python) EDI to XML converter?
> 
> That wouldn't be hard.  If there's already a DTD for edi, that would
> help a lot.  ;-)

EDI syntax doesn't fit into the SGML model, so I don't think it could
be done with an DTD-driven parser.  Parsing and converting EDI is easy.

> > [You have to know that I've never written a parser in Python before.
> >  The last (big) parser I wrote was in C with a few years back the help
> >  of lex and yacc if memory serves me right.]
> >
> > What is the best approach to writing an EDI parser in Python?

The main thing is to understand that it's not complicated.  I worked
on an EDI product years ago, and we hired several programmers with CS
backgrounds who immediately wanted to use stuff like lex and yacc on
it.  Really, EDI was designed to be processed by RPG and COBOL
programs and just doesn't have much syntactic hair.  If you find
yourself reaching for fancy compiler implementation techniques, you're
probably misunderstanding something.

Basically, an EDI document is a sequence of lines called "segments"
where the segments are sequences of "elements".  The spec describes
various document types (purchase order, invoice, etc.) and the types
of segments you expect to find in each one.  For each segment, the
spec describes which elements you expect to find in each one.  You
break a segment down into elements with a simple lexical scanner
(re.split should be more than powerful enough) and then have a
dictionary indexed by segment type of what elements should be in it.
Within documents, there are some simple looping constructs, that are
no big deal.

If you want to write a general EDI parser that understands all the
documets and segments in the spec, 99% of the work is typing in all
the rules from the spec.  The code that follows them should be very
simple.  In reality, nobody uses all the docs or segments.  They pick
a few docs and use some subset of the allowable segments in specific
ways.  So if you're trying to implement EDI because you have some
customer who wants to order widgets from you by EDI, you don't need a
fully general system.  Just ask them for the spec of the docs they
want to use, and code those.

What brought on this question?  I seem to remember this subject came
up before, maybe on sci.crypt.  I don't think there's any free general
EDI software around because the big specication documents (X12,
Edifact) aren't free.  But the code framework is pretty
straightforward to write.  I did it in C in a few weeks, and in Python
it would be a lot simpler.


From jepler at unpythonic.net  Sun Sep  1 17:30:52 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Sun, 1 Sep 2002 16:30:52 -0500
Subject: EOFError using cPickle.load
In-Reply-To: <6avc9.264595$me6.34329@sccrnsc01>
References: <6avc9.264595$me6.34329@sccrnsc01>
Message-ID: <20020901163044.A1401@unpythonic.net>

either do not use Windows, or make sure you open with 'rb' and 'wb'.

Jeff



From marklists at mceahern.com  Thu Sep 12 19:17:31 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 12 Sep 2002 18:17:31 -0500
Subject: Restricting methods in derived classes
In-Reply-To: <slrnao23m1.2ks.huaiyu@gauss.almadan.ibm.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOGEENDOAA.marklists@mceahern.com>

[Huaiyu Zhu]
> I was hoping for a way to confirm wanted methods, because there are more
> methods that I don't want.  But the above is good enough for preventing
> obvious mistakes.  Thank you.

Have you considered a metaclass solution?  Here's a half-baked attempt that
doesn't work and merely serves to demonstrate how weak my grasp of Python
is.  If Alex Martelli is in the house, perhaps he can shed some light.  :D

Here's some classic Martelli speak on metaclasses for ya:


http://groups.google.com/groups?selm=mailman.1025801286.31332.python-list%40
python.org

Meanwhile, my lame attempt...

#!/usr/bin/env python

import inspect

def make_unwanted(cls, method_name):
    def unwanted(self, *args, **kwargs):
        raise AttributeError(cls.__name__, method_name)
    return unwanted

class Restricted(type):

    def __init__(cls, name, bases, dict):
        for base in bases:
            for method in inspect.getmembers(base,
inspect.ismethoddescriptor):
                if not method[0].startswith('__') and method[1] not in
dict['wanted_methods']:
                    dict[method[0]] = make_unwanted(cls, method[0])
        super(Restricted, cls).__init__(name, bases, dict)

class RestrictedDict(dict):

    __metaclass__ = Restricted

    wanted_methods = [dict.update]

d = RestrictedDict()
d['a'] = 1

# I expect the following to blow up, but alas, they don't...
for key in d.keys():
    print key

d.update({'a': 2})

print d.items()

# Will this give me any clue?
print dir(d)

# Time to go listen to some good music...

// m




From igetenoughspamalreadythanksjlr at doctor.com  Tue Sep  3 11:33:33 2002
From: igetenoughspamalreadythanksjlr at doctor.com (John La Rooy)
Date: Wed, 04 Sep 2002 01:33:33 +1000
Subject: efficient list merging
References: <ced73313.0209031012.13624f3@posting.google.com>	<mailman.1031080984.2275.python-list@python.org>	<20020903220944.475d7222.christophe.delord@free.fr> <mailman.1031100605.23330.python-list@python.org>
Message-ID: <1031110376.483649@ampungk.ozonline.com.au>

Fran?ois Pinard wrote:

> Taking advantage of other clever suggestions, I would guess that:
> 
>     both = list1 + list2
>     result = dict(zip(both, both)).keys()
> 
> would not be so bad, speed-wise.
>

result=dict(zip(list1,list1))
result.update(dict(zip(list2,list2)))

John



From steven_shaw at adc.com  Wed Sep 11 20:33:36 2002
From: steven_shaw at adc.com (Steven Shaw)
Date: 11 Sep 2002 17:33:36 -0700
Subject: dynamism
References: <66715c02.0209091923.265fe77e@posting.google.com> <Xns92855D48384F9duncanrcpcouk@127.0.0.1> <3d7e209a_1@news.iprimus.com.au> <Xns92865CE1C916Aduncanrcpcouk@127.0.0.1> <Xns9286606D44C35duncanrcpcouk@127.0.0.1>
Message-ID: <66715c02.0209111633.771aa9d3@posting.google.com>

Duncan Booth <duncan at NOSPAMrcp.co.uk> wrote in message news:<Xns9286606D44C35duncanrcpcouk at 127.0.0.1>...
> Duncan Booth <duncan at NOSPAMrcp.co.uk> wrote in 
> news:Xns92865CE1C916Aduncanrcpcouk at 127.0.0.1:
> > The code below gives these timings on my machine:
> > 3.84s: 1000000 x New()
> > 3.89s: 1000000 x New2()
> > 2.39s: 1000000 x Old()
> > 5.19s: 10000 x attributes(<class '__main__.New'>)
> > 5.81s: 10000 x attributes(<class '__main__.New2'>)
> > 4.17s: 10000 x attributes(__main__.Old)
> 
> Sorry about the bad form of following up to my own post, but I thought I 
> had better add that the above timings are on Python 2.2.1 and I forgot to 
> specify -O.
> 
> The optimised times for 2.2.1 are:
> 3.77s: 1000000 x New()
> 3.81s: 1000000 x New2()
> 2.33s: 1000000 x Old()
> 4.94s: 10000 x attributes(<class '__main__.New'>)
> 5.56s: 10000 x attributes(<class '__main__.New2'>)
> 3.92s: 10000 x attributes(__main__.Old)
> 
>  Using the latest CVS build I get rather different timings:
> 
> 0.94s: 1000000 x New()
> 1.05s: 1000000 x New2()
> 1.39s: 1000000 x Old()
> 3.52s: 10000 x attributes(<class '__main__.New'>)
> 4.36s: 10000 x attributes(<class '__main__.New2'>)
> 3.05s: 10000 x attributes(__main__.Old)
> 
> So it appears that creating object-derived classes is now faster than 
> creating old-style classes, although both tests are much faster than in 
> 2.2. Setting attributes is still somewhat slower. Oh, and the -O flag on 
> the CVS build has no measurable effect.

Hi Duncan,

Thanks alot for your time on this. I hadn't read about the __slots__
mechanism before. I feel that there must be something wrong, though,
because the Old object layout is winning! Maybe the implementation is
finding the index of the variable in the __slots__ array and then
using returning __values__[index] in order to implement the attribute
access. I made up the "__values__" array - there must be a values
array somewhere. Array-based object layout amounts to the same thing
as record-based-layout in Python because all primitives are the same
size (right?). There needs to be a way for Python to remember the
indexes... Hmmm. I think your test code doesn't allow the compiler to
remember the indexes since it creates an instance anonymously (using
klass). I tried the code I "attached" at the end of the post:

$ python f.py
2.2.1 (#4, Aug 21 2002, 14:22:21) 
[GCC 3.2]
0.14s: 10000 x New()
0.12s: 10000 x New2()
0.09s: 10000 x Old()
13.03s: 10000 x attributes(<class '__main__.New'>)
14.04s: 10000 x attributes(<class '__main__.New2'>)
8.97s: 10000 x attributes(__main__.Old)
13.13s: 10000 x attributesNew()
14.12s: 10000 x attributesNew2()
8.32s: 10000 x attributesOld()

$ python -O f.py
2.2.1 (#4, Aug 21 2002, 14:22:21) 
[GCC 3.2]
0.14s: 10000 x New()
0.11s: 10000 x New2()
0.10s: 10000 x Old()
12.18s: 10000 x attributes(<class '__main__.New'>)
12.66s: 10000 x attributes(<class '__main__.New2'>)
7.94s: 10000 x attributes(__main__.Old)
12.26s: 10000 x attributesNew()
12.78s: 10000 x attributesNew2()
7.99s: 10000 x attributesOld()

It made no difference. I checked the bytecode and it is the same for
accessing a slot as for accessing a normal attribute so optimisations
aren't currently possible at that level...

BTW your machine seems very fast. What is it?

Cheers, Steve.

---------------
class Old: pass

class New(object):
    __slots__ = ['a', 'b', 'c', 'd']

# New-style, but not using slots.
class New2(object): pass

def timeit(repeat, fn, *args, **kw):
    import time
    start = time.time()
    for i in xrange(repeat):
        fn(*args, **kw)
    end = time.time()
    # Format arguments for printout
    argstring = ', '.join(
        [str(a) for a in args] +
        [ '%s=%s' % (k, v) for k, v in kw.items()])
    if len(argstring) > 60:
        argstring = argstring[:57] + '...'
    print "%.2fs: %d x %s(%s)" % \
          ((end-start), repeat, fn.__name__, argstring)

def attributes(klass):
    inst = klass()
    inst.a = 1
    for i in xrange(100):
        inst.b = inst.a+1
        inst.c = inst.a+2
        inst.d = inst.a+4
        inst.a += 5

def attributesNew():
    inst = New()
    inst.a = 1
    for i in xrange(100):
        inst.b = inst.a+1
        inst.c = inst.a+2
        inst.d = inst.a+4
        inst.a += 5

def attributesNew2():
    inst = New2()
    inst.a = 1
    for i in xrange(100):
        inst.b = inst.a+1
        inst.c = inst.a+2
        inst.d = inst.a+4
        inst.a += 5

def attributesOld():
    inst = Old()
    inst.a = 1
    for i in xrange(100):
        inst.b = inst.a+1
        inst.c = inst.a+2
        inst.d = inst.a+4
        inst.a += 5

def main():
    import sys
    print sys.version
    timeit(10000, New)
    timeit(10000, New2)
    timeit(10000, Old)
    timeit(10000, attributes, New)
    timeit(10000, attributes, New2)
    timeit(10000, attributes, Old)
    timeit(10000, attributesNew)
    timeit(10000, attributesNew2)
    timeit(10000, attributesOld)

if __name__ == "__main__": main()


From martin at v.loewis.de  Tue Sep 17 01:45:00 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 17 Sep 2002 07:45:00 +0200
Subject: pickle magically not working
References: <3d865dac@nntp.server.uni-frankfurt.de> <mailman.1032224577.11771.python-list@python.org>
Message-ID: <m3it15cgpv.fsf@mira.informatik.hu-berlin.de>

jepler at unpythonic.net writes:

> If I had to guess, I'd say the "fix" was in Pickle itself, you might turn
> something up by investigating Python's CVS tree.  Or somebody else might be
> able to immediately indicate the relevant code...

It's in structseq.c:structseq_reduce (rev. 1.4 and 1.5), also fixed in
Python 2.2.1.

Regards,
Martin


From vincent_a_primavera at netzero.net  Fri Sep 27 09:35:50 2002
From: vincent_a_primavera at netzero.net (Vincent A. Primavera)
Date: 27 Sep 2002 13:35:50 +0000
Subject: comp.lang.python...
In-Reply-To: <slrnap910v.5gd.gerhard.haering@lilith.my-fqdn.de>
References: <mailman.1033054480.24506.python-list@python.org>
	<an1ip7$rgs$00$1@news.t-online.com>
	<mailman.1033131395.24431.python-list@python.org> 
	<slrnap910v.5gd.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <1033133750.1570.22.camel@del-vap-bos.ralphpill.com>

Hello,
	Thank you very much!

	Vincent A. Primavera.

On Fri, 2002-09-27 at 16:15, Gerhard H?ring wrote:
> Vincent A. Primavera wrote in comp.lang.python:
> > Hello,
> > 	I've been subscribed to python-list at python.org for at least a couple of
> > years I would guess.  I have been using a mail client to receive all of
> > the messages, but I figure that using a news client would be a more
> > organized method.  However, I can't seem to find what I need to enter
> > for a server & user/password combination...
> 
> If your ISP doesn't offer an NNTP server, you'll probably have bad
> luck, unless you pay for third-party NNTP access somewhere.
> 
> However, there's a mail-to-news-gateway (and reverse) available at
> http://gmane.org/, which NNTP access for many Python mailing lists.
> The data you'll need are:
> 
> NNTP server: news.gmane.org
> group name: gmane.comp.python.general
> 
> No username or password required for reading. If you post, you'll need
> to answer a one-time confirmation message for every group.
> 
> I'd still prefer to use comp.lang.python diretly. But this is quite
> cool for all the SIG groups, as you can read them without actually
> subscribing.
> 
> -- Gerhard
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 






From max at alcyone.com  Thu Sep 12 16:16:27 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 12 Sep 2002 13:16:27 -0700
Subject: Coding Style: Quotes
References: <bt1g9.17429$Pf7.590136@news1.west.cox.net>
Message-ID: <3D80F61B.A254ED63@alcyone.com>

John Waycott wrote:

> I'm curious if others have adopted any standards for choice of single
> vs.
> double quotation marks to delimit strings. A look through the standard
> library reveals the choice of one over the other is rather arbitrary.
> I
> suspect it really makes no difference, but the question has come up
> during a
> code review.

It really makes no difference (provided similar quotes are escaped in
the string), so it comes down to a simple matter of style.  I personally
tend to use single quotes for simple keywords, the equivalent of
identifiers, or non-English data (such as HTML or XML fragments), and
double quotes for English phrases or text.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ The revolution will not televised
\__/ Public Enemy
    Blackgirl International / http://www.blackgirl.org/
 The Internet resource for black women.


From jdavis at empires.org  Mon Sep 16 05:46:58 2002
From: jdavis at empires.org (Jeff Davis)
Date: Mon, 16 Sep 2002 09:46:58 GMT
Subject: A question about C/API of python.
References: <1818979.1032155554@dbforums.com>
Message-ID: <mMhh9.5390$V7.1451374@twister.socal.rr.com>

Are you sure that you need to access a variable in the .so from the c 
program?

If you really do, then I would recommend using shared memory. It's fast, 
pretty simple to use, and you don't have to worry about changing it to a 
python variable and then back again.

Regards,
        Jeff


hpyhpy wrote:

> 
> I use python to load C module(.so),then use C load python,like this:
> C(main)--->python--->C(.so)
> But now my trouble is :
> In C(main), I have some Global variable that depend on process of
> C(.so).
> But because python is in middle of them,so I can't use variable of
> C(.so) in C(main) dexterously.
> I see some C/API reference but it still pluzzle me.Anybody can tell me
> some advise?
> Thanks.
> 
> --
> Posted via http://dbforums.com



From aleax at aleax.it  Fri Sep 27 03:23:53 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 27 Sep 2002 07:23:53 GMT
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <amvcev$cf6$1@s1.read.news.oleane.net> <7nHk9.172208$pX1.6117207@news2.tin.it> <ofLk9.304267$z91.13145351@bin3.nnrp.aus1.giganews.com>
Message-ID: <dITk9.175519$pX1.6209886@news2.tin.it>

Terry Reedy wrote:

> 
> "Alex Martelli" <aleax at aleax.it> wrote in message
> news:7nHk9.172208$pX1.6117207 at news2.tin.it...
>> The reason
>>         zz = a and b or c
>> is not the same thing as
>>         zz = a ? b : c
>> is that, when a is true and b is false, the former returns c,
>> while the latter returns b.
> 
> In almost all the practical cases where I have thought about using
> this,  b is *not* false or cannot become false, in which important and
> real case the two expressions *are* equivalent.

A typical, frequent, practical case is classification, e.g.:

"Append bleep to list truebleeps when bleep is true, append it
to list falsebleeps instead when false".

If we had a ternary operator, a natural way of expression would be:

    (bleep?truebleeps:falsebleeps).append(bleep)

Of course this would be a disaster with the and/or idiom instead:

    (bleep and truebleeps or falsebleeps).append(bleep)

As list truebleeps starts out empty, thus false, this buggy code
would NEVER append anything to it.

In this case, like in most others (but far from all!) we don't
need short-circuiting, so we can perfectly well code:

    (truebleeps, falsebleeps)[not bleep].append(bleep)

This is now correct -- but still nowhere as clear as good old:

    if bleep: truebleeps.append(bleep)
    else: falsebleeps.append(bleep)

which to me remains the canonical way to code this in Python.
Yes, it IS fractionally more verbose, but, so what?  It's
immediately obvious, simple, and not clever.


> I think this is an elegant hack, and so I shared it in response to the
> OP.  Others think it ugly.  So be it.

It's _dangerous_ -- to use it, one always has to be fully aware
of the risk that subexpression b might be false.  I think this
is a pitfall just waiting for the unwary to stumble.  Not being
an aesthete, I consider that more important than issues of beauty.


Alex



From claird at starbase.neosoft.com  Tue Sep 17 17:49:15 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 17 Sep 2002 16:49:15 -0500
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <Xns928CED692541cliechtigmxnet@62.2.16.82>
Message-ID: <C05DA40F31B09329.0165201CAF094801.13AF9332EB64347A@lp.airnews.net>

In article <Xns928CED692541cliechtigmxnet at 62.2.16.82>,
Chris Liechti  <cliechti at gmx.net> wrote:
>lpgray at uop.com (Lee Gray) wrote in 
>news:96904b50.0209171305.6d65f574 at posting.google.com:
>> My boss is pretty open-minded, but also needs to see something working
>> to be convinced Python is even a viable platform (he had never heard
>> of it).  Otherwise, .Net is a given, whether it's any good or not,
>> since corporate is MS all the way.
>
>with python you have to choice of diffrent GUI libs. if you consider 
>wxWindows you can play around with the wxPython demo application that comes 
>with the download. on the other hand when you support win32 only the 
>win32all package has support for all the windows APIs and it has some 
>samples that get installed.
			.
			.
			.
You can have Python *and* .Net.  Well, probably; it's
been promised, and, to the extent anyone knew a year
ago what .Net is, Python's part of it.

While we're pointing out individual applications that
might be of interest, please keep in mind that Python
is in serious production use at tens of THOUSANDS of
sites.  Don't judge it solely by what you see of the
first two demonstrations someone happens to push your
way.

Incidentally, Python *is* a "Microsoft solution".
There are plenty of people within Microsoft who use
Python because of its productivity in Win* contexts.
It's hard for me to give concrete references for this
claim; essentially all the people doing so describe
their efforts only "off the record".  I'm willing to
bet, though, that Python will be around for longer
than VBScript.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From wlfraed at ix.netcom.com  Sat Sep  7 13:53:57 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Sat, 07 Sep 2002 10:53:57 -0700
Subject: Larry Wall's comment on python...
References: <mailman.1031370910.1620.python-list@python.org>
Message-ID: <lfedla.pu2.ln@ix.netcom.com>

David LeBlanc fed this fish to the penguins on Friday 06 September 2002 
08:56 pm:

> One could hardly expect Larry to say "I have seen Pearl 6 - and it is
> Python!" :)
> 

        "The serpent has tempted me"

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From max at alcyone.com  Wed Sep 11 13:28:28 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 11 Sep 2002 10:28:28 -0700
Subject: function pointer
References: <ffcedbc9.0209110215.644c1788@posting.google.com>
Message-ID: <3D7F7D3C.7F47993C@alcyone.com>

Christina wrote:

> Does Python have function pointers? If not, is there a possibility to
> emulate them?

Functions and methods (both bound and unbound) are first-class objects
in Python.  Python doesn't have function pointers, but it doesn't need
them.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Procrastination is the thief of time.
\__/ Edward Young
    WebVal / http://www.alcyone.com/pyos/webval/
 URL scanner, maintainer, and validator in Python.


From bob at eleaston.com  Fri Sep  6 13:26:27 2002
From: bob at eleaston.com (Bob Easton)
Date: Fri, 6 Sep 2002 13:26:27 -0400
Subject: Interrupting Python
Message-ID: <yz5e9.4237$rU2.145619@news4.srv.hcvlny.cv.net>

I have a script that can run, accessing network resources, for several days.
Since the script does not normally need keyed input, exception processing
does not raise the keyboard exception until after the program ends normally.
I would like to be able to interrupt if from the keyboard, but have not
learned the trick.  How can I do this?

--
Bob Easton






From op73418 at mail.telepac.pt  Wed Sep  4 16:25:45 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Wed, 04 Sep 2002 21:25:45 +0100
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com> <yu99u1l5bkra.fsf@europa.research.att.com>
Message-ID: <dqqcnu8ebc0bcma85urdjn1hdsemsft3n4@4ax.com>

On Wed, 4 Sep 2002 19:46:49 GMT, Andrew Koenig <ark at research.att.com>
wrote:

>Gon?alo> Anyway, I've been directed to PEP 227, cheerfully obeyed, and
>Gon?alo> from what I have understood having read-only nested scopes
>Gon?alo> was more of a design decision than something else. And while
>Gon?alo> I can live happily with the decision (some would even say
>Gon?alo> happier) there is something about the "You can look at it but
>Gon?alo> not touch it" attitude that I dislike ;-)
>
>You can do anything to the object that the object support.
>What you can't do is change the binding of the name.

Yes, I am aware of this. You can even wrap x in a list and then modify
the list to indirectly change x, but it was the rebinding that I was
refering to, so let me rephrase the above to "You can look at it AND
even touch it but ONLY if suitably dressed" :-)

>
>    x = []
>    def foo(y):
>        x.append(y);
>    foo(3)
>    foo(4)
>    foo(5)
>
>After executing this code, x is [3, 4, 5].

All the best,
Gon?alo Rodrigues


From jubafre at brturbo.com  Wed Sep 25 19:29:47 2002
From: jubafre at brturbo.com (jubafre at brturbo.com)
Date: Wed, 25 Sep 2002 20:29:47 -0300 (GMT-03:00)
Subject: transform strings list in a int lists or other type???
Message-ID: <872170378.1032996587856.JavaMail.nobody@webmail2.brturbo.com>

                  Can i tranform a list of strings in a list of int, or other type in python???
and also write in a file? I think the open(path,'wb') in __builtin__ dont?t write in a file other types, just strings, rigth???
	
Juliano Freitas
www.gebrasil.hpg.com.br 

From erikprice at mac.com  Tue Sep 10 12:01:00 2002
From: erikprice at mac.com (Erik Price)
Date: Tue, 10 Sep 2002 12:01:00 -0400
Subject: Python iCalendar API?
Message-ID: <80494127-C4D6-11D6-A54B-00039351FE6A@mac.com>

Is there a Python module that provides capability of processing files 
in .ics (iCalendar) format?  I found nothing concrete on Google or at 
python.org ...





Erik





--
Erik Price                                                  (zombies 
roam)

email: erikprice at mac.com
jabber: erikprice at jabber.org




From chris.gonnerman at newcenturycomputers.net  Mon Sep 30 01:20:40 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Mon, 30 Sep 2002 00:20:40 -0500
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05> <an7bvd$burft$1@ID-125932.news.dfncis.de> <an7p7t$cpg$0@216.39.172.122> <mailman.1033338759.6775.python-list@python.org> <7xd6qwqtgz.fsf@ruckus.brouhaha.com> <mailman.1033352259.30842.python-list@python.org> <7xelbcgo56.fsf@ruckus.brouhaha.com> <mailman.1033357298.16341.python-list@python.org> <7x65woks0i.fsf@ruckus.brouhaha.com>
Message-ID: <006d01c26841$1fa587a0$ba01010a@local>

----- Original Message -----
From: "Paul Rubin" <phr-n2002b at NOSPAMnightsong.com>


> No, I'm still not convinced.  Is business arithmetic well-
> defined enough that you can never do the same calculation two
> ways and get different results?
>
> Example: bunches are three for a dollar.  How much do you pay
> if you buy nine bananas?
>
>     Answer #1:
>       price_per_banana = 1.00 / 3
>       number_of_bananas = 9
>       total_price = number_of_bananas * price_per_banana
>       print total_price
>
>     Answer #2:
>       price_per_bunch = 1.00   # three bananas
>       number_of_bunches = 3
>       total_price = number_of_bunches * price_per_banana
>       print total_price
>
> With floating point and rounding on output, you get the same
> answer both ways.
>
> With decimal arithmetic, you get $2.97 the first way and
> $3.00 the second way.

Excellent example, wrong focus.  The business user (in this
case, the retail store owner/manager) would define which
method, and which results, are right.  The second mode is
the "canonical" way in the manual system, but lately I've
realized that many (most?) retail POS systems can do it "both"
ways, handling single items by a variation of method 1 which gives answers
consistent with method 2.

I'm glad I didn't have to write and debug that code.

> For "business arithmetic" to always give the same answer, it
> has to forbid at least one of those methods of doing the
> problem.

Except for the hack I described above, you're right.  *The
business user will define the rules* regarding when and to
how much you round.  You seem to think that I mean to round
to two places (as in US dollars and cents) when what I mean
is to handle numbers in (or as if in) base 10 all the time.

It isn't a question of when you round, so much as it is a
question of what *happens* when you round.  Proper floating
decimal numbers would also solve the problem; it's floating
binary that causes unexpected anomalies.

The banana problem might be solved in a given POS (point of
sale) system by using four or five decimals, then rounding to
two at the line-item level.  That would give correct results
for both methods above.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From aleax at aleax.it  Thu Sep 19 07:27:53 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 19 Sep 2002 11:27:53 GMT
Subject: lists in dictonaries ?
References: <mailman.1032425484.16513.python-list@python.org>
Message-ID: <Zwii9.113647$ub2.2417956@news1.tin.it>

Manuel Hendel wrote:

> Is it possible to put lists in dictionaries? And is it also possible
> to put multiple lists in a dictionary for one key?

Yes to the first, no to the second (stricto sensu).  Lists are OK
as values in dictionaries.  But a dictionary has ONE value "for
one key".  You can make that value a list of lists, a tuple of
lists, and so on, but not, strictly speaking, "multiple" anythings.


Alex



From eric.brunel at pragmadev.com  Mon Sep 16 03:48:29 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 16 Sep 2002 09:48:29 +0200
Subject: Tkinter listbox clicking problem
References: <sGog9.307183$aA.53945@sccrnsc02> <altfrk$1eva$1@nntp6.u.washington.edu> <2D7E3697CF5D1521.5D9711540E657ED8.7EF0451C598438FC@lp.airnews.net> <9rCg9.262546$kp.888557@rwcrnsc52.ops.asp.att.net>
Message-ID: <am41tg$s6q$1@wanadoo.fr>

Matthew Bonig wrote:
> I'm confused... I looked at these examples but I didn't see anything pop
> out that answered my question... or was this mostly for the other post?

Binding to <Button-1> is equivalent to binding to <ButtonPress-1>. When 
your button is pressed, the list has actually not been updated yet. But if 
you bind to <ButtonRelease-1>, you'll get what you want.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From johnroth at ameritech.net  Sun Sep  1 19:40:36 2002
From: johnroth at ameritech.net (John Roth)
Date: Sun, 1 Sep 2002 19:40:36 -0400
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <un50ccotaak86b@corp.supernews.com>
Message-ID: <un594pno1gf7b7@news.supernews.com>

"A.M. Kuchling" <amk at amk.ca> wrote in message
news:un50ccotaak86b at corp.supernews.com...
> In article <3d725881.345921 at news.t-online.de>, Gerson Kurz wrote:
> > in the match or find function. So basically, couldn't one come up
with
> > a *human readable* syntax for re, and compile that instead? Python
>
> Maybe Ka-Ping Yee's rxb?  http://web.lfw.org/python/rxb15.py
>
> The problem with a new syntax is that no one else would be using it,
so
> you'd still need to learn the existing syntax for use with grep, vi,
Perl,
> &c.  (It wouldn't surprise me if Perl 6's revised regexes run into
this very
> difficulty and don't gain much adoption.)

I see you've already gotten to my suggestion - look at Apocalypse 5
and Exegesis 5 on the O'Reilly Perl page and see what Larry Wall
has done.

I tend to agree that Larry is going out on a limb. However, since he's
done so, if we are thinking about a new regex syntax, we really
should consider doing it the same way. As you point out, there's going
to be enough difficulty getting one radical revision accepted. Getting
two different ones accepted is likely to sink everyone's effort.

Also, doing it the same way will simplify the Python effort in Parrot
(assuming anyone cares about that any more...)

And as far as grep, Vi and so forth are concerned (Perl isn't
an issue - that's the way it's going,) doing something to fix them
isn't all that hard - at least for the GNU versions of the programs.

>
> --amk




From pearu at cens.ioc.ee  Mon Sep  9 04:52:38 2002
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Mon, 9 Sep 2002 11:52:38 +0300 (EEST)
Subject: returning values from FORTRAN with f2py
In-Reply-To: <c3835e5f.0209071122.338a787d@posting.google.com>
Message-ID: <Pine.LNX.4.21.0209091147470.2238-100000@cens.kybi>

On 7 Sep 2002, Chris Fonnesbeck wrote:

> I am wondering how to persuade some fortran code that I have compiled
> with f2py to return its calculated values to me in python.  I have
> some code that does gaussian elimination (mostly from Numerical
> Reciopes), and is supposed to return an inverted matrix so that I can
> solve a system of linear equations.  However, when run in python, the
> function runs without error but returns no values:
> 
> >>> a = [[2,1,3],[1,-1,1],[3,6,-1]]
> >>> b = [7,2,10]
> >>> import gauss
> >>> gauss.gauss(a,3,b,1)
> >>>
> 
> Is there something that I have to change in my fortran code to make
> this work?

You can add the following comment lines to your Fortran code:

cf2py  intent(in,copy,out) a
cf2py  intent(in,copy,out) b

(provided that you want the arguments a and b returned to Python).

If you cannot modify Fortran codes then you should create a .pyf file and
insert the above lines there (without cf2py prefix, of course).

HTH,
	Pearu




From storedmails at gmx.net  Tue Sep 24 10:43:59 2002
From: storedmails at gmx.net (beachboy)
Date: 24 Sep 2002 07:43:59 -0700
Subject: Sessionhandling
References: <c5b3549a.0209230650.34a30fca@posting.google.com> <f72dac0d.0209231211.5614b398@posting.google.com>
Message-ID: <c5b3549a.0209240643.2fdc5941@posting.google.com>

> Are you aware that a users language preference is encoded in every
> request that they make to a web server, through the "Accept-Language"
> HTTP header?

Thank you very much for your response. This could be a very good
solution, but the problem is that the company which i work for want
that the user has to choose which language he want.
So I still have the problem with Sessions.


From fpm at u.washington.edu  Sun Sep 15 23:15:48 2002
From: fpm at u.washington.edu (Frank Miles)
Date: 16 Sep 2002 03:15:48 GMT
Subject: pypgsql -- not preserving transactions ?
Message-ID: <am3id4$1epu$1@nntp6.u.washington.edu>

I've discovered that the 2.0 version of pyPgSQL packaged with Debian/woody
doesn't preserve transactions (at least with Postgres).  For example, if
a transaction consists of a series of INSERTs, and an error occurs partway
through the series, a COMMIT at the end of the series will result in the
INSERTs up to the erroneous one being thrown out (rollback), but unfortunately
those following the error being committed.  In looking at the imported code,
there are comments about trying to recover from a defective state, with
the class variable 'inTransaction' being cleared.

Does any newer version fix this nonsense?  Or perhaps someone can recommend
an alternative module, preferably compatible with DB-API 2.0?

TIA!

	-frank
-- 


From peter at engcorp.com  Fri Sep  6 22:47:23 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 06 Sep 2002 22:47:23 -0400
Subject: py2exe and a single exe file?
References: <3D792351.3070008@motorola.com>
Message-ID: <3d7968a2@news.sentex.net>

Stephen Boulet wrote:
> I'm trying to get py2exe to work for me, and I must be doing something 
> wrong.
> 
> I have a script that uses easygui (which uses Tkinter) and Numeric. It 
> also imports a small module from the same directory that the script 
> resides in.
> 
> I try building it like this:
> 
>    python setup.py py2exe -d N:\ -i easygui
> 
> It builds and runs, but only with the other stuff in the build directory 
> with the generaged exe. The other stuff is:
> 
>   tcl directory containing tcl8.3 and tk8.3 directories
>   _numpy.pyd
>   _sre.pyd
>   _tkinter.pyd
>   multiarray.pyd
>   umath.pyd
> 
> What I would like to accomplish is just having a single exe file lumping 
> everything together. Is this possible?

In addition to Gerhard's answer: I've successfully used PowerArchiver
(and I'm sure there are a dozen other similar possibilities) to package
up a py2exe'd program plus all the DLLs into a single .EXE which is
really just a self-extracting-and-executing ZIP file.  You run it, it
unpacks itself to a temporary directory, runs the main program, and
then cleans up everything (or not, optionally) when it's done.

-Peter



From cliechti at gmx.net  Sat Sep  7 12:54:50 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 7 Sep 2002 18:54:50 +0200
Subject: getting input in a embeded GUI app ...
References: <b2df190a.0209070836.9713b08@posting.google.com>
Message-ID: <Xns9282C063D9443cliechtigmxnet@62.2.16.82>

rgkindya at yahoo.co.in (rgkindya) wrote in 
news:b2df190a.0209070836.9713b08 at posting.google.com:

> Dear Newser,
> 
>      I'm a newbie for python. I wish to embed Python interpreter into my
> vc++ GUI app(for ms-windows only). when executing input statements (eg
> raw_input()) , in pythonwin environment (activepython) , it gets input
> using a new small input window, in IDLE , it gets input in its own
> window (next line). please help me to do these two type of inputs in
> my embeded GUI app. my app is only GUI , there is no console mode.

tha pythonwin way: you can write your own input function in c/c++ that gets 
the input from a message box. it should return Python string. (i assume 
you've found out how to write python funcs in c/c++, i.e. like those for in 
a module)
then you can replace the builtin function raw_input with your function 
(assign it to __builtin__.raw_input). take a look at the python win sources 
on how to do this in python.
 
> please debug my engliz bugs ...

:-) 

chris



-- 
Chris <cliechti at gmx.net>



From fbarbuto at telusplanet.net  Fri Sep 20 05:11:01 2002
From: fbarbuto at telusplanet.net (Fausto Arinos de A. Barbuto)
Date: Fri, 20 Sep 2002 09:11:01 GMT
Subject: What the heck? open() requires an integer?
Message-ID: <FCBi9.13802$Zk4.708378@news2.telusplanet.net>

     
Hi,

    The open() function is returning me a strange error.
    An attempt to create an output file with

            f = open("output.txt","w")

    yields the following error:

    Traceback (most recent call last):
    File "<interactive input>", line 1, in ?
    TypeError: an integer is required

    That happens when I am using either Pythonwin or
    IDLE 0.8, but not in interactive mode. In this latter
    case I can open, write to and close a file OK.

    What's the matter? Thanks in advance!

---Fausto





From mhammond at skippinet.com.au  Fri Sep  6 23:23:54 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 07 Sep 2002 03:23:54 GMT
Subject: COM/Active X newbie question
References: <5523ec5.0209061446.1404cd10@posting.google.com>
Message-ID: <ejee9.1387$_l1.3889@news-server.bigpond.net.au>

Gabe Gittings wrote:
> I am a newbie to prgramming in Python and programming under windows. I
> am trying to access an .ocx file. I can access the methods in python
> after running makePy.py but when I run a test script and try to access
> a method from the COM object I get this error.
> 
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
>   File "C:\Python22\Lib\site-packages\win32com\gen_py\7ADCDC5B-D0B6-11D1-AA76-002078136998x0x1x0.py",
> line 57, in GetRunManagerStatus
>     return self._oleobj_.InvokeTypes(0xa, LCID, 1, (8, 0), (),)
> com_error: (-2147418113, 'Catastrophic failure', None, None)
> 
If it is a true OCX, then it probably needs to be hosted in a GUI 
environment, such as pythonwin.  See pywin\demos\ocx for examples.

If you can find any VBScript or JScript code that works with the object, 
it should be usable the way you are trying.  If the only examples you 
can find are hosted inside a form in a full-blown VB sample, then the 
Pythonwin thang should work.

Mark.



From ht at cogsci.ed.ac.uk  Mon Sep 16 13:18:19 2002
From: ht at cogsci.ed.ac.uk (Henry S. Thompson)
Date: 16 Sep 2002 18:18:19 +0100
Subject: Accessing Windows Registry under cygwin
In-Reply-To: <20020916164259.GG424@tishler.net>
References: <f5b1y8298w2.fsf@cogsci.ed.ac.uk>
	<20020916164259.GG424@tishler.net>
Message-ID: <f5bbs6x7t0k.fsf@cogsci.ed.ac.uk>

Jason Tishler <jason at tishler.net> writes:

> Sorry for the sluggish response time, but I just returned from
> vacation...]
> 
> On Mon, Sep 09, 2002 at 09:59:41PM +0100, Henry S. Thompson wrote:
> > On 11 September 2001 Norman Vine said "I have managed to get the
> > '_winreg' module working with Cygwin < _winreg = Python access to the
> > Windows Registry >" [1].  Assiduous searching does not reveal that
> > this work ever surfaced -- where might I find it?
> 
> See the following:
> 
>     http://sourceforge.net/tracker/?group_id=5470&atid=305470&func=detail&aid=554807
> 
> Are you willing to help?

I wish I were competent to help -- I have no experience with
programming to WIN32 APIs.  I'll download the package and have a look.

It's not entirely clear from the sourceforge log what's actually
needed next.

ht
-- 
  Henry S. Thompson, HCRC Language Technology Group, University of Edinburgh
          W3C Fellow 1999--2002, part-time member of W3C Team
     2 Buccleuch Place, Edinburgh EH8 9LW, SCOTLAND -- (44) 131 650-4440
	    Fax: (44) 131 650-4587, e-mail: ht at cogsci.ed.ac.uk
		     URL: http://www.ltg.ed.ac.uk/~ht/
 [mail really from me _always_ has this .sig -- mail without it is forged spam]



From jepler at unpythonic.net  Tue Sep 17 22:37:29 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Tue, 17 Sep 2002 21:37:29 -0500
Subject: timeoutsocket module help
In-Reply-To: <am7r98.h0.1@kserver.org>
References: <am7r98.h0.1@kserver.org>
Message-ID: <20020917213720.A2554@unpythonic.net>

You won't have a lot of luck using timeoutsocket to influence the timeout
of name lookups -- those are done in C code, not at the level where
timeoutsocket can be useful.

Investigate a pure python DNS setup, like pydns
(http://pydns.sourceforge.net/).

Jeff
PS following for your enjoyment, an inefficient program that attempts to
look up a host in many different RBLs, using threads for parallelism and
the aforementioned pydns for lookups with a defined timeout:

RBL_LIST = (
    'bl.redhatgate.com', 'blackholes.intersil.net', 'dews.qmail.org',
    'dssl.imrss.org', 'mr-out.imrss.org', '3y.spam.mrs.kithrup.com',
    'blackhole.com', 'pu.net', 'blackholes.2mbit.com',
    'blackholes.five-ten-sg.com', 'blackholes.mail-abuse.org',
    'blackholes.wirehub.net', 'blacklist.spambag.org',
    'block.blars.org', 'blocktest.relays.osirusoft.com', 'dev.null.dk',
    'dialup.blacklist.jippg.org', 'dialups.mail-abuse.org',
    'dialups.relays.osirusoft.com', 'dnsbl.njabl.org', 'list.dsbl.org',
    'multihop.dsbl.org', 'unconfirmed.dsbl.org', 'dul.maps.vix.com',
    'dul.orca.bc.ca', 'dynablock.wirehub.net', 'flowgoaway.com',
    'formmail.relays.monkeys.com', 'http.opm.blitzed.org',
    'inputs.relays.osirusoft.com', 'ipwhois.rfc-ignorant.org',
    'opm.blitzed.org', 'orbs.dorkslayers.com', 'outputs.orbz.org',
    'relays.ordb.org', 'pm0-no-more.com', 'pu.net',
    'proxies.relays.monkeys.com', 'rbl-plus.mail-abuse.org',
    'rbl.maps.vix.com', 'rbl.spam.org.tr', 'relays.dorkslayers.com',
    'relays.mail-abuse.org', 'relays.osirusoft.com',
    'relays.radparker.com', 'relays.visi.com',
    'sbl.spamhaus.org', 'socks.opm.blitzed.org',
    'socks.relays.osirusoft.com', 'bl.spamcop.net',
    'spamguard.leadmon.net', 'spamhaus.relays.osirusoft.com',
    'spammers.v6net.org', 'spamsites.relays.osirusoft.com',
    'spamsources.fabel.dkspews.relays.osirusoft.com',
    'wingate.opm.blitzed.org', 'xbl.selwerd.cx', 'ztl.dorkslayers.com',
    'or.orbl.org', 'spamips.rbl.shub-inter.net',
)

import sys, socket, threading, DNS

DNS.DiscoverNameServers()

def check_one(i, j):
    dns = DNS.Request(timeout=5)
    k = "%s.%s" % (i, j)
    try:
	res = dns.req(k)
    except DNS.DNSError, reason:
	l = reason
    else:
	status = res.header['status']
	if status == 'NOERROR' and res.answers:
	    l = res.answers[0]['data']
	elif status == 'NOERROR':
	    l = "NOERROR but no answer?"
	elif status == 'NXDOMAIN':
	    return
	else:
	    l = status
    print "%-50s %s" % (k, l)
	
for i in sys.argv[1:]:
    i = socket.gethostbyname(i).split(".")
    i.reverse()
    i = ".".join(i)
    tl = []
    for j in RBL_LIST:
	t = threading.Thread(target=check_one, args=(i, j))
	t.start()
	tl.append(t)

    while tl:
	tl.pop().join()




From kkto at csis.hku.hk  Fri Sep  6 08:11:15 2002
From: kkto at csis.hku.hk (Isaac To)
Date: 06 Sep 2002 20:11:15 +0800
Subject: Dictionary...
Message-ID: <7ir8g7xqqk.fsf@enark.csis.hku.hk>

Hi,

Some time ago I want to have a dictionary-like type that can support
dictionary operation, plus that given any key, I want to know the
predecessor and successor of any given key efficiently.  I've thought of
using a Python list, but it is impossible to update quickly.  The closest
thing I can think of is an auxillary map that implements the successor and
predecessor functionalities.  But I'm wondering there is anything better.
After all, Python dictionaries are implemented by height balancing tree, so
internally the needed data can be retrieved efficiently.  Any hint?

Regards,
Isaac.


From sschwarzer at sschwarzer.net  Wed Sep 11 16:23:02 2002
From: sschwarzer at sschwarzer.net (Stefan Schwarzer)
Date: Wed, 11 Sep 2002 22:23:02 +0200
Subject: "str.contains(part)" or alternatives?
Message-ID: <alo8jr$l04$1@news.online.de>

Hello all

I would like to get your thoughts on testing if a string contains another.

To do something dependent on whether a part is contained in another string, the
following code is familiar to me:

 >>> s = 'Hello'
 >>> if s.find('ll') != -1:
...     do_something_if_found()
... else:
...     do_something_if_not_found()

However, I dislike the test for the special value -1; that reminds me of error
code checking. On the other hand, using exceptions could look like

 >>> try:
...     s.index('ll')
...     do_something_if_found()
... except ValueError:
...     do_something_if_not_found()

This has the problem that a ValueError raised in do_something_if_found() may give
the false impression that the substring isn't contained in the string s. Moreover,
that code doesn't represent my thoughts very well. (Maybe I could say that better
if I were a native English speaker :) ).

I would like something analogous to .startswith and .endswith:

 >>> s = 'Hello'
 >>> if s.contains('ll'):
...     do_something_if_found()
... else:
...     do_something_if_not_found()

Do you have other suggestions how to do the test easily without using special
values? Should a .contains method be part of string objects in future Python
versions? What do you think?

Stefan



From mike at bindkey.com  Fri Sep 13 15:14:54 2002
From: mike at bindkey.com (Mike Rovner)
Date: Fri, 13 Sep 2002 12:14:54 -0700
Subject: squeeze with packages
Message-ID: <altdet$602$1@main.gmane.org>

Hi all,

I'm trying to use squeeze
(http://www.pythonware.com/products/python/squeeze/) and run into couple
of problems:

- I could not squeeze packages (subdirs with __init__.py) and
- when I try to make archive (squeeze -1x ...), I got an error "name
'explode' is not defined".

Is it possible to bypass those obstacles?

Thanks in advance,
Mike







From andres.meza at nuevosmedios.ws  Wed Sep 18 12:01:09 2002
From: andres.meza at nuevosmedios.ws (Andres Meza)
Date: Wed, 18 Sep 2002 11:01:09 -0500
Subject: problems getting a single item out of a list
Message-ID: <1032364878.3088.64.camel@coraje>

Manuel Hendel wrote:
> I try to get a single item out of a list, but after this, the single
> item is stil a list item.
> 
> domainfield = fields[3:4]
> 
> I just want to get a new variable with a the value of the 3rd item of
> the list fields.
> 

If I understand your question, you have two ways to make it work, from
my point of view:
a) Get the value from domainfield:

    domainfields = fields[3:4]
    value = domainfields[0]

b) Get the value directly from the list:

    value = fields[3]

I hope it helps.

Bye,
-- 
   Andr?s Meza
   Technical Manager
 __Nuevos Medios - NM S.A.____________________
   e-mail:         andres.meza at nuevosmedios.ws
   day time phone: (+572)6827794 ext.102
   url:            www.nuevosmedios.ws
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020918/c6f7455c/attachment.html>

From luc.saffre at gmx.net  Mon Sep 16 09:37:21 2002
From: luc.saffre at gmx.net (Luc Saffre)
Date: Mon, 16 Sep 2002 16:37:21 +0300
Subject: "if x == None" raises "'NoneType' object is not callable"
References: <JHEOKEOOLIGLDHCMAHMOKEPADOAA.marklists@mceahern.com>
Message-ID: <3D85DE91.8010102@gmx.net>

Thank you, Mark. Now I will remeber the difference between "is" and "=="!

There is another big bug in my code snippet. If I write::
      try:
         return self._values[name]
      except KeyError,e:
         raise AttributeError,str(e)

instead of::
      try:
         return self._values[name]
      except KeyError,e:
         AttributeError,str(e)

then the results are also much more as expected :-)
For every undefined attribute I told Python "yes, I have this attribute 
defined, and its value is None".

The explanation for this strange message is now clear (at least for my 
simplistic thinking):
- Because I used "==" (and not "is"), Python made a deep look into my 
object and looked into each attribute, not only the hash()
- One of these attributes was some of those magic functions, perhaps 
__cmp__(). And my __getattr__() returned None instead of raising 
AttributeError for these attributes which, I admit, if you define them, 
should be callable...

Okay... Python is great!
Luc



On 9/16/2002 4:14 PM, Mark McEahern wrote:

>Comparison to None should generally be comparison by identity rather than
>equality:
>
>  if x is None:
>
>or:
>
>  if x is not None:
>
>not:
>
>  if x == None:
>
>nor:
>
>  if x != None:
>
>What you might consider is adding a __len__ magic method to your Row class
>like so:
>
>  def __len__(self):
>    return len(self.values)
>
>Then you can do this:
>
>  r = Row()
>  if r:
>    ...
>  else:
>    ...
>
>Cheers,
>
>// mark
>
>-
>
>
>  
>






From jb at cascade-sys.com  Sat Sep 28 04:07:16 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sat, 28 Sep 2002 01:07:16 -0700
Subject: combining mailbox and email modules
References: <NPck9.581867$UU1.104088@sccrnsc03> <slrnap41re.1mh.matt@happy-hour.t-island.com> <LAxk9.79207$gA4.27843@sccrnsc02> <1fj6jb2.62mx08f47amsN%fgranger@alussinan.org>
Message-ID: <3D956334.3020804@cascade-sys.com>

Fran?ois Granger wrote:

>Bob Horvath <nospam at horvath.com> wrote:
>
>  
>
>>The problem is that it looks good to me.  I've played around with 
>>it, and if I take out the MIME gunk (It actually is a Klez virus), 
>>it parses ok.
>>    
>>
>
>I have had difficulties with the handling of multipart message with the
>email package. I did not found a solution yet.
>
My experience is that a lot of Spam violates Mime rules.  Often they 
omit the end marker, I guess relying on EOF to terminate the message. 
 Then the python packages can't parse any of the remaining messages.

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From tjlavoie at mb.sympatico.ca  Tue Sep 17 09:53:33 2002
From: tjlavoie at mb.sympatico.ca (Tim Lavoie)
Date: 17 Sep 2002 08:53:33 -0500
Subject: tkinter asyncore.loop
References: <am73d3$o4v$1@oslo-nntp.eunet.no>
Message-ID: <87znugbu3m.fsf@theasylum.dyndns.org>

>>>>> "Rune" == Rune Hansen <rune.hansen at viventus.no> writes:

    Rune> Hello, Some months back I sendt a question to the list where
    Rune> I asked if anyone knew how to catch errors in an asyncore
    Rune> loop.  I got an answer to override the "handle_error" in the
    Rune> client. I still can't get

    Rune> it to trap any errors though. The error I need to trap is
    Rune> "Connection lost". If I remove the network connection and
    Rune> try to update the stream

    Rune> or post to the socket nothing happens. When I reconect the
    Rune> network my application just continues to serve like nothing
    Rune> has happened. This is
    Rune> of course nice, but not the functionallity I'm looking for.

I tried much the same thing, and never did see much in useful errors
hitting my handle_error, but catching exceptions in handle_read did
the trick. My little app also uses tkinter and asyncore, along with
timeoutsocket. Anyway, the exceptions were caught and used to trigger
the error-handling callback to do the work.

  


From mbonig at hotmail.com  Fri Sep 13 12:49:28 2002
From: mbonig at hotmail.com (Matthew Bonig)
Date: Fri, 13 Sep 2002 16:49:28 GMT
Subject: Tkinter listbox clicking problem
Message-ID: <sGog9.307183$aA.53945@sccrnsc02>

Right now I have it bound so when a Tkinter listbox gets a <Button-1>
(lb.bind("<Button-1>", someFunc)) it calls function someFunc. Now,
someFunc() simply prints the curselection(). The problem is it's lagged...
if I first select list item 1 (index 0), and then 2, and then 3 (array
indexes 0,1,2) it prints out themout incorrectly.
(I have someFunc actually do some try, except stuff so I don't get uncaught
exceptions) Instead of printing 0,1,2 it prints "nothing selected", 0, 1.
The function get's called before the listbox updates it's selection markers!

Does anybody know how I could force listbox to update it's selection markers
so instead of printing "nothing selected, 0,1  it prints 0,1,2 like it
should?

Thanks
Matthew Bonig




From grahamd at dscpl.com.au  Sun Sep 15 03:06:36 2002
From: grahamd at dscpl.com.au (Graham Dumpleton)
Date: 15 Sep 2002 00:06:36 -0700
Subject: Pmw.ScrolledText programatically scrolling
References: <3D81D44E.C5FC4A8B@mill.co.uk>
Message-ID: <dc6f5c99.0209142306.68368441@posting.google.com>

Joe Connellan <joec at mill.co.uk> wrote in message news:<3D81D44E.C5FC4A8B at mill.co.uk>...
> I've got a Pmw.ScrolledText() that I'm using as a log and I've written
> an addToLog() method like so:
> 
>  def addToLog(self, logString):
>   self.logScr.configure(text_state = 'normal')
>   self.logScr.insert(END, logString)
>   self.logScr.configure(text_state = 'disabled')
> 
> I now want it to scroll when I add a line so the user doesn't have to
> constantly scroll down to view the most recent lines.
> 
> I can't find any info on doing this with ScrolledText or its text
> component - do I need to reposition the scrollbars for it to scroll?

Another slightly different way to what others have suggested is the
following:

  def __init__(self,...):
    ...
    self._eventView = Pmw.ScrolledText(g1.interior(),text_padx=4,
        text_pady=4,text_wrap="char",text_state="disabled",
        vscrollmode="static")
    self._eventView.pack(side="top",padx=8,pady=8,fill="both",expand="true")
    self._eventView.yview("moveto",1.0)

  def _clearLog(self):
    self._eventView.configure(text_state="normal")
    self._eventView.clear()
    self._eventView.configure(text_state="disabled")
    self._eventView.yview("moveto",1.0)
 
  def _appendLog(self,message):
    p1,p2 = self._eventView.yview()
    self._eventView.configure(text_state="normal")
    self._eventView.insert("end",message)
    self._eventView.configure(text_state="disabled")
    if (p1 == 0.0 and p2 == 0.0) or p2 == 1.0:
      self._eventView.yview("moveto",1.0)

A trick in this code is that one uses yview() to determine what of the
text is being shown and will only move to last text inserted if the scroll
area was already at the end.

The reason for this is that it can be mighty annoying to have the scrolled
area always go to the end if you have used the scroll bar to explicitly
go to an earlier part of the text to look at it.


From jason at atdesk.com  Fri Sep 27 09:38:20 2002
From: jason at atdesk.com (Jason D. Smith)
Date: Fri, 27 Sep 2002 13:38:20 +0000 (UTC)
Subject: How do you pronounce "__" (double underscore)?
References: <525eaba.0209260956.72bf6786@posting.google.com> <20020926164641.1753ebbb.ckasso@sprynet.com>
Message-ID: <an1n0c$3sda$1@news3.infoave.net>

Chris Kassopulo <ckasso at sprynet.com> writes:

> On 26 Sep 2002 10:56:45 -0700
> pknotz at sandia.gov (Pat Notz) wrote:
> 
> > @ = "at"
> > ! = "bang"
> > . = "dot"
> > * = "star"

I'm for "George" and "Penelope"
-- 
Jason.


Hi, I'm Bob. I'm a tomato and I'm here to help you.


From peter at engcorp.com  Fri Sep  6 23:15:54 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 06 Sep 2002 23:15:54 -0400
Subject: Timothy Rue
References: <898662b7d85d2c966c0bd45d563d5d2d@xganon.com> <alakj8$20qo$1@nntp6.u.washington.edu> <alb4m8$ksc$1@newshost.accu.uu.nl>
Message-ID: <3d796f52@news.sentex.net>

Martijn Faassen wrote:
> Which group is that anyway? :) Seems like you and Tim have been posting
> on comp.sys.amiga misc. Apparently you were looking for a Tim Rue 
> FAQ at the time, and the thread has a link, FAQ posted in '99!
> 
> http://groups.google.com/groups?q=tim.rue.faq&start=10&hl=en&lr=&ie=UTF-8&scoring=d&selm=7r5lrp%24ftj%241%40ns2.dialnet.net&rnum=15

Thanks for the link, Martijn!  Humorous reading at this point.

I find the revelations about the Team AMIGA conspiracy especially
intriguing, considering the past connections between certain
PSU members and the



From wurmy at earthlink.net  Thu Sep 12 22:11:53 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Fri, 13 Sep 2002 02:11:53 GMT
Subject: Expect() from telnetlib giving me problems.
References: <4378fa6f.0209121325.130766fc@posting.google.com>
Message-ID: <3D814A31.2010909@earthlink.net>

Marc wrote:

[snip]
>   File "C:\python22\lib\telnetlib.py", line 541, in expect
>     list[i] = re.compile(list[i])
>   File "C:\Python22\lib\sre.py", line 178, in compile
>     return _compile(pattern, flags)
>   File "C:\Python22\lib\sre.py", line 228, in _compile
>     raise error, v # invalid expression
> error: nothing to repeat

Note that this is an error in sre.py, not in telnetlib. (A clue! ;-)

> I've tried several different ways to get the expect command to work. I
> can't find any examples and only limited documentation. Here is how
> I'm trying to implement it:
> 
>     alist = [ 'COMPLD', '*/' ]
>     buffer = conn.expect(alist, 5)
>  
> Can someone please tell me what is wrong or provide me with some
> working examples of using the command.

expect() expects a list of regular expressions. '*/' probably isn't a valid 
regex. See also

   http://www.python.org\doc\lib\telnet-objects.html

"""
expect(list[, timeout])

Read until one from a list of a regular expressions matches.

The first argument is a list of regular expressions, either compiled 
(re.RegexObject instances) or uncompiled (strings). The optional second 
argument is a timeout, in seconds; the default is to block indefinitely.
"""

HTH,

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html



From ajw126NO at SPAMyork.ac.uk  Wed Sep 11 09:48:56 2002
From: ajw126NO at SPAMyork.ac.uk (Andrew Wilkinson)
Date: Wed, 11 Sep 2002 14:48:56 +0100
Subject: Abstract Syntax Trees
References: <GuFf9.1028$XE6.54057@newsfep1-win.server.ntli.net> <mailman.1031743334.4015.python-list@python.org>
Message-ID: <kPHf9.1648$XE6.71805@newsfep1-win.server.ntli.net>

What an idiot I am! It was the fact that it didn't have a new line at the
end that was causing the problem.

It does seem strange that the file will run fine, but you can't access it's
syntax tree. That looks suspiciously like a bug to me...

Thanks,
Andrew

"holger krekel" <pyth at devel.trillke.net> wrote in message
news:mailman.1031743334.4015.python-list at python.org...
> Andrew Wilkinson wrote:
> > Hi,
> >
> > I'm having a lot of trouble trying to get the parser and compiler
modules to
> > properly parse some files.
> >
> > At the moment the trouble seems to be with any loop construct (i.e. for
or
> > while).
> >
> > If you create a test file like this...
> >
> > ---
> > for i in range(10):
> >     print i
> > ---
>
> make sure that this file has a newline at the end.
>
> others would say, get a better OS :-)
>
> regards,
>
>     holger
>




From leafnode-usenet at schabi.de  Thu Sep 19 09:28:34 2002
From: leafnode-usenet at schabi.de (Markus Schaber)
Date: Thu, 19 Sep 2002 15:28:34 +0200
Subject: Remarks to Python 2.3
References: <20020918222421.328088eb.use-net@schabi.de>
 <amannn$p4b$1@panix1.panix.com>
 <m3u1knj8j9.fsf@mira.informatik.hu-berlin.de>
Message-ID: <20020919152834.592eaa39.leafnode-usenet@schabi.de>

Hallo,

martin at v.loewis.de (Martin v. Loewis) schrieb:

> Or, better yet, submit bug reports.

I'm not really shure whether my remarks are about "bugs" - I just wanted
to collect other opportunities here to help me decide whether one can
regard this as bugs or not, and eventually submit bug reports.

Gru?,
Markus


From cjw at sympatico.ca  Wed Sep 25 08:53:43 2002
From: cjw at sympatico.ca (Colin J. Williams)
Date: Wed, 25 Sep 2002 08:53:43 -0400
Subject: Attempt to install and test gnuplot fails
Message-ID: <3D91B1D7.A5AB6F85@sympatico.ca>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020925/3704f28e/attachment.html>

From martin at v.loewis.de  Tue Sep 10 14:42:34 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 10 Sep 2002 20:42:34 +0200
Subject: shared libraries
References: <xYpf9.1$wH3.129@client>
Message-ID: <m3fzwhit45.fsf@mira.informatik.hu-berlin.de>

"Jim Taylor" <jimt at aiinet.com> writes:

> The loader is loading the module, however, during the init function when
> Py_InitModule() is called, a segmentation fault occurs indicating an address
> out of bounds error. When I try to print the contents of the module table
> from within the initialization function, I also get a segmentation fault, so
> the problem is the way in which the module has been loaded. It appears there
> is an address space issue, but I don't know how to resolve it.
> 
> Ideas anyone?

I recall the problem, but I don't think it was ever solved. "A gcc
bug" was the theory that was most widely used to explain the
phenomenon.

Of course, this was all a few years ago when Python 1.5 was still in
use :-)

Regards,
Martin



From bokr at oz.net  Mon Sep  9 17:58:38 2002
From: bokr at oz.net (Bengt Richter)
Date: 9 Sep 2002 21:58:38 GMT
Subject: Embedded Python implementation
References: <MAfe9.266658$v53.14092531@news3.calgary.shaw.ca>
Message-ID: <alj5ie$bl9$0@216.39.172.122>

On Sat, 07 Sep 2002 04:50:52 GMT, Evan <tagith1394 at hotmail.com> wrote:

>Hello,
>
>I am currently trying to embed Python in an application I am writing, 
>and all is going well, except for one inexplicable bug..
>
>When I pass a variable (pointer to a string) to PyRun_String, it will 
>always error on line 2, character 5.. regardless of what is actually on 
>the line in that spot (for that matter, even if there is no line 2)..
>
>The script works perfectly when I place it in a file, and use 
>PyRun_File, and it also works perfectly when I place the string in 
>quotations, and pass it to PyRun_String directly inside the C code..
>
>The C code in question is as follows:
>if ( PyRun_String( com_list, Py_file_input, pGlobals, pLocals ) == NULL )
>         PyErr_Print();
>
>I am confident the problem is not with pGlobals or pLocals, as those 
>arguments are used with PyRun_File, as well as with the quoted string..
>
Looking at \python-2.1\python\pythonrun.c:
--
int
PyRun_SimpleString(char *command)
{
	PyObject *m, *d, *v;
	m = PyImport_AddModule("__main__");
	if (m == NULL)
		return -1;
	d = PyModule_GetDict(m);
	v = PyRun_String(command, Py_file_input, d, d);
	if (v == NULL) {
		PyErr_Print();
		return -1;
	}
	Py_DECREF(v);
	if (Py_FlushLine())
		PyErr_Clear();
	return 0;
}

--
I would model the code after that. You say you have the directories, and Py_file_input
is the  constant from Python.h, so capture v and do the right thing (note the decref), IWG ;-)

Regards,
Bengt Richter


From jb at cascade-sys.com  Wed Sep 25 21:45:20 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Wed, 25 Sep 2002 18:45:20 -0700
Subject: Larry Wall's comment on python...
References: <F7DB8D13DB61D511B6FF00B0D0F06233025D48F5@mail.hebmex.com>
Message-ID: <3D9266B0.7090906@cascade-sys.com>

sismex01 at hebmex.com wrote:

>>Ironically, Guido originally ruled (in the style guide, IIRC) 
>>that thou shalt use tabs and tabs are 8 spaces, "as god
>>intentended them to be."
>>    
>>
>
>Yeesh! Why in God's name would he say such a dumb thing?
>I mean, even ancient mechanical typewriters have movable
>tab settings. Rigid, 8-char tabs is just so... neolithical.
>

Once upon a time, a large class of computers, and their attendant 
peripherals, OSs and utility programs universally implemented tabs 
hard-wired at 8 stops.  In some circles this heritage extended back 30 
or 40 years.  Unix originally followed this convention, as did IIRC all 
Dec computers and that of most other "minicomputer" vendors.  Ed and the 
original VI editors only had fixed tab stops consistent with terminal 
drivers and all the unit record equipment at the time.  IIRC, even Emacs 
originally only had tab stops at every 8 columns.  The original PC also 
followed this convention, though not all applications did.  I don't know 
for sure but I bet $1 that Linux tty drivers today also are hard wired 
at 8 char tab stops.

Now, if you've ever worked in an environment where EVERYBODY uses the 
same tab stops then you'd probably agree that tabs are vastly superior 
to spaces.  I don't think there's any question.  I presume it's from 
this perspective that the right honorable BDFL issued his original 
proclamation, which is actually quite reasonable, especially from my old 
fart's perspective.

In fact, problems only arise when alternative, competing tab stop 
conventions try to coexist, which I agree is not  generally solvable. 
 Closest solution I've seen is some sites where everybody uses Emacs and 
employ coding conventions so that on startup Emacs can read a specially 
coded comment and automatically switch to the proper indent for that 
file.  This can sort of work but your printer and other rendering 
utilities also all have to recognize the codes.  Spaces are a more 
practical solution.  

I myself still prefer tabs as I find "indenting" and "out denting" 
sections of code to be a common operation (in Python and in other 
languages).  Inserting or removing an "if" or changing inline code to a 
function makes indenting and out denting a common operation.  In the 
environments I use, this works well with tabs and not so well or at all 
with spaces.  By default, I still prefer 8 char stops but I also find I 
can live just fine with 4.  But I can't stand it without tabs.  I'll 
convert to spaces if I need to when delivering code but I can't work on 
it that way.

Anyway, what's wrong with the "neolithical" 8 char stops?  For several 
decades, all the C code ever written was tab indented at 8 columns, 
along with most of the text and other programming languages on Unix. 
 Some people complained that in practice 8 char tabs forced them to 
indent too much and their programs got squashed off the right edge of 
the page.  I have little sympathy for that argument, as I feel if your 
program is properly modularized then then this should not be much of a 
problem as individual modules should not have to be indented all that 
much.  Also, I suspect part of the pressure came from Pascal users, 
where some coding conventions require indenting two levels, once for the 
begin/end pairs and again a second time for the code contained between 
them.  What a crock.  

For what it's worth, Python's C implementation today appears to be tab 
indented to 8 columns, at least all the parts I've looked at.  The main 
interpreter loop is a good example of a "complex" piece of code.  The 
innermost loop includes a single switch statement that's over 1400 lines 
long!  Surrounding logic require that the switch statement and case 
labels start 2 tab stops in and thus the body of each case starts 3 
stops in or in column 24.  Yikes, that's like 1/3 of the total.  And yet 
none of the case bodies extend past column 73 or so.  Turns out that in 
practice most code lines are pretty short.  The ones that aren't 
typically are function calls with multiple arguments, and these may 
easily be split onto multiple lines (often thereby increasing 
readability).  Since Guido personally wrote the bulk of this system as a 
labor of love, I surmise that he's still pretty sentimental about the 
old ways.

Regards

--jb


-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From SBrunning at trisystems.co.uk  Tue Sep 10 09:55:23 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 10 Sep 2002 14:55:23 +0100
Subject: [UK] Reminder - the next Python South-East UK booze-up is on Thur
	sday, September 12th, 2002
Message-ID: <31575A892FF6D1118F5800600846864DCBD77C@intrepid>

The next Python South-East UK booze-up is on Thursday, September 12th, 2002,
at The Lamb and Flag, Covent Garden, London, at seven p.m.  See
<http://www.brunningonline.net/simon/python/boozeup.cgi> and 
<http://www.brunningonline.net/simon/blog/archives/000320.html> for details.
All welcome!

No PSU business to be




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



From pyth at devel.trillke.net  Mon Sep  2 07:23:48 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 2 Sep 2002 13:23:48 +0200
Subject: Why it is so dramatical?
In-Reply-To: <200209021251.17165.b.maryniuk@forbis.lt>; from b.maryniuk@forbis.lt on Mon, Sep 02, 2002 at 12:51:17PM +0200
References: <mailman.1030957743.4021.python-list@python.org> <200209021228.58453.b.maryniuk@forbis.lt> <20020902124505.J3893@prim.han.de> <200209021251.17165.b.maryniuk@forbis.lt>
Message-ID: <20020902132348.K3893@prim.han.de>

Bo M. Maryniuck wrote:
> On Monday 02 September 2002 12:45, holger krekel wrote:
> > It's much more interesting to look at real performance limitations
> > instead of some theoretical pusing-the-limits loops.
> Hmmm... I've raised this, `coz I have a problem with a perfomance in Python 
> app, which generates XML for XSLT. I found, that 
> 
> 	string1 += string2
> 
> ...sucks and just for experiment I got rewrote XML generating piece of code to 
> Perl and I've found, that Perl does the same *much* faster than Python. :( 

Did you try using cStringIO? With your first example

[hpk at cobra /tmp]$ python slow.py
<function func1 at 0x80fda84> took 4.68154907227 seconds
<function func2 at 0x810be84> took 0.00871503353119 seconds

where func2 is:

def func2():
    _body = cStringIO.StringIO()
    _data = 'A'*81
    for a in xrange(0, 0xfff):
        _body.write(_data)
    return _body.getvalue()

(you need to do 'import cStringIO' at the top of the file.)
and func1 is your code.

regards,

    holger


P.S: the full script

import cStringIO

_data = 'A'*81

def func1():
    _body = ''
    _data = 'A'*81

    for a in xrange(0, 0xfff):
        _body += _data
    return _body

def func2():
    _body = cStringIO.StringIO()
    _data = 'A'*81
    for a in xrange(0, 0xfff):
        _body.write(_data)
    return _body.getvalue()

def drive(func):
    import time
    start = time.time()
    res = func()
    end = time.time()
    print func, "took",end-start,"seconds"
    return res


print len(drive(func1))
print len(drive(func2))




From eva_ag at eresmas.com  Sun Sep 15 10:35:26 2002
From: eva_ag at eresmas.com (Eva)
Date: 15 Sep 2002 07:35:26 -0700
Subject: Good book choice?
Message-ID: <e6c4872c.0209150635.4816ef98@posting.google.com>

Hello all!

I am new to Python but an experienced programmer in Fortran95 (not too
object
oriented). I am looking for a suited Python book. I'd like it to have
clear ideas about OO, as well as an straight pointing into the core
language.

It seems that 'Python Essential Reference (2nd Edition)' is a good
choice. There is also the book 'Making use of Python' (by Rashi
Gupta). Is there any user that bought both? Is it good the treatment
of OO in the former? Is there a lack of applications point of view in
the former? Could anybody compare them? pros/cons?

Thanks in advance,
Eva.


From marklists at mceahern.com  Sun Sep 29 13:52:27 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sun, 29 Sep 2002 12:52:27 -0500
Subject: Python's import: why doesn't it work?
In-Reply-To: <3d9739c7$0$19818$91cee783@newsreader01.highway.telekom.at>
Message-ID: <JHEOKEOOLIGLDHCMAHMOCEGMECAA.marklists@mceahern.com>

[Jurie Horneman]
> However, I do stand by my criticisms of the Python docs.

Feel free to improve them then.

I didn't read every word of your post.  It seems you're having trouble
importing.  Do you still need help with that or have you found the help you
need?

Cheers,

// mark




From pedro_rodriguez at club-internet.fr  Thu Sep 26 12:30:47 2002
From: pedro_rodriguez at club-internet.fr (Pedro RODRIGUEZ)
Date: Thu, 26 Sep 2002 18:30:47 +0200
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <%8ci9.138477$5r1.5541622@bin5.nnrp.aus1.giganews.com> <amltd3$3k$1@glue.ucr.edu> <3D911FE1.1050801@something.invalid>
Message-ID: <amvcev$cf6$1@s1.read.news.oleane.net>

On Wed, 25 Sep 2002 04:30:57 +0200, Greg Ewing wrote:

> thp at cs.ucr.edu wrote:
> 
>> But why use grotesque hacks in the first place?  "a ? b : c" is a
>> well-established syntax for conditional expressions.  Why doesn't
>> Python simply adopt it?
> 
> 
> I believe because Guido (a) thinks it's ugly and (b) doesn't think it
> would add enough to the language to be worth including.
> 
> Personally, I rather like
> 
>     a if b else c
> 
> but Guido doesn't like that either, unfortunately.
> 
> Perhaps I'll slip it into Pyrex when no-one's looking...:-)


Ada has two short-circuits operators : "and then" and "or else" that
do the trick :
    if A and then B then
        ...
    end if;

Looks cleaner to me.

Pedro


From martin at v.loewis.de  Mon Sep 16 01:37:23 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Sep 2002 07:37:23 +0200
Subject: Pickling slotted instances
References: <mailman.1031805375.9030.python-list@python.org> <j4fzwf349u.fsf@informatik.hu-berlin.de> <mailman.1032144560.28257.python-list@python.org>
Message-ID: <m3it16ze98.fsf@mira.informatik.hu-berlin.de>

Andrew McNamara <andrewm at object-craft.com.au> writes:

> I'm really just wondering what makes the solution space large: my
> superficial understanding suggests all the necessary information is
> readily available for serialising the object's state (but I can see
> metaclasses make it complex).

I don't know what issues Guido had in mind, but here are a few for
consideration:

1. How to determine all slots? It has been proposed that __slots__
   should return a tuple that includes base slots as well;
   alternatively, iterate over base types yourself. Alternatively,
   return slots in __dict__.

2. Should all slots be pickled unconditionally? You seem to expect
   this, but Samuele Pedroni points out that, in Java, some slots can
   be marked transient, to avoid pickling them.

3. How to pickle slots? You are suggesting that the list of slot names
   is included, but that appears to be redundant - if you know the type,
   you also know what slots it has.

4. How to pickle unset slots? If a type has a slot but the instance
   has not assigned the slot, requesting the slot value raises an
   AttributeError.

If you are in doubt that this is really as tricky as presented, I'd
encourage you to come up with a patch. In the process of implementing
the patch, you may either find that all issues have a straight-forward
solution, or that there are really serious issues remaining.

Regards,
Martin


From mjais at web.de  Wed Sep  4 05:36:10 2002
From: mjais at web.de (Markus Jais)
Date: Wed, 04 Sep 2002 11:36:10 +0200
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com>
Message-ID: <oukd9.7$4E3.806@news.ecrc.de>

Benjamin wrote:

> Hi guys (and girls)
> I just begun to learn how to program. And I choosed Python for
> doing that. surfing around the web i found an article about this
> language, they said it's the ideal language for beginners to the
> world of programming. and i admit it, the syntax is very clear making
> the code very readable, i learn easily the concepts of programming
> and i soon wrote some nice programs. but since i finished my
> newbie-tutorial, i am wondering about one question: is it worth to go
> deeper into python? i talked to a friend of mine, he told me that i
> should try php..
> what should i do?? ^_^ in other words: what is python good for?
> why should i continue with python?
> thank you!
> 
> _____________________________________________________
> I'm going to prove the impossible really exists


hi
one reason is that with every programming language 
you can learn new concepts.

when I started with perl I learned a lot about string processing
and regular expressions and this knowledge is also useful
with python

with java I learned object oriented programming and the concepts
are also useful for python

with Ruby (my favourite, http://www.rubycentral.com) I learned
more about OOP and other cool stuff

with python (and ruby) I learned to appreciate clean syntax and
how to use OOP with a very dynamic language.

what I want to say:
you should always learn several languages. every one has its advantages
and if you know more you can decide which is best for you problems.

for example I try to use Ruby most of the time
but when there are good modules in python or perl for my problem but
ruby has no such module I swith to python or perl

the pragmatic programmers (http://www.pragmaticprogrammer.com - very good, 
you should visit it) recommend to learn at least one new language every 
year.

so maybe you should learn both PHP and Python (and Ruby, of course :-)   )

markus






From imbosol at vt.edu  Tue Sep 10 17:57:17 2002
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 10 Sep 2002 17:57:17 -0400
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <bd751532.0209082142.54236307@posting.google.com>
Message-ID: <allrrs$6hm$2@solaris.cc.vt.edu>

Brian Slesinsky wrote:
> I've long thought that Python would be improved by an optional "end"
> statement.

Python already has that.  It's called a comment.


-- 
CARL BANKS
http://www.aerojockey.com


From cmay4 at yahoo.com  Tue Sep 24 14:29:00 2002
From: cmay4 at yahoo.com (Chuck May)
Date: Tue, 24 Sep 2002 18:29:00 GMT
Subject: Limit number of concurrent threads
Message-ID: <Xns929393554CAB0cmay4yahoocom@165.112.130.8>

I am writing a database utility that needs to run a particular process 
on every table in the database.  Currently, it is something like this:

    def do_table(table):
        # handle table operations here

    for table in all_tables:
        do_table(table)

    # summarize results

Since some tables take much longer than others, and since the table are 
not in contention with each other, I'd like to thread the operation.  
What I would ideally like is to have a set number of threads, say 10, 
running at the same time (my database has a concurrent connection 
limit).  Since there are about 80 tables, once there are 10 threads 
running I'd like to add another one as soon as one finishes.  Then I 
need to wait at the end until all the threads finish, so that I can 
summarize results.  One other nice feature would be to exit the 
application if any thread fails.

I've searching for examples that fit my problem, and I haven't found 
any.  Most examples I've found look something like this:

    threads = []
    for i in range(number_of_tasks):
        thread.append(thread_for_task)
    
    # let the threads run - just wait for shutdown
    for t in threads:
        t.join() 

which whould work fine if I wanted to start all 80 threads at once.

I've read about the Queue, but I'm not sure if it's what I want, and I 
can't seem to find any good examples on it.

Any help would be appreciated.  Thanks.

-- 
Chuck May
IMS, Inc.



From jepler at unpythonic.net  Tue Sep 17 16:59:40 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 17 Sep 2002 15:59:40 -0500
Subject: subclassing tuple
In-Reply-To: <H2LMGE.1JBF@beaver.cs.washington.edu>
References: <H2LMGE.1JBF@beaver.cs.washington.edu>
Message-ID: <20020917205939.GE18327@unpythonic.net>

On Tue, Sep 17, 2002 at 08:05:02PM +0000, Douglas Zongker wrote:
> Is there any way to create a subclass of 'tuple' that has a customized
> constructor?  I tried doing the obvious thing, but it doesn't work --
> it seems the tuple initializer gets called *before* the one for my
> subclass.  This:

You must override __new__.  See the "docs".

http://www.python.org/2.2/descrintro.html

Jeff



From akakakhel at attbi.com  Sat Sep 14 18:52:11 2002
From: akakakhel at attbi.com (Ali K)
Date: Sat, 14 Sep 2002 22:52:11 GMT
Subject: IDLE
References: <wpbf9.264320$aA.45691@sccrnsc02> <3D7D4F84.318CDB0F@bellatlantic.net>
Message-ID: <v4Pg9.428631$me6.49396@sccrnsc01>

Hardware: 386 enhanced, 75MHz, 32MB RAM.
OS: Win98
Copied files from other computer to mine.
It doesn't open when I click on ide.pyw or when I choose to edit a py file
with it.




From zoltan at trados.com  Tue Sep 17 05:00:49 2002
From: zoltan at trados.com (Zoltan Sekeres)
Date: Tue, 17 Sep 2002 11:00:49 +0200
Subject: More on Protecting Source Code
References: <mailman.1032247760.982.python-list@python.org>
Message-ID: <3D86EF41.5090808@trados.com>

David LeBlanc wrote:
> Well, I think of it this way: machine coded binaries are more like a 128 bit
> key and Python is more like a 40 bit key. I agree that nothing is safe from
> reverse engineering; it's a matter of how much pain and money it takes to
> unscrut the inscrutable. Python could do better - how I'm not totally sure.

Why should do Python better wrt obscurity ? I don't think that's the
intention behind the language. As always: Use the right tool.
(Probably INTERCAL http://www.ugcs.caltech.edu/~jlm/intercal.html is the 
right one for you :-)

If you develop an app you have the copyright. If somebody uses it 
without your approval the law is on your side. But using obscurity to me 
always looks like starting a race which you cannot win. If you need a 
headstart use dongles and something heavily compiled.

The last time I looked SAP R/3 came with the ABAP source code for the 
business logic. And haven't heard that SAP has much of a problem with 
piracy.

-- 
Zoltan



From ponderor at lycos.com  Tue Sep 24 12:22:55 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: Tue, 24 Sep 2002 16:22:55 -0000
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Sep 24)
Message-ID: <up14av386chdda@corp.supernews.com>

QOTW:  "When something "is *EVERYWHERE*", how can it be "clearly VISIBLE"
at the same time?" - Alex Martelli 
    http://mail.python.org/pipermail/python-list/2002-September/123526.html

"... [A]necdotally speaking I find Python development goes faster more as
a result of its very high readability than as a result of low loc/FP."
				 Peter Hansen     


Threads
    Another good summary regarding a Python to C++ conversion
        http://mail.python.org/pipermail/python-list/2002-September/121194.html
    
    Insight into the Python shipped with Mac OS X 10.2 (Jaguar)
        http://mail.python.org/pipermail/python-list/2002-September/122912.html
    
    Rob Andrews requests ideas for Python advantages on Linux, with many
    resources in response
        http://groups.google.com/groups?selm=cBci9.43632%24U7.15670948%40twister.socal.rr.com
    
    Where's shared memory functionality in the standard python distribution?
        http://mail.python.org/pipermail/python-list/2002-September/123431.html
     
    Classifying Python's typing evokes question and discussion of terms
        http://mail.python.org/pipermail/python-list/2002-September/123509.html
    
    Serial port programming tips, including: get a breakout box
        http://mail.python.org/pipermail/python-list/2002-September/123523.html
    
    Mathematical perspecives on Programming, ended in the reality of references
        http://mail.python.org/pipermail/python-list/2002-September/123526.html
    
    Getting started with Internationalization / il8n of a wxWindows app
        http://mail.python.org/pipermail/python-list/2002-September/123656.html
    
    Reviews and perspective on core programming language books
        http://mail.python.org/pipermail/python-list/2002-September/123676.html 
    
    Is Python 2.2.1 compatible with Tcl/Tk 8.4? Yes, but you must recompile
    Python.
        http://mail.python.org/pipermail/python-list/2002-September/123770.html
    
    Discussion regarding Python as a compiled language
        http://mail.python.org/pipermail/python-list/2002-September/123776.html
    
    Mark Hammond details the steps for MS Outlook plug-in, finishing with
    the Python connection
        http://mail.python.org/pipermail/python-list/2002-September/124112.html
    
Misc.
    New Python-XML column  by Uche Ogbuji at xml.com. First article:
        http://www.xml.com/pub/a/2002/09/18/py.html
    
    O'Reilly offers Python Success Stories
        http://mail.python.org/pipermail/python-announce-list/2002-September/001694.html
    
    Lightweight Languages Worshop 2002 (LL2) Saturday, November 9,
    2002 - Cambridge, MA, USA
        http://ll2.ai.mit.edu/
    
    3-day Python training session with Mark Lutz in Colorado, on October 28-30
        http://www.rmi.net/~lutz/oct02-longmont-class.html
    
    Brett Cannon's Python-dev summary for 2002-09-01 to 2002-09-15 
        http://mail.python.org/pipermail/python-announce-list/2002-September/001695.html
    
    Call for Participation: UK Python Conference 2003, April 2-3 - Oxford.
        http://mail.python.org/pipermail/python-announce-list/2002-September/001692.html
    
Software
    eGenix.com's mx Python extension packages mxODBC 2.0.5 and BASE 2.0.4
    versions released.
        http://www.egenix.com 
    
    CilentForm HTML forms handler 0.0.2a alpha released
        http://mail.python.org/pipermail/python-announce-list/2002-September/001705.html
    
    PyReverse 0.4 reverse code generator (plus) released including bug
    fixes, pyvcg support and improved documentation
        http://mail.python.org/pipermail/python-announce-list/2002-September/001702.html
    
    RuneBlade 0.3.2 release includes updates for wxPython 2.3.3.1 and
    2.3.2.1, aspect oriented support code, and optimizations and bug fixes
        http://mail.python.org/pipermail/python-announce-list/2002-September/001701.html
    
    wxPython 2.3.3.1 released - Unicode and many other improvements for
    this wxWindows wrapper
        http://mail.python.org/pipermail/python-announce-list/2002-September/001700.html
    
    Gordon McMillan releases version 5b4 of Installer
        http://mail.python.org/pipermail/python-announce-list/2002-September/001699.html
    
    The bugfix-release, 0.5.1, of pyOpenSSL is out, and introduces basic
    PKCS12 support
        http://mail.python.org/pipermail/python-announce-list/2002-September/001693.html
    
    Version 0.8.1 of the Python/XML distribution is now available, and
    should be considered a beta release
        http://mail.python.org/pipermail/python-announce-list/2002-September/001691.html
    
    DocUtil's gets a DocFactory: a wxPython based "kind of integrated
    publishing environment for documentation"
        http://mail.python.org/pipermail/python-announce-list/2002-September/001690.html
    
    pyblosxom is a weblogging tool written in Python and modelled after blosxom.
        http://mail.python.org/pipermail/python-announce-list/2002-September/001689.html
    
    FontTools/TTX 2.0 beta 1, an Open Source Python library and toolkit to
    manipulate fonts.
        http://mail.python.org/pipermail/python-announce-list/2002-September/001688.html
    
    
========================================================================

Everything you want is probably one or two clicks away in these pages:

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
         http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Brett Cannon continues the marvelous tradition established by 
    Andrew Kuchling and Michael Hudson of summarizing action on the
    python-dev mailing list once every other week.
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

    Much of Python's real work takes place on Special-Interest Group
    mailing lists
        http://www.python.org/sigs/

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
        http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
        http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
        http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

    Links2Go is a new semi-automated link collection; it's impressive
    what AI can generate
        http://www.links2go.com/search?search=python

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning
(approximately), ask <claird at phaseit.net> to subscribe.  Mention
"Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.



From h_schneider at marketmix.com  Sun Sep 29 08:20:50 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Sun, 29 Sep 2002 14:20:50 +0200
Subject: SOLVED: _winreg.DeleteValue - Permission denied ...
References: <mailman.1033133498.15589.python-list@python.org>
Message-ID: <an6r1c$hnp$04$1@news.t-online.com>

Thanks a lot Dave! It was right before my eyes ...

Bye,
Harald

> Perhaps you need to open the registry key with write permissions, e.g.:
>
> key = _winreg.OpenKey(_winreg.HKEY_LOCAL_MACHINE, \
>                       'SOFTWARE\\TEST\\SUBTEST', \
>                       0, _winreg.KEY_ALL_ACCESS)
>
> (Also note that you should use double backslashes in your strings or use
> raw strings)
>
> HTH,
> Dave
>
>




From henrik.motakef at web.de  Thu Sep 26 16:59:19 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 26 Sep 2002 22:59:19 +0200
Subject: How do you pronounce "__" (double underscore)?
References: <525eaba.0209260956.72bf6786@posting.google.com> <20020926164641.1753ebbb.ckasso@sprynet.com>
Message-ID: <873crwfowo.fsf@pokey.henrik-motakef.de>

Chris Kassopulo <ckasso at sprynet.com> writes:

> How about:
> 
>  _  = "un"
>  __ = "dun"

"dun-init-dun"? Sounds too much like "done" for me... :-)


From quiteblack at yahoo.com  Tue Sep 24 03:52:33 2002
From: quiteblack at yahoo.com (black)
Date: Tue, 24 Sep 2002 00:52:33 -0700 (PDT)
Subject: how to build up a guest book with python ?
Message-ID: <20020924075233.91441.qmail@web21306.mail.yahoo.com>


Hi,all~

    I'm learning python for almost 1 month and decide to build up a guest book with it, but I really know nothing about web programming, any tutorials for me please ?



---------------------------------
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020924/02215c97/attachment.html>

From bobx at linuxmail.org  Wed Sep 11 22:17:40 2002
From: bobx at linuxmail.org (Bob X)
Date: Thu, 12 Sep 2002 02:17:40 GMT
Subject: [FEEDBACK] Is this script efficient...is there a better way?
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <L9Rf9.168314$l_4.127784@atlpnn01.usenetserver.com> <4ZRf9.1782$Lg2.248733@news2.news.adelphia.net> <3hSf9.168947$l_4.98525@atlpnn01.usenetserver.com>
Message-ID: <8PSf9.1796$Lg2.257378@news2.news.adelphia.net>

Steve Holden wrote:
> "Bob X" <bobx at linuxmail.org> wrote in message
> news:4ZRf9.1782$Lg2.248733 at news2.news.adelphia.net...
> [ ... ]
> 
>>Thank you very much...as a follow up:
>>
>>#1 Can I use the same sys.exit("message here") on the IO errors as well?
> 
> 
> Sure. I assumed you would generalize the first example to the rest ...
> 
> 
>>#2 For the "print looks like a debugging statement"...how would it be
>>done different...in the context of my script and not a total re-write?
> 
> 
> I simply thought that since you were writing the output to the outFile
> anyway it seems a bit redundant to be printing it. Plus the readlines()
> includes the line ending, so you'd get double-spaced output fom the print
> statements anyway.
> 
> 
>>#3 I will test your shorter one as I have stayed away from RE till now.
> 
> 
> Well, obviously there may be problems if the keywords contains any
> characters special to regular expressions. Nut my example whould wqork for
> testing purposes.
> 
> 
>>#4 On your shorter one should xreadlines still be done for memory sake?
> 
> 
> Yes. This is a general improvement.
> 
> regards
> -----------------------------------------------------------------------
> Steve Holden                                  http://www.holdenweb.com/
> Python Web Programming                 http://pydish.holdenweb.com/pwp/
> Previous .sig file retired to                    www.homeforoldsigs.com
> -----------------------------------------------------------------------
> 
> 
> 
Thank you very much...your shorter script did work. The keywords are all 
"normal" english words that would not have any special characters. I 
only added a re.IGNORECASE (since case doesn't matter to me) after a 
quick search of the python docs.

Domo!

Bob



From Omri at atavox.spam  Tue Sep 10 04:58:48 2002
From: Omri at atavox.spam (Omri at atavox.spam)
Date: Tue, 10 Sep 2002 08:58:48 +0000 (UTC)
Subject: Howto debug self?
References: <alhivr$fom$04$1@news.t-online.com>
Message-ID: <alkc87$2ar$1@reader-00.news.insnet.cw.net>

Do you mean something like:

(Pdb) print self.__dict__
{'a': 5, 'c': 16.6999999999999, 'b': 'good'}

?



From member at dbforums.com  Mon Sep 16 01:52:34 2002
From: member at dbforums.com (hpyhpy)
Date: Mon, 16 Sep 2002 05:52:34 +0000
Subject: A question about C/API of python.
Message-ID: <1818979.1032155554@dbforums.com>

I use python to load C module(.so),then use C load python,like this:
C(main)--->python--->C(.so)
But now my trouble is :
In C(main), I have some Global variable that depend on process of
C(.so).
But because python is in middle of them,so I can't use variable of
C(.so) in C(main) dexterously.
I see some C/API reference but it still pluzzle me.Anybody can tell me
some advise?
Thanks.

--
Posted via http://dbforums.com



From paulmg at digitalbrain.com  Tue Sep 24 04:52:20 2002
From: paulmg at digitalbrain.com (Paul MG)
Date: 24 Sep 2002 01:52:20 -0700
Subject: instance creation
Message-ID: <265e380f.0209240052.4c2988f3@posting.google.com>

Hi

I am dabbling with Python and have run into a bit of an oddness. I
have a class Appointment, with fields Title and Details. I want to
allow two means of construction:

1) Providing Title and Details strings explicitly, to create a new
instance.
2) Providing an ID, which causes that Appointment to be loaded from a
backing store.

Now in C++ or smalltalk, I would do one of

- Provide two constructors:

    def __init__(self, name, details):
      self.name = name
      self.details = details

    def __init__(self, id):
      load(id)

    def load(self, id):
      i = open("appointment"+str(id))
      self = pickle.load(i)

- Provide a class method (creation method) to instantiate:

    def create(id):
      i = open("appointment"+str(id))
      return pickle.load(i)

However, much of this seems not to work.

- Having two constructors seems to be illegal - or at least the second
overwrites the first.
- Assigning to 'self' in the load method seems to really not work!
- Putting the create() method outside of the class works fine - but I
want it to be a class method on Appointment, goddammit!:) Is there no
way to do this?

Any help would be appreciated, basically i am looking for the python
idiom to do this, as my C++ and Smalltalk idiom's clearly aren't in
the grain of the language!

Thanks very much
Paul MG


From whisper at oz.net  Tue Sep 24 22:45:30 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 24 Sep 2002 19:45:30 -0700
Subject: PythonWin CVS problem
Message-ID: <GCEDKONBLEFPPADDJCOEIENFFCAA.whisper@oz.net>

Second day it's inaccessable with "permission denied"

David LeBlanc
Seattle, WA USA 



From skip at pobox.com  Tue Sep  3 10:02:05 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 3 Sep 2002 09:02:05 -0500
Subject: How to avoid freeing memory on exit?
In-Reply-To: <31575A892FF6D1118F5800600846864DCBD710@intrepid>
References: <31575A892FF6D1118F5800600846864DCBD710@intrepid>
Message-ID: <15732.49373.250008.442195@12-248-11-90.client.attbi.com>

    Roy> Is there any way to get the interpreter to just exit without doing
    Roy> all that memory cleanup?

    Gerhard> os._exit

    Gumuz> so what does 'one' underscore prefixing a methodname mean?
 
    Simon> Proceed with caution.

In particular, in this case it means that Roy should make sure he *really*
doesn't care if __del__ methods associated with those 200,000 objects don't
get called.

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From Andreas.Leitgeb at siemens.at  Thu Sep 19 06:55:44 2002
From: Andreas.Leitgeb at siemens.at (Andreas Leitgeb)
Date: Thu, 19 Sep 2002 10:55:44 +0000 (UTC)
Subject: lists in dictonaries ?
References: <mailman.1032426629.1933.python-list@python.org>
Message-ID: <slrnaojc79.c05.Andreas.Leitgeb@pc7499.gud.siemens.at>

Andrew Thompson <andrew.thompson at ashecastle.com> wrote:
> You can put any python objects into a python dictionary, even the
> dictionary itself...

Oh, can I create a dictionary in Python that contains all those
dictionaries that don't contain themselves ... ?-)

*scnr*'ly yours.


From gvdbogae at vub.ac.be  Tue Sep  3 11:41:02 2002
From: gvdbogae at vub.ac.be (Geert-Jan Van den Bogaerde)
Date: 03 Sep 2002 17:41:02 +0200
Subject: Design question.... parent/child class
In-Reply-To: <3d74b021$0$234$4d4ebb8e@news.nl.uu.net>
References: <3d74b021$0$234$4d4ebb8e@news.nl.uu.net>
Message-ID: <1031067663.914.137.camel@gandalf>

On Tue, 2002-09-03 at 14:56, Gumuz wrote:
> hi all,
> 
> i have a SessionManager class, which contains a list of Session
> objects(instantiated from a Session class). I have a design problem.
> 
> Each session object has a small 'queue' of messages for the specific
> session. How can a session send a message to another session without having
> some kind of parent-reference to the SessionManager object?
> 
> i am a bit puzzled, or is this parent-reference not such a bad idea after
> all?
> 

There have been other replies which might be more suited to your
situation, but if you want a very loose coupling I've found using a
signal dispatching system very effective. There is a python
implementation here:
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/87056/

This is similar to Qt's signal/slot system, if you're familiar with
that. You have your session objects send a signal whenevever some
interesting event happens, and your other objects listening for the
signals they're interested in.

Best Regards,

Geert-Jan Van den Bogaerde




From thorsten at thorstenkampe.de  Fri Sep 20 03:49:58 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Fri, 20 Sep 2002 09:49:58 +0200
Subject: Avoiding referencing (was: Python sequences reference - problem?)
References: <mailman.1032443760.5578.python-list@python.org> <YZli9.114624$ub2.2439313@news1.tin.it>
Message-ID: <amejvb$55qsi$1@ID-77524.news.dfncis.de>

* Alex Martelli
> Just remembed that EVERYTHING works by reference, and what
> could possibly be the problem...?

The problem is that we newbies sometimes don't want referencing:

#v+
def part(seq, indices):
    """ chop seq by indices """
    partition = []
    
    for slice in indices:
        partition.append(seq[:slice])
        del seq[:slice]
    partition.append(seq)
    return partition
#v-

>>> foo = [11, 22, 33, 44, 55, 66, 77]
>>> part(foo, [2, 3])
[[11, 22], [33, 44, 55], [66, 77]]  # yes
>>> foo
[66, 77]                            # aaarrrghhh

Obviously I did not pass 'foo' "by value" to function 'part' but by 
reference and obviously not only local variable 'seq' was modified in 
the function but "outer" variable 'foo', too.

So how to deal with that?

1. "Just don't call function 'part' by with a variable, call it by 
value - part([11, 22, 33, 44, 55, 66, 77], [2, 3]) - and everything's 
okay".

Obviously no solution to me.

2. "If you don't want referencing, call it with a copy of 'foo' like: 
part(foo[:], [2, 3])"

Yes, but how do I know in advance that 'part' modifies 'foo'? I might 
not be the author of 'part'.

3. "Rewrite function 'part' so that it works with a tuple as argument. 
Thereby you're avoiding all possible modifying of existing objects."

Yes, but I'm also losing the "extended" possibilities of lists: 
"list.append", "list.reverse", "list.index" etc. My code will be twice 
as long as it would be with list methods.


Is there a way to "bypass" referencing _without_ disadvantages? I'm 
sure there is a "canonical" way to do this in Python!

Thorsten


From danb_83 at yahoo.com  Fri Sep 20 15:28:44 2002
From: danb_83 at yahoo.com (Dan Bishop)
Date: 20 Sep 2002 12:28:44 -0700
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amabpb$5rm$1@news.nems.noaa.gov> <3D8B1537.4010803@thinkware.se>
Message-ID: <ad052e5c.0209201128.1a53108@posting.google.com>

Magnus Lyck? <magnus at thinkware.se> wrote in message news:<3D8B1537.4010803 at thinkware.se>...
> ChrisBarker wrote:
> > The only reason I can see
> > to not writing your class in Python is performance, and unless you are 
> > doing a lot of calculations, that is probably a non-issue.
> 
> Performance issues might well matter in large financial
> systems, but that's not the only reason.

Please name one other real-world situation in which decimal arithmetic
is actually *needed*.

> [snip]
> And if someone would write
> 
> x = Decimal(9.999999999999999) - Decimal(9.9999)
> 
> they would not get what they expected (unless they
> expected 0.000099999999998). Ok, they could write

But how often does an error of 10 parts per trillion really matter?

Afaict, not nearly enough to justify having a default number
representation that works against the hardware.

> [major snippage]

> But perhaps a command line argument or something like
> "import exact_arithmetic" could change the behaviour to
> using exact instead of floating point arithmetic. I guess
> we'd need a module anyway, to specify things. Like:
> 
>  >>> import exact_arithmetics as exact
>  >>> exact.default_rounding(exact.UP)
>  >>> exact.default_fractions(2)
>  >>> exact.max_precision = 20
>  >>> 7.5/20
>  0.38
>  >>>

In your proposal, is 1/3 exact?  How about sqrt(2)?

What's so important about terminating decimal fractions that they
should get special treatment?  It's true that they're useful for
currency calculations, but isn't a Money class enough?


From teaandbikkie at aol.com  Fri Sep 20 07:16:02 2002
From: teaandbikkie at aol.com (TeaAndBikkie)
Date: 20 Sep 2002 11:16:02 GMT
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com>
Message-ID: <20020920071602.29041.00000016@mb-bg.aol.com>

>From: Benjamin phncontact.libero.it@ 

>but now i don't know
>what to programm. sounds weird, doesn't it? :) what
>should i program??

For a fun programming experience, I suggest a collaboration with other
programmers, its more fun than on your own and you learn different things. Open
source programming can be very rewarding.

You could look on sourceforge <www.sourceforge.net> for a project to join, or
for ideas, and you can limit the selection to those using Python.

Kind regards,
Misha


From jhorneman at pobox.com  Sun Sep 29 13:36:39 2002
From: jhorneman at pobox.com (Jurie Horneman)
Date: Sun, 29 Sep 2002 19:36:39 +0200
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com>
Message-ID: <3d9739c7$0$19818$91cee783@newsreader01.highway.telekom.at>

"Jurie Horneman" <jhorneman at pobox.com> wrote in message
news:9ac02e81.0209290839.4f4d6475 at posting.google.com...
> rant

My thanks to both of the people who have ignored my ranting and tried to
help me. I believe I did what you suggested, but it didn't work, hence my
frustration.

Right after posting my original problem, I completely removed and
reinstalled Python. And hey, now it works, with solution 4, ie packages as
they are described in the documentation. I don't know what I did to my
installation to make it do that, but my problem appears to be solved.

For what it's worth - I've already been called on this once in a private
email: although my post was meant to illustrate my frustration and criticism
of the Python docs, it was also meant to be at least partly
self-deprecatory. I should have known better: not because nobody out there
gets that, but because text, and especially text sent out to millions of
people, just is not a good place for irony - especially when one is asking
for help. So if you felt offended: sorry, that wasn't what I intended.

However, I do stand by my criticisms of the Python docs.

Jurie.





From aleax at aleax.it  Wed Sep 18 08:45:41 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 18 Sep 2002 12:45:41 GMT
Subject: minimizing Tk programs to the taskbar
References: <3D8872F1.87850856@optushome.com.au>
Message-ID: <Vz_h9.126208$pX1.4528221@news2.tin.it>

Richard Tardif wrote:

> Is there any way to minimize programs to the taskbar? I'm talking about
> windows here. By taskbar I mean the area down in the bottom right
> corner, near the clock, that shows icons for what proggies you've got
> running. Thanks for any help.

I don't think Tkinter offers such an intrinsically non-cross-platform 
functionality. I suspect therefore that you'd need to work this part 
through the win32all add-ons, win32gui.Shell_NotifyIcon &tc.


Alex



From itamar at itamarst.org_NOSPAM  Sun Sep  1 17:15:50 2002
From: itamar at itamarst.org_NOSPAM (Itamar Shtull-Trauring)
Date: Sun, 01 Sep 2002 17:15:50 -0400
Subject: Twisted docs (was Re: Python IPC options)
References: <1030829572.943495@irys.nyx.net> <3D715160.2020109@itamarst.org_NOSPAM> <cb035744.0209011308.7b4b3ae2@posting.google.com>
Message-ID: <3D728386.2060908@itamarst.org_NOSPAM>

Will Stuyvesant wrote:

> Please stop coding and working on functionality until you have good
> docs.  You have a doc explaining the difference between a framework
> and a library, that is good.  More of that!  About daemons, servers,
> internet protocols, tutorials, whatever.

We don't have enough documentation, that's true.

But see http://www.twistedmatrix.com/documents/dev - I count 14 docs there 
that have immediately useful info. They cover writing servers, clients, 
threading, proceses, web developlement, scheduled execution, database 
integration and more.

> Open source is great but weak without good docs.

That is very true. I take it you're going to volunteer some time to write 
docs? We'd really appreciate it! ;)

-- 
Itamar Shtull-Trauring    http://itamarst.org/
Available for Python, Twisted, Zope and Java consulting



From DeepBleu at DeepBleu.org  Fri Sep 20 19:15:49 2002
From: DeepBleu at DeepBleu.org (DeepBleu)
Date: Fri, 20 Sep 2002 23:15:49 GMT
Subject: How to upload a file using http protocol
References: <mailman.1032552637.6156.python-list@python.org>
Message-ID: <F_Ni9.284613$eK6.8277278@twister.austin.rr.com>

Yes it is.  Assume that the user has to enter the file name (with the path)
on a form.  You use File type form element which allows the user to browse
for the file.
Using CGI module, and a Submit button (or some javascript with a submit
function) this form element has a KEY.  The VALUE of this key is the Byte
content of the file, NOT the file name.  Thus, by using the VALUE KEY, the
server is actually receiving the byte content.  From there on, the sky is
the limit.
DeepBleu

"A" <export at hope.cz> wrote in message
news:mailman.1032552637.6156.python-list at python.org...
> Hi,
> Does anyone know how to upload a file through HTTP protocol?Is that a
possible?
> Thank you for help.
> ladislav
>
>
>




From grante at visi.com  Thu Sep 26 17:27:17 2002
From: grante at visi.com (Grant Edwards)
Date: Thu, 26 Sep 2002 21:27:17 GMT
Subject: How do you pronounce "__" (double underscore)?
References: <525eaba.0209260956.72bf6786@posting.google.com>
Message-ID: <VYKk9.1167$w7.1497161@ruti.visi.com>

In article <525eaba.0209260956.72bf6786 at posting.google.com>, Pat Notz wrote:
> @ = "at"
> ! = "bang"
> . = "dot"
> * = "star"
> 
> Is there a short (one or two syllable) pronounciation for "__" (double
> underscore).  "Underscore underscore" and "double underscore" are kind
> of long.  Somthing like "splat"?  Anyone?

I've heard "splat" used for both '#' and '*', so I'd cross it
off the list...

-- 
Grant Edwards                   grante             Yow!  Sign my PETITION.
                                  at               
                               visi.com            


From lbrannma at cablespeed.com  Sat Sep 28 13:25:46 2002
From: lbrannma at cablespeed.com (Lance)
Date: Sat, 28 Sep 2002 10:25:46 -0700
Subject: How to persist IDLE Window state?
Message-ID: <upbpdql5vskib9@corp.supernews.com>

Hi,

I'd like to have the IDLE window state persist across IDLE sessions. Has
anyone done this?

I'm new to Python and open source. It looks like I might be able to edit one
of the idle.py files, rather than changing C code. Any suggestions would be
appreciated.

Thanks,
Lance




From gerhard.haering at opus-gmbh.net  Wed Sep  4 05:47:02 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 4 Sep 2002 09:47:02 GMT
Subject: using python to call java code??
References: <3cc90cd2.0209040122.17242a04@posting.google.com>
Message-ID: <slrnanbli0.1rs.gerhard.haering@haering.opus-gmbh.net>

Tim Schulz wrote:
> I am considering to use zope in order to develop an content management
> portal. My problem is, that I have to integrate some ready-built java
> code. Is there any possibility to call java code from python? Is there
> a API for this kind problem?

http://jpe.sourceforge.net/

I'm not sure if that's stable yet. I also don't know if I'd bet a project
on it, but probably not. At least not without some thorough testing.

-- Gerhard


From anthony at computronix.com  Thu Sep 12 12:06:59 2002
From: anthony at computronix.com (Anthony Tuininga)
Date: 12 Sep 2002 10:06:59 -0600
Subject: How to upload files using mod_python?
In-Reply-To: <1031845024.2957.154.camel@coraje>
References: <1031845024.2957.154.camel@coraje>
Message-ID: <1031846819.472.26.camel@chl0151.edmonton.computronix.com>

You may wish to report this to the mod_python developers, as __methods__
was removed in Python 2.2, right?

On Thu, 2002-09-12 at 09:37, Andres Meza wrote:
> Greetings. 
> 
> I have created a python CGI script that can upload files sucessfully.
> Now, I want to create a mod_python script with the same feature using
> the "publisher" handler and I get this error:   
> 
> 
> [Thu Sep 12 08:53:41 2002] [error] PythonHandler mod_python.publisher:
> Traceback (most recent call last): 
> [Thu Sep 12 08:53:41 2002] [error] PythonHandler mod_python.publisher:
> File
> "/home/mrnet/rpm/tmp/mod_python-root/usr/lib/python2.2/site-packages/mod
> _python/apache.py", line 193, in Dispatch 
> [Thu Sep 12 08:53:41 2002] [error] PythonHandler mod_python.publisher:
> File
> "/home/mrnet/rpm/tmp/mod_python-root/usr/lib/python2.2/site-packages/mod
> _python/publisher.py", line 104, in handler 
> [Thu Sep 12 08:53:41 2002] [error] PythonHandler mod_python.publisher:
> File
> "/home/mrnet/rpm/tmp/mod_python-root/usr/lib/python2.2/site-packages/mod
> _python/publisher.py", line 308, in __init__ 
> [Thu Sep 12 08:53:41 2002] [error] PythonHandler mod_python.publisher:
> AttributeError: 'file' object has no attribute '__methods__' 
> 
> Anyone have an idea about why I get this error? 
> 
> Thanks for your time, 
> 
> -- 
> 
>    Andr?s Meza
> 
>    Technical Manager
> 
>  __Nuevos Medios - NM S.A.____________________
> 
>    e-mail:         andres.meza at nuevosmedios.ws
> 
>    day time phone: (+572)6827794 ext.102
> 
>    url:            www.nuevosmedios.ws
-- 
Anthony Tuininga
anthony at computronix.com
 
Computronix
Distinctive Software. Real People.
Suite 200, 10216 - 124 Street NW
Edmonton, AB, Canada  T5N 4A3
Phone:	(780) 454-3700
Fax:	(780) 454-3838
http://www.computronix.com




From rsrchstr at msn.com  Sun Sep 29 20:38:51 2002
From: rsrchstr at msn.com (mike henley)
Date: Mon, 30 Sep 2002 01:38:51 +0100
Subject: idea for a much needed application
References: <576c1752.0209291326.56dcdf31@posting.google.com> <pasepuc0r73mh5vlcsu1l7erl4e2b17qvi@4ax.com>
Message-ID: <an86f7$c2veb$1@ID-157863.news.dfncis.de>

"TGOS" <tgos at invalid.invalid> wrote in message
news:pasepuc0r73mh5vlcsu1l7erl4e2b17qvi at 4ax.com...
> >
> Use an IRC bot on your system, that does the following after each
> dial-up:
>
> - get your new IP address
> - make a connection to a specific IRC server
> - authenticate you as a specific user
>   (and obtain possession of a specific nickname if currently taken by
>   someone else)
> - enter a specific channel
> - authenticate as channel operator to gain control over the channel
>   (remove operator status of all useres that are currently operators
>   of the channel)
> - regularly announce your current IP address as HTTP link on the
>   channel, e.g.:
>   http://1.2.3.4/
>
> Users that know the IRC network and channel can go there and wait in
> your channel till you log on (they can tell their IRC client to notify
> them if you enter the channel or play a certain sound or perform some
> other actions, like executing a specific script on their PC).

that's the whole idea of giving a personal webserver "presence", so that
others don't have to "go there and wait in my channel till i log on". Also,
i want to be able to give a simple AOL or MSN username to anyone without
assuming they can write specific scripts.

> Once you
> log on your IP address will appear in the channel in regular intervals
> (e.g. once every 5 minutes or so), as well as it appears every time a
> new user is joining the channel as welcome message. Users then just have
> to click on it in their IRC client and their default browser will open
> and take them to your page.
>
> You will need to select an IRC network where people can register their
> nicknames and a channels, so nobody can prevent your bot from working
> correctly by gaining access to your nickname or channel and to make sure
> your channel stays alive, even if no user is lurking inside and you are
> not online (your bot isn't there either, the channel is empty).
>
too messy. IRC is a messy medium.

>
> Problem solved with technology around already since 1988.
>

yes, this is 2002, almost 2003, it should be possible to do it in a much
easier and automated way without all the 1988 mess above, even assuming i
can do it, i do not wish to make it as difficult as it sounds above for
others. I'll be looking at jabber, xml-rpc, and relevant technologies. if
anyone can suggest anything, please let me know.




From quiteblack at yahoo.com  Tue Sep  3 02:55:06 2002
From: quiteblack at yahoo.com (black)
Date: 2 Sep 2002 23:55:06 -0700
Subject: The Module Search Path is ???
Message-ID: <7153651c.0209022255.7fd50ba0@posting.google.com>

tutorial told me I could found it in the folder:/usr/local/lib/python.
but my OS is windows so folders are different. I searched the folder
"lib" but find no file named "python", any ideas ?


From kccw at well.ox.ac.uk  Fri Sep 13 04:19:20 2002
From: kccw at well.ox.ac.uk (Kenny)
Date: Fri, 13 Sep 2002 09:19:20 +0100
Subject: Practical Python Re: [newbie] Buying an introductory/reference text
In-Reply-To: <a38d43b0.0209121413.1bf274c9@posting.google.com>
References: <Mu5g9.1702$U_.1021@news04.bloor.is.net.cable.rogers.com>
 <a38d43b0.0209121413.1bf274c9@posting.google.com>
Message-ID: <Pine.LNX.4.43.0209130914180.28386-100000@zeon.well.ox.ac.uk>

Hello,
	Just wondered has anyone taken a look at Magnus Hetland's book
Pratical Python. I sort of saw a description while randomly surfing
through the book sites.

Kenny



From leafnode-usenet at schabi.de  Thu Sep 19 16:20:35 2002
From: leafnode-usenet at schabi.de (Markus Schaber)
Date: Thu, 19 Sep 2002 22:20:35 +0200
Subject: Remarks to Python 2.3
References: <20020918222421.328088eb.use-net@schabi.de>
 <amannn$p4b$1@panix1.panix.com>
 <m3u1knj8j9.fsf@mira.informatik.hu-berlin.de>
 <20020919152834.592eaa39.leafnode-usenet@schabi.de>
 <j4r8fqq6nf.fsf@informatik.hu-berlin.de>
Message-ID: <20020919222035.185daa67.leafnode-usenet@schabi.de>

Hello,

loewis at informatik.hu-berlin.de (Martin v. L?wis) schrieb:

> There is always the potential danger that a newsgroup message gets
> ignored, and the poster gets the impression that nobody cares about
> the (potentially significant) problem.

I know this and I'm pretty shure I'd take other steps in this case (like
PM to some gurus or using the bug tracker).

> Of course, bug reports are only the right thing if the poster feels
> that Python behaves "bad", where "bad" means either incorrect,
> non-intuitive, or in an otherwise undesirable way. For a bug report,
> one should also research whether this is a well-known issue.

And my original post was my first step in research whether this is a
well known issue and whether it is considered "bad" by others. I learned
that one of my points was already corrected, and simply is a
misdocumentation - so I'll try do submit a documentation correction
patch.

Gru?,
Markus


From dwelch at nospam.vcd.hp.com  Fri Sep  6 18:38:30 2002
From: dwelch at nospam.vcd.hp.com (djw)
Date: Fri, 06 Sep 2002 15:38:30 -0700
Subject: Larry Wall's comment on python...
References: <mailman.1031349792.2374.python-list@python.org>
Message-ID: <3D792E66.4080304@nospam.vcd.hp.com>

[snip]
> structures are not completely analogous, they BOTH burden the programmer
> with arbitrary rules to govern the physical layout of the code. Although
[snip]

I completely disagree that Python's rules for code layout are arbitrary.

D



From marklists at mceahern.com  Tue Sep 24 18:22:02 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 24 Sep 2002 17:22:02 -0500
Subject: List/Tuple bug or feature?
In-Reply-To: <8765wv6o0d.fsf@nospam.eml.cc>
Message-ID: <JHEOKEOOLIGLDHCMAHMOMEDNEBAA.marklists@mceahern.com>

[Mahesh Padmanabhan]
> I spent a lot of frustrating hours tracking this down and I am
> confused as to why I see this behavior.
>
> >>> x = ('a' 'b')
> >>> x
> 'ab'
>
> >>> x = ['a' 'b']
> >>> x
> ['ab']
>
> As you might have guessed, I missed a comma between two elements in
> a fairly large list.
>
> If this is a concatenation feature, how can I prevent something like
> this happening in the future ?

This will sound flippant, but I don't really mean it that way:

	Put a comma in there next time.

You don't really say much about how you generated this large list, but one
idea would be to automate the process of generating it and/or fix whatever
bug led to the omission of the comma.

I don't have anything more clever than that to offer; perhaps someone else
does.

Cheers,

// m




From david at no.westcontrol.spam.com  Wed Sep 11 06:26:44 2002
From: david at no.westcontrol.spam.com (David Brown)
Date: Wed, 11 Sep 2002 12:26:44 +0200
Subject: function pointer
References: <ffcedbc9.0209110215.644c1788@posting.google.com>
Message-ID: <aln5ce$5kt$1@news.netpower.no>

"Christina" <christina at iaeste.at> wrote in message
news:ffcedbc9.0209110215.644c1788 at posting.google.com...
> Hi!
>
> Does Python have function pointers? If not, is there a possibility to
emulate them?
>
> Christina

Functions are first-order objects (is that the right term?) in Python, so
you can just assign them to a variable.  For example:

def foo(x):
    print x

bar = foo

bar(10)






From irmen at NOSPAMREMOVETHISxs4all.nl  Mon Sep 16 15:11:37 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Mon, 16 Sep 2002 21:11:37 +0200
Subject: pyro & parallel processing
References: <rLOb1LAK5fh9EwRi@jessikat.demon.co.uk> <uy9a1ritm.fsf@online.no> <am53og$hdl$1@news1.xs4all.nl> <IWzDioA8Sih9Ewla@jessikat.demon.co.uk>
Message-ID: <am5ad9$2dh$1@news1.xs4all.nl>

Robin Becker wrote:

> The master can presumably pass the new global cutoff to the slaves, but
> I'm not sure yet how interruptible I can make them, so I guess my second
> question is can we have asynchronous calls into the workers or do I need
> to have multiple published classes with a common thread protected state?

You almost certainly want to use Pyro's oneway call mechanism.
This is very handy to use with callback objects, just read the
Pyro manual on this subject.

Irmen de Jong



From nas at python.ca  Sat Sep  7 17:51:20 2002
From: nas at python.ca (Neil Schemenauer)
Date: Sat, 7 Sep 2002 14:51:20 -0700
Subject: Daemonize?
In-Reply-To: <20020907201022.GA93398@hishome.net>
References: <F7DB8D13DB61D511B6FF00B0D0F06233022B66EC@mail.hebmex.com> <mailman.1031184848.32071.python-list@python.org> <unjt47pc5r8671@corp.supernews.com> <20020907161037.GA70503@hishome.net> <20020907174957.GA25058@glacier.arctrix.com> <20020907201022.GA93398@hishome.net>
Message-ID: <20020907215120.GA25431@glacier.arctrix.com>

Oren Tirosh wrote:
> Any particular reason why you think this is better?

One less system call. :-)

  Neil



From jbublitzNO at SPAMnwinternet.com  Sun Sep  8 01:02:38 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Sun, 08 Sep 2002 05:02:38 GMT
Subject: PyQT picture display
References: <albp27030dq@enews4.newsguy.com> <3D799F42.2020903@SPAMnwinternet.com> <alebe2021ft@enews4.newsguy.com>
Message-ID: <3D7AD8EF.9050306@SPAMnwinternet.com>

Jim wrote:
> Jim wrote:

>>As far as I can tell, nothing's wrong. It works fine
>>for me. Are you sure the jpg files actually exist and
>>the path to them is correct? The only thing I changed
>>in your program was the filenames for the image files.

>>I used PyQt-3.3.2 with Qt 2.3.0 on Linux.

>>(another) Jim

> The jpg files are in the same directory as the program. I'm using Qt 3.0.5 
> with PyQt 3.3.2 on Linux. I'm sure that JPEG support is compiled in the Qt 
> library as I got this to work using C++. Does JPEG support need to be built 
> in when compiling PyQt?  if so, what must I do to accomplish this? Thanks,

As far as I know, PyQt has no options for enabling or
disabling jpeg support - the jpeg handling appears to
be done at the C++ level, so if Qt has jpeg enabled,
that should be sufficient.

I also tested your program on a different machine
running Qt 3.03 - works fine there also. I have no
idea why you're having problems. Some of the things
you might try are full explicit paths to the jpg files,
different jpg files, make sure the jpg files you're
referencing are actually readable. Do you have
multiple versions of Qt running? If so, be sure that
Qt 3.0.5 libs are being found by PyQt at run time.
Run ldconfig just to be sure. Do the PyQt example
programs (in PyQt-3.3.2/examples3) work? Delete any
older PyQt files and re-run 'make install'.

Sorry I can't be of more help.

Jim



From joconnor at cybermesa.com  Wed Sep  4 20:43:06 2002
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Thu, 05 Sep 2002 07:43:06 +0700
Subject: Privacy and Inheritance
References: <mailman.1031112136.4594.python-list@python.org>
Message-ID: <20020905.074306.188213258.2038@cybermesa.com>

In article <mailman.1031112136.4594.python-list at python.org>, "Dennis B."
<dcswest at flashmail.com> wrote:

> Greetings;
> 
> Have enough confidence in Python to realize this is probably just my
> misunderstanding, but wondering why Python mangles private, double
> underscore attributes with the name of the base class even if they're
> inherited and the instance is that of the sub class.
> 
> Like the following doesn't even seem to work:
> 
> class StandardNormal:
> 
>     def letEventZScore(self, eventZScore):
>         self.__eventZScore = float(eventZScore)
> 
> class DerivativeNormal(StandardNormal):
> 
>     def getCentral(self):
>         if self.__eventZScore < 0:
> 
> Unless I change that last line to something like:
> 
>         if self._StandardNormal__eventZScore < 0:
> 
> Maybe just missing the point all together, but wondering if anyone'll
> point out anything a little more elegant and/or intuitive.
> 
> Thank you and have a great day,

Coming from a Smalltalk background, myself, which does not have anything akin to
'protected' and 'private', rather than a C++ background I would say the
real solution is "don't bother" :)

Code is communication...between you and the machine, but also between you
and other developers.  Sometimes communicating your intent through
convention to other developers is more important than communicating your
intent through syntax to the compiler.  Idon't really care as much about
getting the compiler to *enforce* what I want in ptotected or private
access as I do in telling other developers my intention of protectedness.

The convention seems to be that a leading underscore (_foo) implies that
the variable or function is not meant for public access.  You're still
free to reference it if you want, but you've been warned that it's not
recommended and you are on your own.

Like calling sys.exit() vs sys._exit(), both are legal, but you shouldn't
use the second flippantly as it has certain ramifications you shold be
aware of.

This allows developers to communicate their intention of how their code
should be used, without mucking up the language syntax and rules with a
lot of requirements for how to handle 'protectedness'

It's a different philosophy than in C++ or Java, but I think to be
effective in Python, it would be good to adopt that philosophy

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.r4h.org/r4hsoftware


From michaels at one.net  Thu Sep 19 08:45:16 2002
From: michaels at one.net (Michael Schneider)
Date: Thu, 19 Sep 2002 08:45:16 -0400
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <mailman.1032329366.6813.python-list@python.org> <j4ptvbaapm.fsf@informatik.hu-berlin.de> <3D890FB0.4030408@something.invalid> <m3admeo4u1.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3D89C6DC.5050801@one.net>

If stackless makes it into the core.  Couldn't the "dangerious modules" 
check if the
current program is running in stackless mode.  

If stackless mode + dangerious module => exception



ex: (not the right syntax, just the symantics)


#################### start of my dangerNonStacklessModule.py

require NoStackless

import sys

#############################################################






Martin v. Loewis wrote:

>Greg Ewing <see_reply_address at something.invalid> writes:
>
>>So, the only way to find out whether a particular
>>extension is safe to use in Stackless is to try
>>it. 
>>
>
>That is not true. It is more reliable to inspect the source code.
>
>In general, it is very easy to do so: if there are no callbacks to
>Python, the module is stackless-safe.
>
>>That's why I say that I would be nervous about
>>using Stackless. Not to say that would stop me
>>from using it, but I'd feel nervous while I was
>>doing it! :-)
>>
>
>That is perfectly fine with me. The question is whether incorporation
>of Stackless Python in the core distribution would already make you
>nervous. I can see no rational reason for that.
>
>>Yes, I understand that, and I acknowledge that it
>>won't cause any harm to have Stackless in the core
>>if the Stackless stuff is switched off by default.
>>
>
>It is better than that: it does not need to be "switched off". Even if
>switched on (at build time), it won't start copying slices of the
>stack. An application actively has to import one of the modules, and
>to call a function in those modules, before you need to get nervous.
>
>Regards,
>Martin
>



From nospam at bigfoot.com  Fri Sep 20 16:39:02 2002
From: nospam at bigfoot.com (Gillou)
Date: Fri, 20 Sep 2002 22:39:02 +0200
Subject: urllib2 ?
Message-ID: <amg08t$glh$1@norfair.nerim.net>

Hi,

I'm stuck reading the doc of urllib2. When urllib seemed so easy.

Are there examples somewhere ?

TIA

--Gilles





From NineOfSix at gmx.de  Mon Sep  2 06:56:08 2002
From: NineOfSix at gmx.de (Axel Grune)
Date: Mon, 02 Sep 2002 12:56:08 +0200
Subject: file processing question
Message-ID: <akvg3g$5bn$04$1@news.t-online.com>

Hi, Im trying to read the data of a file with following method

file = open( inputfile, "r")
data = file.read()
file.close()

but 'data' is always smaller then the file size, e.g. I have a file 
which is 25677824 bytes but file.read() reads only 4295 (file.tell() = 
25677824).
Im using python 2.2 under win2k.

BTW,

file = open( inputfile, "r")
lines = file.readlines()
data = ''join( lines)
file.close()

has the same effect.

TIA,
Axel



From see_reply_address at something.invalid  Sun Sep 29 20:57:18 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Mon, 30 Sep 2002 12:57:18 +1200
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <mailman.1031377930.4225.python-list@python.org> <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com> <3D8E6353.7030303@something.invalid> <dvbtouciidaufjneja3s4t89ds0r7jd76k@4ax.com> <3D911A1B.1000109@something.invalid> <a6f9355c.0209250720.4512b3c3@posting.google.com>
Message-ID: <3D97A16E.5060801@something.invalid>

Alan Daniels wrote:

> Unfortunately, that fails when one person writes their code with tabs
> set to 8, and another person reads the code with tabs set to 4. Code
> that was hand-aligned with a mix of tabs and spaces doesn't align
> anymore.


If you're trying to line up things on two lines
having different indentation levels, that's true.
So you don't do that -- you only try to line
things up at the same indentation level.

e.g.

<tab><tab>fooblarg(x, y,
<tab><tab>         z)

Note the absence of tabs underneath the "fooblarg(".

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From mcfletch at rogers.com  Tue Sep 10 10:05:20 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 10 Sep 2002 10:05:20 -0400
Subject: Graph Layout Algorithms
References: <alksio$d2l$1@newsg3.svr.pol.co.uk>
Message-ID: <3D7DFC20.9010008@rogers.com>

Boa Constructor has some code in their UML view that I contributed to do 
this using topological sorts.  The topological sorting algos (one by me, 
one by Tim) are available seperately here (as toposort.py):

	http://members.rogers.com/mcfletch/programming/

if you don't want the GPL contamination of using Boa's copy.

Basically, you do the topo-sort, then lay out each generation 
above/below next/previous generations.  For really nice-looking graphs, 
you'd want to do some post-processing to make children show up 
underneath parents as much as possible (i.e. minimise number of crossing 
arcs in the graph), and maybe do some work with spacing so that each 
generation takes up approx. the same amount of space.

HTH,
Mike

Duncan Smith wrote:
> Hello,
>          Just wondering if anyone has implemented anything in Python.  I
> need to lay out my graphs (directed acyclic / undirected / trees) in a
> reasonably clear way.  So if anyone has already implemented anything, or if
> anyone has advice on appropriate algorithms, I'd be glad to hear about it.
> Cheers.
> 
> Duncan
> 
> 
_______________________________________
   Mike C. Fletcher
   Designer, VR Plumber, Coder
   http://members.rogers.com/mcfletch/





From erwin at andreasen.com  Tue Sep  3 13:12:21 2002
From: erwin at andreasen.com (Erwin S. Andreasen)
Date: 03 Sep 2002 19:12:21 +0200
Subject: Better tracebacks for exec and execfile
References: <3i99nusjbgpiepva9pkuv3mtl7equ7h88h@4ax.com>
Message-ID: <87elcbm1ze.fsf@andreasen.org>

Stefan Franke <franke at ableton.com> writes:

> 1. I'm toying araound with a text templating engine containing
> embedded Python scripts. These scripts are run with exec.
> If they throw an exception, the traceback looks like
> 
>   File "<string>", line 5
>      e = a+c+b+
>                ^
>  SyntaxError: invalid syntax

I don't know how to make SyntaxError appear correctly, but for the
others you can use compile which lets you pass in a file name. You can
then exec the compiled string.

compile won't let you specify a start line number: for that I simply
added "\n" * start_line to the start of the evaluated string :)


-- 
===============================================================
<erwin at andreasen.org>                           Herlev, Denmark     
<URL:http://www.andreasen.org/>                             <*>   
===============================================================



From gerhard.haering at gmx.de  Sat Sep 28 21:15:46 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 29 Sep 2002 01:15:46 GMT
Subject: A simpler singleton pattern?
References: <mailman.1033259618.14506.python-list@python.org>
Message-ID: <slrnapcl23.5re.gerhard.haering@lilith.my-fqdn.de>

lee lion wrote in comp.lang.python:
> singleton pattern is a way to provide one and only one object of a
> particular type. There are many ways to implement a singleton
> pattern, is the below one?

> class OnlyOne:
>     instance = None
>     def __init__(self, arg):
>         OnlyOne.instance = arg
>     def __getattr__(self, name):
>         return getattr(OnlyOne.instance, name)

That's another cool way. Did you read about the Borg pattern from Alex
Martelli, yet? It works similarly, but without the need to override
__getattr__:
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66531

-- Gerhard


From F.L.Burton at bio.gla.ac.uk  Sun Sep  1 16:39:22 2002
From: F.L.Burton at bio.gla.ac.uk (Francis Burton)
Date: Sun, 01 Sep 2002 21:39:22 +0100
Subject: Python IPC options
References: <1030829572.943495@irys.nyx.net> <ddc19db7.0209010836.50242309@posting.google.com>
Message-ID: <3D727AFA.1FE19151@bio.gla.ac.uk>

Armin Steinhoff wrote:
> SPREAD -> http://www.python.org/other/spread/

This seems to be very close to what I am looking for.
Thanks for the pointer!

> Hope this is enough :)

More than enough. :-) I'll investigate Spread first,
then maybe delve into some of the others.

Francis


From djc at object-craft.com.au  Thu Sep 19 23:46:05 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 20 Sep 2002 13:46:05 +1000
Subject: monetary applications [was: Python GUI app to impress the bos s?]
References: <31575A892FF6D1118F5800600846864DCBD817@intrepid> <mailman.1032362843.10509.python-list@python.org> <23891c90.0209190014.26c5bb9@posting.google.com> <s2fdma.pp3.ln@ix.netcom.com>
Message-ID: <m38z1xpblu.fsf@ferret.object-craft.com.au>

>>>>> "Dennis" == Dennis Lee Bieber <wlfraed at ix.netcom.com> writes:

Dennis> Paul Boddie fed this fish to the penguins on Thursday 19
Dennis> September 2002 01:14 am:


>> It could well be that not many people do financial database
>> programming in Python, bizarre and unlikely as that may
>> seem. Taking a look at pyPgSQL, as the only database module I have
>> installed that actually readily documents such things (and that
>> should tell you something), it would seem that monetary amounts are
>> converted to floating point numbers - surely the first thing one
>> would want to avoid in manipulating financial data. Perhaps people
>> use integers, or perhaps other modules provide better decimal
>> support.

Dennis>         If I recall from the c.l.ada group, GNAT's "fixed
Dennis> point" data type is implemented as scaled integers, so even
Dennis> that might not pass the spec for a monetary type. My VB
Dennis> references are on loan (or currently stuck in the Windows
Dennis> environment and unreadable on Linux) so I can't recall what
Dennis> VB6 had for fixed/monetary.

Dennis>         The only language I've ever encountered that had a
Dennis> truely monetary safe fixed point data type is COBOL -- which
Dennis> has always used packed BCD formats (though where the sign
Dennis> "bit" is stored may vary). COBOL was created from the
Dennis> beginning for "business" applications (heck: COmmon Business
Dennis> Oriented Language) and was spec'd by a government team. My
Dennis> college mainframe required /four/ of the sixteen registers to
Dennis> hold /one/ BCD number (32 BCD digits, 2 digits/byte, 16bytes @
Dennis> 4/register).
 
I while ago I worked for a stock broker and developed some Python code
for doing financial post trade calculations.  I sweated quite a bit
about loss of accuracy in stamp duty, GST, and commission calculations
until it was pointed out that errors of a few cents here are there are
considered normal.  For larger amounts of money (100's of 1,000's of
dollars) an error of around a dollar would be more or less
transparently handled within the various back end systems used between
the trading organisations (institutional investors and stock broker).

It might be worth talking to the actual users of the software to see
what margin of error is considered normal before getting too upset
about achieving 100% accurate results.  You may be able to get a
provably correct answer in your code, but you have no control over the
software being used by other people (if there are other people).

I found my largest problem was trying to repeat the calculations being
performed by a third party system and reproduce the truncation and
rounding that was evident in their final numbers.

- Dave

-- 
http://www.object-craft.com.au


From marklists at mceahern.com  Mon Sep 23 22:44:13 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 23 Sep 2002 21:44:13 -0500
Subject: confusion about opening files
In-Reply-To: <3d8fcda4$0$198$75868355@news.frii.net>
Message-ID: <JHEOKEOOLIGLDHCMAHMOEEONEAAA.marklists@mceahern.com>

[robie1373]
> I am trying to open a file,
>
> >>> import os
> >>> file = os.open("e:\\d12", "r")
> Traceback (most recent call last):
>   File "<pyshell#4>", line 1, in ?
>     file = os.open("e:\\d12", "r")
> TypeError: an integer is required
>
> I get this error if the mode is r, r+, a+ or whatever.  I also get it
> whether the file exists or not. Can someone explain what I am doing wrong?

This is a FAQ.  You should probably find the Python FAQ and read it.  Short
story:  Don't use 'from x import *' unless you know what you're doing.

Cheers,

// m

-




From tmohr at s.netic.de  Wed Sep 18 12:39:24 2002
From: tmohr at s.netic.de (Torsten Mohr)
Date: Wed, 18 Sep 2002 18:39:24 +0200
Subject: OpenGL Bindings for Tcl/Tk 8.4
References: <am87gi$o9$1@schleim.qwe.de>
Message-ID: <amaa7s$46v$1@schleim.qwe.de>

Hi,

i just received a hint by eMail that some more informations would
help:

OS: Linux, Kernel 2.4.19
Compiler: gcc-2.95.3


Thanks for any hints,
Torsten.


> Hi,
> 
> i just downloaded and installed Tcl/Tk 8.4 and PyOpenGL-2.0.44.
> 
> Sadly it doesn't compile, though i edited "togl.c", as mentioned
> in the #error directive.
> 
> Also the latest CVS snapshot doesn't compile.
> 
> 
> Has anybody got some patches/hints that will help me compiling
> and installing OpenGL bindings for python?
> 
> 
> Best regards,
> Torsten.



From gerhard.haering at opus-gmbh.net  Mon Sep 16 05:22:31 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 16 Sep 2002 09:22:31 GMT
Subject: CPAN for Python
References: <mailman.1032164058.1426.python-list@python.org>
Message-ID: <slrnaob8ds.1v8.gerhard.haering@haering.opus-gmbh.net>

Bo M. Maryniuck wrote:
> Hi.
> I "heard" that Python will have something like CPAN for Perl soon.

It's probably pythonsiphon that you've heard of.

> Is there is something new and moving forward in this area?

http://sourceforge.net/mailarchive/forum.php?thread_id=1048175&forum_id=2412

Looks like Suchandra Thapa is still working on it (my patch still hasn't
been integrated, however). So the answer is probably that yes, it's moving
forward, but slooowly.

-- Gerhard


From gerhard.haering at opus-gmbh.net  Thu Sep 12 07:59:11 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 12 Sep 2002 11:59:11 GMT
Subject: Starting external functions in Python applications
References: <1d3ffc6b.0209120345.64cf64a1@posting.google.com>
Message-ID: <slrnao105i.3ak.gerhard.haering@haering.opus-gmbh.net>

Paulus wrote:
> How can I link an external dll in a Python programs and (when possible)
> start any function from this DLL ? 

You don't exactly link to a DLL, but the usual answer is that you can use
calldll - http://www.nightmare.com/software.html

This site has only Python 2.1 binaries - for Python 2.2 binaries, you can
install it using the PPM package manager of the ActivePython distribution.

-- Gerhard


From 2002 at weholt.org  Mon Sep 23 07:17:47 2002
From: 2002 at weholt.org (Thomas Weholt)
Date: Mon, 23 Sep 2002 11:17:47 GMT
Subject: UpToDate SOAP-module ??
References: <ATAj9.21315$0p1.376966@news2.ulv.nextra.no> <j4fzw1ghpv.fsf@informatik.hu-berlin.de>
Message-ID: <vLCj9.24765$sR2.439699@news4.ulv.nextra.no>

> My personal recommendation is to ignore SOAP.

Eh ... ok. I might be the victim of pure hype but I'm developing a
WebService-project, currently supporting XML-RPC only, and I do feel like I
have to support SOAP too if I'm to be taken serious. I like the ease of use
of XML-RPC and there are lots of services available for it allready, but
SOAP seem to be the next, allthough somewhat bloated and complicated, step
for Web-Services.

Anyway, if you want to provide more information about why I safely can
ignore SOAP feel free. It would make my project a bit less complicated so
I'd be glad to skip it. :-)

Best regards,
Thomas




From kseehof at neuralintegrator.com  Tue Sep  3 17:54:43 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Tue, 3 Sep 2002 14:54:43 -0700
Subject: Converting Python app to C++ completely
In-Reply-To: <aklm6c$r1r$1@news.swman.net.uk>
Message-ID: <EFEFIHAINIINNELLHFAIKEFFDFAA.kseehof@neuralintegrator.com>

> Hi all,
... snip ...
> Now, what I want to do is write the thing in C++. The reason? heh,
> because I want to ;) Well, actually I want to learn C++ too, and I would
> like to have the performance of a compiled program for larger data sets
> (just for fun and comparison more than anything).

Hmm.  I guess that could be a good reason.  In practice, the only real
reason to do a complete python to C++ conversion that I have ever seen
is to reduce the applications footprint size (i.e. exclude the python
libraries).  Speed is not a good reason to convert the entire application.
You can get all the speed you need by converting individual components to
C++ as extensions.

I recommend that you learn how to write extensions, or better yet, learn
how to use Boost.Python: http://www.boost.org/libs/python/doc/index.html

I'd doubt that converting the GUI code from wxPython to wxWindows would
make much difference.  Has anyone done wxPython vs wxWindows benchmark
comparisons?  In most cases changes in GUI response time will be irrelivant,
except maybe high frequency events such as mouse move handling when you
want smoothness.

> 1) How easy is it to convert existing Python code to C++? I read that
> each line of C++ code is roughly equivilent to 5-8 lines of C++ code,
> but is the conversion a simple task or is it troublesome?

Actually, you can convert your entire python program into 1 line of C++
code :-)

> 2) Are there any pitfalls to beware of - I understand that dealing with
> memory and pointers might be a problem, but like I said earlier, I just
> don't really know?

Yes, I think your biggest issue will be memory management.

> 3) Does C++ have equivilent functional constructs like reduce and map?
> These helped to make my code much cleaner, tighter and faster, and I
> would miss them dearly.

Don't bother.  Just use for loops in C++.

> btw - the code I am writing doesn't rely upon any platform specific
> stuff at all and is quite portable in Python - I will be aiming for the
> same thing in Python and assume it will be possible.

C++ and wxWindows are almost as portable as Python and wxPython.  C++
varies a bit: http://fusshuhn.ourfamily.com/cppincomp.html.





From mike at bindkey.com  Fri Sep 27 15:20:12 2002
From: mike at bindkey.com (Mike Rovner)
Date: Fri, 27 Sep 2002 12:20:12 -0700
Subject: [XML-SIG] well-formed xml
References: <JHEOKEOOLIGLDHCMAHMOAEOPEBAA.mark@mceahern.com> <E17uppg-0007LF-00@smtp.web.de>
Message-ID: <an2b4m$7g9$1@main.gmane.org>

IIRC, double quotes are the must:
s = '<a href="http://google.com/search?hl=en&amp;q=foobar">search</a>'

Cheers,
Mike

"Juergen Hermann" <jh at web.de> wrote in message
news:E17uppg-0007LF-00 at smtp.web.de...
On Thu, 26 Sep 2002 19:08:48 -0500, Mark McEahern wrote:

>I'm obviously missing something because this seemingly innocent chunk of
>xhtml:
>
>  from xml.dom import minidom
>
>  s = "<a href='http://google.com/search?hl=en&q=foobar'>search</a>"

This is not well-formed and thus not XHTML.

... en&amp;q= ...

is the correct form.







From eworrall at netcomuk.co.uk  Sun Sep 29 12:45:47 2002
From: eworrall at netcomuk.co.uk (Eric Worrall)
Date: Sun, 29 Sep 2002 17:45:47 +0100
Subject: Linker can't find library (PyProlog setup)
References: <mdGl9.14197$Lg2.4064162@news2.news.adelphia.net>
Message-ID: <3D972E3B.C07047E5@netcomuk.co.uk>

Robert Oschler wrote:
> 
> I'm having trouble getting gcc/ld to 'see' the existence of a known shared
> library.  I quadruple-checked and I do indeed have the files 'libtermcap.so'
> and 'libtermcap.a' in my '/usr/lib/termcap' directory.  'gcc' is being
> passed the following library directive '-ltermcap'.  I tried the following
> 'export' commands in my build script (the echoed results do indeed show
> /usr/lib/termcap being properly added to the respective environment
> variables):
> 
>     export LD_LIBRARY_PATH="/usr/lib/termcap:$LD_LIBRARY_PATH"
>     echo "LD_LIBRARY_PATH=$LD_LIBRARY_PATH"
>     # run gcc command here
> 
> and I also tried:
> 
>     export PATH="/usr/lib/termcap:$PATH"
>     echo "PATH=$PATH"
>     # run gcc command here
> 
> and lastly, both:
> 
>     export LD_LIBRARY_PATH="/usr/lib/termcap:$LD_LIBRARY_PATH"
>     export PATH="/usr/lib/termcap:$PATH"
>     # run gcc command here
> 
> In all three cases I still get a complaint from 'ld', 'cannot
> find -ltermcap'.  What am I doing wrong?
> 
> thx

Include the library path using the -L option:

gcc -L/usr/lib/termcap etc

Eric Worrall

-- 
You have just received an Etech Solution
For all your Linux requirements contact
eworrall at netcomuk.co.uk


From llothar at web.de  Tue Sep 17 11:26:22 2002
From: llothar at web.de (Lothar Scholz)
Date: Tue, 17 Sep 2002 17:26:22 +0200
Subject: More on Protecting Source Code
References: <mailman.1032232999.22661.python-list@python.org> <3d86b4a0$0$23173$afc38c87@news.optusnet.com.au>
Message-ID: <icieou0lpp0cgqetbl560airkm5ohuqcn2@4ax.com>

On Tue, 17 Sep 2002 14:50:39 +1000, in comp.lang.python you wrote:

>
>"Security through obscurity is no security against a determined attacker"
>
>So, someone who is willing to hacka nd hack and hack away at your code, can
>get at it, and this is true for things coded in assembler, C, Python or
>Befunge. Most people aren't determined attackers, they're simply Joe Average
>who needs to get a  job done.

And some of them must count the time it takes to crack the program vs
the money they need to buy it.

>Do a very very simple risk analysis. Who is your target audience, what is
>the price of your software, will people bother going through your bytecode,
>or will they just post the registration key to a website? Will stolen
>software lead to a loss in revenue - or will those pirate users be people
>who would never have bought it in the first place? (yes, the pirate users
>are annoying, and you would feel ripped off, but OTOH you haven't lost
>money, and you've gained a bunch of users)

Your comment here is language independent.a

The main problem comes with stealling the intellectual properties (for
the few programs where you can say 'intellectual'). This is the
difference between python and compiled languages like Eiffel.
I agree that this problem is the main reason for so less commerical
python applications.

But it could be changed easily if the core python developer would see
this problem as relevant as commerical users. But at the moment they
don't see the point. Thats the problem.

Until this changed i would recommend using Armadillo and embedd your
python source code inside a C binary wrapper. Armadillo
encrypt/decrypt memory pages that are not in use so its a to hard for
95% of all attackers.




From dsavitsk at e-coli.net  Wed Sep 18 13:45:13 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Wed, 18 Sep 2002 17:45:13 GMT
Subject: COM object registration
References: <afph9.9468$yt3.4669302@newssrv26.news.prodigy.com> <T_Rh9.17227$Ee4.51953@news-server.bigpond.net.au> <uCTh9.10343$yt3.4866032@newssrv26.news.prodigy.com> <4J_h9.19439$Ee4.59690@news-server.bigpond.net.au>
Message-ID: <JY2i9.10441$yt3.4917953@newssrv26.news.prodigy.com>

"Mark Hammond" <mhammond at skippinet.com.au> wrote in message
news:4J_h9.19439$Ee4.59690 at news-server.bigpond.net.au...
> dsavitsk wrote:
> >>>>a = win32com.client.Dispatch('ecp.MODULE_REGISTER2')
> >>>
> > Traceback (most recent call last):
> >   File "<interactive input>", line 1, in ?
> >   File "C:\Python\win32com\client\__init__.py", line 92, in Dispatch
> >     dispatch, userName =
> > dynamic._GetGoodDispatchAndUserName(dispatch,userName,clsctx)
> >   File "C:\Python\win32com\client\dynamic.py", line 81, in
> > _GetGoodDispatchAndUserName
> >     return (_GetGoodDispatch(IDispatch, clsctx), userName)
> >   File "C:\Python\win32com\client\dynamic.py", line 72, in
_GetGoodDispatch
> >     IDispatch = pythoncom.CoCreateInstance(IDispatch, None, clsctx,
> > pythoncom.IID_IDispatch)
> > com_error: (-2147467259, 'Unspecified error', None, None)
>
> It is likely that an exception is being encountered that we can get at
> pretty easy.
>
> Add a named param "debug=1" to all UseCommandLine calls.  Then, on both
> machines, start Pythonwin and select the debug trace collector.  Rerun
> your tests and see if any additional interesting exceptions are reported.
>
> Mark.
>

This is the traceback, but I'm afraid it doesn't mean much to me.  It is the
error that I got when using a VB client and where I get the 'localserver'
notion.

in _Invoke_ with 1000 0 2 ()
in _Invoke_ with 1000 0 1 (u'MODULE_REGISTER2',)
Registered: ecp.MODULE_REGISTER2 (for debugging)
Object with win32trace dispatcher created (object=None)
Traceback (most recent call last):
  File "C:\Python\win32com\server\dispatcher.py", line 27, in
_CreateInstance_
    self.policy._CreateInstance_(clsid, reqIID)
  File "C:\Python\win32com\server\policy.py", line 199, in _CreateInstance_
    myob = call_func(classSpec)
  File "C:\Python\win32com\server\policy.py", line 634, in call_func
    return apply(resolve_func(spec), args)
  File "C:\Python\win32com\server\policy.py", line 624, in resolve_func
    return getattr(module, fname)
AttributeError: 'localserver' module has no attribute 'regmodule2'

-doug




From fperez528 at yahoo.com  Wed Sep  4 17:38:31 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 04 Sep 2002 15:38:31 -0600
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Sep  4)
References: <448A0F008E0AFE11.19B9EF86BEE2B191.107D5020B234F743@lp.airnews.net> <al5tvu$1mtjjr$1@ID-71831.news.dfncis.de> <al5ue4$sj2$1@peabody.colorado.edu>
Message-ID: <al5ugn$sj2$2@peabody.colorado.edu>

Fernando P?rez wrote:

> James Kew wrote:
> 
>> I can't persuade Google to find this message-id; any chance of a link?
> 
> google_groups('comp.lang.python anansispaceworks.com')[0]

Oops. Make that 

google_groups('comp.lang.python anansispaceworks.com scientific')[0]

Sorry,

f.


From tdelaney at avaya.com  Wed Sep 18 21:12:43 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 19 Sep 2002 11:12:43 +1000
Subject: Python comment stripper
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5AE@natasha.auslabs.avaya.com>

> From: Dennis Reinhardt [mailto:DennisR at dair.com]
>
> At the time extraction is going on, it is impractical for 
> Python to be doing
> the extraction -- there is the Python DLL (v2.0: 653 Kb and 
> v2.2: 821 Kb)
> for starters.  I am aiming for a small footprint.  The 361 Kb 
> download has
> Python22.dll inside and that DLL  is used when the hello 
> program runs. :

Um ... if you are including Python22.dll in the download, surely that is
dwarfing your code size?

Using your numbers ...

>                       raw Kb          compressed Kb
>                     ----------    --------------------
> original              10.44               2.98
> 4 spaces -> 1 space    8.85               2.86
> + strip comments       4.65               1.34

So you would have:

821 -> 365 (I zipped it up ;)

Therefore your package sizes would go from 368 KB -> 366 KB.

Even assuming you have 100 times as much (compressed) code, and the comment
savings are identical:

663 KB -> 499 KB

Is the savings worth the hassle?

Tim Delaney



From gerhard.haering at opus-gmbh.net  Tue Sep  3 03:32:06 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 3 Sep 2002 07:32:06 GMT
Subject: The Module Search Path is ???
References: <7153651c.0209022255.7fd50ba0@posting.google.com>
Message-ID: <slrnan8pa3.2ds.gerhard.haering@haering.opus-gmbh.net>

In article <7153651c.0209022255.7fd50ba0 at posting.google.com>, black wrote:
> tutorial told me I could found it in the folder:/usr/local/lib/python.
> but my OS is windows so folders are different.

Yep.

The module search path is dynamically constructed during Python startup, but
can be changed even at a later point.

It's in sys.path:

>>> import sys
>>> sys.path
['', 'C:\\Python22\\Lib\\site-packages\\Pythonwin', 'C:\\Python22\\Lib\\site-pac
kages\\win32', 'C:\\Python22\\Lib\\site-packages\\win32\\lib', 'C:\\Python22\\Li
b\\site-packages', 'C:\\Python22\\DLLs', 'C:\\Python22\\lib', 'C:\\Python22\\lib
\\lib-tk', 'C:\\Python22', 'C:\\Python22\\lib\\site-packages\\PIL']

-- Gerhard


From phoebe_1 at att.net  Tue Sep 24 13:41:21 2002
From: phoebe_1 at att.net (Holden Caulfield)
Date: 24 Sep 2002 10:41:21 -0700
Subject: Solaris 64 bit compilation
References: <c2595393.0209201844.2ed62265@posting.google.com> <m3sn03k44m.fsf@mira.informatik.hu-berlin.de> <c2595393.0209210627.44f672a8@posting.google.com> <m3d6r7i9o0.fsf@mira.informatik.hu-berlin.de>
Message-ID: <c2595393.0209240941.5f3fc565@posting.google.com>

martin at v.loewis.de (Martin v. Loewis) wrote in message news:<m3d6r7i9o0.fsf at mira.informatik.hu-berlin.de>...
> phoebe_1 at att.net (Holden Caulfield) writes:
> 
> > > Please try to report failures carefully. It does not say sizeof(LONG),
> > > instead, it says SIZEOF_LONG. Which gives rise to the question: What
> > > value does SIZEOF_LONG have, in pyconfig.h?
> > > 
> > cc -c -DNDEBUG -g -Xc xarch=v9 -xcode=pic32 -I. -I./Include
> > -DHAVE_CONFIG_H  -o Modules/p
> > ython.o Modules/python.c
> > "./Include/pyport.h", line 480: #error: "LONG_BIT definition appears
> > wrong for platform (bad gcc/glibc config?)."
> 
> So what value has SIZEOF_LONG in pyconfig.h?
> 
> How did you invoke configure? How did you manage to get -xarch=v9 into
> the compiler command line?
> 
> Regards,
> Martin

My initial post was just a query to find out if python in 64bit mode
was available/compiled successfully by anyone.
To answer your questions:
  - pyconfig.h does say 8 as SIZEOF_LONG
  - I had hacked the Makefile as a initial pass to see where things
were and I
    posted to get an idea about it.

Since then, I have hacked the configure.in and added the option
--enable-solaris-64bit and regenerated the configure script. Now,
good news, it all builds fine. Except I am having a few problems with
extension
modules these are:
  - 'curses' modules does not compile. 
  - 'dbm' modules compiles fine but core dumps during test.
So, I disabled these modules in 'setup.py' and built it and ran the
make test.
I had only two failures these are:

1)
test test_time failed -- Traceback (most recent call last):
  File "./Lib/test/test_time.py", line 46, in test_mktime
    time.mktime, (999999, 999999, 999999, 999999,
  File "/apps/python/build/Python-2.2.1/Lib/unittest.py", line 279, in
failUnlessRaises
    raise self.failureException, excName
AssertionError: OverflowError
2)
test_gettext
test test_gettext crashed -- exceptions.OSError: [Errno 20] Not a
directory: 'xx/LC_MESSAGES/gettext.mo'

I could probably live without them for now.

Are there any plans to support 64-bit python executable/module on
solaris platforms for future versions?
Is anyone even interested in the configure.in patches?

Thanks


From sismex01 at hebmex.com  Thu Sep 26 10:01:00 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Thu, 26 Sep 2002 09:01:00 -0500
Subject: Larry Wall's comment on python...
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062330260462A@mail.hebmex.com>

> 
> sismex01 at hebmex.com wrote:
> 
> >>Ironically, Guido originally ruled (in the style guide, IIRC) 
> >>that thou shalt use tabs and tabs are 8 spaces, "as god
> >>intentended them to be."
> >
> >Yeesh! Why in God's name would he say such a dumb thing?
> >I mean, even ancient mechanical typewriters have movable
> >tab settings. Rigid, 8-char tabs is just so... neolithical.
> >
> 
> Once upon a time, a large class of computers, and their attendant 
> peripherals, OSs and utility programs universally implemented tabs 
> hard-wired at 8 stops.  In some circles this heritage extended
> back 30 or 40 years.  Unix originally followed this convention,
> as did IIRC all Dec computers and that of most other "minicomputer"
> vendors.
> Ed and the original VI editors only had fixed tab stops consistent
> with terminal drivers and all the unit record equipment at the time.
> IIRC, even Emacs originally only had tab stops at every 8 columns.
> The original PC also followed this convention, though not all
> applications did.  I don't know for sure but I bet $1 that Linux
> tty drivers today also are hard wired at 8 char tab stops.
> 

Thanks for a beautiful reply.

Well, I might not be an old fart :-), but I *was* kinda present
at that time, I started work in highschool hacking on CP/M machines,
writing drivers for diverse printers and terminals for some apps
(ahh, customizable terminal control... we lost that during the
PC era, everybody wanted CGA, MDA or Hercules).

AND, I distinctly remember that most printers could customize
their tab stops via escape codes, same as some (not all) terminals;
I worked a lot on some Xoroc terminals, and lots on a Northstar
Advantage (ahh, beautiful hardware... *snif*) and they also had
customizable tabstops.

Hmmm... maybe, what I'm getting at is, why did terminal-designers
everywhere toss out years of research in typewriter technology
(at the time) and ignore customizable tab-stops?  It's simply not
sane, I think.

> Now, if you've ever worked in an environment where EVERYBODY uses the 
> same tab stops then you'd probably agree that tabs are vastly 
> superior to spaces.  I don't think there's any question.  I presume
> it's from this perspective that the right honorable BDFL issued his
> original proclamation, which is actually quite reasonable, especially 
> from my old fart's perspective.

Sounds quite likely.

> 
> In fact, problems only arise when alternative, competing tab stop 
> conventions try to coexist, which I agree is not  generally solvable. 
> Closest solution I've seen is some sites where everybody uses Emacs
> and employ coding conventions so that on startup Emacs can read a 
> specially coded comment and automatically switch to the proper indent
> for that file.  This can sort of work but your printer and other
> rendering utilities also all have to recognize the codes.  Spaces
> are a more practical solution.

Indeed.  I *do* recognize that using actual tabs is a much better
solution, but having hard-stops each 8 chars is a show-stopper,
so if my editor can't customize it's tab width, or customize it's
stops (say, when in asm, on columns 12, 18 and 32) then I'm never
going to use it willingly (you *could* whip me into submission,
but I'll curse your name for the rest of my natural life ;-)

> I myself still prefer tabs as I find "indenting" and "out denting" 
> sections of code to be a common operation (in Python and in other 
> languages).  Inserting or removing an "if" or changing inline code
> to a function makes indenting and out denting a common operation.
> In the environments I use, this works well with tabs and not so well 
> or at all with spaces.  By default, I still prefer 8 char stops but
> I also find I can live just fine with 4.  But I can't stand it without
> tabs.  I'll convert to spaces if I need to when delivering code but I 
> can't work on it that way.

That's what I mean.  An editor should be smart enough to handle tab
stops at arbitrary columns, and when using tabbed text, jump between
those columns.  Best of both worlds.

Problems arise when mixing tabs and spaces, because you can't ask
a program to read your mind; so, is move-left a "jump to previous indent
level" or a "move to left character position"?  That's the users's
responsability to define, but with flexible tab stops, or tab positions,
it becomes much easier.

> 
> Anyway, what's wrong with the "neolithical" 8 char stops?  
>

I find them hmmm... philosophically [sp] "wrong", because computers,
being programmable and customizable ("terminals" being specialized
computers) and so, should _at_least_ do what their mechanical
counterparts are able to do, and then some.

Would an electronic calculator survive, if it couldn't do the same
operations as a mechanical adding machine?

> For several decades, all the C code ever written was tab indented
> at 8 columns, along with most of the text and other programming
> languages on Unix.

Yup, I did the same also, but maybe it was because there was no
alternative?

> Some people complained that in practice 8 char tabs forced them
> to indent too much and their programs got squashed off the right
> edge of the page.  I have little sympathy for that argument, as
> I feel if your program is properly modularized then this should
> not be much of a problem as individual modules should not have
> to be indented all that much.

Same here, I find that kind of argument a falacy: "my program is
badly modularized and I can't read it clearly, it's the editor's
fault".  Nope, sorry.

> Also, I suspect part of the pressure came from Pascal users, 
> where some coding conventions require indenting two levels, 
> once for the begin/end pairs and again a second time for the
> code contained between them.  What a crock.

If it was corporate standard, it sux.  But, as for personal taste,
I never could decide which looks better, either:

if something > another
begin
  blah blah
end;

--- or ---

if something > another
  begin
  blah blah
  end;

--- or ---

if something > another
 begin
  blah blah
 end;


I'm still somewhat ambiguous over which is more readable,
maybe it depends on the context, the surrounding code,
the tab width, the amount of spaces one uses for the
begin/end indent and the actual code indent, I don't
know and I don't really care to do reasearch about it.

> 
> For what it's worth, Python's C implementation today appears 
> to be tab indented to 8 columns, at least all the parts I've
> looked at.  The main interpreter loop is a good example of a
> "complex" piece of code.  The innermost loop includes a single
> switch statement that's over 1400 lines long!  Surrounding logic
> require that the switch statement and case labels start 2 tab
> stops in and thus the body of each case starts 3 stops in or in
> column 24.  Yikes, that's like 1/3 of the total.  And yet 
> none of the case bodies extend past column 73 or so.  Turns 
> out that in practice most code lines are pretty short.  The ones
> that aren't typically are function calls with multiple arguments,
> and these may easily be split onto multiple lines (often thereby
> increasing readability).  Since Guido personally wrote the bulk
> of this system as a labor of love, I surmise that he's still pretty
> sentimental about the old ways.

I suppose you're right.

It's not the same to write something for love of the work,
as when it's for just work.

> 
> Regards
> 
> --jb
> 
> 
> -- 
> James J. Besemer		503-280-0838 voice
> 2727 NE Skidmore St.		503-280-0375 fax
> Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
> 				http://cascade-sys.com	
> 

It's been nice tossing ideas around :-)

-gustavo

Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From b.maryniuk at forbis.lt  Fri Sep  6 11:57:04 2002
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Fri, 6 Sep 2002 17:57:04 +0200
Subject: ttySx
In-Reply-To: <20020906152418.GA6937@sferacarta.com>
References: <200209061708.25677.b.maryniuk@forbis.lt> <20020906152418.GA6937@sferacarta.com>
Message-ID: <200209061757.04427.b.maryniuk@forbis.lt>

On Friday 06 September 2002 17:24, Marco Mariani wrote:
> Here it is.
> Of course, the code is old and sucks. 
Yes, it sucks. Because it is GSM modem, and not Hayes compatible (I forgot 
this note post here). Anyway, thank you much and I already had solved this 
problem in other Python way. :)

-- 
Regards, Bogdan

If you are good, you will be assigned all the work.  If you are real
good, you will get out of it.




From gregory.p.green at boeing.com  Wed Sep  4 13:05:14 2002
From: gregory.p.green at boeing.com (Greg Green)
Date: Wed, 4 Sep 2002 17:05:14 GMT
Subject: Daemonize?
References: <mailman.1031146755.28515.python-list@python.org>
Message-ID: <1y89n0s5.fsf@boeing.com>

sismex01 at hebmex.com writes:

> Sorry to ask here, but I don't have any "unixy" python
> documentation at hand.
> 
> Normally, I put an app into "daemon" mode, by first
> forking, then reopening stdin/out/err to /dev/null,
> and then forking again. Works good enough for my
> purposes.
> 
> But, am I right? Is this the correct way to do this?
> Also, I can't find any "reopen" in python to apply
> to sys.stdin/out/err.
> 
> Any help will be most appreciated.
> 
> Thanks.
> 
> -gus

The best place to look for this is either Stevens book "Advanced unix
programming" or the unix programming faq online:
http://www.erlenstar.demon.co.uk/unix/faq_toc.html see section 1.7
-- 

Greg Green



From marklists at mceahern.com  Fri Sep 20 10:15:55 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 20 Sep 2002 09:15:55 -0500
Subject: isFloat: Without Exception-Handling
In-Reply-To: <3D8B23E5.3040005@thinkware.se>
Message-ID: <JHEOKEOOLIGLDHCMAHMOKECJEAAA.marklists@mceahern.com>

[Magnus Lycka]
> Something like this perhaps:
>
> import re
>
> def isFloat(S):
>      floatRE = r"^[-+]?(\d+\.?\d*|\.\d+)([eE][-+]?\d+)?$"
>      return re.match(floatRE, str(S)) is not None
>
> I'm not sure it's faster though.

If you were going to run this lots of times, you might want to compile the
pattern.  Anyway, the exception based code is faster, by my testing (harness
at bottom):

       isFloatExcept 0.220
           isFloatRE 0.521
   isFloatRECompiled 0.400

Again, I wonder, why avoid the exception test?

Cheers,

// mark

Harness:

#!/usr/bin/env python

import re
from time import clock

# Compile up front.
floatPattern = re.compile(r"^[-+]?(\d+\.?\d*|\.\d+)([eE][-+]?\d+)?$")

def isFloatExcept(s):
    is_float = True
    try:
        float(s)
    except (ValueError, TypeError), e:
        is_float = False
    return is_float

def isFloatRE(s):
    floatRE = r"^[-+]?(\d+\.?\d*|\.\d+)([eE][-+]?\d+)?$"
    return re.match(floatRE, str(s)) is not None

def isFloatRECompiled(s):
    return floatPattern.match(str(s)) is not None

def time_it(fn, test_data, n):
    time_in = clock()
    for x in xrange(n):
        for (s, expected) in test_data:
            actual = fn(s)
            assert actual == expected
    time_out = clock()
    return time_out - time_in

def show(fn, r):
    print "%20s %1.3f" % (fn.func_name, r)

def test():
    test_data = [(1.2, True),
                 (1, True),
                 (-1, True),
                 (1e+10, True),
                 ("1e-10", True),
                 ("banana", False),
                 ("5j", False),
                 (4+0j, False),
                 (isFloatExcept, False),
                 (__import__, False),
                 (clock, False)]
    n = 1000
    results = {}
    for fn in (isFloatExcept, isFloatRE, isFloatRECompiled):
        r = time_it(fn, test_data, n)
        results[fn] = r
    for fn in results:
        show(fn, results[fn])

if __name__ == "__main__":
    test()

-




From reuben_stump at hotmail.com  Tue Sep 10 09:14:56 2002
From: reuben_stump at hotmail.com (Reuben Stump)
Date: Tue, 10 Sep 2002 09:14:56 -0400
Subject: UNIX processes
References: <mailman.1031654172.15956.python-list@python.org>
Message-ID: <mbmf9.64934$2L.3399713@e3500-atl2.usenetserver.com>

Try "ps -Alf".

Probably more information than you need.  Check the man page for more
options, including output formating.

R.

"Bo M. Maryniuck" <b.maryniuk at forbis.lt> wrote in message
news:mailman.1031654172.15956.python-list at python.org...
Hi, all.
Somebody know how to list all the processes in RIGHT way on Linux/UNIX?
I need to find the zombies, stopped processes etc.

--
Regards, Bogdan

Linux: The Ultimate NT Service Pack







From ponderor at lycos.com  Mon Sep  9 00:39:16 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 8 Sep 2002 21:39:16 -0700
Subject: Newby Question: os.rename
References: <AcGc9.68214$Xa.3523883@twister.southeast.rr.com> <tLGc9.68215$Xa.3526150@twister.southeast.rr.com> <fDdd9.6312$6i4.491988@newsread1.prod.itd.earthlink.net>
Message-ID: <e81be8b2.0209082039.67f33ab4@posting.google.com>

> >
> > "Kevin Bass" <KBASS1 at nc.rr.com> wrote in message
> > news:AcGc9.68214$Xa.3523883 at twister.southeast.rr.com...
...
> >
> > I have figured it out. Thanks!
> >
> >
> >

"CotaBas" <cotabas at earthlink.net> wrote in message news:<fDdd9.6312$6i4.491988 at newsread1.prod.itd.earthlink.net>...

> 
> LOL if you wait long enough people solve there own problems....

If you're skimming or Google'ing for a similar problem...you'll find
the problem listed 3 times and no answer in the follow-up!


From martin at v.loewis.de  Wed Sep  4 17:42:55 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 04 Sep 2002 23:42:55 +0200
Subject: str() of a tuple
References: <H1xnHJ.1zE1@beaver.cs.washington.edu>
Message-ID: <m37ki1a0tc.fsf@mira.informatik.hu-berlin.de>

Douglas Zongker <dougz at cs.washington.edu> writes:

> Why, oh why does the str() of a container use the repr()s of the
> objects inside?  This seems to violate what the docs claim str() does
> -- "return a nicely printable representation of an object."  Would a
> sane implementation break something?

Yes. If you have [''], then recursively printing this using str would
produce '[]', which would be confusing.

Regards,
Martin


From jepler at unpythonic.net  Sat Sep 21 23:01:11 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Sat, 21 Sep 2002 22:01:11 -0500
Subject: Math optimization
In-Reply-To: <3D8D0612.A6CE7011@bellatlantic.net>
References: <6T5j9.145247$pX1.5195840@news2.tin.it> <3D8D02A0.FF572334@bellatlantic.net> <3D8D0612.A6CE7011@bellatlantic.net>
Message-ID: <20020921220111.A2274@unpythonic.net>

... and your results will change again if you work at function scope, not
global scope.  Lookups of locals or constants are faster than lookups of globals.

I still doubt you'll measure a substantial difference between * and / on
ints, though.  I am getting 1.94 vs 2.08 seconds on 1,000,000 iterations,
which is actually bigger than I expected.

Comparing a*b to a*3 (and likewise for division) also surprised me:  a*b is
faster in function scope (and likewise for division).

If you are planning to do a lot of operations of this kind, for instance on
long lists, then what you're actually looking for is the "Numeric"
extension, which can perform these tasks at C speeds.

Jeff

r = range(1000000)

def f(x):
    for i in r:
	res = x * 3

def g(x):
    for i in r:
	res = x / 3

def h(x):
    for i in r:
	pass

def j(x, y):
    for i in r:
	res = x * y

def k(x, y):
    for i in r:
	res = x / y

import time

for fun in (f, g, h):
    t0 = time.time()
    fun(7)
    t1 = time.time()
    print fun.__name__, t1-t0 

for fun in (j, k):
    t0 = time.time()
    fun(7, 3)
    t1 = time.time()
    print fun.__name__, t1-t0 



From jepler at unpythonic.net  Wed Sep 11 11:10:00 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Wed, 11 Sep 2002 10:10:00 -0500
Subject: ANN: Thinking in Tkinter
In-Reply-To: <b16e4ef7.0209110641.256799a0@posting.google.com>
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid> <alln71$1s2s$1@nntp6.u.washington.edu> <b16e4ef7.0209110641.256799a0@posting.google.com>
Message-ID: <20020911150949.GB28720@unpythonic.net>

On Wed, Sep 11, 2002 at 07:41:29AM -0700, Stephen Ferg wrote:
> So no combination of "bind()" statements (well, no *easily coded*
> combination of bind() and callback logic) can emulate this behavior.

Sure it is.  The code comes to < 50 lines of tcl, surely a similar
number in Python.  See my other message.

Jeff



From inyeol_lee at yahoo.com  Thu Sep  5 05:04:35 2002
From: inyeol_lee at yahoo.com (Inyeol Lee)
Date: 5 Sep 2002 02:04:35 -0700
Subject: How to get function name which raised IOError/OSError?
Message-ID: <cf701bae.0209050104.170ad403@posting.google.com>

Is there any way to get function name which raised IO/OSError?
I want to write an error message formatter like;

def print_oserror():
  e = sys.exc_info()
  func = get_func(e)   # I don't know how to implement this.
  print "cannot %s %s: %s." % (func, e[1].filename, e[1].strerror)

try:
  os.mkdir("foo")
  os.chdir("bar")
  os.rmdir("baz")
except OSError:
  print_oserror()  # print message like 'cannot chdir bar: blabla...'

Thanks,
Inyeol Lee


From pyth at devel.trillke.net  Sun Sep  1 19:10:14 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 2 Sep 2002 01:10:14 +0200
Subject: Add methods to a class at runtime?
In-Reply-To: <20020901214438.GA4349@lilith.ghaering.test>; from gerhard.haering@gmx.de on Sun, Sep 01, 2002 at 11:44:38PM +0200
References: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com> <slrnan4dl3.18h.gerhard.haering@lilith.my-fqdn.de> <wmvc9.21757$FJ1.3181934@e3500-atl1.usenetserver.com> <20020901214438.GA4349@lilith.ghaering.test>
Message-ID: <20020902011014.H3893@prim.han.de>

Gerhard H?ring wrote:
> * Robert Oschler <Oschler at earthlink.net> [2002-09-01 17:19 -0400]:
> > Ok bear with me guys (Gerhard and Peter).  I'm not lazy, the reason I
> > don't test certain things I should is because Python is capable of
> > such amazing introspective dynamism that, having a _very_ long
> > experience bank in strongly typed "static" languages like C++ and
> > Pascal, I find I don't even consider yet even trying certain
> > possibilies that Python is capable of.
> > Some interesting tidbits I found while trying out your suggestions, as
> > you will see in the two ".py" files that follow, I discovered that to
> > create a method you can dynamically add to a class _instance_, you
> > need to omit the 'self' parameter in the method declaration.
> 
> Yep, I found that out today too, when I wrote my last reply. Though I've
> never used that in real life.
> 
> > [...] I'm running 2.1.1, anybody know if 2.2.2 is going "disable"
> > either of these abilities?
> 
> It will be strictly (even binary) compatible with Python 2.2.{0|1}, so
> the answer is that these features will stay.

Even with future version you should be able to rely on this. 
Attribute names are first looked up in the instance and if they are bound 
to a function object you can call it. 

> >>> class Foo:
> ...   def action(self): print "foo"
> ... 
> >>> class Bar:
> ...   def action(self): print "bar"
> ... 
> >>> foo = Foo()
> >>> foo.__class__
> <class __main__.Foo at 0x80d39ac>
> >>> foo.action()
> foo
> >>> foo.__class__ = Bar
> >>> foo.action()
> bar
> >>> foo.__class__
> <class __main__.Bar at 0x80e75c4>
> >>> 
> 
> AFAIK Guido isn't too fond of this feature. I guess it might go away
> some day.

With new-style classes (python 2.2) it's already somewhat restricted:

>>> class a(object):
        pass

>>> class b(dict):
        pass

>>> a1=a()
>>> a1.__class__=b
------------------------------------------------------------
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: __class__ assignment: 'b' object layout differs from 'a'
>>>

regards,

    holger



From eugene at ulka.com  Wed Sep 18 10:07:18 2002
From: eugene at ulka.com (Eugene Pervago)
Date: 18 Sep 2002 07:07:18 -0700
Subject: Loading modules from several directories
References: <a1247c5b.0209171610.243f3c02@posting.google.com> <33803989.0209172247.2e6afc1f@posting.google.com>
Message-ID: <a1247c5b.0209180607.843526e@posting.google.com>

tebeka at cs.bgu.ac.il (Miki Tebeka) wrote in message news:<33803989.0209172247.2e6afc1f at posting.google.com>...
> Hello Eugene,
> Doesn't
> 
> sys.path.append('/somedir/foo')
> sys.path.append('/somedir/foo/project_a')
> import foostuff
> import projectstuff
> 
> works?
> 
> Miki

Miki,
The problem is with following:
>>> sys.path.append('/onedir/foo')
(where /onedir/foo/utils.py exists)
>>> sys.path.append('/anotherdir/project_a/')
(where /aotherdir/project_a/foo/bar.py exists)

>>> import foo.utils
>>> import foo.projectA.bar
ImportError: No module named projectA.bar

If I put the project_a path first, it loads its modules but can't
import the /onedir/foo modules. It seems package can only be in one
place, once Python finds it, it just stops looking elsewhere.

Thanks,
Eugene Pervago


From debl2absolutelynospammy at bellatlantic.net  Sun Sep 15 18:43:30 2002
From: debl2absolutelynospammy at bellatlantic.net (David Lees)
Date: Sun, 15 Sep 2002 22:43:30 GMT
Subject: Good book choice?
References: <e6c4872c.0209150635.4816ef98@posting.google.com>
Message-ID: <3D850D21.9E34DDEF@bellatlantic.net>

I looked at the Rashi Gupta book, but purchased 'Practical Python' by
Magnus Lie Hetland, which I have read about the first 80 pages.  I like
the book, which is written in a conversational style.  There are many
clear short examples of each of the language elements.  If you have a
bookstore you might go browse it and compare the two.

David Lees

Eva wrote:
> 
> Hello all!
> 
> I am new to Python but an experienced programmer in Fortran95 (not too
> object
> oriented). I am looking for a suited Python book. I'd like it to have
> clear ideas about OO, as well as an straight pointing into the core
> language.
> 
> It seems that 'Python Essential Reference (2nd Edition)' is a good
> choice. There is also the book 'Making use of Python' (by Rashi
> Gupta). Is there any user that bought both? Is it good the treatment
> of OO in the former? Is there a lack of applications point of view in
> the former? Could anybody compare them? pros/cons?
> 
> Thanks in advance,
> Eva.


From theller at python.net  Fri Sep 20 14:52:10 2002
From: theller at python.net (Thomas Heller)
Date: Fri, 20 Sep 2002 20:52:10 +0200
Subject: Destructor never called ???
References: <mailman.1032506132.22396.python-list@python.org>
Message-ID: <amfqoq$5iect$1@ID-59885.news.dfncis.de>

Tim Peters wrote:
> If an object is destroyed by GC, and the object has a __del__, the __del__
> is called.
> 
[...]
> 
> What is true is that if an object *in* a trash cycle has a __del__, that
> cycle will not be collected, but rather shuffled into gc.garbage for the
> programmer to do something sane with it explicitly.  There may be any number
> of objects with __del__s not in cycles themselves but reachable only from
> cycles; they don't inhibit collection, and if collection occurs their
> __del__s will be invoked.  It's only the cases where Python can't possibly
> guess a safe order in which to call destructors where both __del__ and
> collection are inhibited.
> 
So what are the recommended practices if my program creates datastructures
containing cyclic references?

- avoid using __del__ in these objects by  cleaning them 'manually'?
- only using weak references to break these cycles?

- and finally using the debugflags of the gc module to ensure everything
is cleaned correctly?

Thomas



From mcherm at destiny.com  Tue Sep 17 09:08:44 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Tue, 17 Sep 2002 09:08:44 -0400
Subject: Python.jobmart.com Server Upgrade
References: <20020916221626.13483.93466.Mailman@mail.python.org> <1020916154114.212AAC/w.lozinski@maya>
Message-ID: <am79g8$gac$1@main.gmane.org>

lozinski at openstepnews.com wrote:
> The python jobmart now allows for full text searches of the embedded resumes.
> So hiring managers can find just the right person with whatever specialty skill
> is required.

For those who, like me, weren't familiar with "the python jobmart", an 
instant's Googling results in:

                 http://python.jobmart.com/

Hey, perhaps everyone else already knows about this one, but *I* hadn't, 
and I wish I had since I just finished a job search in which one of the 
few criteria that I wasn't able to achieve was a job which used Python.

-- Michael Chermside






From buzzard at urubu.freeserve.co.uk  Sat Sep  7 09:44:51 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Sat, 7 Sep 2002 14:44:51 +0100
Subject: Smoothing a discrete set of data
References: <3csmdso5.fsf@morpheus.demon.co.uk>
Message-ID: <alcvvr$882$1@news8.svr.pol.co.uk>

"Paul Moore" <gustav at morpheus.demon.co.uk> wrote in message
news:3csmdso5.fsf at morpheus.demon.co.uk...
> This isn't so much a Python question, as an algorithms question. I
> couldn't find a newsgroup for discussing algorithms (is there a good
> one?) so as I'm implementing the code in Python I thought I'd try here
> - I hope that's OK.
>
> I have a set of data, basically a histogram. The data is pretty
> variable, and I'd like to "smooth" it to find trends. Actually, this
> comes up a *lot* for me - some examples: sampled IO rates on a machine
> - I want to look for trends (is IO higher overnight or during the day,
> etc) or fuel consumption for my car (do I use less fuel since I had
> the service).
>
> Normally, what I do with things like this is draw a graph, and try to
> spot the trends "by eyeball". But it feels to me like I should be able
> to write something which smooths the data out. I just don't see how
> :-)
>
> Take an example - here's a rough ASCII-art diagram
>
>   9     XX                                  XX
>   8     XX XX                               XX
>   7  XX XX XX                               XX XX
>   6  XX XX XX                            XX XX XX
>   5  XX XX XX                            XX XX XX
>   4  XX XX XX                            XX XX XX
>   3  XX XX XX XX          XX          XX XX XX XX
>   2  XX XX XX XX XX    XX XX XX    XX XX XX XX XX
>   1  XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX
>
> This seems to me to have very clear peaks at the left and right, and a
> shallow point in the middle. I'd say that you could "smooth" this to
> something that averaged 8 for the first 3 bars, then 2 for the next 9,
> then 7 for the last 3. But try to make that precise...
>

You could change the bin sizes, but I'd generally keep them equal.

> When I've tried to formalise this, I've always stumbled over the fact
> that it's a trade-off exercise (there most be some sort of "tolerance"
> parameter in the algorithm) - on the one hand, the original data is
> "best" (in the sense that it matches exactly) whereas on the other
> hand, a single horizontal line at the average is "best" (fewest number
> of steps).
>
> My instinct is that you're looking for something that balances number
> of steps vs difference from the original data. But for the life of me
> I can't see how to make this intuition precise.
>
> Can anyone help me? I can't believe that this problem has never come
> up before, but I can't find any literature on it.
>
> Thanks in advance,
> Paul Moore.

You could use something like a moving average, but I'm not sure that will
help you answer any specific questions.  It sounds like you might be
interested in comparing means (for fuel consumption), in which case you'd do
something like a (paired?) t-test (and you wouldn't generally throw away
information by smoothing the data).

IO rates and fuel consumption would generally be approached differently
because the data generating processes have different characteristics.  (I
wouldn't try to answer them both using a single method.)  This really sounds
like a job that needs a statistical approach.

Duncan




From llothar at web.de  Tue Sep  3 09:49:31 2002
From: llothar at web.de (Lothar Scholz)
Date: Tue, 03 Sep 2002 15:49:31 +0200
Subject: Python to surpass C performance by 2030
References: <kuptw4o65q.fsf@lasipalatsi.fi>
Message-ID: <vbf9nusj5sa6b024vt5ouhtcmifrumlt44@4ax.com>

On 27 Aug 2002 15:08:49 +0300, Erno Kuusela <erno-news at erno.iki.fi>
wrote:

>learn python today, and in 28 years you will have a job writing
>the speed-critical parts of c programs in python :)

Cool, lets go to the next bank and start a python .com company.
You can get the CEO of the ysear with this idea.

Please tell me when i can buy stock options from your company :-)


From erikprice at mac.com  Tue Sep 10 17:09:42 2002
From: erikprice at mac.com (Erik Price)
Date: Tue, 10 Sep 2002 17:09:42 -0400
Subject: Python iCalendar API?
In-Reply-To: <unslu0cjragpf0@corp.supernews.com>
Message-ID: <A0229C10-C501-11D6-A54B-00039351FE6A@mac.com>

On Tuesday, September 10, 2002, at 04:36  PM, A.M. Kuchling wrote:

> In article <slrnans8f6.jc.gerhard.haering at lilith.my-fqdn.de>,
>    Gerhard H?ring wrote:
>> Looks like there is:
>> http://webcvs.kde.org/cgi-bin/cvsweb.cgi/kdepim/libical/src/python/
>
> This seems to be a copy of the libical library; its official home page
> is at http://softwarestudio.org/libical/ .

Thanks, Gerhard and amk -- I'll have to check it out.  New project!!!


Erik





--
Erik Price                                                  (zombies 
roam)

email: erikprice at mac.com
jabber: erikprice at jabber.org




From bobx at linuxmail.org  Wed Sep 11 21:20:00 2002
From: bobx at linuxmail.org (Bob X)
Date: Thu, 12 Sep 2002 01:20:00 GMT
Subject: [FEEDBACK] Is this script efficient...is there a better way?
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <L9Rf9.168314$l_4.127784@atlpnn01.usenetserver.com>
Message-ID: <4ZRf9.1782$Lg2.248733@news2.news.adelphia.net>

Steve Holden wrote:
<snip>
>>#!/usr/local/bin/python -w
>>
>># import the needed libs
>>import sys, string
>>
>># make sure the command line arguments are there
>>if len(sys.argv) < 3:
>>     print "usage: fread.py [log file] [hit file]"
>>     sys.exit(1)
>>
> 
> Unless the return is important, you'll find
> 
>     sys.exit("usage: fread.py [log file] [hit file]")
> 
> more convenient.
> 
> 
>># open the files with some error checking
>>try:
>>     inFile = open(sys.argv[1],"r")
>>except IOError:
>>     print "Cannot open log file!\n"
>>     sys.exit(1)
>>
>>try:
>>     outFile = open(sys.argv[2],"w")
>>except IOError:
>>     print "Cannot open hits file!\n"
>>     sys.exit(1)
>>
> 
> Your error checking is exemplary. Normally I'm happy to have a Python
> exception explain the error condition, but of course your explicit messages
> are great.
> 
> 
>># build list of keywords
>>kw = [ "some", "words" ]
>>
>># loop through the list and print the lines to a file
>>for line in inFile.readlines():
>>     for badword in kw:
>>         if line.find(badword) > -1:
>>             result = '%s %s' % (badword, line)
>>             print result            # Print the result
>>             outFile.write(result)   # Write the result
>>
> 
> If you only wanted to know the lines that match at least one of the words
> you could follow the write() with a "break", since there's no need to
> continue the loop then.
> 
> Also, the "print" looks like a debugging statement.
> 
> 
>># close the files
>>inFile.close()
>>outFile.close()
>>
>># let me know when it's done
>>print "Finished processing file..."
>>
> 
> Highly commendable. Another approach would be to build a regular expression
> from the keywords list and search for the r.e. in each line. The following
> code is untested...
> 
> kw = ["some", "words"]
> kwpat = "|".join(kw)
> pat = re.compile(kwpat)
> for line in inFile.readlines():
>     if pat.search(line):
>         outFile.write(line)
> 
> for example.
> 
> regards
> -----------------------------------------------------------------------
> Steve Holden                                  http://www.holdenweb.com/
> Python Web Programming                 http://pydish.holdenweb.com/pwp/
> Previous .sig file retired to                    www.homeforoldsigs.com
> -----------------------------------------------------------------------
> 
> 
> 
Thank you very much...as a follow up:

#1 Can I use the same sys.exit("message here") on the IO errors as well?
#2 For the "print looks like a debugging statement"...how would it be 
done different...in the context of my script and not a total re-write?
#3 I will test your shorter one as I have stayed away from RE till now.
#4 On your shorter one should xreadlines still be done for memory sake?

Thanks!

Bob



From robin at jessikat.fsnet.co.uk  Wed Sep  4 17:17:30 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 4 Sep 2002 22:17:30 +0100
Subject: jython memory usage
References: <hVm5ydAY2hd9Eww$@jessikat.demon.co.uk>
 <3d76366a$1_5@news.bluewin.ch> <rnLbuKAv5jd9Ewzz@jessikat.demon.co.uk>
 <3d76494d$1_5@news.bluewin.ch> <YR13DVAUNld9EwXH@jessikat.demon.co.uk>
 <3d765ce6_4@news.bluewin.ch>
Message-ID: <zzKd7KAqhnd9EwxT@jessikat.fsnet.co.uk>

.....
it, test_pdfbase_ttfonts.py doesn't look as though it's doing anything
particularly odd, certainly I see no references to system/spawn in
there. 
-- 
Robin Becker


From melanie at hotmail.com  Sun Sep 22 10:17:37 2002
From: melanie at hotmail.com (melanie at hotmail.com)
Date: 22 Sep 2002 14:17:37 GMT
Subject: School Corporal Punishment
Message-ID: <amkje1$j85$11471@news5.svr.pol.co.uk>

--_NextPart_00009291-000018EB-00838C64-53A0
Content-Type: text/plain
Content-Transfer-Encoding: 7bit

There's a great new forum called "School Corporal Punishment" at http://www.network54.com/Forum/198833
--_NextPart_00009291-000018EB-00838C64-53A0--


From aleax at aleax.it  Fri Sep 27 03:47:09 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 27 Sep 2002 07:47:09 GMT
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <amvcev$cf6$1@s1.read.news.oleane.net> <7nHk9.172208$pX1.6117207@news2.tin.it> <amvjr0$f1v$1@s1.read.news.oleane.net>
Message-ID: <12Uk9.175615$pX1.6212603@news2.tin.it>

Pedro RODRIGUEZ wrote:
        ...
>> The tricks deal with ways to work around this problem, by ensuring that
>> and's RHO can never be false, e.g.
>>         zz = (a and [b] or [c])[0]
> 
> If python had short-circuiting operators ('and then', 'or else'), this
> will be the ternary operator :
>     zz = (a and then [b] or else [c])[0]

Python does have short-circuiting operators, and they're spelled 'and'
and 'or', exactly like I coded in the snippet you quote.


> The problem when talking about conditionnal expression should be to know
> what really matters in the context :
> - the short-circuit evaluation (only (a,b) or (a,c) are evaluated)
> - or the ability to pick a value in a list
> 
> The first one isn't possible in a Python expression AFAIK,

Of course it is -- you seem to keep ignoring the semantics of Python's
"and" and "or" operators, which ARE short-circuiting (and return one of
the operands, differently from C/C++'s short-circuiting && and ||,
which return 0 or 1 whatever operands they're given).

It's just an ugly hack, that's all.


Alex



From gua81 at XXXyahoo.com  Thu Sep 26 04:24:18 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Thu, 26 Sep 2002 18:24:18 +1000
Subject: LInking cgi error
Message-ID: <amug6m$ivm$1@lust.ihug.co.nz>

This thing worked before.... I dont know why it didn't work now.

this code is in the other page with the link:
'<a href="parser.cgi?link=%s/%s">%s</a>'% (dirname, filename, filename)

This code is the one that crash:
all_values=cgi.FieldStorage()
kkk=all_values["link"].value
start_parse(kkk)

when I run it, it gives this error:
Traceback (most recent call last):
  File "parser.cgi", line 53, in ?
    direc=all_values["link"].value
  File "/usr/local/lib/python2.1/cgi.py", line 550, in __getitem__
    raise KeyError, key
KeyError: link




From ahindra at hotmail.com  Sun Sep 15 19:12:41 2002
From: ahindra at hotmail.com (=?ISO-8859-1?Q?Andr=E9s?=)
Date: 15 Sep 2002 16:12:41 -0700
Subject: Python "compiler" is  too slow for processing large data files???
References: <3d6d097d$0$321$e2e8da3@nntp.cts.com>
Message-ID: <5cc917de.0209151512.1a9e4941@posting.google.com>

If you don't need the data files to be human readable you could
improve speed saving the lists as python objects with pickle, cpickle
or marshal modules, then exec would be faster.


From lee100 at sport.rr.com  Tue Sep 17 20:55:16 2002
From: lee100 at sport.rr.com (Lee Gray)
Date: Wed, 18 Sep 2002 00:55:16 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <mailman.1032300272.15953.python-list@python.org>
Message-ID: <U9Qh9.479731$q53.16320265@twister.austin.rr.com>

Phil Thompson wrote:

> On Tuesday 17 September 2002 10:05 pm, Lee Gray wrote:
>>
>> Can someone point me to a good Python GUI app to demonstrate (other
>> than IDLE)?
> 
> http://www.abo.fi/~iporres/Projects/fog0000000023.html
> 
> For SQL based applications PyQt contains a set of data aware widgets. PyQt
> includes an example SQL browser, query, reporter in 180 lines of Python.
> 
> Phil

This looks very interesting!

Thanks,
Lee


From peeryc at hotmail.com  Fri Sep 27 15:11:54 2002
From: peeryc at hotmail.com (Christopher Peery)
Date: 27 Sep 2002 12:11:54 -0700
Subject: Python Process that can survive
References: <mailman.1033137937.24205.python-list@python.org>
Message-ID: <571739c3.0209271111.6514c32f@posting.google.com>

sismex01 at hebmex.com wrote in message news:<mailman.1033137937.24205.python-list at python.org>...
> > From: peeryc at hotmail.com [mailto:peeryc at hotmail.com]
> > 
> > Alright... I love python but I think I may have found a definite
> > problem. I can't seem to run a python process, kick it to the
> > back-ground, and then kill the terminal that started it. With most
> > other things, the python process would keep running in the
> > back-ground regardless. For me, the process always dies with the
> > terminal.
> > 
> > So my question: is there a way to separate a python script from the
> > underlying terminal so that this will not happen.
> > 
> > I'm hoping for some scheme that's supported by the language. I've
> > found a few hacks on-line where the python process is daemonized
> > but this requires shutting down all the standard io streams.
> > There's got to be a better way especially since it can be done in
> > Perl very easily with no special tricks.
> > 
> > By the way, I'm doing this on a linux box.
> > 
> > Chris
> > 
> 
> This same question I asked a few weeks ago, and got a few
> answers back from lotsa nice peepl, so I pass this function
> on to you for it's safekeeping (tag, your it ;-)
> 
> def Daemonize():
>    "Detach the current process from the console."
>    import os, sys
>    if os.fork(): os. exit(0)
> 
>    # Close connection to current console.
>    os.close(sys.  stdin  .fileno())
>    os.close(sys.  stdout  .fileno())
>    os.close(sys.  stderr  .fileno())
> 
>    # Reopen std handles to /dev/null.
>    dev null = os.open("/dev/null", 0)
>    os.dup2(dev null, 0)
>    os.dup2(dev null, 1)
>    os.dup2(dev null, 2)
> 
>    # Create new OS session.
>    os.setsid()
> 
>    # Reset UMASK.
>    os.umask(0)
> 
>    # Refork to have init as parent.
>    if os.fork(): os. exit(0)
> 
> 
> This is a synthesis of several replies I received, and it
> seems to work, but others might have some observations
> to make to this little function.
> 
> HTH!
> 
> -gustavo
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Advertencia: 
> La informacion contenida en este mensaje es confidencial y restringida 
> y
> esta destinada unicamente para el uso de la persona arriba indicada, 
> Esta
> comunicacion representa la opinion personal del remitente y no refleja
> necesariamente la opinion de la Compa ia. Se le notifica que esta
> estrictamente prohibida cualquier difusion, distribucion o copia de 
> este
> mensaje. Si ha recibido esta comunicacion o copia de este mensaje por 
> error,
> o si hay problemas en la transmision, favor de comunicarse con el 
> remitente.
> 
> 
> Todo el correo electr nico enviado para o desde esta direcci n 
> ser 
> procesado por el sistema de correo corporativo de HEB. Tal correo
> electr nico esta sujeto a ser almacenado y puede ser revisado por 
> alguien
> ajeno al recipiente autorizado con el prop sito de monitorear que se 
> cumplan
> las normas de seguridad de la empresa.


Yeah I found this hack on the web and I can't believe that this is
only way to do this in python. You're basically stripping the process
of all it's stream and sending it on its way. This is really messy.

Plus in my lab there are a bunch of perl guys and a few python people.
To date, what one can do in their script the python guys can do to
(only nicer and much more readible). Perl has the ability to have a
process continue after the underlying terminal is killed if the
process is in the background. What is python doing that doesn't allow
this?

Chris


From heikowu at ceosg.de  Sat Sep  7 13:14:04 2002
From: heikowu at ceosg.de (Heiko Wundram)
Date: 07 Sep 2002 19:14:04 +0200
Subject: Killing/Terminating a running thread
Message-ID: <1031418844.17923.94.camel@d168.stw.stud.uni-saarland.de>

Hi to all!

I've looked over google groups, but I've found no real answer that
satisfied the needs I have. Just a little explanation:

I have a Queue module, which stores method-calls and delegates them to
threads, which take over running the call. Now it happens that a method
takes much longer to complete than I would like it to, and in this case,
the thread running the code should just be terminated, discarding all
values computed so far.

The basic problem I have that it is pretty much impossible to get at the
PID of the thread running the code in question; but I have a reference
to the thread class in a central "garbage-collector" (the queue object).
And anyway, this approach would only work on Unix (I guess), when using
pthreads, which create a "real" subprocess with its own PID for a
thread.

Has anyone had a similar problem, and know of a solution?

Yours,

	Heiko Wundram
	Netzwart Wohnheim-D
	Zimmer 2206 - Universit?t 18 - 66123 Saarbr?cken
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 196 bytes
Desc: Dies ist ein digital signierter Nachrichtenteil
URL: <http://mail.python.org/pipermail/python-list/attachments/20020907/32863056/attachment.sig>

From magnus at thinkware.se  Tue Sep 17 15:03:34 2002
From: magnus at thinkware.se (=?ISO-8859-1?Q?Magnus_Lyck=E5?=)
Date: Tue, 17 Sep 2002 21:03:34 +0200
Subject: Thinking of joining
References: <3D7970DC.5115BCC8@earthlink.net> <20020907095211.24938.00000006@mb-mt.aol.com>
Message-ID: <3D877C86.4090403@thinkware.se>

TeaAndBikkie wrote:
> and won't leave
> you pining for the fields. 

Or fjords!



From roy at panix.com  Mon Sep 16 15:16:16 2002
From: roy at panix.com (Roy Smith)
Date: 16 Sep 2002 15:16:16 -0400
Subject: Converting Python app to C++ completely
References: <aklm6c$r1r$1@news.swman.net.uk> <3D85D9EB.9020102@thinkware.se> <roy-2A7271.09563416092002@reader1.panix.com> <3D86209F.4080501@earthlink.nospam>
Message-ID: <am5am0$fi6$1@panix2.panix.com>

Joe Knapka  <jknapka at earthlink.nospam> wrote:
> You might try rewriting it in Python, using lists of characters
> instead of strings for the stuff that's causing performance
> trouble. Given str, lst = list(str), then do all the processing
> (in place) on lst, and ''.join(lst) to turn it back into a string.

An interesting idea, I'll try it!

The only problem is, I really had two reasons for converting it
in the first place.  One was to improve performance, the other was to
learn C++.  If I solve the first problem in Python, I won't solve the
second problem at all :-(



From henrik.motakef at web.de  Tue Sep 17 15:24:30 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 17 Sep 2002 21:24:30 +0200
Subject: CGI question
References: <mailman.1032289042.7889.python-list@python.org>
Message-ID: <873cs8l8r5.fsf@pokey.henrik-motakef.de>

"David LeBlanc" <whisper at oz.net> writes:

>>> When I want to present another
>>> page, I want the url to look like this: http://my.domain.name/?page.
{...]
>>> Now, my problem is to find out what follows behind the question mark in
>>> my CGI script.

Everything after the question mark will end up in the QUERY_STRING
environment variable (as a raw, URL encoded string).

> From what I've heard, the difficulties of writing mod_rewrite rules is
> enough to make even seasoned apache warriors shudder.

Maybe that's the point where one should mention mod_rewrite's
documentation:
http://httpd.apache.org/docs/mod/mod_rewrite.html

There is also a "rewrite guide" with lots of practical examples:
http://httpd.apache.org/docs/misc/rewriteguide.html

hth
Henrik


From nhodgson at bigpond.net.au  Wed Sep  4 19:19:05 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Wed, 04 Sep 2002 23:19:05 GMT
Subject: embeddable ide?
References: <Ltpd9.6669$yt3.3168986@newssrv26.news.prodigy.com>
Message-ID: <Jxwd9.15932$Dr5.66078@news-server.bigpond.net.au>

Stevek wrote:
> We're planning on making extensive use of Python to provide our users with a
> powerful scripting mechanism.  I've successfully embedded and extended
> Python but now I want to give the user a decent development environment for
> writing their scripts.

    The approach that is most likely to succeed is to take one of the 
current IDEs and reshape it to fit into your application. Full IDEs are 
quite complex beasts and so this will require quite a bit of effort and 
learning. Open source IDEs like PythonWin or Boa Constructor would be 
reasonable candidates although the commercial IDE vendors are also 
likely to be open to approaches. As Gerhard says a major criterion is 
that you should choose IDE code that uses the same UI toolkit as your 
application.

    Creating a Python IDE component would be a very interesting but 
difficult project. There are Visual Basic IDE components such as
http://msdn.microsoft.com/vba/default.asp
    An important issue in producing an IDE component is in defining a 
reasonable scope for the component so that it allows a great deal of 
flexibility but also allows the application to define limitations like 
source code being stored within the application's database rather than 
in files.

    Neil



From tjreedy at udel.edu  Sat Sep 14 11:58:58 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 14 Sep 2002 15:58:58 GMT
Subject: question about dictionary type..
References: <alt22f$pf8$1@newsreader.mailgate.org> <alv4op$m67$1@newsreader2.netcologne.de>
Message-ID: <61Jg9.16334$5r1.665293@bin5.nnrp.aus1.giganews.com>

> Interesting is the line "start[key] = start = {}", because first
> "start[key] = {}" is evaluated, then "start = {}" (with "{}" being
the
> same object as in the first assignment). You can check this with
"dis", as
> I've read in another posting in this newsgroup:

Since this is documented in Ref Manual 6.3

"An assignment statement evaluates the expression list (remember that
this can be a single expression or a comma-separated list, the latter
yielding a tuple) and assigns the single resulting object to each of
the target lists, from left to right."

you can count on this behaviour persisting in future versions.

TJR







From snacht at xs4all.nl  Thu Sep 26 05:10:12 2002
From: snacht at xs4all.nl (Stefaan Nachtergaele)
Date: Thu, 26 Sep 2002 11:10:12 +0200
Subject: jython web scripting options
Message-ID: <amuj0p$orl$1@news1.xs4all.nl>

Hi all,
I'm prototyping a three tier web application in java. I've looked at a lot
of java web frameworks but for development speed nothing compares to web
scripting languages like php, etc. But that's not an option for me cause I
need to talk with my server side java code (php-java integration is dinky).
My question is, what options are there for jython.

I've looked around and I found PSP (http://www.ciobriefings.com/psp/). It
actually seems perferct for me. Has anyone had any experience with this? Is
it any good?

Are there any such other options than this? I notice there's a lot of
solutions for Python, but I need the java integration so that does not seem
to be an option for me.

Advise welcome.

Regards,
Stefaan Nachtergaele





From marklists at mceahern.com  Wed Sep 18 07:25:09 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 18 Sep 2002 06:25:09 -0500
Subject: monetary applications [was: Python GUI app to impress the boss?]
In-Reply-To: <20020918052920.GA81158@hishome.net>
Message-ID: <JHEOKEOOLIGLDHCMAHMOCEHLDPAA.marklists@mceahern.com>

[Oren Tirosh]
> I don't understand what exactly you are complaining about but is sounds
> interesting - can you elaborate?

Perhaps the lack of a FixedPoint type?

// m
-




From mertz at gnosis.cx  Mon Sep 16 20:42:17 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Mon, 16 Sep 2002 20:42:17 -0400
Subject: [OT?] Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de>
 <pUdh9.9377$yt3.4604908@newssrv26.news.prodigy.com>
 <mailman.1032184821.5180.python-list@python.org>
 <c%qh9.102794$ub2.2120685@news1.tin.it>
 <mailman.1032212180.11973.python-list@python.org>
Message-ID: <ppnh9kKkXUBX092yn@gnosis.cx>         

Lulu of the Lotus-Eaters wrote:
> I would reformulate the issue slightly:  If you BELIEVE your code is
> that valuable, seek psychiatric help.

BT (or a billion other folks with the same neurosis) retorted:
|> I really don't understand posts like these.
|> You think I'm fool to protect my work?

Cliff Wells <LogiplexSoftware at earthlink.net> wrote previously:
|Obscuring the code will dissuade only the most casual or incompetent
|from decompiling/disassembling your code.  Sort of like hiding a spare
|key to your house under the "Welcome" mat.  And that would be illegal,
|hence the suggestion to *copyright* your work.

>From a factual perspective, Cliff is right.  Copyright is the means by
which source code can be protected (not that I think copyright laws are
good ones, but that's how things actually are).  Obscurity isn't.

But even though my prior remark was mostly meant as a little quip, there
was more to it than just that.  There is a particular neurosis that goes
into the many grammatically flawed, defensive, and self-aggrandizing
posts that one sees so frequently on Usenet.  The obsession with
protecting code that has not yet been written, but whose greatness is
nonetheless certain is just one variant--various other "secrets" and
"insights" work the same way.  I exaggerate by saying "see psychiatric
help"--the neurosis is usually not quite as severe as to require that.
But at the risk of being too vulgarly Freudian, it seems like an anal
retentive obsessive complex that is going on with this phobic fear of
parting with something so very precious.

Yours, Lulu...

P.S. Horkheimer and Adorno say societies can have neuroses too, not
just individuals.  Just look at the RIAA, DVDCCA, Disney (&the Senator
from Disney), MS, and so on.

--
 mertz@  _/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: \_\_\_\_    n o
gnosis  _/_/             Postmodern Enterprises            \_\_
.cx    _/_/                                                 \_\_  d o
      _/_/_/ IN A WORLD W/O WALLS, THERE WOULD BE NO GATES \_\_\_ z e




From aleax at aleax.it  Fri Sep 20 11:05:35 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 15:05:35 GMT
Subject: Destructor never called ???
References: <76710511.0209111450.40999dd8@posting.google.com> <ame8es$ppc$1@panix2.panix.com> <buzi9.136158$pX1.4848554@news2.tin.it> <amfd37$r1a$1@panix2.panix.com>
Message-ID: <3PGi9.119651$ub2.2538597@news1.tin.it>

Aahz wrote:

> In article <buzi9.136158$pX1.4848554 at news2.tin.it>,
> Alex Martelli  <aleax at aleax.it> wrote:
>>Aahz wrote:
>>>
>>> There are several different issues.  First of all, __del__ does not get
>>> called if an object is destroyed by GC (GC gets used to clean up cyclic
>>> garbage).  Secondly, Jython does not use refcounting; it's pure GC, just
>>
>>Right, but, I'd put it differently: GC is inhibited from destroying
>>objects whose types define a __del__ -- rather, it collects them into
>>gc.garbage, because it can't know that finalizing them is safe.
> 
> Thanks (and also to Tim) for correcting my sloppy writing.

Reading Tim's post reminded me that _my own_ expression was quite
imperfect too  -- GC doesn't destroy objects *that are part of the
loop* and have a __del__... it DOES destroy object that have a
__del__ if they're not part of the loop itself, but rather, are
kept alive only by being referenced from objects in the loop, if
and when the loop itself is collected/destroyed.  Phew...


Alex



From narnett at mccmedia.com  Wed Sep  4 12:40:17 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Wed, 4 Sep 2002 09:40:17 -0700
Subject: Bind threads to addresses -- Windows & urllib?
Message-ID: <PHEDLIMPMIAGAEPEFBGCOEDJCHAA.narnett@mccmedia.com>

I can't see an obvious easy way to bind threads to IP addresses on a Win2K
machine.  I have a multi-threaded Web spider, which uses urllib to retrieve
pages, and would like to bind each thread to a different address on the
machine, if possible.  Alternatively, I'd be willing to run separate
processes to accomplish the same.  So far, my searches aren't coming up with
any solution... And it would be wonderful if such an approach were also
portable to Linux...  Pointers?

--
Nick Arnett
Phone/fax: (408) 904-7198
narnett at mccmedia.com





From dsavitsk at e-coli.net  Wed Sep 18 22:54:47 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Thu, 19 Sep 2002 02:54:47 GMT
Subject: Catching results of an exec
References: <3d88f701$0$304$39cecf19@nnrp1.twtelecom.net>
Message-ID: <X%ai9.85$005.52035@newssrv26.news.prodigy.com>

>>> class foo:
...     def do_it(self, arg):
...         exec(arg)
...     def get_it(self, arg):
...         return eval(arg)

>>> f = foo()
>>> f.do_it('self.a = "bar"')
>>> b = f.get_it('self.a')
bar

is this too simple?

-d


"Greg Fortune" <lists at gregfortune.com> wrote in message
news:3d88f701$0$304$39cecf19 at nnrp1.twtelecom.net...
> Is there any way to "catch" the value produced by an exec?  I'm
considering
> things like print statements that don't really assign a value, but do send
> a result to stdout.
>
> For example, if I execute
>
> exec('print dir()')
>
> from an interpreter, there will be some stuff spit out on the console.
> What if I really wanted to display the list from dir() in a different way?
> I'm working on a really dinky server that will sit on a machine and
execute
> arbitrary python code.  I want the results from the code execution to be
> returned to client.
>
> Even just eval('dir()') exhibits the desired behavior, but only from an
> interactive interpreter.
>
> Right now I've got a nasty hack that redirects stdout, uses an
> InteractiveInterpereter from the code module to execute the code, writes
to
> an output file, reads the result from the file, restores stdout and then
> sends the result to the client.  Hopefully, that's a terrible way to do
it.
>  It just occured to me that I can probably use a pipe in place of the
> output file, but still....
>
> It would be great if exec and eval set something somewhere that held the
> results of the code execution.  I looked in the InteractiveInterpreter
> (which, at least, kicks the results of the code execution out to stdout),
> but all it does is
>
> exec code in self.locals
>
> Did I miss something?
>
> Greg Fortune




From klaatu at nospam.invalid  Fri Sep 20 14:18:56 2002
From: klaatu at nospam.invalid (Klaatu)
Date: Fri, 20 Sep 2002 18:18:56 GMT
Subject: win32comext problem w/adsi
Message-ID: <Xns928F730726CC1klaatuburadanikto@64.164.98.7>

Python 2.2.1, win32comext (win32all) 148, Windows 2000 SP3

I've installed the Win32 Extensions to Python from 
http://starship.python.net/crew/skippy/win32/Downloads.html which states 
that the 148 version is for Python 2.2. I also have Python 2.2.1 from 
http://www.python.org/2.2.1

I'm just trying to run test.py stored in (on my system) "D:\Python22\Lib
\site-packages\win32comext\adsi". Executing

python test.py

from the command line gives me a dialog stating that "The dynamic link 
library pythoncom21.dll could not be found in the specified path...", 
followed by this traceback:

Traceback (most recent call last):
  File "test.py", line 5, in ?
    from win32com.adsi import *
  File "D:\Python22\lib\site-packages\win32comext\adsi\__init__.py", line 
25, in
 ?
    from adsi import *
ImportError: DLL load failed: The specified module could not be found.

The string "pythoncom21.dll" is only found within one file: adsi.pyd in 
that same folder. The DLL does not exist on my system. I do have a 
pythoncom22.dll in my system32 folder.

Is this a problem with the Win32 Extensions package that the author of 
the package will have to fix? Or is there something I can do to fix this 
problem, given my fairly limited knowledge of Python. Any other 
suggestions?

TIA


From rjones at ekit-inc.com  Thu Sep 12 17:56:46 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Fri, 13 Sep 2002 07:56:46 +1000
Subject: ANN: NINZ 1.2.1
In-Reply-To: <20020912135926.A7584@ibook.distro.conectiva>
References: <20020912135926.A7584@ibook.distro.conectiva>
Message-ID: <200209130756.46405.rjones@ekit-inc.com>

On Fri, 13 Sep 2002 2:59 am, Gustavo Niemeyer wrote:
> NINZ stands for NINZ Is Not ZSI.

And ZSI would be?


    Richard




From python at conquered.org  Thu Sep  5 09:55:29 2002
From: python at conquered.org (Luis)
Date: Thu, 5 Sep 2002 07:55:29 -0600
Subject: Python class ( Prolog ) ???
Message-ID: <001001c254e3$e5fdcc80$6a01a8c0@LUISLAPTOP>

Does anyone have a class that provides the functionality of prolog ???
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020905/b09f590b/attachment.html>

From fredrik at pythonware.com  Wed Sep  4 03:14:09 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 04 Sep 2002 07:14:09 GMT
Subject: ANN: New PEP Format: reStructuredText
References: <mailman.1030681307.29889.python-list@python.org> <0z+c7LA8mHc9EwYa@jessikat.fsnet.co.uk> <slrnan1b1q.pmk.mlh@furu.idi.ntnu.no>  <mailman.1030796387.5329.python-list@python.org> <slrnan1ead.r05.mlh@furu.idi.ntnu.no>
Message-ID: <5pid9.5884$e5.1033477@newsb.telia.net>

Magnus Lie Hetland wrote:

> I like plain-text formats, but I reStructuredText is too convoluted
> (and has too many strange constructs, like lots of underscores and
> double colons etc) for my taste. I'm sure many others (among them, I'm
> sure, the designers) will find it to be the perfect balance of
> simplicity and functionality.

as long as their logos are always exactly 134 pixels wide,
of course...

</F>




From emile at fenx.com  Mon Sep 16 19:31:52 2002
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 16 Sep 2002 16:31:52 -0700
Subject: Not sure how to phrase this question
References: <000501c25dd4$38711460$9865fea9@arthur>
Message-ID: <am5pi6$55a$1@main.gmane.org>

Arthur:
> Would hate to have to resort to truely understanding
> this stuff to get to where I am hoping to get. 
> 

It would help to know where that is... ;-)


--

Emile van Sebille
emile at fenx.com

---------






From thorsten at thorstenkampe.de  Sat Sep  7 10:46:07 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Sat, 7 Sep 2002 16:46:07 +0200
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <3D7958DB.50903@nospam.free.fr>
Message-ID: <ald3g5$1p8lpp$1@ID-77524.news.dfncis.de>

* laotseu
>Rod Stephenson wrote:
>> Slashdot has a list of questions posed to Larry Wall (perl). When
>> asked his thoughts on other scripting languages, he makes the
>> following observation about python
>> 
>> "Python is cool to look at small bits of, but I think the "outline"
>> syntax breaks down with larger chunks of code. I'm with Aristotle on
>> the structure of discourse--a story should have a beginning, and
>> middle, and an end. So should blocks"
>> [...]
>
> def fun(args):
> #BEGIN
>      if ceci or cela:
>      #BEGIN
>          do_this(with_that)
>          [snip a lot of code]
>      #END
> #END
>
> So what the problem with blocks ?

Nope. Bad, bad readability, sorry.

#||<--- LOOK HERE, LARRY, THIS IS WHERE THE FUNCTION _*BEGINS*_!
def fun(args):
    #||<--- LOOK HERE, LARRY, THIS IS WHERE THE IF-CLAUSE _*BEGINS*_!
    if ceci or cela:
        do_this(with_that)
        [snip a lot of code]
    #||<--- LOOK HERE, LARRY, THIS IS WHERE THE IF-CLAUSE _*ENDS*_!
#||<--- LOOK HERE, LARRY, THIS IS WHERE THE FUNCTION _*ENDS*_!


Thorsten


From eugene1977 at hotmail.com  Fri Sep 13 19:13:23 2002
From: eugene1977 at hotmail.com (eugene kim)
Date: Fri, 13 Sep 2002 23:13:23 +0000
Subject: question about dictionary type..
Message-ID: <alt22f$pf8$1@newsreader.mailgate.org>

hi..

i'd like to have a dictionary which looks like
table = { category1 : { category2 : { category3 : 'garbage value' } } }

category1 has many category2
category2 has many category3

i expected
table[ 'computer' ][ 'language' ][ 'python' ] = 0
to create {computer : { language : { python :0 }}}
table[ 'computer' ][ 'language' ][ 'java' ] = 0
to becomes { 'computer' : { 'language' : { 'python' :0 , 'java' :0 }}}

but python doesn't allow me to do this..
can anyone help me?

thank you


From duncan at NOSPAMrcp.co.uk  Wed Sep  4 08:40:21 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Wed, 4 Sep 2002 12:40:21 +0000 (UTC)
Subject: newsgroup for zope
References: <al4tru$o3j$03$1@news.t-online.com>
Message-ID: <Xns927F8A36AC9Aduncanrcpcouk@127.0.0.1>

stephan <somebody at hotmail.com> wrote in news:al4tru$o3j$03$1 at news.t-
online.com:

> Hi perhaps a stupid question,
> 
> but does there exist a newsgrup for zope centric items?

No newsgroups. Lots of mailing lists though: see the Zope website for 
details: http://www.zope.org/Resources/MailingLists

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From edream at tds.net  Fri Sep 20 11:09:06 2002
From: edream at tds.net (Edward K. Ream)
Date: Fri, 20 Sep 2002 15:09:06 GMT
Subject: Setting cursor colors in tkinter
References: <rvFi9.926$fO.166520@kent.svc.tds.net> <amfbo5$e2r$1@wanadoo.fr>
Message-ID: <mSGi9.1362$fO.179392@kent.svc.tds.net>

> from Tkinter import *
> root = Tk()
> root.configure(cursor='xterm red green')
>
> doesn't produce an error. The colours seem to be taken into account,
but
> the xterm cursor isn't really a good example. I tried
> root.configure(cursor='gumby red green') and the result was what I
expected.
>
> Maybe there are some issues on Windows?

Many thanks, Eric, for this information.

I am indeed running on XP.  Plain "gumby" works on my system, and the color
specs do not.  I ran the following test from wish:

proc createWindowWithCursor {} {
 toplevel .panel
 text     .panel.text
 .panel.text configure -cursor {gumby red green}
 pack .panel.text
}

This yields: bad cursor spec "gumby red green"

-cursor gumby or -cursor {gumby} work as expected, so this looks to me like
a Tk problem on Windows XP.

Edward




From theller at python.net  Thu Sep 26 13:01:09 2002
From: theller at python.net (Thomas Heller)
Date: Thu, 26 Sep 2002 19:01:09 +0200
Subject: Python threading?
References: <mailman.1033054839.28121.python-list@python.org>
Message-ID: <amvegt$9qh1l$1@ID-59885.news.dfncis.de>

Tim Peters wrote:
> [Robert Oschler]
> 
>>Having written _many_ state machines in my life, which async programming
>>boils down to in one form or another, I feel it depends on the task.
>>...
>>Contrast this with a simple procedure call where a linear series of
>>function calls are made, each with appropriate error handling and status
>>adjustment, in a pleasant centrally located manner.
> 
> 
> [Thomas Heller]
> 
>>I wonder if generators can come to a rescue here?
> 
> 
> Absolutely.  Read the start of PEP 255.
> 
> 
>>Still trying to get them into my head...
> 
> 
> Think of them as resumable functions.  You're familiar with functions that
> remember their own data state across calls, via abusing globals, or via
> declaring private static vrbls, in C.  Methods on objects often do the same
> by stuffing their state into instance attributes between calls.  A generator
> remembers its data state across resumptions simply by virtue of that its
> local variables don't vanish, and also remembers its control-flow state
> across resumptions.  Remembering local data state by magic is a real
> convenience, but can be simulated in many other ways with reasonable effort;
> it's the ability to remember and restore control-flow state by magic that
> can make generators more than merely convenient.

I _have_ written kind of hardware descriptions with it - what you do 
with VHDL testbenches normally. This was fun!
There were several generators like this, manipulating clock, reset
and other signals in a counter device. The value yield'ed is simply
the time to wait:

     def run_clock(self):
         while 1:
             self.clock = not self.clock
             yield 50 # wait 50 ns

     def run_reset(self):
         self.reset = 1
         yield 200 # a 200 ns reset pulse
         self.reset = 0

The dispatcher goes like this: Creates the generators,
pushes them onto a priority queue (together with the time
value they will have to be served), serves them by calling
the next() method, and pushes the result back onto the queue:

def run(*gens):
     queue = MyList() # a priority queue
     # insert all generators
     for g in gens:
         queue.put((0, g()))
     # and run them
     while queue:
         now, gen = queue.pop(0)
         try:
             dt = gen.next()
         except StopIteration:
             pass
         else:
             queue.put((now + dt, gen))

--------

OTOH, I've been missing the possibility to pass
values back _to_ the generators - something like maybe
'work = yield event', probably triggered by calling gen.next(work)). 
Wouldn't this be useful?

I fear I cannot explain this clearly, but I think
David Mertz statemachine example would also profit from it.
As I understand it, he passes this 'work' around in global variables.

Thomas



From phil at river-bank.demon.co.uk  Tue Sep 10 18:04:17 2002
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Tue, 10 Sep 2002 23:04:17 +0100
Subject: ANN: PyQt v3.4 Released
Message-ID: <3D7E6C61.7010700@river-bank.demon.co.uk>

Riverbank Computing Limited is pleased to announce the release of v3.4 
of PyQt at http://www.riverbankcomputing.co.uk/pyqt/.

PyQt is the set of Python bindings for Trolltech's GUI (and SQL and XML) 
toolkit for UNIX/Linux and Windows. It contains 300 classes and over 
5,750 functions and methods. It comes with a debugger and a utility for 
generating Python code from GUIs created with Qt Designer.

The main difference in this release is a change in licensing. PyQt is 
now licensed in a similar way to Qt itself with a GPL version and a 
commercial version. The GPL version is only available for UNIX/Linux. If 
you are using the commercial version of Qt then you also need the 
commercial version of PyQt. If you are using the GPL version of Qt then 
you can use the GPL version of PyQt. See the website for more details.

Also released are updated (and free) binary packages for the 
non-commercial edition of Qt under Windows and the GPL version for the 
iPAQ and Zaurus.

SIP v3.4 has also been released. SIP is the C++ bindings generator used 
to create PyQt.

Phil




From NineOfSix at gmx.de  Mon Sep  2 03:37:51 2002
From: NineOfSix at gmx.de (Axel Grune)
Date: Mon, 02 Sep 2002 09:37:51 +0200
Subject: ascii value
Message-ID: <akv4fn$eo2$03$1@news.t-online.com>

How can I get the ascii value of char which is in a string, e.g. getting 
the 97 if the char was 'a'.

TIA,
Axel



From phr-n2002b at NOSPAMnightsong.com  Fri Sep 20 20:43:09 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 20 Sep 2002 17:43:09 -0700
Subject: Python question on printout to HTML
References: <amgeok$5kg5c$1@ID-49758.news.dfncis.de>
Message-ID: <7x8z1w2mw2.fsf@ruckus.brouhaha.com>

"Ken" <ken at hotmail.com> writes:
> Hi, I wonder how you make the program to print out onto HTML such that it
> prints at a time delayed setting of about 5-10 seconds between each line of
> text so that the steps can be more readable to the user of the interface?

HTML isn't really designed to be interactive at that level.

Three ways to do it: 

    1) send as a multipart-MIME file and emit a new part from the server
    every 10 seconds.  Ugh.

    2) Forward from one page to another with an HTML META refresh tag,
    with 10 second refresh time.  Ugh.

    3) Send all the text on one html page as client side javascript.  Have
    javascript code on the client display a new piece of it every 10
    seconds.  Ugh.

Of the three, the last is probably best for JS-enabled browsers.

Better: redesign your user interface to not need these kinds of
timeouts.  If some users need 10 seconds to view something, then
others will need more than 10 seconds and others will get bored
waiting that long.  It's best to put everything on one page so users
can read at their own pace.  Next best is have the user navigate the
browser him/herself, again at their own pace.


From mbonig at hotmail.com  Mon Sep 23 13:28:02 2002
From: mbonig at hotmail.com (Matthew Bonig)
Date: Mon, 23 Sep 2002 17:28:02 GMT
Subject: Jython file manipulation?!
References: <2yej9.32454$gA4.4053@sccrnsc02> <amn87j$73a$1@mamenchi.zrz.TU-Berlin.DE>
Message-ID: <CaIj9.414430$_91.558962@rwcrnsc51.ops.asp.att.net>

There are no exceptions raised...
If you look at the source Jython uses for the os.rename function it just
uses Java's File.renameTo()...
that function returns false... which then the os.rename function throws a
OSError exception, but it's not helpful in any way.
The only thing I can think is happening is that I am doing this renaming on
files that are on a netword drive on a win32 box, but I have double
checked... the new filename is NOT there... (since win32 won't overwrite the
new file if it exists, and therefore return an error, but that's not
happening). I've also double checked to make sure I have write permissions
on the network drive, which I most definetly do (I usually write 3-5 files
onto that drive a day).
This is why I thought there might be some sort of security issue with
Jython... I'm pretty sure it's in the Jython side since when I tried exactly
the same thing (os.rename(oldfilename, newFilename)) in python it worked
perfectly.

Thanks
Matthew

"Andreas Ulbrich" <ulbi at ivs.tu-berlin.de> wrote in message
news:amn87j$73a$1 at mamenchi.zrz.TU-Berlin.DE...
> Matthew Bonig wrote:
> > I am trying to rename a file in Jython...
> > the command I am using is just
> > os.rename(filename, newFilename)
>
> This works for me (Jython 2.1, JDK 1.4.1)
>
> You should give us some details, i.e. exceptions etc., to get some
> better answers here.
>




From tdelaney at avaya.com  Thu Sep 26 19:39:56 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 27 Sep 2002 09:39:56 +1000
Subject: comparing all values of a list to regex
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5EA@natasha.auslabs.avaya.com>

> From: Alex Martelli [mailto:aleax at aleax.it]
> 
> I don't see exactly what value the call to map adds in your case.

None at all in this case - good catch :) I'll blame it on feeling really
buggered yesterday (am today too for some reason - thank goodness it's only
about 7 hours until the weekend :)

> > # Find out which list to append to
> > 
> > match_list = match_all
> > 
> > if not matches:
> >     match_list = match_none
> > elif matches == 1:
> >     match_list = match_one
> > elif matches != len(alist):
> >     match_list = match_some
> > 
> > match_list.append(alist)
> 
> Yes, choosing a list first and then appending to the chosen
> list is nicer, but you cannot do that within the original
> specs (for all I know the match_some list might grow
> unbearably large over time, in this case).

Hmm - why not? Ignoring that the original specs didn't actually *have* that
case, choosing the list to append to, or doing it in-place is merely an
implementation detail. Of course, you would need instead to do something
like:

match_list = None

if not matches:
    match_list = match_none
elif matches == 1:
    match_list = match_one
elif matches == len(alist):
    match_list = match_all

if match_list is not None:
    match_list.append(alist)

Tim Delaney



From sholden at holdenweb.com  Sun Sep 29 07:07:39 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 29 Sep 2002 07:07:39 -0400
Subject: Local Python Server Installation
References: <3d8b47c6.277669@news.jaring.my> <slrnaomn6m.39s.gerhard.haering@haering.opus-gmbh.net> <3d8c993c.2190426@news.jaring.my> <mailman.1032625889.13783.python-list@python.org> <3d8d4009.375344@news.jaring.my> <bQYk9.1630$6g7.1090@fe03> <3d947c26.2656654@news.jaring.my> <JR3l9.2341$79.998@fe03> <3d952602.1791864@news.jaring.my>
Message-ID: <hbBl9.5471$pD4.5271@fe05>

:-)

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------
"Kenneth Gomez" <kennethg at pd.jaring.my> wrote in message
news:3d952602.1791864 at news.jaring.my...
> Dear Steve,
>
> I managed to get it to work!!
>
> I didnt configure the ScriptAlias directory to where my scripts were.
> How silly of me!
>
> Thanks a lot Steve.
>
> Regards,
> Kenneth
> Cheers,
> Kenneth Gomez.





From pyth at devel.trillke.net  Wed Sep 11 07:21:10 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 11 Sep 2002 13:21:10 +0200
Subject: Abstract Syntax Trees
In-Reply-To: <GuFf9.1028$XE6.54057@newsfep1-win.server.ntli.net>; from ajw126NO@SPAMyork.ac.uk.trillke.net on Wed, Sep 11, 2002 at 12:09:38PM +0100
References: <GuFf9.1028$XE6.54057@newsfep1-win.server.ntli.net>
Message-ID: <20020911132110.S3893@prim.han.de>

Andrew Wilkinson wrote:
> Hi,
> 
> I'm having a lot of trouble trying to get the parser and compiler modules to
> properly parse some files.
> 
> At the moment the trouble seems to be with any loop construct (i.e. for or
> while).
> 
> If you create a test file like this...
> 
> ---
> for i in range(10):
>     print i
> ---

make sure that this file has a newline at the end. 

others would say, get a better OS :-)

regards,

    holger



From wlfraed at ix.netcom.com  Fri Sep 20 17:37:22 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Fri, 20 Sep 2002 14:37:22 -0700
Subject: monetary applications [was: Python GUI app to impress the bos s?]
References: <31575A892FF6D1118F5800600846864DCBD817@intrepid> <s2fdma.pp3.ln@ix.netcom.com> <7xk7lhb943.fsf@ruckus.brouhaha.com> <3D8B18B7.2040203@thinkware.se>
Message-ID: <je4gma.oh3.ln@ix.netcom.com>

Magnus Lyck? fed this fish to the penguins on Friday 20 September 2002 
05:46 am:


> In one system it worked well for years, until
> the application was ported from VMS to HP-UX.
> It would probably have worked on HP-UX as well,
> but values calculated on one platform and stored
> in the database didn't match values calculated
> on the other.
>
        Heck -- a few years back we had that problem going from VAX VMS to 
Alpha VMS. The Alpha's registers used (if I recall the nomenclature) 
G-Float for doubles, where the VAX used D-Float. D-Float doubles are 
essentially a 32-bit float with another 32-bits of mantissa tacked on. 
G-Float, instead, adds a few bits to the exponent, and less than 
32-bits of added mantissa.

        Along with the problems of having a "double" passed to a subroutine 
that only worked "single" (on VAX you essentially were ignoring the 
extended mantissa, on the Alpha the entire exponent was trash).

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From dwelch91 at nospam.attbi.com  Sun Sep 29 13:05:01 2002
From: dwelch91 at nospam.attbi.com (djw)
Date: Sun, 29 Sep 2002 17:05:01 GMT
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com>
Message-ID: <3D9732BD.500@nospam.attbi.com>

Jurie Horneman wrote:
> Once again my attempt to profit from Python's vaunted productivity
> increase has come to a screeching halt because some innocent thing
> that one would reasonably expect to work in one line, intuitively, and
> without even having to think about, turns into a nightmare already
> lasting more than an hour, requiring numerous experiments, searching
> through the documentation - both manually and using my editor's
> find-in-files - finally leaving me no other option but to write a post
> to comp.lang.python, knowing that my programming flow for today is
> well and truly gone. I hate doing this, because I know I cannot keep
> my frustration out of my writing, and also that someone will reply and
> gently or not so gently point out something which actually IS written
> down in some obscure corner of the documentation, despite my attemps
> to RTFM.
> 
> So, what is it this time? No, it's not Unicode, another aspect that
> frequently trips me up, causing my face to meet the rough asphalt of
> the infobahn. It's importing modules. Again.
> 
> (This is an obscure reference to an attempt on my part to implement
> MBCS conversion functions as Python codecs. This failed miserably and
> despite the helpful advice I received here, I never did try again. Too
> bad, because those conversion functions are pretty good. And since I'm
> doing some parenthesized ranting here: does anyone outside of the
> Python world use 'codec' to refer to anything except audio / video
> codecs? I actually had to post here just to find out that what I was
> looking for were *called* codecs. Sheesh.)
> 
> Anyway.
> 
> I have this script, let's call it 'blah.py'. It sits in a directory
> far, far away from where my Python is installed. It wants to include a
> module called 'intro' from a subdirectory called 'templates'.
> 
> This is evidently impossible.
> 
> Among my vain, futile attempts were:
> 
> *** Solution 1 ***
> The World is a Beautiful Place
> 
> 
>>>>import templates.intro
>>>
> 
> Ha! ha! ha! OK, even I knew this was a bit optimistic.
> 
> *** Solution 2 ***
> .PTH file
> 
> Put a templates.pth containing the single line 'templates' in the same
> place as 'blah.py', then
> 
> 
>>>>import templates.intro
>>>
> or
> 
>>>>import intro
>>>
> 
> No dice: 'ImportError: No module named...' etc.
> 
> *** Solution 3 ***
> IMPorting by hand
> 
> 
>>>>import imp
>>>>f = open('templates\intro.py')
>>>
> 
>>>>imp.find_module('intro.py')
>>>
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No module named intro.py
> 
> 
>>>>imp.find_module('intro')
>>>
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No module named intro
> 
> 
>>>>imp.find_module('intro', '\\templates\\')
>>>
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No frozen submodule named .\templates\.intro
> 
> (No WHAT?)
> 
> 
>>>>imp.find_module('intro', '\\templates\\')
>>>
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No frozen submodule named \templates\.intro
> 
> 
>>>>imp.find_module('intro', ['\\templates\\'])
>>>
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No module named intro
> 
> 
>>>>imp.find_module('intro.py', ['\\templates\\'])
>>>
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No module named intro.py
> 
> 
>>>>imp.find_module('intro', ['.\\templates\\'])
>>>
> (<open file '.\templates\intro.py', mode 'r' at 0x01511110>,
> '.\\templates\\intro.py', ('.py', 'r', 1))
> 
> Whoa! I finally managed to call imp.find_module without raising an
> exception! OK, now to call imp.load_module, because the documentation
> says:
> 
> "Load a module that was previously found by find_module()"
> 
> Alright, so there seems to be some connection between find_module and
> load_module... The latter probably takes the return value of the
> former as a parameter, right? Right? Wrong.
> 
> Here are my attemps to call imp.load_module:
> 
> x = imp.find_module('intro', ['.\\templates\\'])
> 
> 
>>>>y = imp.load_module('intro', f, 'intro.py', x)
>>>
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> TypeError: load_module() argument 4, item 0 must be string, not file
> 
> 
>>>>y = imp.load_module('intro', x)
>>>
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> TypeError: load_module() takes exactly 4 arguments (2 given)
> 
> Okay, wait, my assumption was obviously naive, better read the doc...
> aha. Almost the same parameters as find_module... wonder why I even
> needed to call find_module in the first place. OK:
> 
> 
>>>>y = imp.load_module('intro', f, 'intro.py', imp.get_suffixes())
>>>
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> TypeError: load_module() argument 4 must be sequence of length 3, not
> 5
> 
> What? But... but... *splutter* the doc says:
> 
> "The description argument is a tuple, as would be returned by
> get_suffixes(), describing what kind of module must be loaded."
> 
> So why doesn't this work? OK, I have no idea why I'm bothering with
> these kinds of arcana, but for the heck of it I'll try:
> 
> 
>>>>y = imp.load_module('intro', f, 'intro.py', ('.py', 'r', 1))
>>>
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
>   File "intro.py", line 29, in ?
> ImportError: No module named Base
> 
> Yay! What?! Now my IMPORTED module can't find the modules IT is trying
> to import? (It doesn't matter if I put Base next to Blah.py or in
> templates - Intro can't find it.)
> 
> Arrrgghhh!!!! OK, enough time spent on this dead end. (Oh, and by the
> way not all the functions of the imp module are documented.)
> 
> *** Solution 4 ***
> Packages
> 
> OK, more searching, browsing and scouring my hard drive for the word
> 'import'. Of course! I forgot that in the Python docs, important
> information is cunningly hidden in the Tutorial section! (Really, am I
> being unreasonable for assuming that once one has reached a certain
> level of knowledge, one shuns the Tutorial and goes straight for the
> main documentation, because that's where one can find everything there
> is to know? That's only how it works in every language or piece of
> software in general I've ever encountered.)
> 
> So there are these things called 'packages'. Wow! Handy, that! All one
> needs to do is put an empty file called '__init__.py' in a directory,
> and hey! That directory becomes a package and Bob's your uncle.
> 
> So I put an empty file called '__init__.py' in 'templates' and try the
> following:
> 
> 
>>>>import templates.intro
>>>
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No module named templates.intro
> 
> 
>>>>from templates import intro
>>>
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No module named templates
> 
> This is *exactly* how it's written in the Tutorial, and
> it-does-not-work.
> 
> *** Solution 5 ***
> Abject Defeat, Ask on Internet, Be Ridiculed by Peers for Not Managing
> Basic Operation
> 
> Amusing little anecdote: I've performed all of the above steps twice -
> once while growing increasingly frustrated, and a second time while
> writing this post. Twice, while typing in the code above, I at some
> point entered:
> 
> import intro
> 
> ... and it would work. For real. The symbols were there. Not a peep
> out of the interpreter.
> 
> Because I don't trust PythonWin (don't get me started on PythonWin, I
> haven't found a better IDE yet, I may actually succumb and PAY for one
> soon *shock horror*) I quit and restarted and lo and behold! the
> interpreter complained that there was no such module, as it had been
> doing all along, the horrible beast.
> 
> In other words, at random, unreproducable moments, Python perversely
> imported my module, whereas normally it refused to.
> 
> To cut a long story... hm... Please please help me! Stop me before I
> whine again. And please don't be offended by my ranting, I just can't
> hide this much frustration. I'm using Python 2.2 by the way.
> 
> Thanks in advance,
> 
> Jurie.

Hmmmm... I think you need to look into the documentation under 
"packages"... this tripped me up too until I figured out this (somewhat) 
unintuitive mechanism to get modules in subdirs to load.

Here's what I did (on a Win2K box) to show you the steps:

1. Create a c:\blah subdir ("far" from Python directory, but doesn't 
really matter)
2. Create a blah.py in c:\blah that contains 1 line:
import templates.intro
3. Create a c:\blah\templates directory
4. Create a blank (empty) file in c:\blah\templates called __init__.py
5. Create a intro.py in c:\blah\templates that contains the line:
print "intro!" # this will print when we successfully import intro.py
6. Run blah.py:

C:\blah>python blah.py
intro!

C:\blah>

BTW, step #4 is the critical piece.

Hope this helps,

Don





From mnations at airmail.net  Wed Sep 18 16:45:23 2002
From: mnations at airmail.net (Marc)
Date: 18 Sep 2002 13:45:23 -0700
Subject: Strange Telnet error
Message-ID: <4378fa6f.0209181245.7b06124d@posting.google.com>

Hi,

I have been running some code that all the sudden stopped working. I
can not figure out what's happened. I am doing a simple telnet
connection that sends and receives info. Now I am getting this error:


Traceback (most recent call last):
  File "C:\Python22\lib\threading.py", line 408, in __bootstrap
    self.run()
  File "C:\Python22\lib\threading.py", line 396, in run
    apply(self.__target, self.__args, self.__kwargs)
  File "main.py", line 85, in run
    command[0](*command[1:3])
  File "key.py", line 92, in command
    conn.write(com)
  File "C:\python22\lib\telnetlib.py", line 272, in write
    self.sock.sendall(buffer)
AttributeError: 'int' object has no attribute 'sendall'


Has anyone ever seen this before? If so, what could do this be related
to?

Thanks,
Marc


From tjreedy at udel.edu  Wed Sep 11 10:35:44 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 11 Sep 2002 14:35:44 GMT
Subject: Python primer - comments appreciated!
References: <aljdhe$1q63hj$1@ID-77524.news.dfncis.de> <w3cf9.535714$m91.21695020@bin5.nnrp.aus1.giganews.com> <allpms$1r005a$1@ID-77524.news.dfncis.de> <LIwf9.447074$Aw4.18721391@bin2.nnrp.aus1.giganews.com> <alnipo$1pqmu8$1@ID-77524.news.dfncis.de>
Message-ID: <4xIf9.575097$2p2.23759339@bin4.nnrp.aus1.giganews.com>

"Thorsten Kampe" <thorsten at thorstenkampe.de> wrote in message
news:alnipo$1pqmu8$1 at ID-77524.news.dfncis.de...
> *  Terry Reedy
> > Then
> > append to the list.  This working depends on there being *two*
> > references to the list.  The one in the dict (possibly new,
possible
> > not) and the one returned by the method and used by append().

Please reread second sentence!!!

> Okay, let's make this clear:
> D = {'0': [3, 6], '1': [4]}
> D.setdefault('1', [])
>   ...returns: [4]

It returns a reference to the list with current value [4]

> [4].append(7)
>  ...puts [4, 7] on the "stack"

NOOOOO!!!!!  In context of of original expression, it appends 7 to
list pointed to by *both* dict and reference returned by setdefault().

> But where is the instruction saying "store [4, 7] back to D['1']"?

The list is never taken out of D['1'].  The append is 'in place' via
the double reference!
Try
print id(D['1']), id(D.setdefault('1', []))
to see that same list.

> Shouldn't this be: D['1'] = D.setdefault('1', []).append(7) ?

Since append returns None, this will not work.

Terry J. Reedy





From rob  Tue Sep 24 13:24:33 2002
From: rob (Rob Andrews)
Date: Tue, 24 Sep 2002 17:24:33 GMT
Subject: Python highlighting in Kobol
Message-ID: <Xns92937F76BDDB9robuselesspythoncom@24.28.95.158>

This is a rather minor thing to point out, I suppose, but I found it 
interesting. Kobol, the Linux/Windows COBOL IDE from The Kompany, supports 
Python highlighting in the editor.

http://www.thekompany.com/products/kobol/

I don't think they mention it in their promotional pages on thekompany.com, 
but I have the IDE and just noticed their highlighting of Python and 
numerous others.

regards,
Rob


From DennisR at dair.com  Thu Sep 19 17:44:26 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Thu, 19 Sep 2002 21:44:26 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <Xns928CED692541cliechtigmxnet@62.2.16.82> <C05DA40F31B09329.0165201CAF094801.13AF9332EB64347A@lp.airnews.net> <s4Qh9.479726$q53.16320265@twister.austin.rr.com> <vATh9.666$oY5.33977741@newssvr14.news.prodigy.com> <oY+rxnAGxMi9Ew5+@jessikat.demon.co.uk> <rF6i9.686$xY4.42252550@newssvr13.news.prodigy.com>
Message-ID: <_yri9.2058$CR.74235141@newssvr14.news.prodigy.com>

> I have duplicated your experience with both Netscape and Opera on Win98.
I
> have neither a fix nor a facile explanation at this time.  I have
repeatedly
> run both browsers on other CGI code in both Perl and native PLD2.  I
suspect
> a recent change here is at fault.

The server code for implementing CGI style headers was incorrectly
implemented.  This is not a recent change but, rather, incorrect coding.
When I posted the code URLs, I believed that I had "repeatedly run both
browsers on other CGI code...".  At this point, I can only conclude that the
CGI code has been run on IE only (where it does work despite the header
errors).

Rather than rush a fix here, I am going to hold off reposting until I have
done a proper release.  The files sample files whose URLs have previously
been given have been deleted so that no one else serves as an inadvertent
beta tester.

I am sorry for posting immature code.
--

Dennis Reinhardt

DennisR at dair.com
http://www.dair.com




From wurmy at earthlink.net  Wed Sep 18 18:08:58 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Wed, 18 Sep 2002 22:08:58 GMT
Subject: Calling function from command line
References: <amao3t$mug$1@netnews.upenn.edu>
Message-ID: <3D88F9F4.6020306@earthlink.net>

GREGORY KNESER wrote:
> Hi,
> 
> I have a file func.py which includes a working function definition with some 
> inputs and some defaults for those inputs.
> 
> I want to type
> 
> c:\python22\python func.py func1('input1','input2')
> 
> or some such thing and run the function with the inputs I fed it rather 
> than the defaults.  Is such a call possible?

This code may get you started:

---->8----
# cmdline.py

def twice(x):
     return 2 * x

def foo(*args):
     print "Yay! You used these arguments:"
     for arg in args:
         print "-", arg

def bar():
     print "bar!"


if __name__ == "__main__":

     import sys, string

     funcname, args = sys.argv[1], sys.argv[2:]

     arglist = string.join(args, ", ")
     cmd = "%s(%s)" % (funcname, arglist)
     print "Executing:", cmd

     result = eval(cmd)

     print "Result:", result

---->8-----

On my box (Win2K with 4NT), I can call it like this:

(P:\test\general) $ cmdline.py twice 4
Executing: twice(4)
Result: 8

(P:\test\general) $ cmdline.py foo "bar" 'bar' 42
Executing: foo(bar, 'bar', 42)
Yay! You used these arguments:
- <function bar at 0x0076A768>
- bar
- 42
Result: None

(P:\test\general) $ cmdline.py bar
Executing: bar()
bar!
Result: None

Some people may yell at you or me <wink> because the eval() function is used, 
but as long as it's you who executes the code, it's really no biggie. There may 
be a better way to do this, though, e.g. by inspecting the module's globals and 
locals, and maybe use apply()... Even better may be, just firing up the 
interactive interpreter with your module loaded, and type your function calls 
there.

-- 
Hans "so much code so little time" Nowak
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html



From spam at ob_scure.dk  Fri Sep  6 06:55:22 2002
From: spam at ob_scure.dk (Thomas Jensen)
Date: Fri, 06 Sep 2002 12:55:22 +0200
Subject: how could we know which module is imported already ?
References: <7153651c.0209060233.2a72caaa@posting.google.com>
Message-ID: <3D78899A.2010202@ob_scure.dk>

black wrote:
> want to know how many and what the module is imported~

Have a look at sys.modules

import sys
print len(sys.modules)

-- 
Best Regards
Thomas Jensen
(remove underscore in email address to mail me)



From skip at pobox.com  Fri Sep  6 10:48:05 2002
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 6 Sep 2002 09:48:05 -0500
Subject: pygtk
In-Reply-To: <pan.2002.09.06.13.30.48.241597.8751@ns.kleinerdrache.org>
References: <pan.2002.09.06.13.30.48.241597.8751@ns.kleinerdrache.org>
Message-ID: <15736.49189.54748.336080@12-248-11-90.client.attbi.com>

    Martin> Is this a good group to discuss pygtk things or is this just for
    Martin> python?  (I think I have questions in both.)

A better place to discuss pygtk is the pygtk at daa.com.au mailing list.
Comp.lang.python (aka python-list at python.org) is a newsgroup for discussion
of general Python topics.

    Martin> So now this message is also to say 'Hello' to everybody in this
    Martin> room.

Back atcha...

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From brueckd at tbye.com  Fri Sep  6 16:03:59 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Fri, 6 Sep 2002 13:03:59 -0700 (PDT)
Subject: Larry Wall's comment on python...
In-Reply-To: <wkr8g7rn4c.fsf@yahoo.com>
Message-ID: <Pine.LNX.4.44.0209061302050.1523-100000@ns0.tbye.com>

On 6 Sep 2002, Rod Stephenson wrote:

> "Python is cool to look at small bits of, but I think the "outline"
> syntax breaks down with larger chunks of code. I'm with Aristotle on
> the structure of discourse--a story should have a beginning, and
> middle, and an end. So should blocks"
> 
> I'm not quite sure what he's trying to get at here - I guess that for
> a long heavily indented chunk of code, you could lose track of the
> overall structure, but I don't write code this way.
> 
> Any comments?

The very idea of Larry Wall lecturing on code readability is hilarious.

-Dave




From pedronis at bluewin.ch  Wed Sep  4 13:52:15 2002
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Wed, 4 Sep 2002 19:52:15 +0200
Subject: jython memory usage
References: <hVm5ydAY2hd9Eww$@jessikat.demon.co.uk><3d76366a$1_5@news.bluewin.ch> <rnLbuKAv5jd9Ewzz@jessikat.demon.co.uk>
Message-ID: <3d76494d$1_5@news.bluewin.ch>

Robin Becker <robin at jessikat.fsnet.co.uk> wrote in message
rnLbuKAv5jd9Ewzz at jessikat.demon.co.uk...
> In article <3d76366a$1_5 at news.bluewin.ch>, Samuele Pedroni
> <pedronis at bluewin.ch> writes
> >From: Robin Becker <robin at jessikat.fsnet.co.uk>
> >> A colleague has been running tests of reportlab under jython-2.1
several
> >> of the tests use prodigious amounts of memory (one even failed). We're
> >> using win2k+j2sdk1.4.0_01 as the jython engine.
> >>
> >> Anyone know what sorts of things cause x10 memory usage? I find one
> >> particular test needs more that 1Gb and then falls over for lack of
> >> memory.
> >
> >That means you are running jyhton through:
> >
> >java -mx1024m  -cp ...jython.jar org.python.util.jython ...
> >
> >and it uses all that memory and would need more?
> >
> >regards.
> >
> >
> well actually we're running with the default jython.bat which doesn't
> seem to specify a memory, 1Gig is probably the default or maybe I ran
> out of page space. The python process never gets over 90Mb grrrhhhh.

then maybe I'm puzzled,

have you checked process memory usage or is just that you get
an "out of memory" and are estimating based on your total memory.

Because the normal jython.bat uses the JVM default which is
to scream "out of memory" after 64Mb of object memory has
been allocated, see -Xmx here

http://java.sun.com/j2se/1.4/docs/tooldocs/windows/java.html#options

regards.






From chris.gonnerman at newcenturycomputers.net  Thu Sep 26 01:21:46 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Thu, 26 Sep 2002 00:21:46 -0500
Subject: Python on a Novell server
References: <3d928cb5_6@news.newsgroups.com>
Message-ID: <008901c2651c$9d600300$0101010a@local>

----- Original Message ----- 
From: "JB" <jb at yahoo.de>


> We have a Novell net at our school and are trying to install 
> Python 2.2 on a Novell server. This does not work at the 
> moment as we get the error message "python.dll not found" 
> at the moment.
> 
> Is it possible to install Python a a server or has it to be 
> installed locally on each workstation?

I have successfully used the Pythonware distribution stored
on a Linux/Samba server from various Win32 clients... I
assume something like this is what you want?

You need a special startup batch file to run this way, or a
starter executable, which can set the PATH to include the
Python folder.  This should be easy on a Novell system
where you can use the login script to set the path
appropriately.

The python.org version uses the registry to contain some
paths; if you use win32all with the Pythonware version, you
need to set those registry entries manually.  pythonware.com
has a FAQ for this (don't have the URL handy).

Good luck!

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From shalehperry at attbi.com  Thu Sep 12 01:26:53 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 11 Sep 2002 22:26:53 -0700
Subject: [FEEDBACK] Is this script efficient...is there a better way?
In-Reply-To: <3D7FF723.3050005@one.net>
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <HPRf9.1780$Lg2.247316@news2.news.adelphia.net> <3D7FF723.3050005@one.net>
Message-ID: <200209112226.53151.shalehperry@attbi.com>

On Wednesday 11 September 2002 19:08, Michael Schneider wrote:
> Bob X wrote:
> >> 1) readlines() loads the entire file into a list so if you have a 30+
> >> mb file you just ate 30+mb of memory.  Try using xreadlines()
> >> instead, it reads the file line by line and is much more memory
> >> friendly.
> >
> > Very cool...I had missed that!
>
> It is much slower though.  I am not sure about your config, but I am
> running with a Gig of memory.
> I would much rather have the speed then the  30 mb of memory.  Again,
> check your config.

sorry, I am not used to my program being the only piece of software running on 
a system.  Sure his app may only use 30mb out of a 512mb system.  But how 
many other apps are coded just like his and believe all of the machine's 
memory can be used up by one program.  Every time I have made that choice I 
have had to go back and change it.  Plus this allows for the file to not be a 
file anymore.



From spoermeg at voldelig.com  Tue Sep 17 19:38:40 2002
From: spoermeg at voldelig.com (Terje Johan Abrahamsen)
Date: Tue, 17 Sep 2002 18:38:40 -0500
Subject: Source code
References: <mailman.1032152959.16896.python-list@python.org>
Message-ID: <am8eq2$3ml4d$5@ID-53055.news.dfncis.de>

"Tim Peters" <tim.one at comcast.net> wrote in message
news:mailman.1032152959.16896.python-list at python.org...
> [Terje Johan Abrahamsen]
> > If I would like to write some code that would not be possible for
> > others to read, after I distribute the program, would Python be
> > suitable?
>
> No, but I don't know of a language that would be.  Unless you're also
going
> to supply a restricted CPU that prevents stepping through the machine
code,
> you can't really hide anything from a determined reverse-engineerer of
> ordinary skill.
>
> > ...
> > or should I look for a compiled language instead?
>
> It's very easy to reverse-compile Python bytecode.  It's not so easy to
> disassemble compiled machine code, but if you think there aren't lots
people
> who can read machine language almost as easily as high-level source,
you're
> in for a surprise.  Compilers follow predictable patterns too.  You could
> try writing directly in highly obfuscated assembler; that would raise the
> bar high enough to discourage almost everyone, although it would also
> attract a class of cracker who would feel challenged to outwit you.
>
> If you didn't know all that already, you're playing a game you can't win.
A
> better idea is to question why you want to play it at all <wink>.

I am not looking to stop people that really want to copy my code. My program
is not that unique. However, I want them to have to type the code themselves
if they want to duplicate it. Not only to reuse my sourcecode if it can be
picked up from Py2Exe.




From vitus at communiware.ru  Mon Sep  2 04:08:08 2002
From: vitus at communiware.ru (Victor Wagner)
Date: Mon, 2 Sep 2002 08:08:08 +0000 (UTC)
Subject: Could Python supplant Java?
References: <Xns926F8B90671A0AusYourStandingInIt@130.133.1.4> <slrnam3o3f.qo.gerhard.haering@haering.opus-gmbh.net> <8d3f4438.0208202348.578ff9c8@posting.google.com> <H16r29.KxE@bath.ac.uk> <ak025t$p8q$2@srv38.cas.org> <um80n06kf34081@news.supernews.com>
Message-ID: <akv697$qrr$1@wagner.wagner.home>

In comp.lang.tcl Dan Johnson <danieljohnson at vzavenue.net> wrote:
> disposal),
>> or perhaps kindergarten herder ...

> Serious question:

> What is so developer-hostile about Windows?

"Friendly" interface. To be a "friendly" is to consider oneself equal
with his friends. If you want your friend to do something for you,
you ASK him, not ORDER him.

Developer is one who want be a master of his programs, not a friend.
So, best interface for developer machine is slavy one, not friendly.
Programs are to obey orders, not to carry out friendly talks.

Really, it is true for any user, not just developer. But ignorant
users can consider friendlyness helpful. They need even more than 
friendliness, they need patronage.

But developer is by definition a person, who know something about
informational technologies. So, arrogance of "frienly" interface
is unbearable for him. He knows what we wants his machine to do,
and when he sees that machine has its own opinion, as 
"friendly" interface should and do, it becomes very irritating.


-- 
http://www.communiware.ru                     http://www.ice.ru/~vitus


From ned at nedbatchelder.com  Tue Sep 10 07:22:05 2002
From: ned at nedbatchelder.com (Ned Batchelder)
Date: 10 Sep 2002 04:22:05 -0700
Subject: How to choose a package name?
Message-ID: <dcd0187d.0209100322.498b9cf1@posting.google.com>

What's the current thinking on how best to choose package names to
avoid conflicts?  The last discussion about it was four years ago, and
included this message from Guido:
http://groups.google.com/groups?q=g:thl404422857d&selm=199801150144.UAA17896%40eric.CNRI.Reston.Va.US

He mentions a NIST registry, but on a quick look it seems abandoned.

In Java, I know which part of the namespace I already own (based on
domain registration).  I understand Guido's reservations about that
system (code changes ownership).  Do we have anything other than
"choose something reasonably unique, and search Google, Vaults of
Parnassus, Python Starship, and comp.lang.python for conflicts" as a
guideline?

--Ned.
http://www.nedbatchelder.com


From brian at sweetapp.com  Thu Sep 26 00:00:51 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Wed, 25 Sep 2002 21:00:51 -0700
Subject: Getting started
In-Reply-To: <3D928488.3060906@cascade-sys.com>
Message-ID: <00a801c26511$4e084750$df7e4e18@brianspiv1700>

> Some use "weak typing" to mean any type system where the typing can be
> foiled by casts;

What does "foiled by" mean in this context?

Cheers,
Brian




From grabiller at 3dvf.net  Wed Sep 25 11:35:16 2002
From: grabiller at 3dvf.net (Guy Rabiller)
Date: Wed, 25 Sep 2002 17:35:16 +0200
Subject: Newbie References Question -> Exemple
References: <3d91c72f$0$8558$79c14f64@nan-newsreader-02.noos.net> <3d91d2cf$0$1484$79c14f64@nan-newsreader-01.noos.net>
Message-ID: <3d91d7ab$0$13192$79c14f64@nan-newsreader-03.noos.net>

Ok I solved it, thanks to sismex.

If I want to change a coordinate, I have yo write:

p[0][:] = [5,5]

and not

p[0] = [5,5]

I understand that in the first case, the reference pointed ( via
dictionary ) by p[0] is lost, a new value is created and rebinded to p[0],
so quad and p[0] are not synchonised anymore.

In the second case, the reference pointed by p[0] is updated with a new
value. p[0] and quad are still synchronised.

Is my understanding correct ?

Thanks a lot to all.

--
guy rabiller
3d animator / td
grabiller at 3dvf.net
http://grabiller.3dvf.net

Guy Rabiller wrote:
> Ok,
>
> thanks for your answers, I understand now that a variable name is
> 'just' a name in a dictionary, and not a container.
>
> What I want to do is simple, and I suppose that this is my 'way of
> thinking' regarding Python that is wrong.
>
> I have a list of points, and a list of polygons that should contain
> 'references' to points.
>
> So for me:
> -> points list:
> p = []
> p.append([0,0])
> p.append([0,1])
> p.append([1,1])
> p.append([1,0])
> -> polygon:
> quad = []
> quad.append(p[0])
> quad.append(p[1])
> quad.append(p[2])
> quad.append(p[3])
>
> But now, if I change the coordinate of one point:
> p[0] = [5,5]
> I will have:
>>>> p
> [[5, 5], [0, 1], [1, 1], [1, 0]]
> but:
>>>> quad
> [[0, 0], [0, 1], [1, 1], [1, 0]]
>
> For me the quad list of point is not 'updated' with the change of one
> of its point.
>
> What I want to do is that, quad keeps a list of 'references' to points
> coordinates.
> Then if I change one point coordinates, the change is reflected in
> the quad list of points.
>
> I have no clue how to do this.
>
>
> Guy Rabiller wrote:
>> Hi,
>>
>> let say I have:
>> i1 = 1
>> i2 = 2
>> and
>> p = [i1,i2]
>>
>> How can I have:
>> p=[*i1,*i2]
>> rather than
>> p=[**i1,**i2] as it is curently ?
>>
>> ( Sorry for this nasty hybrid syntax )
>>
>> Textualy, how if I want that p[0] refere to the 'i1' container, and
>> not to the refererence it contains ?
>>
>> What I want is that if now I set:
>> i1 = 4
>> that automaticaly:
>> p -> [4,2]
>> and not keeping [1,2]
>>
>> Thanks in advance.




From buzzard at urubu.freeserve.co.uk  Mon Sep  9 20:09:18 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Tue, 10 Sep 2002 01:09:18 +0100
Subject: RPy / Win2k
References: <albgmp$cgv$1@newsg2.svr.pol.co.uk> <mailman.1031613673.1670.python-list@python.org>
Message-ID: <aljd9u$k9n$1@newsg4.svr.pol.co.uk>

"Walter Moreira" <walterm at parque.homelinux.net> wrote in message
news:mailman.1031613673.1670.python-list at python.org...
> On Sat, Sep 07, 2002 at 01:19:13AM +0100, Duncan Smith wrote:
> > Hello,
> >          Is there anyone out there who has successfully used RPy under
> > Win2k?  I get the impression (from Brian Ripley) that it should be
possible
> > to embed R under Win2k, but that it probably isn't straightforward.  Can
> > anyone help?  Cheers.
>
> Well, actually it is possible thanks to a patch from Rene Hogendoorn. I
> have been busy but I will release a Windows version soon. The changes in
> the source are not straightforward, indeed.  I'll try to do it before
> end of september.
>
> Regards:
> Walter
>
>

Thanks Walter, that's good news.  When I spoke to Brian he said that R could
not be built as a shared library under Windows.  Does this mean I won't need
to reinstall R?  Cheers.

Duncan




From lee100 at sport.rr.com  Tue Sep 17 23:20:02 2002
From: lee100 at sport.rr.com (Lee Gray)
Date: Wed, 18 Sep 2002 03:20:02 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <mailman.1032314961.18359.python-list@python.org>
Message-ID: <ChSh9.480276$q53.16363391@twister.austin.rr.com>

Skip Montanaro wrote:

> 
>     Lee> Can someone point me to a good Python GUI app to demonstrate
>     (other Lee> than IDLE)?
> 
> Lee,
> 
> At IPC 10, Michel Sanner from Scripps took home the best paper award for
> his
> ViPEr work.  References:
> 
>     http://www.python10.org/p10-awards.html
>     http://www.python10.org/p10-papers/10/index.htm
>     http://www.scripps.edu/~sanner/python/viper/index.html
> 

Very nice!  Thanks for the ref.

Lee


From johandewitte at hotmail.com  Tue Sep  3 07:52:53 2002
From: johandewitte at hotmail.com (Johan de Witte)
Date: 3 Sep 2002 04:52:53 -0700
Subject: simpleparse example
References: <3d62a770.10000379@newszilla.xs4all.nl>
Message-ID: <cfb4599.0209030352.464a25d1@posting.google.com>

baronvm2 at yahoo.com wrote in message news:<3d62a770.10000379 at newszilla.xs4all.nl>...
> Hi,
> 
> I'm trying to use SimpleParse to process semistructured data. I have
> defined the language and can produce a taglist. 
> 
> How can I use the dispatchprocessor to process the taglist? Does
> anybody have an example.
> 
> Thanx,
> Hans


I have plenty of working code to do such a thing. I use it to parse
weakly structured HTML pages and strip out the unneeded stuff. The
code can however be applied to almost _any_ kind of data.
(Theoreticaly it can be applied to _any_ kind of data). You can have
tons of examples, and you can use all my stuff as long as you don't
sell it.


From peter at engcorp.com  Sun Sep  1 13:36:52 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 01 Sep 2002 13:36:52 -0400
Subject: Add methods to a class at runtime?
References: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com> <slrnan4dl3.18h.gerhard.haering@lilith.my-fqdn.de> <YYrc9.31010$vY2.601601@e3500-atl2.usenetserver.com>
Message-ID: <3D725034.DF95E07D@engcorp.com>

Robert Oschler wrote:
> 
> "Gerhard H?ring" <gerhard.haering at gmx.de> wrote:
> > >>> def bar(self): print "bar"
> > ...
> > >>> Foo.bar = bar
> > >>> foo.bar()
> > bar
> 
> Thanks but I need a little bit more detail.  Suppose I want to do it from
> within code?  For example, I create a function called
> AddBarToFoo(FooObject), that given a object of type Foo (FooObject) adds the
> Bar method to it dynamically, what would be the syntax then?  Would I have
> to put the method declaration code, which you show being typed in to the
> interpreter, in a string and then eval/exec it?  Is there a more convenient
> way than that?

Robert, pretty much anything you can type at the interactive prompt
can be typed as-is into a source file and executed the same way.
That's why the prompt is so good for experimenting and learning.
Try doing what you want interactively, the way Gerhard did, and then
in your text editor window paste in the bits of code as you get
them working.

Basically, to do what you are asking you shouldn't have to do
anything different than what Gerhard did, but there's no substitute
for trying it yourself.  (If you have troubles, show us some 
snippets that you've tried so we can get a better idea where
you are trying to do with this.)

-Peter


From peter at engcorp.com  Tue Sep 24 20:06:11 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 24 Sep 2002 20:06:11 -0400
Subject: List/Tuple bug or feature?
References: <mailman.1032905792.2164.python-list@python.org>
Message-ID: <3d90fdeb$1@news.sentex.net>

sismex01 at hebmex.com wrote:
> It's one of python's features :-)
> 
> For when you wanna write a long string, SQL for example:
> 
> Query = "SELECT what, when, where, why "
>         "FROM t1, t2, t2 "
>         "WHERE t1.id = t2.id_t1 and t3.id_t2 = t2.id and t1.id_t2 = t3.id "
>         "AND t1.answer = 'OK'"
> 
> Kinda neat.

Especially when it actually works... you need the brackets or 
parentheses around the above strings or it will give you a
SyntaxError.

-Peter



From glingl at aon.at  Sun Sep 29 14:52:53 2002
From: glingl at aon.at (Gregor Lingl)
Date: Sun, 29 Sep 2002 20:52:53 +0200
Subject: Python and GUI question
References: <3d96b808_3@news.bluewin.ch>
Message-ID: <3D974C05.4030007@aon.at>

Josef B?rgi schrieb:

> Dear NG
> 
> I designed a CMS using mySQL and PHP. Now I like to write an administrativ
> GUI Client Tool using Python. So I bought a few books about it, saying that
> there is more then one GUI interface (Tk, Qt, xvWindow). So can anyone tell
> me which one you use and why ? And how do you Design GUI interfaces (IDE) ?

There is a nice book (in German): Python and GUI-Toolkits, (mitp), which
contains a thorough evaluation of all those toolkits as well as
a bunch of tipps concerning their respective strengths and weaknesses, 
and when to use which one.

Regards (from Austria)

Gregor

(P.S.: As a teacher I use Tkinter, because it comes with Python and 
everybody who has Python has it. I don't have to delve deep into
GUI-programming)

> 
> Best wishes fromSwitzerland
> 
> Joe
> 
> 




From tismer at tismer.com  Sat Sep 21 23:10:09 2002
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 22 Sep 2002 05:10:09 +0200
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amabpb$5rm$1@news.nems.noaa.gov> <3D8B1537.4010803@thinkware.se> <ad052e5c.0209201128.1a53108@posting.google.com>
Message-ID: <3D8D3491.8010803@tismer.com>

Dan Bishop wrote:
 > Magnus Lyck? <magnus at thinkware.se> wrote in message 
news:<3D8B1537.4010803 at thinkware.se>...
 >
 >>ChrisBarker wrote:
 >>
 >>>The only reason I can see
 >>>to not writing your class in Python is performance, and unless you are
 >>>doing a lot of calculations, that is probably a non-issue.
 >>
 >>Performance issues might well matter in large financial
 >>systems, but that's not the only reason.
 >
 >
 > Please name one other real-world situation in which decimal arithmetic
 > is actually *needed*.

There is no reason to argue about that.
Decimal arithmetic is exact for most of the operations
bankers require. They have been demanding for it since
decades, and they will not drop it. Having "no rounding
issue at all" is better than any consideration about
neglectible errors, it cannot be beaten, since *they*
understand it.
Cannot be discussed. They pay the bill.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
       whom do you want to sponsor today?   http://www.stackless.com/






From mwh at python.net  Wed Sep 25 07:41:29 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 25 Sep 2002 11:41:29 GMT
Subject: Remarks to Python 2.3
References: <20020918222421.328088eb.use-net@schabi.de>
Message-ID: <lkwupa5mix.fsf@pc150.maths.bris.ac.uk>

Markus Schaber <use-net at schabi.de> writes:

> - When replacing list elements using the extended slice syntax - e. G.
> in a[::2] = range(0, -2, -1) - the left side must have exactly the same
> numer of elements as the right side. So, to cite the example, a[::2] =
> range(3) is illegal for a=[0, 1, -1, 3]. 
> 
> Now I could image that it makes sense to allow one single additional
> element to be added to the list in case only the step is specified and
> the operation doesn't produce a gap. So the given example above would be
> valid, giving a sequence of five elements, whereas with a=[0, 1, 2] it
> would stay illegal (as it would produce a gap). a[0:3:2]=range(3) would
> stay illegal as it explicitly specifies the length of the range,
> a[0:3:5]=range(3) could be allowed.

Wanna implement that?  I agree it sounds reasonable.  This sort of
thing is way harder to get right than you might think, though.

If you (or anyone else) submits a patch, I'll review it.

Cheers,
M.

-- 
  Monte Carlo sampling is no way to understand code.
                                  -- Gordon McMillan, comp.lang.python


From fredrik at pythonware.com  Wed Sep  4 03:14:06 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 04 Sep 2002 07:14:06 GMT
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <un50ccotaak86b@corp.supernews.com> <3d72f007.1990343@News.CIS.DFN.DE>
Message-ID: <2pid9.5882$e5.1033577@newsb.telia.net>

Gerson Kurz wrote:

> Seriously, my thinking was, the re.compile function is there to
> compile an expression to a binary representation for optimized
> searching. So maybe, a "clean syntax" -> "ugly re syntax" compiler
> would be good?

note that the SRE engine contains an "ugly syntax" to "internal
data structure" parser, and an "internal data structure" to "engine
code" compiler.

it's probably easier (and definitely more efficient) to turn a clean
syntax into an "internal data structure" than into an ugly syntax.

(the next step is to use python's own parse tree instead of SRE's
internal structure, and use an extension to python's compiler for
the final step.  make it all pluggable, and you have perl6...)

</F>




From eppstein at ics.uci.edu  Thu Sep 26 02:00:36 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Wed, 25 Sep 2002 23:00:36 -0700
Subject: revenge of the python puzzle
References: <mailman.1033015653.17801.python-list@python.org>
Message-ID: <eppstein-9F5AA8.23003625092002@news.service.uci.edu>

In article <mailman.1033015653.17801.python-list at python.org>,
 Michal Wallace <sabren at manifestation.com> wrote:

> Okay, so you guys think you're so smart. Try this one!  No
> contest this time, I've learned my lesson.  (I suppose if
> someone really wants some free hosting bad enough to solve
> this puzzle, you could talk me into it, but I'm just posting
> this for revenge... Um, I mean "fun".)

I'm not going to spoil this, just point out that my cryptogram applet 
http://www.ics.uci.edu/~eppstein/cryptogram/ (sorry, not written in 
Python, but Java source available) solves this really easily.  You'd 
have to make it a lot shorter or a lot less like English to make it 
difficult...  of course to a human the keywords in the lines with colons 
are dead giveaways...

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From pinard at iro.umontreal.ca  Sun Sep  8 17:20:25 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Sun, 08 Sep 2002 17:20:25 -0400
Subject: Unicode -> String problem
In-Reply-To: <3B48D1DB.C3F726C9@ActiveState.com> (Paul Prescod's message of
 "Sun, 08 Jul 2001 14:34:19 -0700")
References: <20010708204337.MJGP11827.femail4.rdc1.on.home.com@jparlar>
	<3B48D1DB.C3F726C9@ActiveState.com>
Message-ID: <oqofb8ji06.fsf@titan.progiciels-bpi.ca>

> Jay Parlar wrote:
>> [...] Is there a simple way to convert the unicode text to StringType,
>> removing the resulting unicode strings for unrepresentable characters?

[Paul Prescod]
>>>> import codecs
>>>> codecs.ascii_encode(u"foo\u3333bar", "ignore")
> ('foobar', 7)

The documentation for Python 2.2.1 does not speak about `ascii_encode',
and the interactive documentation is rather terse:

---------------------------------------------------------------------->
>>> help(codecs.ascii_encode)
Help on built-in function ascii_encode:

ascii_encode(...)

None
----------------------------------------------------------------------<

Where did you find that bit of information? :-)

For now, I would presume that `codecs.ascii_encode(UNICODE_STRING)' does the
same as `UNICODE_STRING.encode("ascii")' and consequently, we do not need to
know about `ascii_encode'.  Is that right?

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



From duncan at NOSPAMrcp.co.uk  Wed Sep 11 09:57:34 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Wed, 11 Sep 2002 13:57:34 +0000 (UTC)
Subject: Regexp Over Multiple Lines
References: <565e316f.0209040740.36573b56@posting.google.com> <565e316f.0209110529.30e61ea4@posting.google.com>
Message-ID: <Xns928696BCE741duncanrcpcouk@127.0.0.1>

vetter at lincom-asg.com (Keith) wrote in 
news:565e316f.0209110529.30e61ea4 at posting.google.com:

> I am sure that this has been discussed... but in a nutshell can you
> tell me where python lacks in power behind perl.  If I go to rewrite
> my Perl code, which has gotten really messy, in Python will I be
> saying over and over again, "Man, I can't do =that= in Python!"
> 
I don't claim to know Perl, but sticking my neck out I would say on the 
basis of long experience of this newsgroup that you should be able to get 
the results you want in Python, but you may find that the idiomatic way to 
achieve it requires a significant mental gearshift.

Once you learn the idioms you should do fine.

BTW, one of the idioms is not to use regular expressions as often as in 
Perl.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From rwaldron at ihug.co.nz  Thu Sep 26 18:20:48 2002
From: rwaldron at ihug.co.nz (Ross Waldron)
Date: 26 Sep 2002 15:20:48 -0700
Subject: Whats the Difference between Windows XP and Windows 98....
References: <3fda493d.0209250933.1a99079f@posting.google.com>
Message-ID: <d4fe4254.0209261420.1ed18d43@posting.google.com>

If the SLP200 is connected via a serial (COM) port it may be to do
with the COM port settings under Windows 98. Make sure the COM port
settings are:

Bits per second:          9600
Data bits:                   8
Parity:                   None
Stop bits:                   1
Flow Control          Xon/Xoff

"Use FIFO buffering" should be enabled

If you are connected via USB, then it may be the difference between
the "dodgy" USB drivers in Win98 and the better USB drivers in XP. Try
updating your USB drivers for Win98.

Regards
Ross Waldron

naveed_iq at hotmail.com (Naveed Iqbal) wrote in message news:<3fda493d.0209250933.1a99079f at posting.google.com>...
> Whats the Difference between Windows XP and Windows 98....when it
> comes to the win32 extensions to python? I used the win32ui and
> win32com.client modules to print to a printer from windows. the output
> from the XP machine comes out fine, but the output from the 98 machine
> looses its formating. I am using a Smart Label Printer 200. this label
> printer has drivers for all windows family OS's.
> does anybody have an idea as to what is going on?
> thanks
> Naveed


From amdescombes at qualicontrol.com  Fri Sep  6 06:30:03 2002
From: amdescombes at qualicontrol.com (Andre Michel Descombes)
Date: Fri, 6 Sep 2002 12:30:03 +0200
Subject: Python and Jython inconsistencies when encoding strings
Message-ID: <ala00d$dij$1@wanadoo.fr>

Hi,

I've noticed the following inconsistency between Python 2.1 and Jython 2.1 :

when I do the following in Python :

>>> s = u'?' # Alt-0163, pound sign
>>> print s

I get the following exception, which is a normal exception:

Traceback (most recent call last):
  File "<stdin>", line 1, in ?
UnicodeError: ASCII encoding error: ordinal not in range(128)
>>>

I corrected this with using the encode function like this:

>>> print s.encode('latin-1')
?
>>>

now when I do the same thing in Jython 2.1 I get this:

>>> s = u'?' # Alt-0163, pound sign
>>> print s
?
>>>

and if I do:
>>> print s.encode('latin-1')

I get:

 Traceback (innermost last):
   File "<console>", line 1, in ?
 UnicodeError: latin-1 encoding error: ordinal not in range(256)
>>>

Does anybody know what is causing this inconsistency? Is there any way to
avoid it?

Thanks,

Andre




From gerhard.haering at gmx.de  Mon Sep  2 01:21:35 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 2 Sep 2002 05:21:35 GMT
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <un50ccotaak86b@corp.supernews.com> <3d72f007.1990343@News.CIS.DFN.DE>
Message-ID: <slrnan5tad.8pn.gerhard.haering@lilith.my-fqdn.de>

Gerson Kurz wrote in comp.lang.python:
> "A.M. Kuchling" <amk at amk.ca> wrote:
>>The problem with a new syntax is that no one else would be using it, so
>>you'd still need to learn the existing syntax for use with grep, vi, Perl,
>>&c.  (It wouldn't surprise me if Perl 6's revised regexes run into this very
>>difficulty and don't gain much adoption.)
> 
> A tounge-in-cheek-answer: I don't use grep, I use python. I don't use
> vi,

vi? Yuck. vim or die :-P

> I use python. I don't use perl, I use python :) 
> 
> Seriously, my thinking was, the re.compile function is there to
> compile an expression to a binary representation for optimized
> searching. So maybe, a "clean syntax" -> "ugly re syntax" compiler
> would be good? 

I think it would. Though the other way round wouldn't be that bad,
either.

-- Gerhard


From hwlgw at hotmail.com  Mon Sep  2 05:51:06 2002
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 2 Sep 2002 02:51:06 -0700
Subject: Twisted docs (was Re: Python IPC options)
References: <1030829572.943495@irys.nyx.net> <3D715160.2020109@itamarst.org_NOSPAM> <cb035744.0209011308.7b4b3ae2@posting.google.com> <3D728386.2060908@itamarst.org_NOSPAM>
Message-ID: <cb035744.0209020151.2f318538@posting.google.com>

Itamar Shtull-Trauring <itamar at itamarst.org_NOSPAM> wrote:

> I take it you're going to volunteer some time to write 
> docs? We'd really appreciate it! ;)

Okay I will try, but my TODO list is already very long.  So it will be
a weekend thingy and could take a while to produce something useful,
besides I will have to learn Twisted first!

What I propose is to show how to develop a medium sized application in
the Twisted framework that uses a lot of Twisted's capabilities.  It
should have asynchronous *and* synchronous communication, *mobile*
components and because of that a dynamic network.  It should also show
how to exchange data with a server / database via its URL on the
internet.  Also to be fun it needs a GUI so I will use Tkinter, I
believe there are examples already how to combine Twisted with
Tkinter.  Tkinter seems to be the best choice since it is one of
Python's included batteries and it is portable and I have some
experience with it.

Now what kind of application would that be?  My idea for the
application is a trader agent and its environment.  It could be
something like a simulation or a game where you have to program an
automated trader agent that buys and sells stocks for you.  The client
should also have an interactive mode so you can do things manually
too.  There is a central market / database / server online all the
time.
First I need to set up a website that simulates a stock market, or
maybe I can just use an existing one, have to see how stable the data
are they produce.  www.nyse.com?

Does this kind of application show enough of Twisted's features?  You
have more ideas?   Do you think some things are too hard or
impossible?


Demonstrated features
---------------------

* Asynchronous communication:    Updates from the market to the client
                                 (if it is online) in a 
                                 send_and_forget fashion.

* Synchronous communication:     Buying and selling transactions

* Mobility:                      The clients should be able to connect
                                 from anywhere and it should maybe be 
                                 possible to relocate the market.

* Dynamicity:                    See mobility.  Also there will be an 
                                 unknown number of clients at any 
                                 time.

* RPC:                           client <-> market, but maybe also 
                                 perhaps client <-> client for forming
                                 trade groups?

* GUI integration                To show what a client does and to be
                                 able to do manual trading.

* Remote debugging?  Remote market profiling?  Market logging?

Have to go back to my real work now.  What surprises me is that the
Twisted people did not come up with a demo apllication like this
already, the more I think about it the more fun it seems.  I did see a
MUD-something-application mentioned on the Twisted site, but could not
find it.  The trader application looks a bit more ``serious'' perhaps,
better for public relations?  Also make absolutely sure people using
Windows can install everything they need with just doubleclicking an
.exe or whatever.  You can do it, Twisted already has that for the
modules.

Writing software for installing things and writing documentation is
generally considered *boring* and uncool work among programmers.  I
know.  Hope this helps inspiring you to do it and to look at it in
another way?

For now good luck with writing good docs about things I do not know
about yet: Twisted deserves it.  So one day it can become one of
Python's included batteries too!

::

A child of five could understand this!  Fetch me a child of five.


From loewis at informatik.hu-berlin.de  Wed Sep 11 12:07:55 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 11 Sep 2002 18:07:55 +0200
Subject: Background programs on Windows XP
References: <3D7F5181.AD09A901@ex.ac.uk>
Message-ID: <j43csg5x2c.fsf@informatik.hu-berlin.de>

"M.J.Fisher" <m.j.fisher at ex.ac.uk> writes:

> However, when one of the users of this program changed from Windows'95
> to WindowsXP (they also changed from Python2.1 to Python2.2 but I didn't
> think this was probably relevant) this stopped working.

Did it give any information of how it stopped? What happens if you do
this in a cmd.exe window?

Regards,
Martin



From aleax at aleax.it  Fri Sep 20 05:11:17 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 09:11:17 GMT
Subject: How to access XML passed from an XForm
References: <mailman.1032506367.3820.python-list@python.org>
Message-ID: <VCBi9.117757$ub2.2505268@news1.tin.it>

Jeremy Rew wrote:

> I dont know if this has been answered before now, but I couldnt find it
> after a fair bit of searching. (is there any way of running a search on
> the archives?)

Google is your friend -- use an advanced groups search, specifying
comp.lang.python as your group of interest.  I just did that looking
for xform but none of the 6 hits appears relevant to your question.

http://groups.google.com/advanced_group_search?hl=en is one way to
access Google's page for advanced groups search, by the way.


Alex



From dubal at khimjiramdas.com  Fri Sep  6 02:22:38 2002
From: dubal at khimjiramdas.com (J Dubal)
Date: 5 Sep 2002 23:22:38 -0700
Subject: How to install informixdb-1.3 for python2.2 on RH7.3?
References: <c345ff6f.0209042255.65dd5824@posting.google.com>
Message-ID: <c345ff6f.0209052222.37221ef1@posting.google.com>

Hello again!

I managed to compile and install informixdb module manually for py221.
However, when I import it in python2, it gives following error:
ImportError: Cannot locate C core module for informix db interface.

If I try import _informixdb, it gives following error:
ImportError: /opt/informix/lib/esql/libifgen.so: undefined symbol:
stat

Please help. If we cannot make this work, we will have to go back to
1.5.2 - not very attractive alternative.

Thanks in advance.
Dubal.

dubal at khimjiramdas.com (J Dubal) wrote in message news:<c345ff6f.0209042255.65dd5824 at posting.google.com>...
> Hello good people!
> 
> As mentioned in its readme, informixdb-1.3 compiles, installs and
> works with the stock python 1.5.2 of RH7.3. However, when we try to
> compile it for python 2.2.1, informixdb's ./configure fails because it
> is looking for Makefile.pre.in which is not there in python 2.2.1. It
> appears that someone has to repackage informixdb with distutils. How
> to do this? In the meanwhile can we install this manually? I do not
> understand Makefile.pre.in. Please help. Thanks in advance.
> Dubal.


From syver-en+usenet at online.no  Thu Sep 19 14:51:17 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Thu, 19 Sep 2002 18:51:17 GMT
Subject: Creating a typelibrary for calling Python COM server from c#
References: <yvoi9.516983$UU1.86221@sccrnsc03>
Message-ID: <uit11stin.fsf@online.no>

"Dave Gustafson" <DGustafson at attbi.com> writes:

> There was a thread recently that suggested creating a typelibrary in
> order
> 
> to access a Python based COM server from C# (or .NET).  Can anyone
> please
> 
> suggest an easy way to create such a library???

Create an .idl file. MSDN has resources for the syntax of this file.

Run the midl compiler on the .idl file (midl.exe), this will generate
a typelibrary (.tlb) file and c source and header files.

If you don't have Visual Studio, I believe that you can get the midl
compiler by downloading the Microsft Platform SDK (also on msdn).

-- 

Vennlig hilsen 

Syver Enstad


From k.robert at gmx.de  Fri Sep 27 16:17:38 2002
From: k.robert at gmx.de (Robert)
Date: Fri, 27 Sep 2002 22:17:38 +0200
Subject: http file upload from Python (multipart/formdata) how-to?
References: <3d937091$0$181$9b622d9e@news.freenet.de> <ujp7pu4osgoagpnqr4ls727cibm2ua5jq1@4ax.com>
Message-ID: <3d957872$0$182$9b622d9e@news.freenet.de>

"Tim Roberts" <timr at probo.com> schrieb im Newsbeitrag
news:ujp7pu4osgoagpnqr4ls727cibm2ua5jq1 at 4ax.com...
> "Robert" <k.robert at gmx.de> wrote:
>
> >want to upload file data (multipart formdata) - same as upload from
> >web-browser <input type=file ..>
> >
> >urllib.urlencode  + urlopen takes only simple name/value tuples
> >
> >how do I render multipart formdata ?
>
> I guess you will have to generate the HTTP/MIME headers and send the file
> in your own code.  See RFC 1867.
>
> Are you trying to interface with a web form that already exists, or are
you
> just trying to do a file transfer from one machine to another?
> --
> - Tim Roberts, timr at probo.com
>   Providenza & Boekelheide, Inc.

yes think I have to render the RFC1867 HTTP/MIME headers manually and put it
as 'data' parameter to urlopen.

found no python code to do that upon the last google session. only strange
perl hieroglyphs $%()%$//%=?!=... and I have no archeologist ;-)

so I am still too lazy to read and implement the RFC and hope that somebody
already had the same problem

Robert





From owen at nospam.invalid  Mon Sep 16 13:01:39 2002
From: owen at nospam.invalid (Russell E. Owen)
Date: Mon, 16 Sep 2002 10:01:39 -0700
Subject: Tkinter listbox clicking problem
References: <sGog9.307183$aA.53945@sccrnsc02> <altfrk$1eva$1@nntp6.u.washington.edu> <2D7E3697CF5D1521.5D9711540E657ED8.7EF0451C598438FC@lp.airnews.net>
Message-ID: <am52pe$1nas$1@nntp6.u.washington.edu>

That's odd, the first thing I looked for in Listbox when trying to 
answer the original poster's question was the ability to tag items, 
since binding events to tagged items would have been a good solution. I 
don't think items in a Listbox can be tagged. I fear I'm missing 
something, but I'm not sure what. (I imagine one could emulate a Listbox 
with a Text widget, tag each line and bind to those tags, but it sounds 
like a fair amount of work to get the selection behavior right.)

Fortunately Eric Brunel's suggestion of binding to Button-Release will 
do the trick. This is the approach taken by Brian Welch in "Practical 
Programming in Tcl and Tk" (which is where I should have looked to start 
with). By the time the mouse button is released, you can safely query 
the list to ask it for the selection and it will match what the user 
sees.

Admittedly the user may be more used to mouse down triggering actions, 
but that model just doesn't seem to mesh well with the standard bindings 
for Listbox.

-- Russell

>Whoa!  There's a far easier solution.  Read up on 'bindtags'
>in <URL: http://web.pydoc.org/2.2/Tkinter.html > and elsewhere.
><URL: http://www.cwi.nl/~jacob/textui.py.html > has a small
>example that almost is pertinent.


From skip at pobox.com  Thu Sep 26 15:43:58 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 26 Sep 2002 14:43:58 -0500
Subject: Strange change in xmlrpclib behavior
In-Reply-To: <200209261857.g8QIvwY07046@universal.dca.net>
References: <200209261857.g8QIvwY07046@universal.dca.net>
Message-ID: <15763.25470.257583.225525@12-248-11-90.client.attbi.com>

>>>>> "Walt" == Walt Leipold <leipold at universal.dca.net> writes:

    Walt> Any ideas?  (BTW, it's xmlrpclib v0.9.8, Win2000 SP2 and NT4 SP6.)

    Skip> Have you tried a more recent version of xmlrpclib and/or Python
    Skip> 2.2.1 or Python from CVS?

    Walt> I am now using Python 2.1.1, have tried xmlrpclib v0.9.9 without
    Walt> success, and I'm just about desperate enough to go to the Python
    Walt> CVS.  I haven't moved to Python 2.2 because I'd have to rebuild my
    Walt> serial library (but I may have to do that).

My thinking goes like this...  Something had to have changed in your system
for it to suddenly stop working.  I briefly tried running your code on my
Python CVS install here, but it failed because there is no xmlrpcserver
module.  (It's SimpleXMLRPCServer, and has a different API.)  so, to debug
this, either you have to move to a newer setup or I have to move to an older
setup.  Guess which one is not going to happen. <wink>

Even if you don't deliver your software on a newer platform, installing
Python 2.2.1 (which you can do alongside 2.1 on Windows I believe) and
porting your example code to use SimpleXMLRPCServer then gives you access to
a lot more people who can help debug your problem.

    Walt> However, what's *really* bugging me is that the code *used* to
    Walt> work; even if moving to a slightly newer version fixes it, how can
    Walt> I be sure it won't fail again?  I shouldn't turn it over to my
    Walt> customer until I understand what broke it in the first place...

Well, if it used to work, you changed nothing, and now it doesn't work, I'd
guess there is some sort of undetected race condition present in your
system.

    Walt> I'll probably end up junking XML-RPC and writing the socket code
    Walt> using UDP or something similar, just so I can trust it.

Changing communication protocols without knowing that xmlrpclib is to blame
won't make the problem go away.  It might hide it, or it might solve the
problem, but you won't know which.

Skip




From gua81 at XXXyahoo.com  Tue Sep 24 03:40:21 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Tue, 24 Sep 2002 17:40:21 +1000
Subject: Python and cgi and <a> problem!
References: <ammn6e$lj7$1@lust.ihug.co.nz> <Xns9292EA042C259cliechtigmxnet@62.2.16.82> <slrnaov5q3.nrp.jon+usenet@snowy.squish.net> <Xns929330F5A61B0cliechtigmxnet@62.2.16.82> <slrnaovmel.dma.jon+usenet@snowy.squish.net>
Message-ID: <amp4se$ik4$1@lust.ihug.co.nz>

thanks dudes,
everything works fine,
im just doing some simple things, nothing that will break the code.

Cheers.

"Jon Ribbens" <jon+usenet at unequivocal.co.uk> wrote in message
news:slrnaovmel.dma.jon+usenet at snowy.squish.net...
> In article <Xns929330F5A61B0cliechtigmxnet at 62.2.16.82>, Chris Liechti
wrote:
> >>> print '<A href="%s">%s</A>' % (filename, filename)
> >>
> >> I hope none of your filenames have '"', ">" or "&" in.
> >
> > yeah i avoid characters that are special to the shell in my filenames -
i
> > even don't use " " very often. <wink>
> >
> > sure there are improvements, but its a start.
>
> I'm not sure recommending completely broken and insecure code is ever
> a good thing. No offence intended.




From g0w1s0 at yahoo.ca  Sun Sep  1 22:22:00 2002
From: g0w1s0 at yahoo.ca (Cesar)
Date: Sun, 01 Sep 2002 22:22:00 -0400
Subject: curses: AttributeError: initscr
References: <un5dhs4f7qlte8@corp.supernews.com>
Message-ID: <un5iq3ht257s22@corp.supernews.com>

Cesar wrote:
> 
> 
> I'm running python 1.5/2.2.1 under redhat linux 7.2.
> 
> I'm trying to use the curses module like this :
> 
> 1: import curses
> 2:
> 3: screen = curses.initscr()
> 4:
> 
> and I get this error msg:
> 
> Traceback (innermost last):
>   File "./curses_test.py", line 6, in ?
>     screen = curses.initscr()
> AttributeError: initscr
> 
> any idea ??

the problem was stupid !!!

in the same directory where my script live, there was an old file named 
curses.py. Then the instruction : "import curses" imported the bad one...

I'm new to python.. so..



From marklists at mceahern.com  Thu Sep 26 14:00:22 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 26 Sep 2002 13:00:22 -0500
Subject: design question
In-Reply-To: <r5g6pug88d7uq5jm8kbckom2l30n6ec9jn@4ax.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOMENEEBAA.marklists@mceahern.com>

[Gon?alo Rodrigues]
> Well, I have thought that somewhere further down the way some (though
> not all) of the wrapping code could be factored out in some kind of
> metaclass, and then I would just set __metaclass__ at module level. But
> since I am still at the entrance, I am kinda polling for ideas.
>
> Assuming that my description of the problem is enough, how would you go
> about writing a metaclass as a general wrapper?

Hey, no fair calling my bluff!  <wink>

For starters, please tell me how come this isn't sufficient:

class Wrapper(object):

    def __init__(self, wrapped_class):
        self._wrapped_class = wrapped_class()

    def __getattribute__(self, name):
        if __debug__:
            print "calling __getattribute__(%s)" % name
        wrapped_class_attr_name = "_wrapped_class"
        wrapped_class = object.__getattribute__(self,
wrapped_class_attr_name)
        if name == wrapped_class_attr_name:
            return wrapped_class
        else:
            return getattr(wrapped_class, name)

    def __setattr__(self, name, value):
        if __debug__:
            print "calling __setattr__(%s, %s)" % (name, value)
        wrapped_class_attr_name = "_wrapped_class"
        if name == wrapped_class_attr_name:
            object.__setattr__(self, name, value)
        else:
            setattr(self._wrapped_class, name, value)

// m




From bogle at ihug.co.nz  Sat Sep  7 20:40:55 2002
From: bogle at ihug.co.nz (Gib Bogle)
Date: Sun, 08 Sep 2002 12:40:55 +1200
Subject: Python from C/SDL
References: <3D793B29.3198AB0E@ihug.co.nz> <006101c25626$56c9b1c0$0101010a@local> <mailman.1031395211.21628.python-list@python.org>
Message-ID: <3D7A9C97.C5FD0817@ihug.co.nz>

Mike,

Thanks very much for such a detailed response.  My programming skills
are a bit out of date (little OO experience, no C++, no Python) so what
is required here might be a bit of a stretch.  I have sent a message to
Warren Delano (pymol author) to see if he is interested in giving some
assistance.  If not I'll probably put this on the back burner until I
have the time to teach myself the basics of Python.  I have been aware
of it for a couple of years now, and if I was 20 years younger ...

regards
Gib

"Mike C. Fletcher" wrote:
> 
> I'll just deal with the OpenGL stuff.  Basically, PyOpenGL-based apps
> should be able to work like this:
> 
> c draw():
>         context = new MyContext;
>         context.set_current();
>         err = Py*_Call*( size );
> 
>         python:
>                 def render( dimensions ):
>                         do_your_rendering_here()
>                         return
> 
> c:
>         swap_buffers();
>         context.unset_current();
> 
> So, you'd likely need to find, somewhere in pymol, a function to call
> which is _after_ the Python context setup and set_current() which you
> can call.  Depending on the architecture of pymol, that's likely to
> require at least sub-classing, and possibly some re-writing, as most
> rendering libs go out of their way to be sure they're using the "right"
> context before any attempt to render, and most apps assume they'll be
> using a context they set up.  You'll likely need to pass size of the
> window from your code, as some renderers assume they know the current
> size of the rendering context (which they won't if they aren't getting
> their resize events).
> 
> For example, in OpenGLContext, you'd subclass a Context and override
> setCurrent and unsetCurrent to do nothing (or more professionally, you'd
> create C "extending" functions that set the appropriate context for your
> app, which would allow the Python system to have its own contexts
> alongside yours if you wished).  In other frameworks those calls aren't
> actually split out into easily overridable methods (it's only required
> if you're targetting multiple GUI libs), so you might have to re-write
> anywhere you see "set current" methods called.    Would suggest
> contacting the creator of PyMol for suggestions where to apply your lever.
> 
> For a "full" job of creating a PyOpenGL embedded system, you'd also
> provide initialisation/creation functions to allow Python code to create
> a context within your app (e.g. setting display modes).
> 
> There are similar questions asked once in a while, but not particularly
> with pymol AFAIK.  "How do I embed PyOpenGL apps" is the basic template.
>   Suppose I should create a tutorial for it, but I've never done it
> directly myself (though conceptually every context in OpenGLContext is
> "embedded" in the GUI library it uses, since those are all C libs, it's
> just that I do the "calling into Python" from the libs own Python wrappers).
> 
> HTH,
> Mike
> BTW: out of town in a few minutes, so won't be available to clarify
> anything 'til Monday, sorry.
> 
> Chris Gonnerman wrote:
> > ----- Original Message -----
> > From: "Gib Bogle" <bogle at ihug.co.nz>
> >
> >
> >>Hi,
> >>
> >>I am a complete beginner with Python.  My interest in it is focussed on
> >>a great piece of open-source software called pymol, a system for
> >>visualization of molecules that uses Python and OpenGL.  I'd like to get
> >>access to pymol's graphical functions from my own program, which is
> >>built from C, SDL and OpenGL.
> >
> ...
> >>I will not want to use the GUI at all.  SDL would
> >>create the surface(s), and handle the buffer swapping.  My goal is to be
> >>able to send a request to the pymol API asking for a frustum view to be
> >>rendered on the provided surface.
> >>
> >>Sorry if my question is one that gets asked and answered frequently.  If
> >>there is a place where this has been dealt with please direct me.
> >
> >
> > Never heard this one before...
> >
> > Good luck!
> >
> >
> > Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
> > http://newcenturycomputers.net
> >
> >
> 
> --
> _______________________________________
>    Mike C. Fletcher
>    Designer, VR Plumber, Coder
>    http://members.rogers.com/mcfletch/

-- 
--------------------------------------------
Gib Bogle              bogle at ihug.co.nz
1/44 Arthur St         Tel: (64-9) 525-6878
Ellerslie, N.Z.        Fax: (64-9) 525-6878
--------------------------------------------


From bokr at oz.net  Sat Sep  7 13:48:11 2002
From: bokr at oz.net (Bengt Richter)
Date: 7 Sep 2002 17:48:11 GMT
Subject: Interrupting Python
References: <mailman.1031345649.9840.python-list@python.org> <Ef9e9.7309$rU2.218110@news4.srv.hcvlny.cv.net> <albel2$7te$0@216.39.172.122> <R3ce9.9745$rU2.313350@news4.srv.hcvlny.cv.net> <3d796d22@news.sentex.net> <OAke9.17281$rU2.525682@news4.srv.hcvlny.cv.net>
Message-ID: <alde4r$sbo$0@216.39.172.122>

On Sat, 7 Sep 2002 06:32:24 -0400, "Bob Easton" <bob at eleaston.com> wrote:

>Yes Peter, it makes sense that PythonWin works that way by design. You are
>also right about running from the command prompt after development/debug is
>done.
>
>For this particular program, using a file semaphore is ideal since I can
>place it at the right point in the program to enable easy restart, something
>not easily done with the random ctrl-break.
You can place a test for Ctrl-C and/or Ctrl-Break at "the right point"(s)
in your program. See below.

>
>Thanks to all who helped clear the fog for me.
>
>--
>Bob Easton
>
>
>"Peter Hansen" <peter at engcorp.com> wrote in message
>news:3d796d22 at news.sentex.net...
>> Bob Easton wrote:
>> > Using catchCtlC.py in a command line environment works as you expect.
>When I
>> > hit Ctrl-Break, Windows displays "^C" and then a command prompt.  If I
>hit
>> > Ctrl-C, the exception handler catches it and displays the results as
>> > specified in catchCtlC/py
>> >
>> > However, my preferred execution environment is PythonWin, not a command
>> > prompt. There, neither Ctrl-break, not Ctrl-C have any effect.  That
>could
>> > easily be a bug with PythonWin.
>>
>> Not a bug, I think.  The whole Ctrl-C/Ctrl-Break concept is only
>> meaningful for command-line ("console") programs, not GUI apps.
>> I don't use PythonWin, but my guess is it does not bother passing
>> ctrl-c on to the application, possibly by design.
>>
>> Can't you just run it under the console?  What advantages do
>> you find by running it under PythonWin other than during
>> development?
>>
>> Also, have you looked at the msvcrt module and kbhit() ?
>>
>> -Peter
>>
>
---< interruptpy.py >------
#!/usr/bin/python
def test():
    import sys, signal
    class SigHandler:
        def __init__(self):
            self.signaled = 0
            self.sn=None
        def __call__(self, sn, sf):
            self.sn = sn 
            self.signaled += 1
    sh = SigHandler()
    old_SIGINT_Handler = signal.signal(signal.SIGINT,sh)
    old_SIGBREAK_Handler = signal.signal(signal.SIGBREAK,sh)
    signames = {
        signal.SIGINT:'SIGINT',
        signal.SIGBREAK:'SIGBREAK'
    }
    print 'Type Ctrl-C or Ctrl-Break to stop'
    while not sh.signaled:   # put this condition in your program loop
        pass                 # (loop body)

    print 'Interrupted loop with %s (not via exception).' % signames[sh.sn]

    signal.signal(signal.SIGINT,old_SIGINT_Handler)
    signal.signal(signal.SIGBREAK,old_SIGBREAK_Handler)

if __name__ == '__main__':
    test()
---------------------------

 >>> from interruptpy import test
 >>> test()
 Type Ctrl-C or Ctrl-Break to stop
 Interrupted loop with SIGINT (not via exception).
 >>> test() # now I'll do ctrl-break
 Type Ctrl-C or Ctrl-Break to stop
 Interrupted loop with SIGBREAK (not via exception).
 >>> ^Z


 [10:55] C:\pywk\pi>interruptpy.py
 Type Ctrl-C or Ctrl-Break to stop
 Interrupted loop with SIGINT (not via exception).

 [10:55] C:\pywk\pi>interruptpy.py
 Type Ctrl-C or Ctrl-Break to stop
 Interrupted loop with SIGBREAK (not via exception).


Obviously you could sprinkle various tests of sh.signaled
into various parts of your code, to check-point variously.
You could resond differently to Ctrl-C vs Ctrl-Break also.

The above is on NT4 Python 2.2 BTW.

BTW2, I suspect I am not the only one who will appreciate it
if you decide to adopt usual ng conventions, and discontinue
top-posting. Consider it a virtual beer for those who paid
attention to your posts anyway ;-)

Regards,
Bengt Richter


From fb at frank-buss.de  Sun Sep  1 13:47:34 2002
From: fb at frank-buss.de (Frank Buss)
Date: Sun, 1 Sep 2002 17:47:34 +0000 (UTC)
Subject: A "for" with "list" question.
References: <3b55ea60.0208311330.4608ae71@posting.google.com> <mailman.1030830407.5536.python-list@python.org> <akthkp$5nb$1@newsreader2.netcologne.de>
Message-ID: <aktjrm$bnv$2@newsreader2.netcologne.de>

BTW: If you want the second case, only, you can write:

>>> from operator import xor
>>> a = [0, 1, 2]
>>> b = [1, 2, 3]
>>> [item for item in a + b if xor(item in a, item in b)]
[0, 3]

-- 
Frank Bu?, fb at frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de


From robin at jessikat.fsnet.co.uk  Mon Sep 16 11:52:42 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Mon, 16 Sep 2002 16:52:42 +0100
Subject: pyro & parallel processing
Message-ID: <rLOb1LAK5fh9EwRi@jessikat.demon.co.uk>

Has anyone used pyro to do symmetric parallel processing. The pyro
framework is nice with master slave type applications, but the remote
objects don't seem to be able to message back to their invoker.

I thought I'd ask if any one has got pyro to work in a more symmetric
way?
-- 
Robin Becker


From martin at v.loewis.de  Tue Sep 17 01:55:22 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 17 Sep 2002 07:55:22 +0200
Subject: More on Protecting Source Code
References: <mailman.1032232999.22661.python-list@python.org>
Message-ID: <m3elbtcg8l.fsf@mira.informatik.hu-berlin.de>

"David LeBlanc" <whisper at oz.net> writes:

> * Owners of property have the right to control it. If you disagree and own a
> nice car or house, I'll be right over :->

What are you protecting your car against? Somebody taking it and using
it without your permission, right? Not against somebody analysing it
and building a similar car. 

That would be for the car makers to protect; I believe they don't
protect it by technical means against a determined engineer. Instead,
they protect it with patents and trademarks (and, admittedly, by the
fact that you need significant facilities to copy a car).

For software, I think the economical risk is that somebody uses your
software without giving you the money you want - not that somebody
sits down and studies your code for days, to write a program that he
could have written himself from scratch in the same time.

Regards,
Martin


From jonathan at onegoodidea.com  Sun Sep 29 05:43:59 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Sun, 29 Sep 2002 10:43:59 +0100
Subject: [Mac] Editor on Macintosh?
References: <1fj7za2.1erbqps1uvzpj8N%news@agapow.net>
Message-ID: <B9BC89EF.12189%jonathan@onegoodidea.com>

On 28/9/2002 12:19, in article 1fj7za2.1erbqps1uvzpj8N%news at agapow.net,
"Paul-Michael Agapow" <news at agapow.net> wrote:

> Sadly, it seems like the excellent editor Pepper is going away. (See
> <http://www.hekkelman.com>.) I found it the best editor for Python on
> the Macintosh, having never been really happy with the alternatives.
> 
> So what are other MacPython using? There are some Java-based editors,
> but I'm not persuaded that that they'll perform adequately. There's are
> numerous X11 based options, but who wants to swap into X everytime they
> need to edit a file? The once excellent editor Alpha seems to have
> disappeared permanently into testing for its next release.

I'm quite happy with BBEdit.

    <http://www.bbedit.com/>

One feature I'm growing particularly fond of is the new worksheet windows.
These are like MPW worksheets if you're familiar with MPW (Ah! thems were
the days). Basically you get an editor window that has a live shell
underneath it. By hitting the ENTER key instead of return, the last line
typed (or currently selected text) is executed and the stdout inserted at
the cursor point in the window.

This isn't so useful for very interactive stuff and I wouldn't use it as a
substitute for a terminal window, but for recording and monitoring the
output of programs it's fantastic. Basically I open a new worksheet window
and type something like this at the top:

    cd /to/project/directory
    ./start-program --non-interactive --with-lots-of-debugging-output

Then I just select the lines, hit ENTER, and watch the output fill the
window. Then I can scroll up and down and use BBEdit's grep searching
facilities on the output as it's running.

Yes, I'm sure Emacs can do all of this, but I don't want a terminal-based
editor pretending to be a GUI application.

Oh, another thing. BBEdit has a command-line tool now for firing up the
editor allowing you to type:

    % bbedit foo.py bar.py

and have them open up in BBEdit.

Jonathan (no stock options, just a happy user)



From gumuz at looze.net  Tue Sep  3 09:25:04 2002
From: gumuz at looze.net (Gumuz)
Date: Tue, 3 Sep 2002 15:25:04 +0200
Subject: How to avoid freeing memory on exit?
References: <roy-54DDA9.09043103092002@reader1.panix.com> <slrnan9dhv.1vs.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <3d74b6dc$0$231$4d4ebb8e@news.nl.uu.net>

so what does 'one' underscore prefixing a methodname mean?


"Gerhard H?ring" <gerhard.haering at opus-gmbh.net> wrote in message
news:slrnan9dhv.1vs.gerhard.haering at haering.opus-gmbh.net...
> Roy Smith wrote:
> > [...] Is there any way to get the interpreter to just exit without doing
all
> > that memory cleanup?
>
> os._exit
>
> -- Gerhard




From leipold at universal.dca.net  Thu Sep 26 14:57:58 2002
From: leipold at universal.dca.net (Walt Leipold)
Date: Thu, 26 Sep 2002 14:57:58 -0400 (EDT)
Subject: Strange change in xmlrpclib behavior
Message-ID: <200209261857.g8QIvwY07046@universal.dca.net>

Skip:

> Walt> Any ideas?  (BTW, it's xmlrpclib v0.9.8, Win2000 SP2 and NT4 SP6.)
>
> Have you tried a more recent version of xmlrpclib and/or Python 2.2.1 or
> Python from CVS?

I am now using Python 2.1.1, have tried xmlrpclib v0.9.9 without 
success, and  I'm just about desperate enough to go to the Python 
CVS.  I haven't moved to Python 2.2 because I'd have to rebuild
my serial library (but I may have to do that).

However, what's *really* bugging me is that the code *used* to 
work; even if moving to a slightly newer version fixes it, how can
I be sure it won't fail again?  I shouldn't turn it over to my 
customer until I understand what broke it in the first place...  

I'll probably end up junking XML-RPC and writing the socket code
using UDP or something similar, just so I can trust it.

Regards,

-- Walt

------------------------------------------------------------------
"If ignorance is bliss, why                           Walt Leipold
aren't there more happy people?"                   leipold at acm.org
------------------------------------------------------------------



From sholden at holdenweb.com  Mon Sep  9 15:59:08 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 9 Sep 2002 15:59:08 -0400
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid> <b16e4ef7.0209090552.3ad7691b@posting.google.com>
Message-ID: <V57f9.77078$l_4.29066@atlpnn01.usenetserver.com>

"Stephen Ferg" <steve at ferg.org> wrote in message
news:b16e4ef7.0209090552.3ad7691b at posting.google.com...
> > you'd better include a clear warning...
> > that you'll explain how to do it correctly
> > later...
>
> Can you say a little more?  What is it that makes using "bind()"
> incorrect?   And what is it about using "command" that avoids the
> problems?
>
> This is a genuine request for information, not a challenge to your
> statement.  I am still learning about Tkinter features.

The problem, from a learner's point of view, is that most Tkinter widgets
need to be bound to a number of different events. Consider, for example, how
the default button in an interface is triggered by an Enter keystroke.

The command option is specifically provided to allow the programmer to hook
their code in whenever any relevant bound event triggers that widget in the
interface.

Consequently, which bind() is a valuable technique for Tkinter programmers
to know about, it's rather like teaching someone to drive by explaining the
mechanical details of the steering and engine.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From david.abrahams at rcn.com  Thu Sep 12 00:36:38 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Thu, 12 Sep 2002 00:36:38 -0400
Subject: Any C++/Python sig members here?
References: <l1Kf9.1470$Lg2.157618@news2.news.adelphia.net>
Message-ID: <alp5rc$luo$1@bob.news.rcn.net>

"Robert Oschler" <Oschler at earthlink.net> wrote in message
news:l1Kf9.1470$Lg2.157618 at news2.news.adelphia.net...

> I joined the C++ sig for Python, to talk with others about BoostPython.  I
> joined more than a day ago and got my subscription confirmation request
from
> the sig server.  I have not seen any messages yet, is it a very low
traffic
> sig or something?

See http://mail.python.org/pipermail/c++-sig/

To answer your question from
http://mail.python.org/pipermail/c++-sig/2002-September/002055.html:
Yes, Boost.Python works on Linux. See a list of supported platforms and
compilers for Boost.Python v1 at www.boost.org/libs/python/doc/index.html.
Boost.Python v2 is compatible with even more platforms and compilers.

-Dave

--
-----------------------------------------------------------
           David Abrahams * Boost Consulting
dave at boost-consulting.com * http://www.boost-consulting.com





From hst at empolis.co.uk  Mon Sep  9 10:39:45 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Mon, 9 Sep 2002 15:39:45 +0100
Subject: xml.minidom is stripping out my CRLF's in attrib values!!
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB04F412@hendrix.empolisuk.com>

sismex01 at hebmex.com wrote:
> 
> Probably not sripped, but normalized:
> 
> for example, your original string:
> >>> s = "spotted\x0Awith black and white"
> 
> becomes the normalized string:
> >>> s = "spotted with black and white"
> 
> Is something like this happening?
> 
> If it is, then it's the correct behaviour, according
> to the XML standard: all text should be normalized,
> except if it's in a CDATA block.
> 
> HTH
> 
> -gus

White space is usually normalized, but it can be preserved (if your parser supports it) with the xml:space attribute.
If an element has the attribute xml:space="preserve" then white space in that element should not be normalized. If you are parsng with a DTD and your root element is called ROOT then change/create the !ATTLIST declaration to be

<attlist ROOT xml:space #FIXED "preserve"
...other attribute definitions...>

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.



From andrew.henshaw at mail.com  Wed Sep 11 08:34:42 2002
From: andrew.henshaw at mail.com (Andrew Henshaw)
Date: Wed, 11 Sep 2002 08:34:42 -0400
Subject: deleting the first and the last character of a string
References: <mailman.1031741971.20624.python-list@python.org>
Message-ID: <unue39cv3l2ud6@corp.supernews.com>

Manuel Hendel wrote:

> I got a file with fields seperated by a "|". The problem is that there
> is also a "|" at the beginning and at the end of each line although
> there no fields in front or behind these "|". Now I want to delete
> these two "|". Is there anything which says delete the first or last
> character of a string? I checked the string module on the web, but I
> couldn't find anything.
> 
> I solved it like this, but I honestly don't like it that way.
> 
> #!/usr/bin/env python
> #
> import sys
> import string
> 
> inputfile = open(sys.argv[-1], "r")
> 
> for line in inputfile.readlines():
>         fields = string.split(line, "|")
>         lineN = string.join(fields[1:-1], "|")
> 
> Isn't there any better solution? ;-)
> 
> Manuel
> 

Just like your slicing of the list, you can slice a string:

string[1:-1]

Although, I believe that each of your strings will have a \n that
needs to be deleted; therefore,

string[1:-2]

-- 
Andrew Henshaw



From sadams123 at optushome.com.au  Mon Sep 16 21:03:09 2002
From: sadams123 at optushome.com.au (Steven)
Date: Tue, 17 Sep 2002 11:03:09 +1000
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de><pUdh9.9377$yt3.4604908@newssrv26.news.prodigy.com><mailman.1032184821.5180.python-list@python.org> <c%qh9.102794$ub2.2120685@news1.tin.it> <mailman.1032212180.11973.python-list@python.org> <pKth9.118182$pX1.4250378@news2.tin.it>
Message-ID: <3d867f4d$0$23173$afc38c87@news.optusnet.com.au>

> > This question ("how can I obscure Python byte-code?") comes up at least
> > once a month on this list.  It's fairly clear that any scheme you come
> > up with will take longer to implement than it will to break by a
> > competent hacker.
>
> True.. that's why I'm planning to obfuscate the code then encrypt it.

you don't want to go overboard with this, people still have to use the
program - where are you going to keep the code that is 'now' (once
decrypted) in plaintext?

At some stage your code must be in the clear, and at that stage, anyone can
get at it.

e.g. running it on linux, just start it up, get past whatever encryption
you've put on there, when the program is running - and you've either got
unencrypted files sittng on disk, or, (you decideed to be tricky and keep it
all in memory) just 'cat /proc/kcore > some_file.txt' and you've got a
complete memory dump to look through.

As everyone has said, you're running against diminishing returns on any
security through obscurity.

> A hacker (not every programmer) is specialized in modify jmp instructions
> and could find the password
> to decrypt the .pyc in less than 20 minutes, but maybe 5% of them knows
> about python.

if they're able to run the code, it'd be a fairly solid bet that they'd know
about Python, if not, a quick search on Google would probaly net you all the
info you need to find out just what that file is. After all, in order to run
the program they'd need to at least install Python somehow - unless you're
planning to embed Python in some other app, and run your code in that...

I'd say, don't bother, copyright it, and only distribute the compiled python
files under a restrictive license.

Steven




From jorgenhansen at hotmail.com  Thu Sep  5 03:25:20 2002
From: jorgenhansen at hotmail.com (=?ISO-8859-1?Q?J=F8rgen_Hansen?=)
Date: 5 Sep 2002 00:25:20 -0700
Subject: Max line coords canvas.create_line() can display?
References: <fa06e058.0208291055.4b0dd338@posting.google.com> <5pid9.5885$e5.1033551@newsb.telia.net>
Message-ID: <117880a1.0209042325.68193ca7@posting.google.com>

"Fredrik Lundh" <fredrik at pythonware.com> wrote in message news:<5pid9.5885$e5.1033551 at newsb.telia.net>...
> "revyakin" wrote:
> 
> > Is there a limit of line coords (x1,y1,...xn,yn) which a Tk Canvas
> > create_line method can display? I found that I can display a line
> > consisting of 5000 x,y points, but a line of 20 000 points does not
> > get displayed.
> 
> there shouldn't be a limit.
> 
> can you perhaps post a (short) code snippet that illustrates
> the problem?

Hi 

the small program below fails to draw the line, when it is called with
an integer larger than or equal to 16384. I only tested the script on
a Windows 98 machine with Python 2.1.3.

Regards
Jorgen


--- Code start ---
import random, sys
from Tkinter import *

WIDTH = 300
HEIGHT = 200

def randomcoor(n, xmin, xmax, ymin, ymax): 
    """ Generate n sets of random coordinates within the bounds. """
    x = []
    y = []
    for i in range(n):
         x.append(random.randrange(xmin, xmax, 1))
         y.append(random.randrange(ymin, ymax, 1))
    return zip(x,y)         
             

def draw(coor):
    """ Draw a line for the list of coordinates. """    
    root = Tk()
    canvas = Canvas(root, width=WIDTH, height=HEIGHT)
    canvas.pack()
    canvas.create_line(coor)
    root.mainloop()
    

if __name__ == '__main__':
    """ Call the script with an integer number. """
    coor = randomcoor(int(sys.argv[1]), 0, WIDTH-1, 0, HEIGHT-1) 
    draw(coor)   
--- code end ---


From tim.one at comcast.net  Fri Sep 20 03:13:52 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 20 Sep 2002 03:13:52 -0400
Subject: Destructor never called ???
In-Reply-To: <ame8es$ppc$1@panix2.panix.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEGCBFAB.tim.one@comcast.net>

[Aahz]
> There are several different issues.  First of all, __del__ does not get
> called if an object is destroyed by GC (GC gets used to clean up cyclic
> garbage).

If an object is destroyed by GC, and the object has a __del__, the __del__
is called.

>>> class adel:
...     def __del__(self):
...         print "I've been del'ed!"
...
>>> class nodel:
...     pass
...
>>> A = nodel()
>>> B = nodel()
>>> C = adel()
>>> A.B = B  # half of a cycle
>>> B.A = A  # the other half
>>> A.C = C  # C is not in the cycle, but ...
>>> B.C = C  # ... will become reachable only from the cycle
>>> del C    # can only get to C via the cycle now
>>> del A    # break half the cycle
>>> del B    # break the other half
>>> import gc
>>> gc.collect()  # force a collection
I've been del'ed!
4
>>>

What is true is that if an object *in* a trash cycle has a __del__, that
cycle will not be collected, but rather shuffled into gc.garbage for the
programmer to do something sane with it explicitly.  There may be any number
of objects with __del__s not in cycles themselves but reachable only from
cycles; they don't inhibit collection, and if collection occurs their
__del__s will be invoked.  It's only the cases where Python can't possibly
guess a safe order in which to call destructors where both __del__ and
collection are inhibited.

__del__-and-gc-go-together-like-windows-and-fork-ly y'rs  - tim




From emile at fenx.com  Sun Sep  1 12:55:04 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 1 Sep 2002 09:55:04 -0700
Subject: Add methods to a class at runtime?
References: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com> <slrnan4dl3.18h.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <aktgo4$h6$1@main.gmane.org>

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote in message
news:slrnan4dl3.18h.gerhard.haering at lilith.my-fqdn.de...
> Robert Oschler wrote in comp.lang.python:
> > [Glyph said:]
> > "One of the eye-popping cool features of Python is the ability to
> > change code and have existing instances update to use the new
> > methods automatically. "
> >
> > Can someone point me to a code snippet shows how to do this?
>
> >>> class Foo: pass
> ...
> >>> foo = Foo()
> >>> foo.bar()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> AttributeError: Foo instance has no attribute 'bar'
> >>> def bar(self): print "bar"
> ...
> >>> Foo.bar = bar
> >>> foo.bar()
> bar
>

You can also use new.instancemethod to add a method to an specific instance:

Python 2.1.3 (#35, Apr  8 2002, 17:47:50) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> class Foo: pass
...
>>> foo=Foo()
>>> fee=Foo()
>>> def bar(self): print 'bar'
...
>>> import new
>>> fee.bar = new.instancemethod(bar, fee, Foo)
>>> fee.bar
<method Foo.bar of Foo instance at 0088018C>
>>> fee.bar()
bar
>>> foo.bar
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: Foo instance has no attribute 'bar'


Emile van Sebille
emile at fenx.com








From lordmacro at hotmail.com  Sat Sep 21 03:09:19 2002
From: lordmacro at hotmail.com (Tom Widrick)
Date: Sat, 21 Sep 2002 02:09:19 -0500
Subject: implementing reduce protocol in C
References: <3D8A7DA1.7090506@hotmail.com> <62Ai9.136323$pX1.4852975@news2.tin.it> <3D8AF51D.5040305@hotmail.com> <dkGi9.138788$pX1.4922161@news2.tin.it> <UlNi9.250007$2L.9285322@e3500-atl2.usenetserver.com>
Message-ID: <uIUi9.137606$AY5.54537145@e3500-atl1.usenetserver.com>

Well, I finally got it working with a small amount of contortion. I made an
init function for pickle support, called from my module init


static void
initpickle()
{
 PyObject *builtins, *globals, *build, *copy_reg, *constructor;
 /* register the proto builder for cPickle */
 globals = PyEval_GetGlobals();
 if(globals == NULL)
  return;
 builtins = PyDict_GetItemString(globals, "__builtins__");
 if(builtins == NULL)
  return;
 builtins = PyModule_GetDict(builtins);

 build = Py_FindMethod(mod_methodlist, NULL, "__proto_build__");
 copy_reg = PyImport_ImportModule("copy_reg");
 constructor = PyDict_GetItemString(PyModule_GetDict(copy_reg),
"constructor");
 PyObject_CallFunction(constructor, "(O)", build);
 PyDict_SetItemString(builtins, "__proto_build__", build);
}

The trick was that I had to make a file-level variable to hold the module
object and get the
function from it in my reduce function, not Py_FindMethod(). I'm still
trying to learn about
the internal workings of Python, but it seems that the function object is
"changed" when the
module is initialized.

static PyObject *
proto_reduce(HyProtoObject *self)
{
 PyObject *args;
 PyObject *result;
 PyObject *builder;

 if(HyProtoModule == NULL || !PyModule_Check(HyProtoModule))
  Py_FatalError("proto extension module not initialized properly.");

 builder = PyDict_GetItemString(PyModule_GetDict(HyProtoModule),
"__proto_build__");
 args = PyTuple_New(5);
 if(args == NULL)
  return NULL;

 PyTuple_SET_ITEM(args, 0, self->po_name);
 Py_INCREF(self->po_name);
 PyTuple_SET_ITEM(args, 1, self->po_parents);
 Py_INCREF(self->po_parents);
 PyTuple_SET_ITEM(args, 2, self->po_children);
 Py_INCREF(self->po_children);
 PyTuple_SET_ITEM(args, 3, self->po_doc);
 Py_INCREF(self->po_doc);
 PyTuple_SET_ITEM(args, 4, self->po_dict);
 Py_INCREF(self->po_dict);

 result = Py_BuildValue("(OO)", builder, args);
 Py_DECREF(args);
 return result;
};

Who knows if this is "proper" but it works.

Tom







From michael at stroeder.com  Tue Sep 24 12:53:50 2002
From: michael at stroeder.com (=?ISO-8859-1?Q?Michael_Str=F6der?=)
Date: Tue, 24 Sep 2002 18:53:50 +0200
Subject: Strange values in Accept-Language header?
References: <f72dac0d.0209231303.12349f2e@posting.google.com> <8FWj9.39597$V7.10466682@twister.socal.rr.com> <f72dac0d.0209240840.20fafc70@posting.google.com>
Message-ID: <uth66-1h9.ln1@junker.stroeder.com>

Alan Kennedy wrote:
> Accept-Language: x-ns16W5tM7A2Nh6,x-ns2U210btwUq5f
> Accept-Language: en-us,x-ns1jZKcD2t3NhQ,x-ns2r3509OnmPe2

For what it's worth:

Python 2.2.1 (#2, May  4 2002, 19:50:26)
Type "help", "copyright", "credits" or "license" for more information.
 >>> import base64
 >>> base64.decodestring('ns16W5tM7A2Nh6')
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "/usr/lib/python2.2/base64.py", line 44, in decodestring
     return binascii.a2b_base64(s)
binascii.Error: Incorrect padding
 >>> base64.decodestring('16W5tM7A2Nh6')
'\xd7\xa5\xb9\xb4\xce\xc0\xd8\xd8z'
 >>> base64.decodestring('s16W5tM7A2Nh6')
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "/usr/lib/python2.2/base64.py", line 44, in decodestring
     return binascii.a2b_base64(s)
binascii.Error: Incorrect padding
 >>> base64.decodestring('16W5tM7A2Nh6')
'\xd7\xa5\xb9\xb4\xce\xc0\xd8\xd8z'
 >>> base64.decodestring('2U210btwUq5f')
'\xd9M\xb5\xd1\xbbpR\xae_'
 >>>

...

Ciao, Michael.



From phr-n2002b at NOSPAMnightsong.com  Sun Sep 29 19:14:20 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Sep 2002 16:14:20 -0700
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05> <an7bvd$burft$1@ID-125932.news.dfncis.de> <an7p7t$cpg$0@216.39.172.122> <mailman.1033338759.6775.python-list@python.org>
Message-ID: <7xd6qwqtgz.fsf@ruckus.brouhaha.com>

"Chris Gonnerman" <chris.gonnerman at newcenturycomputers.net> writes:
> >>> .70 * .05
> 0.034999999999999996
> 
> Rounded to two decimals, that's 0.03; but done in proper
> decimal numbers, the answer should be 0.035, rounded to 0.04
> using either "classic" or "banker's" decimal rounding.
> ...
> So, Bengt, how do I do this right using only floats?  I can't.
> I have to have decimal arithmetic to get the right answer.

How about just adding a small amount before rounding at the end?

>>> (.70 * .05) + .000000001

rounds to the right thing.


From a at b.c  Tue Sep 24 21:41:36 2002
From: a at b.c (Daniel T.)
Date: Wed, 25 Sep 2002 01:41:36 GMT
Subject: Tkinter status bar?
Message-ID: <a-A27A4D.21421424092002@nnrp01.earthlink.net>

I'm going through a pdf file I found on the internet ("An Introduction 
to Tkinter" by Fredrik Lundh)

In chapter 8 he talks about how to put a status bar at the bottom of a 
window:

# File: tkSimpleStatusBar.py
class StatusBar(Frame):
   def __init__(self, master):
      Frame.__init__(self, master)
      self.label = Label(self, bd=1, relief=SUNKEN, anchor=W)
      self.label.pack(fill=X)
   
   def set(self, format, *args):
      self.label.config(text=format % args)
      self.label.update_idletasks()
   
   def clear(self):
      self.label.config(text="")
      self.label.update_idletasks()

He goes on to say:

We could have inherited from the Label widget itself, and just extended 
it with set and clear methods. This approach have a few drawbacks, 
though:
* It makes it harder to maintain the status bar's integrity. Some team 
members may cheat, and use config instead of set. That's not a big deal, 
until the day you decide to do some extra processing in the set method. 
Or the day you decide to use a Canvas widget to implement a fancier 
status bar.
* It increases the risk that your additional methods conflict with 
attributes or methods used by Tkinter. While the Frame and Toplevel 
widgets have relatively few methods, other widgets can have several 
dozens of widget specific attributes and methods.
* Future versions of Tkinter may use factory functions rather than class 
constructors for most widgets. However, it's more or less guaranteed 
that such versions will still provide Frame and Toplevel classes. Better 
safe than sorry, in other words.

Now my question. I didn't think to inherit from Label, instead I changed 
it to:

class StatusBar:
   def __init__(self, master):
      self.label = Label(master, text="", bd=1, relief=SUNKEN, anchor=W)
      self.label.pack(side=BOTTOM, fill=X)
 
   def set(self, format, *args):
      self.label.config(text=format % args)
      self.label.update_idletasks()
   
   def clear(self):
      self.label.config(text="")
      self.label.update_idletasks()

In other words, why inherit from any of the Tkinter classes?


From bobx at linuxmail.org  Mon Sep 16 21:30:35 2002
From: bobx at linuxmail.org (Bob X)
Date: Tue, 17 Sep 2002 01:30:35 GMT
Subject: Python equivalent to perldoc?
References: <Olvh9.7898$Lg2.1389903@news2.news.adelphia.net>
Message-ID: <%Avh9.7933$Lg2.1392332@news2.news.adelphia.net>

Bob X wrote:
> In Perl you can run "perldoc warnings" and get the definition of what 
> "warnings" does. Is there a equivalent in Python?
> 
Well...I found "pydoc". However I have to explicitly tell python to 
execute it on Windows XP. I cannot type "pydoc sys" nor "python pydoc 
sys" but I have to type "c:\python22\libs\pydoc sys" to get it to work. 
Shouldn't Python know its own libs? Or at least put the "libs" in the 
ENV variable?

Bob



From jb at cascade-sys.com  Fri Sep 27 05:19:22 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Fri, 27 Sep 2002 02:19:22 -0700
Subject: Getting started
References: <00a801c26511$4e084750$df7e4e18@brianspiv1700> <mailman.1033021239.13549.python-list@python.org> <iaDk9.145545$ub2.3207724@news1.tin.it>
Message-ID: <3D94229A.6030207@cascade-sys.com>


Alex Martelli wrote:

>James J. Besemer wrote:
>        ...
>  
>
>>Types are otherwise pretty secure in C++ except the "cast" mechanism
>>"foils" the type system and thus allows type errors to occur.
>>    
>>
>
>Actually, unions are at least as bad -- they're not often used
>for type-punning because casts are handier for that, but if you
>took cast aways, people would (ab-)use unions (just like they
>abuse EQUIVALENCE in Fortran for similar purposes).
>
You're right.  I didn't mean to suggest that casts were the only thing 
that could make C++ programs type unsafe.  Should have inserted an 
"e.g.," in the sentence, as it was intended only as an example.

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From jb at cascade-sys.com  Tue Sep 17 01:48:40 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 16 Sep 2002 22:48:40 -0700
Subject: threads or queue for this task
References: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com> <mailman.1032076038.5654.python-list@python.org> <h2jbouo4j9f8ee646bhfejt3erqku76d61@4ax.com> <mailman.1032202279.8360.python-list@python.org> <Qgqh9.117134$pX1.4203010@news2.tin.it>
Message-ID: <3D86C238.9040306@cascade-sys.com>

Alex Martelli wrote:

>James J. Besemer wrote:
>  
>
>>Generally, Queues never get "full" and you generally don't care if
>>they're empty.
>>    
>>
>
>You can define Queues that can hold a maximum number of items; [...]
>
>Setting size is just an optional embellishment, though
>sometimes it does pay dividends.
>

Setting size has it's place but otherwise introduces unnecessary 
complications.  Among other things, it forces the queue writer to deal 
with the situation of what to do if the queue is full.  Signal error 
back to its client?   Sleep and try again later?  Silently discard the 
request?  Abort?

"Generally," NOT having a limit is by far the more elegant solution.

Even in the general case, the queue may eventually grow to consume all 
available memory if one doesn't have the CPU power to keep up with 
requests.  But, like most Python errors, this eventually will raise a 
specific exception.

Anyway, I felt the "general" case was more applicable to this novice's 
line of questioning.  He/she already was confounded by some 
non-essential details.

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From lyle at users.sourceforge.net  Mon Sep  9 14:34:35 2002
From: lyle at users.sourceforge.net (Lyle Johnson)
Date: Mon, 09 Sep 2002 13:34:35 -0500
Subject: compile error on SWIG file
References: <3D7C55EE.D9278B1C@ipm.fhg.de> <3D7CB26B.2040706@users.sourceforge.net> <3D7CBFD2.4E7DFD1F@ipm.fhg.de>
Message-ID: <3D7CE9BB.7030703@users.sourceforge.net>

Markus von Ehr wrote:

> I only extended the std_vector example from the distribution with
> one function, see example.i included (there are templates inside).
> When I run SWIG with -c++ option, the warnings disappear in fact.
> But the error from the compiler is the same...

<sigh>

If you named the output file "mvelib.c", you are telling your compiler 
that it's C source code (and it isn't). It's C++ source code, so give it 
a C++ extension, e.g.

	swig -c++ -python -o mvelib.cpp mvelib.i

You will also need to update your setup.py script to use the corrected 
file name ("mvelib.cpp" instead of "mvelib.c").

Lyle



From buzzard at urubu.freeserve.co.uk  Thu Sep 26 20:47:02 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Fri, 27 Sep 2002 01:47:02 +0100
Subject: This is driving me insane!
References: <mailman.1033084697.25598.python-list@python.org>
Message-ID: <an09rk$p9e$1@newsg4.svr.pol.co.uk>

"Derek Basch" <dbasch at yahoo.com> wrote in message
news:mailman.1033084697.25598.python-list at python.org...
> Hello,
> I created a very simple class that totals some data
> but I am not getting the output from it that I would
> expect at all.
> Here is the class:
> http://www.geocities.com/dbasch/insane.txt
>
> I would expect to get:
> [0, 0]
> [0, 0]
> [[0, 0]]
>
> but I am getting:
> [0, 0]
> [0, 0]
> [[1, 1]]
>
> I dont understand how the 'grouptotals' list is
> getting the incremented values prior to the
> 'grouptotalcounter' function being called.

It isn't.  self.errorlog is a list containing self.grouptotals.  When you
increment self.grouptotals and return self.errorlog you will get [[1,1]].
Changing the appropriate line to 'self.errorlog.append(self.grouptotals[:])'
will append a copy of self.grouptotals and your function will then return
[[0,0]].

Duncan





From alanmk at hotmail.com  Fri Sep 27 06:41:14 2002
From: alanmk at hotmail.com (Alan Kennedy)
Date: 27 Sep 2002 03:41:14 -0700
Subject: Integrate with Flash MX
References: <322e5959.0209261444.797a936e@posting.google.com>
Message-ID: <f72dac0d.0209270241.402ac4f4@posting.google.com>

Peter Bengtsson wrote:
> I want to use Flash MX to build a GUI for use on the desktop on
> windows, not embedded in a webpage.
> ActionScripts don't do it for me so I want to use Python for the
> business logic.
> 
> Is there some way to enable this connection? Maybe this is more of a
> Flash question.

Hmm, I haven't actually done this, but there are two possible approaches
I can think of.

Method 1. Use the python win32 extensions to control the Flash Player
COM object directly. You create the COM object as follows

code----
import from win32com.client import Dispatch
fl = Dispatch('ShockwaveFlash.ShockwaveFlash.6')
code----

>From now you control the Flash COM object 'fl' from python. I don't
know what the methods of the Flash COM object are, nor do I know
a reference where you can find them. If you do come across such a
reference, I'd love to hear about it.

Also, I'm not sure how you would achieve interaction between the
UI elements in the Flash movie and your python objects. I'm not sure
if this can be done without resorting to ActionScript. I suppose it
depends on the interface presented by the Flash COM object. The nature
of the communication *should* be similar to the two-way communication
between DHTML scripting languages and Flash which is described below.

Method 2. 

I know you said that you don't want to embed the GUI inside a web page.
However doing that would make you task so very much simpler, that it
might be worth considering. Also, perhaps it might be possible to control
Internet Explorer behind the scenes with win32com?

Embed the Flash movie representing your GUI inside a HTML page, and
use Internet Explorer as the "intermediator" between your Flash GUI
and your python logic. 

On the Python side, you write your Python as a DHTML scripting
language. The Python win32 extensions permit you to do this. I think
it goes something like

<script language="python">
    def petshop():
        print "Man:        Good morning, I'd like to buy a cat. "
        print "Shopkeeper: Certainly sir. I've got a lovely terrier."
        print "            [indicates a box on the counter]"
        print "Man:        No, I want a cat really."
        print "Shopkeeper: [taking box off counter and then putting"
        print "            it back on counter as if it is a different"
        print "            box] Oh yeah, how about that?"
        print "Man:        [looking in box] No, that's the terrier."
        print "Shopkeeper: Well, it's as near as dammit."
        print "Man:        Well what do you mean? I want a cat."
        print "Shopkeeper: Listen, tell you what. I'll file its legs"
        print "            down a bit, take its snout out, stick a"
        print "            few wires through its cheeks. There you"
        print "            are, a lovely pussy cat."
        print "Man:        Its not a proper cat."
        print "Shopkeeper: What do you mean?"
        print "Man:        Well it wouldn't meow."
        print "Shopkeeper: Well it would howl a bit."
        print "Man:        No, no, no, no. Er, have you got a parrot?"
</script>

On the Flash side, you will need to use "fscommand"s to communicate out
to the (D)HTML thatr contains your python objects.

For an example of this kind of communication, see the following
Macromedia reference. Wherever you see "Javascript", substitute
that for "PythonScript", i.e. Python running as a DHTML scripting
language.

http://www.macromedia.com/support/flash/ts/documents/java_script_comm.htm

> Anybody got any experience with this?

Sorry, it's an unusual approach that you're using, i.e. integrating
python and Flash. I've never even considered it (until now), let alone
implemented it. I've never heard of anyone doing it.

Is there any particular reason why you don't like ActionScript? I
know version 4 was fairly limiting, but version 5 has really cleaned
up the language a lot, and you can achieve quite a bit.

I'd love to know how you get on with this.

Best of luck,

alan kennedy
----------------------------------------------
check http headers here: xhaus.com/headers
email alan:              xhaus.com/mailto/alan


From jimt at aiinet.com  Tue Sep 10 13:27:38 2002
From: jimt at aiinet.com (Jim Taylor)
Date: Tue, 10 Sep 2002 13:27:38 -0400
Subject: shared libraries
Message-ID: <xYpf9.1$wH3.129@client>

Host: Solaris 2.5.1
Compiler: gcc 2.95

Problem:
When Python loads the extension and the module goes to create the method
table by calling Py_InitModule(), the information passed as parameters is
not correct. I end up getting an address out of bounds error.

Compilation of the module is done with:
gcc  -shared -fPIC -DPIC -I/usr/local/python1.5/SunOS5/include/python1.5 -I/
core -I. -c  -o ctlr.o ctlr.c
gcc -lposix4  -G -Xlinker --export-dynamic  -shared -o ctlr.so ctlr.o

The loader is loading the module, however, during the init function when
Py_InitModule() is called, a segmentation fault occurs indicating an address
out of bounds error. When I try to print the contents of the module table
from within the initialization function, I also get a segmentation fault, so
the problem is the way in which the module has been loaded. It appears there
is an address space issue, but I don't know how to resolve it.

Ideas anyone?
Thanks
Jim Taylor
jimt at aiinet.com




From gerhard.haering at gmx.de  Tue Sep 10 13:09:02 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 10 Sep 2002 17:09:02 GMT
Subject: Trouble upgrading to 2.2.1
References: <aXof9.29813$w51.7932884@news2.news.adelphia.net> <slrnans81k.jc.gerhard.haering@lilith.my-fqdn.de> <_qpf9.29862$w51.7937790@news2.news.adelphia.net>
Message-ID: <slrnans9pd.jc.gerhard.haering@lilith.my-fqdn.de>

Robert Oschler wrote in comp.lang.python:
> "Gerhard H?ring" <gerhard.haering at gmx.de> wrote:
>> Robert Oschler wrote in comp.lang.python:
>> > I'm trying to upgrade from 2.1.1 to 2.2.1.  I'm using kpackage (KDE's
> rpm
^^

Your newsreader is broken. Please fix it or use something better than
this crapware to post to usenet.

>> > wrapper).
>>
>> You should install the OpenSSL libraries in order to fulfill these
>> dependencies. If you do have the OpenSSL libraries installed, but
>> the libcrypto or libssl version numbers are different on your
>> machine, you can try to provide the necessary symlinks so rpm will
>> shut up.
> 
> I have 0.9.6b-34 of OpenSLL installed.  Is this the wrong version or
> do I have a different problem?

I have no idea which Linux distribution you're using or which
libraries your package installs. To fulfill the dependencies, I'd do
this on my box (as root):

$ ln -s /usr/lib/libcrypto.so.0.9.6 /usr/lib/libcrypto.so.2

The version number might be different from 0.9.6 on your box, check
with ...

$ ls -al /usr/lib/libcrypto.so.* /usr/lib/libssl.so.*

... and adjust accordingly if needed.

Ditto for libssl.

HTH,

-- Gerhard


From aahz at pythoncraft.com  Mon Sep 23 21:55:30 2002
From: aahz at pythoncraft.com (Aahz)
Date: 23 Sep 2002 21:55:30 -0400
Subject: Python threading?
References: <f9Pi9.10275$Lg2.2191294@news2.news.adelphia.net>
Message-ID: <amogmi$r1l$1@panix1.panix.com>

In article <f9Pi9.10275$Lg2.2191294 at news2.news.adelphia.net>,
Robert Oschler <Oschler at earthlink.net> wrote:
>
>Also, what is a reasonable number of threads to expect to be able to run
>before context switching overhead becomes a problem (I'm using a PIII 500
>Mhz with 512MB ram if that helps).

It depends.  For pure Python-based computation, more than one thread
will cost you more than you gain.  If it's almost pure I/O (file or
socket stuff), you might be able to use as many as a couple of hundred
threads.  Mostly, you're probably best off somewhere between five and
thirty threads, depending on what you're doing.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From fperez528 at yahoo.com  Mon Sep  9 17:54:47 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 09 Sep 2002 15:54:47 -0600
Subject: Save session in python interactive console
References: <a50631f6.0209091334.7d59476c@posting.google.com>
Message-ID: <alj5b7$8nd$1@peabody.colorado.edu>

newgene wrote:

> Hi, group,
>     After I was working at python interactive console(activepython
> v2.2 at Win2000 platform), I am wondering if possible to save the
> current session, so that I can load it again and continue the work
> next time. Just like R or Matlab, etc. Although python is a
> programming language, one of the best features I like is its
> interactive console. By writting my own module, I can make it as my
> personal working console. If saving session is possible, it will be
> much more convinient.
>     The pickle module provided by python can save any object,but how
> can I save all objects in current session?

This problem is non-trivial and a plain pickle solution will fail. You may 
want to try ipython, which is a better interactive command-line environment 
than the default one and has _limited_ save/restore abilities. I tried using 
pickle but failed, so until I have time to dig deeper into the problem I 
implemented a log save/replay option which while admittedly imperfect, in 
many cases is a reasonable solution.

You'll find ipython at: http://www-hep.colorado.edu/~fperez/ipython/

Cheers,

f.

ps. Some of ipython's nicest features only work in a proper terminal which 
Windows lacks, so you'll need either a Unix environment or the Cygwin system 
under Windows. However the basics stuff is all there under windows.


From max at alcyone.com  Wed Sep  4 18:00:57 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 04 Sep 2002 15:00:57 -0700
Subject: str() of a tuple
References: <H1xnHJ.1zE1@beaver.cs.washington.edu>
Message-ID: <3D768299.30517666@alcyone.com>

Douglas Zongker wrote:

> Why, oh why does the str() of a container use the repr()s of the
> objects inside?  This seems to violate what the docs claim str() does
> -- "return a nicely printable representation of an object."  Would a
> sane implementation break something?

I presume this is an historical issue; perhaps if it were planned again
from scratch maybe it would have been done differently.  It certainly
does seem to cause a fair amount of confusion.

> Is there a convenient way to print out a tuple of floats so that the
> output is actually readable?

You can do it manually yourself:

	'(%s)' % ', '.join(map(str, T))

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ There is nothing so subject to the inconstancy of fortune as war.
\__/ Miguel de Cervantes
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From ponderor at lycos.com  Tue Sep 10 11:42:30 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 10 Sep 2002 08:42:30 -0700
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Sep  9)
References: <D5F9F6F5A0C46CAC.9DCA346C4331EBAF.B9AA42F7FC35A33A@lp.airnews.net>
Message-ID: <e81be8b2.0209100742.30c295b4@posting.google.com>

> 
> "This is actually the first new program I've coded in C (rather than
> Python) in a good four years or so." - Eric S. Raymond 
> http://mail.python.org/pipermail/python-list/2002-September/121533.html
> 

The direct link to this quote is:

http://mail.python.org/pipermail/python-dev/2002-August/028106.html


From tdelaney at avaya.com  Wed Sep 18 23:25:39 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 19 Sep 2002 13:25:39 +1000
Subject: Minimal debug/rep functionality
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5B2@natasha.auslabs.avaya.com>

> From: Peter Hansen [mailto:peter at engcorp.com]
> 
> Mark wrote:
> > Hello all,
> > 
> > In a nutshell, what I'd like to do is develop a utility 
> that would print out 
> > every line in a python program as the program executes (I 
> may even just 
> > want to print line numbers).  I can see hints of how to do 
> this in pdb (and 
> > more directly in the underlying bdb class) and also in the 
> code module.
> > 
> > So, 
> > 
> > 1) Should I try and inherit a class and override some 
> methods or should I 
> > just pull out the relevant code and hand spin a small interpreter.
> > 
> > 2) Anyone have a 20 liner solution for this?
> 
> Sorry, I couldn't find a solution that was larger than six lines:
> 
>  >>> def trace(frame, event, arg):
> ...   if event == 'line':
> ...      print '%s: %s' % (frame.f_code.co_filename, frame.f_lineno)
> ...   return trace
> ...
>  >>> import sys
>  >>> sys.settrace(trace)

Having been writing a fairly comprehensive code and path coverage tool
(incomplete - currently working on determining what *is* an executable line
... think I need to work with the AST), I'd like to point out a couple of
things with the above:

1. frame.f_code.co_filename will return the file that the .pyc was compiled
from. Particularly in the case of standard libraries, this will often not be
the path to the file in *your* filesystem. There is a way to get the
corresponding file in your filesystem (basically, determine the module the
frame comes from, then get the module filename using inspect).

2. The above will be *slow*. Very very slow. For a short program, this may
not matter. For a long program, it will (I'm talking order of magnitude
slower).

The approach I have taken is to simply stored the line numbers and file info
in a list, then dump it out at the end. Well, that's not *precisely* what I
do, but that's the gist of it. I actually do lots of caching of file info,
compression of the data, etc ... but you don't want to see that code ;)



From dance_code at hotmail.com  Tue Sep 24 11:40:49 2002
From: dance_code at hotmail.com (lion)
Date: 24 Sep 2002 08:40:49 -0700
Subject: What is the mechanism behide?
Message-ID: <e895ad50.0209240740.6b59f426@posting.google.com>

I'm studying the singleton pattern, there are something I don't
understand in the sample code below:
class OnlyOne:
  class __OnlyOne:
    def __init__(self, arg):
      self.val = arg
    def __str__(self):
      return `self` + self.val
  instance = None
  def __init__(self, arg):
    if not OnlyOne.instance:
      OnlyOne.instance = OnlyOne.__OnlyOne(arg)
    else:
      OnlyOne.instance.val = arg
  def __getattr__(self, name):
    return getattr(self.instance, name)

print "The output:"
x = OnlyOne('sausage')
print x
y = OnlyOne('eggs')
print y
z = OnlyOne('spam')
print z
print x
print y
print `x`
print `y`
print `z`

Here are the output:
The output:
<__main__.__OnlyOne instance at 0x00C65AE0>sausage
<__main__.__OnlyOne instance at 0x00C65AE0>eggs
<__main__.__OnlyOne instance at 0x00C65AE0>spam
<__main__.__OnlyOne instance at 0x00C65AE0>spam
<__main__.__OnlyOne instance at 0x00C65AE0>spam
<__main__.OnlyOne instance at 0x00C62C58>
<__main__.OnlyOne instance at 0x00C64400>
<__main__.OnlyOne instance at 0x00C5DC20>

    I'm puzzled that what is behide the statement "print x" ? It seem
that the "print x" is implicitly turned into "print x.instance". It
was said that the redirection is steered by the function
__getattr__(self, name), but the function need two parameter, there is
only one, I mean, x. Additionally, the function __str__(self) must
also be invoked implicitly in the print-statement. Finally, the
difference  between statement "print x" and "print `x`" is also
unknown for me.
    It all like magic! Could anyone explain the mechanism for me?

Thank in advance,
Lion


From dyoo at hkn.eecs.berkeley.edu  Mon Sep 16 20:31:42 2002
From: dyoo at hkn.eecs.berkeley.edu (Danny Yoo)
Date: Mon, 16 Sep 2002 17:31:42 -0700 (PDT)
Subject: [Tutor] Re: Not sure how to phrase this question
In-Reply-To: <000501c25dde$79a8be60$9865fea9@arthur>
Message-ID: <Pine.LNX.4.44.0209161717040.18670-100000@hkn.eecs.berkeley.edu>


> In short, I want to avoid a lot of named arguments, want to have the
> user interface with "argument smart" classes (in the manner of how I
> think method overloading generally works), going so far as wanting
> positional arguments that are in effect order insensitive.  Give a
> Intersection class a plane and a line in any order and it will know what
> you are trying to get at. No Intersection(plane=planeinstance,
> line=lineinstance). No error on Intersection(planeinstance,lineinstance)
> because the class has be set up to receive only (lineinstance,
> planeinstance).


Hi Arthur,

This link might give some more ideas on doing this kind of generalized
operations.  The code is in Scheme, but the ideas should transfer over
gracefully:

    http://mitpress.mit.edu/sicp/full-text/book/book-Z-H-18.html#%_sec_2.5

Can you give a list of operations that you want to provide as well as
"Intersection"?  Will order always not matter, or are there operations
like 'subtract', where order does matter?  Perhaps your program might be
amendable to a generic-operations approach.


What you're doing sounds reasonable: I think the main problem here is that
it's out of the experience of most mortals!  *grin*


Best of wishes to you!




From mjackson at alumni.caltech.edu  Fri Sep 27 08:01:07 2002
From: mjackson at alumni.caltech.edu (Mark Jackson)
Date: 27 Sep 2002 12:01:07 GMT
Subject: How do you pronounce "__" (double underscore)?
References: <mailman.1033098816.3362.python-list@python.org>
Message-ID: <an1ha3$u4$2@news.wrc.xerox.com>

brueckd at tbye.com writes:
> On 26 Sep 2002, Mark Jackson wrote:
> 
> > > > Pat Notz wrote:
> > > >> @ = "at"
> > > >> ! = "bang"
> > > >> . = "dot"
> > > >> * = "star"
> > > >> 
> > > >> Is there a short (one or two syllable) pronounciation for "__" (double
> > > >> underscore).  "Underscore underscore" and "double underscore" are kind
> > > >> of long.  Somthing like "splat"?  Anyone?
> > > > 
> > > > _ = 'under'
> > > > 
> > > > __ = 'under-under'
> > > 
> > > 
> > > How about "underder"??
> > 
> > How about "dunder" (double under)?
> 
> How about "score" and "scorescore"?

That would make __seven__ "Gettysburg"!

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	Yes, but what if it were a parrot?
			- Graham Chapman, to John Cleese




From aleax at aleax.it  Tue Sep 24 10:13:41 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 24 Sep 2002 14:13:41 GMT
Subject: list problem
References: <mailman.1032818129.8126.python-list@python.org> <Xns92938CC236BA2duncanrcpcouk@127.0.0.1> <_FZj9.158106$pX1.5692316@news2.tin.it> <Xns92939499298F5duncanrcpcouk@127.0.0.1>
Message-ID: <pq_j9.158510$pX1.5701647@news2.tin.it>

Duncan Booth wrote:

> Alex Martelli <aleax at aleax.it> wrote in
> news:_FZj9.158106$pX1.5692316 at news2.tin.it:
> 
>> I _do_ think that reduce(operator.__add__, ...) is the only
>> really frequent use of reduce, with such dominance that it might
>> even be worth having as a built-in -- clearest and fastest.  It
>> might even specialcase _strings_ in particular, turning the op
>> into a ''.join in this case... just musing...
> 
> The simplest thing could be to modify operator.add to accept any number of
> arguments. Then you could use operator.add(*aList) to add up or
> concatenate as appropriate all the elements of the list.

I suspect that would convert aList to a tuple (but I'd have to
double check the internals to make sure) thereby negating any
performance benefits.  Maybe it would be better to allow operator.add
to be called with ONE argument (must be iterable and not empty), OR two 
(working just like it works now, in this case).  This should avoid
any issue with type-conversion list-to-tuple or whatever.


> It still wouldn't be as fast as ''.join though unless you special cased
> it, and it might be hard to special case it effectively: you would have to
> check the types of all the objects in the list before deciding whether or
> not to use join.

Yes, I could write a class X, with an __add__ that's able to accept a
string argument (thus an instance of X would work with a 
reduce(operator.add ...) but is not a string (thus would fail if one of the 
items in a list passed to ''.join was an instance class X).  But I could 
also simply check the first argument: if a string, FIRST try the equivalent 
of ''.join -- if  that raises a TypeError, fall back to looping on 
addition.  A tiny price to pay, I think, for the rare/weird case of a class 
with an __add__ that, for the benefit of helping newbies avoid the
single most likely performance pitfall of Python.  But I'm not sure
it's such a good idea anyway.  Mostly, the hypothetical builtin sum
or one-argument operator.add would be used to tot up numbers, though
of course as a nice side effect of polymorphism it would also work
for other types (maybe horribly slow for big lists of small strings:-).


Alex



From magnus at thinkware.se  Tue Sep 17 10:46:06 2002
From: magnus at thinkware.se (=?ISO-8859-1?Q?Magnus_Lyck=E5?=)
Date: Tue, 17 Sep 2002 16:46:06 +0200
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <3D7958DB.50903@nospam.free.fr>
Message-ID: <3D87402E.6020700@thinkware.se>

laotseu wrote:
> Rod Stephenson wrote:
> 
>> Slashdot has a list of questions posed to Larry Wall (perl). When
>> asked his thoughts on other scripting languages, he makes the
>> following observation about python
>>
>> "Python is cool to look at small bits of, but I think the "outline"
>> syntax breaks down with larger chunks of code. I'm with Aristotle on
>> the structure of discourse--a story should have a beginning, and
>> middle, and an end. So should blocks"

Don Knuth said:
"We will perhaps eventually be writing only small modules
which are identified by name as they are used to build
larger ones, so that devices like indentation, rather than
delimiters, might become feasible for expressing local
structure in the source language."
See http://www.thinkware.se/cgi-bin/thinki.cgi/PythonQuotes

Stephen Edwards said:
Python: Perl designed by a sane man.
http://www.cs.columbia.edu/~sedwards/classes/2002/w4115/scripting.pdf

Of course there is an end for every Python block. The
Zen thing to consider is the old: "What is the sound of
one hand clapping?" Consider that, and return when you
are enlightened. ;)

My personal experience from many years of coding C++ etc
with curlies and Python without, is that I tend to get
lost a lot more in my C++ code than in my Python code.
Not to mention the code of others... I've certainly had
a number of debugging sessions where I had to help some
mediocre C++ coder who couldn't match his own braces...
The problem is usually that the blocks get to big... The
fact that you can get away with incorrect indentation in
the brace-languages makes it worse.

Most modern editors have tools like indentation guides and
folding that can make it easier to see block structures.

Was it one of Beethoven's symphonies that ended so abruptly
that the audience didn't understand when to applaud? It was
considered a big failure at it's time, and it's considered
one of the best works of a genious today...

>> I'm not quite sure what he's trying to get at here - I guess that for
>> a long heavily indented chunk of code, you could lose track of the
>> overall structure, but I don't write code this way.
>>
>> Any comments?
> 
> 
> def fun(args):
> #BEGIN
>     if ceci or cela:
>     #BEGIN
>         do_this(with_that)
>         [snip a lot of code]
>     #END
> #END

I would recommend against such a coding style.

First of all, if it's difficult to understand the code
(even after the first 10 minutes of Python exposure)
you probably have a bad structure in your code. Too
big blocks, or too much nesting in the structure. The
solution there is refactoring, not comments.

Secondly, you are adding something redundant and foreign
to your code. Getting used to how Python actaully looks
is a better idea than trying to make it look like Pascal.

Thirdly you are creating a maintenace problaem. The main
problem with all comments (except that they often don't
get written) is that they tend to get out of sync with
the code. No syntax checks or test script will detect if
your #END markers are incorrect. If the code is bad enough
to warrant #END markers, it will be easy to write them
incorrectly, and even easier to make mistakes when you
change your code, adding or removing a loop or an if-
statement in the middle of the structure.

If you still feel that you want an explicit end marker to
your blocks, do it like this.

  * Forget about #BEGIN, if you are unable to see where the
    Python blocks begin, you might as well give up.

  * Please forget about the ugly capitals.

  * Add some meaningful information to the end marker so that
    you see which block you are ending.

  * Make sure that you have a program that will check / generate
    the end markers.

It could look like this:

def fun(args):
     if ceci or cela:
         do_this(with_that)
         [snip a lot of code]
     #end: if ceci or cela
#end: fun(args)

It shouldn't be so difficult to write a python script that
will remove all lines beginning with #end: and then regenerate
them correctly for all blocks. (Not that I find it worth
while...)

These kinds of scripts are good to run as pre-checkin triggers
if your develpment environment supports that kind of routines.



From skip at pobox.com  Thu Sep 19 15:26:12 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 19 Sep 2002 14:26:12 -0500
Subject: linking extension modules w/ "ld" on Solaris - to be expected?
In-Reply-To: <8d148763.0209190901.6276b37e@posting.google.com>
References: <mailman.1032393445.324.python-list@python.org>
        <8d148763.0209190901.6276b37e@posting.google.com>
Message-ID: <15754.9428.370794.424364@12-248-11-90.client.attbi.com>

    Alan> I've tried compiling both Python 2.1.1 and 2.2.x on Solaris 8 and
    Alan> 9 with gcc 2.9x and 3.x in my path. I consistently get the same
    Alan> build error each time, so it is likely that there is some *switch*
    Alan> I am not setting somewhere.  

    ...

    Alan> gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
    Alan> -DHAVE_CONFIG_H  -DBUILD=`cat buildno` -o Modules/getbuildinfo.o
    Alan> ./Modules/getbuildinfo.c
    Alan> rm -f libpython2.2.a
    Alan> ar cr libpython2.2.a Modules/getbuildinfo.o
    Alan> gmake: ar: Command not found

You need /usr/ccs/bin in your PATH.  That's where ar and several other
compilation- and link-related commands live, not just compilers proper.

-- 
Skip Montanaro - skip at pobox.com
The need for gutters to be cleaned is directly proportional to how hard it
happens to be raining at the moment.



From Oschler at earthlink.net  Sat Sep 28 14:38:27 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sat, 28 Sep 2002 18:38:27 GMT
Subject: Best way to create a class dynamically on the fly?
References: <tFml9.13043$Lg2.3865279@news2.news.adelphia.net>
Message-ID: <DGml9.13046$Lg2.3865365@news2.news.adelphia.net>

"Robert Oschler" <Oschler at earthlink.net> wrote in message
news:tFml9.13043$Lg2.3865279 at news2.news.adelphia.net...
> What is the best way to create a Python class dynamically at runtime?  A
> sample or doc would be great.  I'm a veteran C++ programmer so this
concept
> is very alien to me as you can imagine (dynamically created classes).  An
> example that would show the creation of a descendant (sub-class), and a
> another example showing the creation of an ancestor class would be
perfect.
>
> thx

I forgot to mention, I would like the classes I create dynamically to be
fully "pickle-able".  If there are any caveats or limitations in this area
please let me know.

thx





From cbbrowne at acm.org  Fri Sep 20 19:54:55 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: 20 Sep 2002 23:54:55 GMT
Subject: Decimal arithmetic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amabpb$5rm$1@news.nems.noaa.gov> <3D8B1537.4010803@thinkware.se> <ad052e5c.0209201128.1a53108@posting.google.com>
Message-ID: <amgcgf$5e1fd$1@ID-125932.news.dfncis.de>

In an attempt to throw the authorities off his trail, danb_83 at yahoo.com (Dan Bishop) transmitted:
> Magnus Lyck? <magnus at thinkware.se> wrote in message news:<3D8B1537.4010803 at thinkware.se>...
>> ChrisBarker wrote:
>> > The only reason I can see
>> > to not writing your class in Python is performance, and unless you are 
>> > doing a lot of calculations, that is probably a non-issue.
>> 
>> Performance issues might well matter in large financial
>> systems, but that's not the only reason.
>
> Please name one other real-world situation in which decimal arithmetic
> is actually *needed*.
>
>> [snip]
>> And if someone would write
>> 
>> x = Decimal(9.999999999999999) - Decimal(9.9999)
>> 
>> they would not get what they expected (unless they expected
>> 0.000099999999998). Ok, they could write

> But how often does an error of 10 parts per trillion really matter?

> Afaict, not nearly enough to justify having a default number
> representation that works against the hardware.

This isn't thinking like an ex-COBOL programmer.

An ex-COBOL programmer will look at this and say something like:

  "You stupid buffoons!  You're arguing over how to deal with some
  abstruse fifteenth decimal place.  In COBOL, we never had to worry
  about such idiocy concerning the numeric /error/ that you're
  describing.

  If we're working with dollars and cents, we define the variable as:

  X PIC 999999999V99

  And we will have none of these idiotic rounding errors."

Add X to a variable with the same format, and there's NO issue of bad
rounding.  The behaviour of BCD arithmetic is perfectly well defined,
and _doesn't_ depend on hoping and praying that the IEEE FP operations
will do approximately the right thing.

The COBOL (or PL/I, or anyone else accustomed to BCD data types) guy
will justifiably look at the bizarro FP rounding as being, well,
bizarre.

There's _no_ good reason for subtracting two perfectly good decimal
values from one another to result in having weird artifacts appear in
some 17th decimal place, aside from having a willingness to admit
fairly deranged arithmetic.

People use FP arithmetic making the ludicrous assumption that FP
values are "Real numbers."  They aren't, despite some languages that
call them REALs.  They also generally aren't decimal values, but
rather binary fractions, from whence enters the bizarre artifacts of
FP.

I think there's a good argument for the idea that people should
generally be flogged mercilessly if they use FP values without having
had a /good/ education in numerical analysis which is /necessary/ in
order to be able to competently use them.
-- 
(concatenate 'string "cbbrowne" "@cbbrowne.com")
http://cbbrowne.com/info/unix.html
"It don't mean a thing, if it ain't got that swing..." 


From fperez528 at yahoo.com  Wed Sep 18 16:59:57 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 18 Sep 2002 14:59:57 -0600
Subject: linux python ideas
References: <3D88E578.6010001@diespammerdieuselesspython.com>
Message-ID: <amapge$22r$1@peabody.colorado.edu>

Rob Andrews wrote:

> I'll be giving a presentation on Python to my local LUG
> (http://lugoj.org) in a few weeks, and would like to point out at least
> a few things that would be of particular interest to linux users.
> 
> The idea is not "how to program in Python", which would take more than
> one quick presentation, but a one-off demonstration of ways in which
> Python is particularly useful for such a group.
> 
> Any suggestions would be appreciated, since my notes are pretty generic
> so far.

<blatant plug>
Check out ipython at http://www-hep.colorado.edu/~fperez/ipython

It's not linux-specific, but its best features only work in a linux/unix 
environment, and it will appeal very much to command-line junkies. Take a 
look at the screenshots to get a feel for its features and feel free to drop 
me a line if you have a question.

It allows you to use the python shell almost as a system shell, not so much to 
replace the regular shell for day to day stuff, but rather to have the power 
of file and directory management when you need to do things with python. For 
me it's very useful in processing scientific data, since I have both access 
to Python/Numeric/Gnuplot and at the same time have essentially a system 
shell.

The introspection features (name? and <tab>) also make it very useful for 
regular programming tasks.
</blatant plug>

Well, check for yourself. In general, python+Numeric+Gnuplot/Grace makes a 
pretty strong stand against the $$$ likes of Matlab. And in many cases beats 
them (in others we're still behind, but catching up quickly).

Cheers,

f.


From manuel at hendel.net  Tue Sep 17 03:19:00 2002
From: manuel at hendel.net (Manuel Hendel)
Date: Tue, 17 Sep 2002 09:19:00 +0200
Subject: problems getting a single item out of a list
Message-ID: <20020917071900.GB23629@partagas.as.de.cw.net>

I try to get a single item out of a list, but after this, the single
item is stil a list item.

domainfield = fields[3:4]

I just want to get a new variable with a the value of the 3rd item of
the list fields.

Can anyone help?

Thanks,
    Manuel

-- 
Wisdom eventually comes to all of us. Someday it might even be your turn. 
-David & Leigh Eddings, "Polgara, The Sorceress" 



From stuart at bmsi.com  Mon Sep 23 17:16:10 2002
From: stuart at bmsi.com (Stuart D. Gathman)
Date: Mon, 23 Sep 2002 17:16:10 -0400
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amabpb$5rm$1@news.nems.noaa.gov> <3D8B1537.4010803@thinkware.se> <ad052e5c.0209201128.1a53108@posting.google.com> <mailman.1032664289.615.python-list@python.org>
Message-ID: <amo0aq$5nf$1@nntp2-cm.news.eni.net>

On Sat, 21 Sep 2002 23:10:09 -0400, Christian Tismer wrote:

> Dan Bishop wrote:

>  > Please name one other real-world situation in which decimal
>  > arithmetic is actually *needed*.
> 
> There is no reason to argue about that. Decimal arithmetic is exact for
> most of the operations bankers require. They have been demanding for it
> since decades, and they will not drop it. 

Let me add a horror story of what happens when financial
applications do *not* demand fixed point/decimal arithmetic.

We took on a client whose accounting system used binary floats for all
dollar amounts.  And I mean *dollar* amounts, not penney amounts.  Since
all pennies were approximate, the total due at the bottom of the invoices
would rarely match what the customer would get adding up the line items
on a calculator.  It was only off by a few cents, but it tended to make
customers lose faith in the system.  Statements also did not add up, and
the Balance Sheet didn't balance.

We converted their system to use fixed point, and now everything balances
and totals to the penney.  Unfortunately, their CPA keeps his own version
of their General Ledger on Excel with binary floating point.  The CPA
balance sheets do not balance.  The CPA General Journal transactions do
not balance.  I demonstrated with a calculator that the CPA figures did
not balance, but they insisted that we had to go with the CPA figures
because he is a CPA, and I'm just a programmer.
This is tricky since their new General ledger does not accept unbalanced
transactions.

My solution was to create a 'CPA adjustment' account to contain
all the discrepancies, and manually decide how to allocate the
extra/missing pennies for the CPA transactions.

There ought to be severe penalties for idiots that use floating point
dollars for financial applications.  If forced to use floating point
(e.g. because customer demands BASIC), then keep money amounts in
whole pennies (or whatever the smallest currency unit for the country
is), and divide by 100 (or whatever) for printing only (or just add the
decimal point yourself).

-- 
Stuart D. Gathman <stuart at bmsi.com>
Business Management Systems Inc.  Phone: 703 591-0911 Fax: 703 591-6154
"Confutatis maledictis, flamis acribus addictis" - Mozart background
song for the Microsoft "Where do you want to go from here?" commercial.


From jdhunter at ace.bsd.uchicago.edu  Thu Sep  5 13:10:24 2002
From: jdhunter at ace.bsd.uchicago.edu (John Hunter)
Date: Thu, 05 Sep 2002 12:10:24 -0500
Subject: plex 
Message-ID: <m2u1l4qs5b.fsf@mother.paradise.lost>

I am writing a plex lexer/scanner to parse a pdf file.  Here is the
first part, which extracts the streams.

I would like to do this a bit more efficiently, namely, to read the
streams in multicharacter chunks rather than one character at a time.
As it is, the function add_stream has to be called for every character
in the stream.  

Any advice how to do this?

Here is the code:

from Plex import *

def add_stream(scanner, text):
    scanner.thisStream += text

def end_stream(scanner, text):
    print 'BeginStream: ', scanner.thisStream, 'EndStream:'  # do something with the stream here
    scanner.thisStream = ''
    scanner.begin('')

lexicon = Lexicon([
  (AnyChar, IGNORE),
  (Bol + Str("stream
") + Eol, Begin('stream')),
  State('stream' , [
    ( Bol + Str("endstream") + Eol, end_stream ), 
    ( AnyChar, add_stream),
    ]),
  ])

filename = "test.pdf"
f = open(filename, "r")
scanner = Scanner(lexicon, f, filename)
scanner.thisStream = ''


while 1:
  token = scanner.read()
  if token[0] is None:
    break


Thanks,
John Hunter



From aahz at pythoncraft.com  Thu Sep 12 10:49:19 2002
From: aahz at pythoncraft.com (Aahz)
Date: 12 Sep 2002 10:49:19 -0400
Subject: class variable won't icrement!
References: <3d8054ef$0$234$4d4ebb8e@news.nl.uu.net> <alpner$bm$1@wanadoo.fr>
Message-ID: <alq9hf$bn$1@panix1.panix.com>

In article <alpner$bm$1 at wanadoo.fr>,
Eric Brunel  <eric.brunel at pragmadev.com> wrote:
>
>I always found that inheritance of class attributes to instance attributes 
>was really misleading, and I even think it would be better if it were 
>simply removed from Python. IMHO, it would far more conform to the 
>"explicit is better than implicit" rule that is one of the basis of Python.

If I think of it as "scope resolution" rather than "inheritance", I find
that it makes more sense.  It works exactly the same way as hiding a
global with a local.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From debl2nononspammmy at bellatlantic.net  Sat Sep 21 19:51:16 2002
From: debl2nononspammmy at bellatlantic.net (David Lees)
Date: Sat, 21 Sep 2002 23:51:16 GMT
Subject: Math optimization
References: <6T5j9.145247$pX1.5195840@news2.tin.it> <3D8D02A0.FF572334@bellatlantic.net>
Message-ID: <3D8D0612.A6CE7011@bellatlantic.net>

Using the quick and dirty code below and just changing line 9 between
prod = n * b and prod = n*7, running on my Win98SE, Python 2.2.1 I get a
13% difference between using a constant and a variable.  I don't know
much about what Python is doing in the way of optimization or if it is
even keeping things as integer arithmetic, so take it with a grain of
salt.

David Lees



import time,string

n = string.atoi(sys.argv[1])
total = n
prod = 1
b = 7
t0 = time.clock()
while (n):
   prod = n*b
   n -= 1
t1 = time.clock()

n=total
t2 = time.clock()
while (n):
   pass
   n -= 1    
t3 = time.clock()

A = t1-t0
B = t3-t2

print total, A-B

David Lees wrote:
> 
> This depends on your hardware, OS and probably Python version. Further
> you need to specify the data type.  Integer, floating point or infinite
> precision integers.  It is quite easy to run benchmarks, which you
> should do yourself.
> 
> David Lees
> 
> Zed wrote:
> >
> > Hi!
> > What's faster in python? Divide or multiply?
> > (a= b*c  is faster than   a=b/c)?
> >
> > Math expressions with constants are calcolated during .pyc creation or real
> > time?
> > (a=b*5   is faster than    a=b*5*2/2 )?
> >
> > Thanks


From mennosimons at gmx.net  Tue Sep 24 12:00:39 2002
From: mennosimons at gmx.net (Ulli Stein)
Date: Tue, 24 Sep 2002 18:00:39 +0200
Subject: PicklingError with md5.md5() object Von: Ulli Stein <mennosimons@gmx.net> Datum: Dienstag, 24. September 2002 17:03:09 Gruppen: comp.lang.python Folgenachricht an: mennosimons@gmx.net Keine Referenzen
Message-ID: <amq1td$eli$07$1@news.t-online.com>

Hi,

while trying to pickle a md5 instance I get the following error:

>>> import cPickle, md5
>>> f=open("/tmp/t", "w")
>>> p = cPickle.Pickler(f)
>>> m=md5.md5()
>>> p.dump(m)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
cPickle.UnpickleableError: Cannot pickle <type 'md5.md5'> objects


What should I do? The marshal module is also not recommended for this task.

Thanks.


From ckasso at sprynet.com  Wed Sep 25 18:02:16 2002
From: ckasso at sprynet.com (Chris Kassopulo)
Date: Wed, 25 Sep 2002 18:02:16 -0400
Subject: how to change font and its size in IDLE (in UNIX)?
References: <mailman.1032931413.10497.python-list@python.org>
Message-ID: <20020925180216.0dce7381.ckasso@sprynet.com>

On Tue, 24 Sep 2002 22:22:35 -0700
Kow K <kowk at earthlink.net> wrote:

> Hi all,
> 
> I'm trying to figure out how to change the font and its size in IDLE.
> It's Courier in 10pt. I don't like Courier, and what's worse,
> 10pt is too small for me. I wish I could change it to Fixed 7x13.
> 
> Any input will be appreciated.
> 
> Kow
> 
> 

>From the file ~/.idle:

# IDLE reads several config files to determine user preferences.  This 
# file is the default config file.  When IDLE starts, it will look in
# the following four files in order:
#     config.txt                      the default config file
#     config-[win/unix/mac].txt       the generic platform config file
#     config-[sys.platform].txt       the specific platform config file
#     ~/.idle                         the user config file

--
Chris Kassopulo


From dcswest at flashmail.com  Wed Sep  4 00:01:41 2002
From: dcswest at flashmail.com (Dennis B.)
Date: 03 Sep 2002 21:01:41 -0700
Subject: Privacy and Inheritance
Message-ID: <1031112108.240.28.camel@localhost>

Greetings; 

Have enough confidence in Python to realize this is probably just my
misunderstanding, but wondering why Python mangles private, double
underscore attributes with the name of the base class even if they're
inherited and the instance is that of the sub class. 

Like the following doesn't even seem to work: 

class StandardNormal: 

    def letEventZScore(self, eventZScore): 
        self.__eventZScore = float(eventZScore) 

class DerivativeNormal(StandardNormal): 

    def getCentral(self): 
        if self.__eventZScore < 0: 

Unless I change that last line to something like: 

        if self._StandardNormal__eventZScore < 0: 

Maybe just missing the point all together, but wondering if anyone'll
point out anything a little more elegant and/or intuitive. 

Thank you and have a great day,




From altis at semi-retired.com  Sat Sep 14 19:58:46 2002
From: altis at semi-retired.com (Kevin Altis)
Date: Sat, 14 Sep 2002 16:58:46 -0700
Subject: Python11
References: <xwMg9.1433$Dy4.464@nwrddc03.gnilink.net> <m3admk47pf.fsf@mira.informatik.hu-berlin.de>
Message-ID: <o2Qg9.53$uo5.66284@news.uswest.net>

Is Portland, OR the official location, I didn't think that had been decided
yet? I haven't seen anything posted on the O'Reilly conference site. I hope
it will be in Portland, that would be great.

ka
"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m3admk47pf.fsf at mira.informatik.hu-berlin.de...
> "Jeff Sasmor" <jeff at sasmor.com> writes:
>
> > Is there any definitive information on the next Python conference,
> > Python 11?  I've read the conferences-discuss archive on
> > Python.org (http://mail.python.org/pipermail/conferences-discuss/)
> > and the discussion seems to be tapering off.  IIRC, by last year
> > at this time there was already a call for papers.
>
> Python11 will be colocated with the O'Reilly Open Source Convention
> (OSCON), which will take place July 7-11, 2003 in Portland, USA.
>
> I believe O'Reilly will organize the specific schedule, including
> setting dates for the call for papers, submission deadlines, etc.
> Given that this is still ten months away, I don't see any urgency.
>
> HTH,
> Martin




From loewis at informatik.hu-berlin.de  Mon Sep 16 04:06:55 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 16 Sep 2002 10:06:55 +0200
Subject: Book for win32?
References: <am430k$9l3$03$1@news.t-online.com>
Message-ID: <j4elbunysg.fsf@informatik.hu-berlin.de>

Thomas Guettler <zopestoller at thomas-guettler.de> writes:

> I am interested in book about Python on win32.
> 
> I found one from o'reilly: "Python on win32"
> 
> Are there better books or alternatives?

No.

Regards,
Martin


From andrewm at object-craft.com.au  Mon Sep 16 20:51:18 2002
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Tue, 17 Sep 2002 10:51:18 +1000
Subject: Pickling slotted instances 
In-Reply-To: Message from loewis@informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=) 
   of "16 Sep 2002 09:25:05 +0200." <j44rcqpfam.fsf@informatik.hu-berlin.de> 
References: <mailman.1031805375.9030.python-list@python.org> <j4fzwf349u.fsf@informatik.hu-berlin.de> <mailman.1032144560.28257.python-list@python.org> <m3it16ze98.fsf@mira.informatik.hu-berlin.de> <mailman.1032158478.7315.python-list@python.org>  <j44rcqpfam.fsf@informatik.hu-berlin.de> 
Message-ID: <20020917005118.81D3C3C8FF@coffee.object-craft.com.au>

>Or, do you mean, by "instance", an object of a certain type which has
>slots? If so, inst.attr is the same as 
>
>getattr(inst.__class__, 'attr').__get__(inst)
>
>(if not inst.__dict__.has_key('attr')). So 'attr' will be searched in
>the class, using the method resolution order. This happens to return 
>a member descriptor, namely <member 'attr' of `inst.__class` objects>.

Yes, that's what I meant - I'd been struggling to work out where attribute
descriptors resided, and you've just added the missing detail - thanks
(I'd been fiddling with "property()" and not getting the results I
expected).

That certainly makes pickling harder.

>"The bases", for me, would put "proper syntax" before support for
>automatic pickling, e.g. via
>
>class X(object):
>  def foo              # defines a slot foo.
>  def bar[transient]   # defines a transient slot.

I'd agree with this. The __slots__ notation is a bit of a wart, but then
again, it's consistent with several other tricks.

>New-style classes are really the same as traditional types. In the
>past, you could not pickle arbitrary types unless you have special
>machinery (such as copy_reg manipulation); so it is not surprising
>that you can't pickle certain cases of new-style classes
>automatically. Notice that you *can* pickle them - just define the
>__getstate__.

I see your point, but it seems like most of the machinery is already
there.  It's a shame that slotted classes aren't more "complete", because
they are the perfect answer to people who use tuples instead of classes
for performance reasons (they are only slightly slower than a tuple).

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/



From duncan at NOSPAMrcp.co.uk  Tue Sep 10 10:00:34 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 10 Sep 2002 14:00:34 +0000 (UTC)
Subject: dynamism
References: <mailman.1031660952.2330.python-list@python.org>
Message-ID: <Xns928597A0CEC6Dduncanrcpcouk@127.0.0.1>

"Mark McEahern" <marklists at mceahern.com> wrote in
news:mailman.1031660952.2330.python-list at python.org: 

> [Duncan Booth]
>> New style classes (which are those with 'object' or other builtin
>> classes as a base class) do not all allow you to add arbitrary
                                   ^^^
>> attributes: 
> 
> Au contraire:
> 
> $ python
> Python 2.2.1 (#1, Jun 25 2002, 10:55:46)
> [GCC 2.95.3-5 (cygwin special)] on cygwin
> Type "help", "copyright", "credits" or "license" for more information.
>>>> class foo(object):pass
> ...
>>>> f = foo()
>>>> f.x = 1
>>>> f.x
> 1
>>>>

Read what I wrote? The word 'all' was supposed to be a clue.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From sam at localhost.localdomain  Wed Sep 11 18:55:46 2002
From: sam at localhost.localdomain (newbie)
Date: Wed, 11 Sep 2002 23:55:46 +0100
Subject: I must be an idiot, please pity me!
References: <mailman.1031782274.18676.python-list@python.org>
Message-ID: <oPPf9.2866$XE6.173592@newsfep1-win.server.ntli.net>

Perhaps I wasn't clear, but I don't believe I mentioned anything not
functioning properly. I did say "All works fine". I don't have failing
code, I simply suspect (hope) the current implentation isn't very good.
The pseudo code listed was what I was epecting python to allow me to do,
but due to the namespace implementation, it doesn't. I have to do what I
considered as a work around, and that is by passing a db connection to
pretty much everything to get seperate modules/functions to share a db
connection. Maybe that's just the way things are done in python, modules
and databases? I don't know, that's why I'm asking.

Putting everything into one module works as expected too. But then it's
not very good for code reusage, as importing that module brings me back to
passing a db connection to everything.

I was hoping that I had missunderstood a basic concept, and that there was
a nice way of perform something along the lines of the pseudo code listed.



On Wed, 11 Sep 2002 18:10:42 +0100, Mark McEahern wrote:

> [newbie]
>> But it doesn't seem to work like that.
> 
> I don't know about all the mind readers out there, but you really
> haven't given us any helpful information.  Are you getting an error?
> What are the symptoms of failure?  What are you expecting to happen? Why
> not try to simplify what it is you're trying to do and share *that*
> code?
> 
> For instance, the functions in each of your modules--do they take a
> database connection as a parameter?
> 
> Have you tried stuffing everything into one module--just to see if it
> works?
> 
> Etc.
> 
> // m
> 
> -


From Michael.Sparks at rd.bbc.co.uk  Fri Sep 13 09:18:54 2002
From: Michael.Sparks at rd.bbc.co.uk (Michael Sparks)
Date: Fri, 13 Sep 2002 14:18:54 +0100
Subject: wrapping yield ?
References: <alqehl$9qv$1@nntp0.reith.bbc.co.uk> <Xns92885F082A21Cduncanrcpcouk@127.0.0.1> <alsdpt$2nq$1@nntp0.reith.bbc.co.uk> <Xns9288759449799duncanrcpcouk@127.0.0.1> <alsg06$3e2$1@nntp0.reith.bbc.co.uk> <Xns92888DEA3BB2duncanrcpcouk@127.0.0.1>
Message-ID: <alsods$5vo$1@nntp0.reith.bbc.co.uk>

> Don't forget that iterating with a generator is *much* more efficient than
> calling a Python functions thousands of times.
...
> the subsequent iterations have virtually no overhead at all.

Can't forget something you don't already know - very much appreciated...

Best Regards,


Michael.




From marklists at mceahern.com  Fri Sep 13 12:18:40 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 13 Sep 2002 11:18:40 -0500
Subject: question about dictionary type..
In-Reply-To: <alt22f$pf8$1@newsreader.mailgate.org>
Message-ID: <JHEOKEOOLIGLDHCMAHMOAEIJDOAA.marklists@mceahern.com>

[eugene kim]
> i'd like to have a dictionary which looks like
> table = { category1 : { category2 : { category3 : 'garbage value' } } }
>
> category1 has many category2
> category2 has many category3
>
> i expected
> table[ 'computer' ][ 'language' ][ 'python' ] = 0
> to create {computer : { language : { python :0 }}}
> table[ 'computer' ][ 'language' ][ 'java' ] = 0
> to becomes { 'computer' : { 'language' : { 'python' :0 , 'java' :0 }}}
>
> but python doesn't allow me to do this..
> can anyone help me?

In this case, the problem is obvious, but rather than "python doesn't allow
me to do this?" it's generally more helpful to provide exact tracebacks or
more explicit information about what's not working.

The problem?  You're not telling Python to create the sub dictionaries...

$ python
Python 2.2.1 (#1, Jun 25 2002, 10:55:46)
[GCC 2.95.3-5 (cygwin special)] on cygwin
Type "help", "copyright", "credits" or "license" for more information.
>>> t = {}
>>> t['computer'] = {}
>>> t['computer']['language'] = {}
>>> t
{'computer': {'language': {}}}
>>>

etc.

// m




From aahz at pythoncraft.com  Sat Sep 14 10:36:38 2002
From: aahz at pythoncraft.com (Aahz)
Date: 14 Sep 2002 10:36:38 -0400
Subject: threads or queue for this task
References: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com> <GmCg9.2354$Ee4.6690@news-server.bigpond.net.au>
Message-ID: <alvhhm$8eu$1@panix1.panix.com>

In article <GmCg9.2354$Ee4.6690 at news-server.bigpond.net.au>,
Mark Hammond  <mhammond at skippinet.com.au> wrote:
>
>Look for Aahz's threading tutorial - URL not at hand, but it can't be 
>too hard to find.  Indeed, I predict he will tell you exactly where it is ;)

On my web page, of course.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From schneiker at athenalab.com  Thu Sep 26 20:32:08 2002
From: schneiker at athenalab.com (Conrad Schneiker)
Date: Fri, 27 Sep 2002 00:32:08 GMT
Subject: http://www.python-in-business.org
Message-ID: <cGNk9.237$2Q6.169@newssvr16.news.prodigy.com>

Chris,

Do you have anything to do with these people?

Is there any sort of analogous North American organization?

Conrad




From whisper at oz.net  Wed Sep 18 00:34:38 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 17 Sep 2002 21:34:38 -0700
Subject: PythonScript
In-Reply-To: <362133AB12B2CE4EA3C01E12CD6CCC6E0254E08B@osprey.corp.resmed.org>
Message-ID: <GCEDKONBLEFPPADDJCOEKEKGFAAA.whisper@oz.net>

<snip>
> The problem I seem to be hitting is utilising python in applications that
> embed an ActiveScript Engine. To date I am unable to get one python script
> to execute. Other languages vbscript, jscript,perl etc, work in the same
> environment so I'm left wondering if its a configuration problem or
> something that I might be overlooking. If anyone could give me pointers,
> url's, snippets etc. it would be interested. The product I specifically
> having problems is FinalBuilder (http://www.atozedsoftware.com)
> though have
> seen this problem in other applications as well.
>
> I've tried both ActiveState and Python distribution 2.2.1 with Win32
> 1.48all.
>
> Regards
>
> Gavin
>

Did you register Python as an ActiveScript host? I believe this is done for
you automatically if you use the installer and not done for you if you
don't. It's possible that it gets done for you the first time you run the
PythonWin IDE too.

Have a look at this file: \Python22\Lib\site-packages\win32com\readme.htm

HTH,

Dave LeBlanc
Seattle, WA USA





From brian at sweetapp.com  Wed Sep 25 20:10:35 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Wed, 25 Sep 2002 17:10:35 -0700
Subject: "which" command?
In-Reply-To: <yibznu5fxvx.fsf@coxeter.math.toronto.edu>
Message-ID: <009c01c264f1$23333690$df7e4e18@brianspiv1700>

> A few days ago there was a discussion here about a
> "which" written in python (as opposed to running
> the OS "which" command). Any pointers to that thread?
> As you can imagine it's been difficult to make a
> sensible google groups query since I can't think
> of any reasonably specific terms to search for...

The actual software is available here:
http://starship.python.net/~tmick/

Cheers,
Brian




From skip at pobox.com  Thu Sep 19 12:22:49 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 19 Sep 2002 11:22:49 -0500
Subject: Remarks to Python 2.3
In-Reply-To: <20020919152834.592eaa39.leafnode-usenet@schabi.de>
References: <20020918222421.328088eb.use-net@schabi.de>
        <amannn$p4b$1@panix1.panix.com>
        <m3u1knj8j9.fsf@mira.informatik.hu-berlin.de>
        <20020919152834.592eaa39.leafnode-usenet@schabi.de>
Message-ID: <15753.63961.863682.282501@12-248-11-90.client.attbi.com>

    Markus> martin at v.loewis.de (Martin v. Loewis) schrieb:

    >> Or, better yet, submit bug reports.

    Markus> I'm not really shure whether my remarks are about "bugs" ...

By submitting to the bug tracker you provide a single place to gather
commentary about an issue.  If it's determined its not a bug, it will be
marked as such and closed.  By carrying out the discussion here, you run the
risk that someone interested in the issue will miss one or more messages
about it because they got lost within the sea of other messages.

-- 
Skip Montanaro - skip at pobox.com
The need for gutters to be cleaned is directly proportional to how hard it
happens to be raining at the moment.



From marklists at mceahern.com  Tue Sep 17 20:34:04 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 17 Sep 2002 19:34:04 -0500
Subject: Splitting a list of strings
In-Reply-To: <am8flq$23i$1@driftwood.ccs.carleton.ca>
Message-ID: <JHEOKEOOLIGLDHCMAHMOMEGDDPAA.marklists@mceahern.com>

> I need to read a list of strings from a file[...]

For those of us who are lazy and lacking in imagination, why not share a few
lines of representative sample data?

// m

p.s.  Don't use str as the name of a variable, you'll clobber the builtin
str.

-




From Padraig at Linux.ie  Tue Sep 10 09:56:37 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Tue, 10 Sep 2002 13:56:37 GMT
Subject: Graph Layout Algorithms
References: <alksio$d2l$1@newsg3.svr.pol.co.uk>
Message-ID: <pSmf9.3468$cP3.6944@news.iol.ie>

Duncan Smith wrote:
> Hello,
>          Just wondering if anyone has implemented anything in Python.  I
> need to lay out my graphs (directed acyclic / undirected / trees) in a
> reasonably clear way.  So if anyone has already implemented anything, or if
> anyone has advice on appropriate algorithms, I'd be glad to hear about it.
> Cheers.
> 
> Duncan

Do you mean graphically represent or internally in data structures?
If graphical you could generate the simple text
files required by:
http://www.research.att.com/sw/tools/graphviz/
Alternatively you could do a python wrapper for
the libgraph of the same toolkit? Actually there
is one done already! http://www.cs.brown.edu/~er/software/

P?draig.



From cjw at sympatico.ca  Fri Sep 27 12:32:57 2002
From: cjw at sympatico.ca (Colin J. Williams)
Date: Fri, 27 Sep 2002 12:32:57 -0400
Subject: Attempt to install and test gnuplot fails
Message-ID: <3D948839.4617A278@sympatico.ca>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020927/7d2ff8ca/attachment.html>

From gerhard.haering at opus-gmbh.net  Wed Sep  4 05:54:47 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 4 Sep 2002 09:54:47 GMT
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com> <oukd9.7$4E3.806@news.ecrc.de>
Message-ID: <slrnanbm0h.1rs.gerhard.haering@haering.opus-gmbh.net>

Markus Jais wrote:
> hi
> one reason is that with every programming language you can learn new
> concepts.
> [...]
> the pragmatic programmers (http://www.pragmaticprogrammer.com - very good, 
> you should visit it) recommend to learn at least one new language every 
> year.

I used to do that, too. But the last ones I tried to check out (Ocaml and
Dylan) didn't solve any of my real problems, so I came back to Python very
quickly. Ocaml didn't have enough libraries for me. And Dylan just doesn't
look mature enough.

> so maybe you should learn both PHP and Python (and Ruby, of course :-)

I'd skip PHP. But then, I'm biased. And proud of it :-P

-- Gerhard


From mcherm at destiny.com  Wed Sep 11 15:07:08 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Wed, 11 Sep 2002 15:07:08 -0400
Subject: Variables vs. names
References: <3D7F8132.2D5CEC24@earthlink.net>
Message-ID: <alo48l$c97$1@main.gmane.org>

Joseph A. Knapka wrote:
> Are there languages that
> achieve Python's semantics using "real"
> variables? Java's semantics seem pretty much
> identical to Python's in this respect, yet
> nobody in the Java community will object if
> you call a thing-with-a-name-and-a-value a
> "variable".

I can't answer your other questions, but I may have some insight on this 
one. I agree that the semantics are basically the same (at least for 
Java objects). But Java doesn't have the ability to access a namespace 
as a dictionary ( locals(), globals(), .__dict__, etc ). Also, people 
who talk about Java have a certain tendency to treat Java as the 
One-True-Way for a language to work, while people who talk about Python 
try to place it in context of other languages.

Of course, maybe there are more concrete reasons that I'm missing.

-- Michael Chermside






From roy at panix.com  Mon Sep  9 06:47:49 2002
From: roy at panix.com (Roy Smith)
Date: Mon, 09 Sep 2002 06:47:49 -0400
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <bd751532.0209082142.54236307@posting.google.com>
Message-ID: <roy-FAC2A7.06474809092002@reader1.panix.com>

bslesins at best.com (Brian Slesinsky) wrote:
> Of course it's unnecessary and should be entirely optional.

One of the biggest impediments to learning Perl is that so much stuff is 
optional.  You learn a concept and think you've got it.  Then, you look 
at a different example of code that purports to do the same thing and it 
looks entirely different.


From robin at jessikat.fsnet.co.uk  Thu Sep  5 05:23:21 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 5 Sep 2002 10:23:21 +0100
Subject: jython memory usage
References: <hVm5ydAY2hd9Eww$@jessikat.demon.co.uk>
 <3d76366a$1_5@news.bluewin.ch> <rnLbuKAv5jd9Ewzz@jessikat.demon.co.uk>
 <3d76494d$1_5@news.bluewin.ch> <YR13DVAUNld9EwXH@jessikat.demon.co.uk>
 <3d765ce6_4@news.bluewin.ch> <zzKd7KAqhnd9EwxT@jessikat.fsnet.co.uk>
 <3d767c4a$1_1@news.bluewin.ch>
Message-ID: <YWyHyIAJKyd9EwHA@jessikat.demon.co.uk>

In article <3d767c4a$1_1 at news.bluewin.ch>, Samuele Pedroni
<pedronis at bluewin.ch> writes
>
>Robin Becker <robin at jessikat.fsnet.co.uk> wrote in message
>zzKd7KAqhnd9EwxT at jessikat.fsnet.co.uk...
>> .....
>> it, test_pdfbase_ttfonts.py doesn't look as though it's doing anything
>> particularly odd, certainly I see no references to system/spawn in
>> there.
>> --
>
>have you tried my little program with your JVM.
>
>If my program work as for me, that means that your py plus jython trigger a
>bug of the JVM, because anyway and in any case  the JVM ought to scream "out
>of memory" well before eating a 1GB of memory.
>
>regards.
>
>
OK here's the results with outofmem

C:\jython-2.1\reportlab\test>javac outofmem.java

C:\jython-2.1\reportlab\test>java outofmem
1
2
3
4
5
6
7
8
9
10
11
Exception in thread "main" java.lang.OutOfMemoryError


Is it possible I'm using a wrong java engine somewhere?
-- 
Robin Becker


From tismer at tismer.com  Wed Sep 18 16:14:11 2002
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 18 Sep 2002 22:14:11 +0200
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <96904b50.0209181149.429e8533@posting.google.com>
Message-ID: <3D88DE93.6080004@tismer.com>

Lee Gray wrote:
...

> Thanks for all the suggestions, folks!
> 
> I've collected lots of links to a variety of very nice screenshots,
> and I've been playing with the wxWindows demo and Boa Constructor. Boa
> was a bit disappointing, not as an app, but because of its current
> alpha status - but it's got great potential and I'll definitely look
> forward to future releases. I'll check out PythonCard too.

Boa is in fact quite a bit alpha-ish, but after you're
used to the quirks and limitations, it is really great.
Also note that you should use the version from CVS
which is *far* more developed than the downloadable.

I've written a quite big wizard for 3D image processing
with wxPython, PIL, Stackless. If you like, I can prepare
some screen shots.

ciao - chris
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?=  Fri Sep 13 12:22:04 2002
From: =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?= (=?iso-8859-5?Q?=B4=D0=DC=F8=D0=DD_=B3=2E?=)
Date: Fri, 13 Sep 2002 18:22:04 +0200
Subject: calling Oracle functions
References: <3d81f811$1@news.mt.net.mk> <mailman.1031929598.17983.python-list@python.org>
Message-ID: <3d8210a3@news.mt.net.mk>

> Oracle differentiates between functions and procedures so you must
> identify that you are calling a function, not a procedure. The DB API
> does not allow for this, so you must use an anonymous PL/SQL block and
> some non DB API code. The way to do this is as follows:

ok. I also tried that, I mean a PL/SQL block... but

> Replace the data types as appropriate.

> v_Vars = v_Cursor.setinputsizes(p_Result = cx_Oracle.NUMBER)

didn't know about this :)) thanks.
...
> If anyone has suggestions about how to improve this, fire away... :-)

yes, put it in the documentation. 
Actually is there any documentation about cx_Oracle, past the README
file?

Now a simpler question:
	is it possible to specify the data source name directly not through
	the alias defined in tnsnames.ora. In perl DBD I can specify a
	connection string like cs = "dbi:Oracle:host=1.10.100.200;sid=TEST"?
	



-- 
??????

         ???????? ?? ????...
                ...????????? ?? ????.



From maxm at mxm.dk  Sun Sep 29 15:02:18 2002
From: maxm at mxm.dk (Max M)
Date: Sun, 29 Sep 2002 21:02:18 +0200
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com> <TFGl9.642$XX3.454818@newssrv26.news.prodigy.com>
Message-ID: <3D974E3A.3000809@mxm.dk>

dsavitsk wrote:
> "Jurie Horneman" <jhorneman at pobox.com> wrote in message

<250 lines and 8.5 KB of quote snipped ...>

Please do not quote the full message you are replying to, just to add a 
few short lines of your own. We can allready read the first message in 
that post.

It is a pretty shure way not to get read.


kind regards Max M



From aleax at aleax.it  Mon Sep 16 15:27:44 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 16 Sep 2002 19:27:44 GMT
Subject: threads or queue for this task
References: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com> <mailman.1032076038.5654.python-list@python.org> <h2jbouo4j9f8ee646bhfejt3erqku76d61@4ax.com> <mailman.1032202279.8360.python-list@python.org>
Message-ID: <Qgqh9.117134$pX1.4203010@news2.tin.it>

James J. Besemer wrote:
    ...
>>How can one code if you don't
>>reliably know when a Queue is empty or full?
> 
> Generally, Queues never get "full" and you generally don't care if
> they're empty.

You can define Queues that can hold a maximum number of items; in
this case, they do get full if they're being filled faster than
they're being emptied.  Sometimes, this IS what you want -- and
generally in such cases you just call method put and let it sleep
until a slot frees up, should that be needed -- just the converse
of the normal case for empty.

For example, suppose an architecture such as:

                        --> W1
   supervisor -> Q ---/---> W2
                      \---> W3

i.e., one supervisor thread that produces work requests and
places them in a single queue Q, from which three worker
threads peel work requests and produce results (say the W
threads place their results in another queue Q1 from which
another thread X peels them for output or whatever -- it
doesn't really matter).

Say that producing a work request takes CPU time T, that
processing the work request and producing the result takes
10T, and that 3 or 4 simultaneous threads is ideal to keep
the system maximally busy.  Round-robin scheduling assumed.

Now, if Q is unlimited, the supervisor thread will soon
fill it up.  In, say, the first 40T or so time units, threads
W1, W2 and W3 will each have gotten and processed one work
unit, while the supervisor has received 10T time units and
thus has produced 10 work units: there are now 7 units in
Q.  After another 40T have elapsed, Q has grown to now hold
14 units.  And so on: Q grows without bounds, assuming there
is always more work to perform.  Eventually Q overflows
available memory, and boom, the whole process goes haywire.
Even if the total number of work units to perform is bounded,
say N, with this arrangement Q must still grow to hold 0.7N
work units at the peak.  This might force pagination and at
the very least slow things down noticeably.

The solution is trivial: just set Q to hold, at most, 3 items.
A little bit more, if you want -- doesn't matter.  In any
case, soon the supervisor thread will find Q full and wait
until one of the W threads is done with its work and peels
off another work unit, thus freeing a slot in Q.  Basically,
with this simple device, we have changed the way processing
time is being allocated: instead of each 40T units going,
30 to processing 3 work units and 10 to producing 10 work
units, giving Q a maximum size automatically balances the
number of work units produced and processed when the
system is working in steady-state, so we know by this that
each 33T units will go, 30 to process 3 work units, 3 to
produce 3 more work units.  Besides avoiding wasting too
much memory on Q, this also speeds up the rate at which
results are output in steady state: 3 results come out
every 33T time units, instead of every 40T time units.
That's an average wait of 11T time units per result produced
(best possible, since it takes 1T to produce the work
request and 10T to process it!) rather than 13.3T -- in
this sense, the system with an unlimited-size Q is about
20% slowes than the variant with a Q of limited size (during
the system's overall steady-state).

You're welcome to apply queueing theory to analyze variants
where times to produce and process working units are not
deterministic (or, probably easier, simulate them:-) to
determine the effects of Q's size depending on these key
system parameters.  E.g., if sometimes processing time is
very short, we want to set Q's size large enough to
ensure there is just about more work for the W thread that
got free unexpectedly fast -- and similarly, if sometimes
time to produce the next working unit is occasionally quite
long.  For each assumption about times there will be
optimal sizes for Q to optimize system throughput or
response times.  Rarely will optimal sizes be infinite.



> The import thing about Queues is that the reader thread will
> automatically be suspended when reading from an empty queue and
> furthermore it will automatically wake up when the queue becomes non
> empty.
> 
> All you need from Queues is to create them and to get and put data.

Yep.  Setting size is just an optional embellishment, though
sometimes it does pay dividends.


> You probably should look for a book on concurrent sequential processes.
>  Most of the paradigms and pitfalls are pretty standard and would be
> spelled out in the book.  Unfortunately, I am at a loss to make a
> specific recommendation.  Perhaps somebody else has an idea.

I think the standard reference on this nowadays might be
Gregory Andrews' "Foundations of Multithreaded, Parallel, and 
Distributed Programming", Addison-Wesley.


Alex



From marklists at mceahern.com  Fri Sep 20 16:50:58 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 20 Sep 2002 15:50:58 -0500
Subject: average?
In-Reply-To: <amfscs$paa$1@solaris.cc.vt.edu>
Message-ID: <JHEOKEOOLIGLDHCMAHMOOEEFEAAA.marklists@mceahern.com>

[Carl Banks]
> What I can't figure out is how you got from this to what you
> typed. :-)

Yeah, holy cow, me neither!  <wink>

Thanks for the reply.

Cheers,

// mark



From warkid at storm.ru  Tue Sep 17 06:47:37 2002
From: warkid at storm.ru (Kerim Borchaev)
Date: Tue, 17 Sep 2002 14:47:37 +0400
Subject: More on Protecting Source Code
In-Reply-To: <GCEDKONBLEFPPADDJCOEAEDDFAAA.whisper@oz.net>
References: <GCEDKONBLEFPPADDJCOEAEDDFAAA.whisper@oz.net>
Message-ID: <1134805499.20020917144737@storm.ru>

Hello David,

Tuesday, September 17, 2002, 7:26:06 AM, you wrote:

DL> I suppose one solution is to modify the Python interpreter with different
DL> op-codes and that ought to make it somewhat painful for the average hacker.
DL> A better solution is to make a .pyc file approximately as hard as a binary
DL> .exe file to decompile - however that could be done.

Just a thought.
Keeping your python modules encripted(e.g. with rotor module?) you can
reimplement __import__ builtin to properly handle importing these modules.

Best regards,
 Kerim                            mailto:warkid at storm.ru





From davbucko at deletethis.yahoo.com  Sun Sep  1 20:33:44 2002
From: davbucko at deletethis.yahoo.com (davbucko)
Date: Mon, 2 Sep 2002 01:33:44 +0100
Subject: Keyerror problem
Message-ID: <_lyc9.7164$7N.29584@newsfep3-gui.server.ntli.net>

Hi guys,

I have a feeling this is going to turn out to be a pretty stupid question,
but here goes anyway...

I have a class Flags as follows:

class Flags:
    def __init__(self):
        self.flags = {}

    def getInverseFlags(self):
        for i in self.flags.keys():
            #switch some keys/values around

    def setFlag(self,name,value):
        self.flags[`name`] = value
#End of class Flags

I then call that class and use it:
flags = Flags()
powers = Powers() #This class is working fine elsewhere, so I don't think
that it is the problem

def getComplimentaryGrid():
    flags.getInverseFlags()
    for j in powers.dict[flags.flags['row1']]: #This line doesn't work
For some reason, on the line above, I am getting the following error
message:

Traceback (most recent call last):
  File "C:\Dave\python\squaresprint.py", line 188, in ?
    iterateGrids()
  File "C:\Dave\python\squaresprint.py", line 163, in iterateGrids
    grid.checkGridCols()
  File "C:\Dave\python\squaresprint.py", line 78, in checkGridCols
    getComplimentaryGrid()
  File "C:\Dave\python\squaresprint.py", line 167, in getComplimentaryGrid
    for j in powers.dict[flags.flags['row1']]:
KeyError: row1

I don't know why this is, as 'row1' is clearly a key.  If I enter the
following just above the line that is giving me trouble above:

for i in flags.flags.keys():
    print i,

I get the output: 'col1' 'col3' 'col2' 'row3' 'row2' 'sum' 'row1' which
shows that 'row1' is a key.

I am entirely new to python, so if this is an idiotic question, please say.
If you need more code, there (obviously) is more, but imvho it is not
causing any problems.

Cheers
Dave




From see_reply_address at something.invalid  Wed Sep 18 02:14:08 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Wed, 18 Sep 2002 18:14:08 +1200
Subject: Splitting a list of strings
References: <am8flq$23i$1@driftwood.ccs.carleton.ca>
Message-ID: <3D8819B0.8000702@something.invalid>

Sean Ross wrote:

> 
> What I'm interested in knowing is, can I extract all of the "attribute"
> strings from lines in such a way that I get my list of "attribute" strings,
> but all of those strings have been removed from lines as well, all at the
> same time?


     attributes = []
     data = []
     for line in lines:
       c = line[0]
       if c == "@":
         attributes.append(line)
       elif c <> "%":
         data.append(line)


-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From syver-en+usenet at online.no  Sun Sep  8 11:43:36 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Sun, 08 Sep 2002 15:43:36 GMT
Subject: Using Python to call a DCOM server - "'Class not registered"
References: <3D787D95.9000907@ob_scure.dk> <5%ce9.1272$_l1.2960@news-server.bigpond.net.au> <3D7A1359.4010607@ob_scure.dk> <Ytwe9.5037$_l1.12309@news-server.bigpond.net.au> <3D7AFF61.7070005@ob_scure.dk>
Message-ID: <u3cskcwrz.fsf@online.no>

Thomas Jensen <spam at ob_scure.dk> writes:
 
> Searching for "unregister localserver" I found your book "Python
> Programming on Win32" on O'Reilly :-)
> If I can find it in Denmark, I think I'll buy it!

You should, was my first python book, and really catapulted me into
the language. It was easy to obtain in norway, so shouldn't be hard in
Denmark either.
-- 

Vennlig hilsen 

Syver Enstad


From jepler at unpythonic.net  Wed Sep 11 22:09:44 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Wed, 11 Sep 2002 21:09:44 -0500
Subject: ANN: Thinking in Tkinter
In-Reply-To: <20020911150511.GA28720@unpythonic.net>
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid> <alln71$1s2s$1@nntp6.u.washington.edu> <b16e4ef7.0209110616.48dfe04c@posting.google.com> <20020911150511.GA28720@unpythonic.net>
Message-ID: <20020911210943.D1555@unpythonic.net>

(whoops, I had intended to send this to python-list)

On Wed, Sep 11, 2002 at 10:05:53AM -0500, Jeff Epler wrote:
> You'd have to read the tcl source for the button bindings to see what's
> going on.  I'll walk you through them ...
> 
> 
> bind Button <1> {
>     tkButtonDown %W
> }
> bind Button <ButtonRelease-1> {
>     tkButtonUp %W
> }
> bind Button <Enter> {
>     tkButtonEnter %W
> }
> bind Button <Leave> {
>     tkButtonLeave %W
> }
> 
> proc tkButtonDown w {
>     global tkPriv
>     set tkPriv(relief) [$w cget -relief]
>     if {[string compare [$w cget -state] "disabled"]} {
>         set tkPriv(buttonWindow) $w
>         $w configure -relief sunken -state active
>     }
> }
> 
> proc tkButtonUp w {
>     global tkPriv
>     if {[string equal $tkPriv(buttonWindow) $w]} {
>         set tkPriv(buttonWindow) ""
>         $w configure -relief $tkPriv(relief)
>         if {[string equal $tkPriv(window) $w]
>               && [string compare [$w cget -state] "disabled"]} {
>             $w configure -state normal
>             uplevel #0 [list $w invoke]
>         }
>     }
> }
> 
> proc tkButtonEnter w {
>     global tkPriv
>     if {[string compare [$w cget -state] "disabled"] \
>             && [string equal $tkPriv(buttonWindow) $w]} {
>         $w configure -state active -relief sunken
>     }
>     set tkPriv(window) $w
> }
> 
> proc tkButtonLeave w {
>     global tkPriv
>     if {[string compare [$w cget -state] "disabled"]} {
>         $w configure -state normal
>     }
>     if {[string equal $tkPriv(buttonWindow) $w]} {
>         $w configure -relief $tkPriv(relief)
>     }
>     set tkPriv(window) ""
> }
> 
> When the pointer leaves or enters the button, tkPriv(window) is
> set to that widget.  When a mouse button is pressed in a button,
> tkPriv(buttonWindow) is set.  When the mouse button is released, and
> window and buttonWindow both match the widget where the release took
> place, the button's -command is invoked.
> 
> These routines are actually slightly different for each of unix, mac,
> and Microsoft Windows in my copy of tk8.3.
> 
> In my opinion, it's great to understand this, and maybe to reimplement
> it in Python as an exercise, but since the vast majority of any GUI 
> application should use the standard behavior for its components as
> defined by the operating system/desktop environment, it makes much
> more sense to me to start teaching Tk with basics like -command, -text,
> -textvariable, the grid manager, and (in Python) things like Toplevel
> subclasses with standard buttons and the like.
> 
> Later, I might have a unit that "teaches" a compound widget like the
> "shufflebox", which does require custom bindings such as doubleclick to
> move selected items to the opposite box.
> 
> Jeff



From eugene1977 at hotmail.com  Tue Sep 17 10:15:59 2002
From: eugene1977 at hotmail.com (eugene kim)
Date: Tue, 17 Sep 2002 14:15:59 +0000
Subject: question about dictionary type..
References: <alt22f$pf8$1@newsreader.mailgate.org> <alv4op$m67$1@newsreader2.netcologne.de> <61Jg9.16334$5r1.665293@bin5.nnrp.aus1.giganews.com> <am0390$cvf$1@newsreader.mailgate.org> <am09qg$78d$1@newsreader2.netcologne.de>
Message-ID: <am6k2q$edf$2@newsreader.mailgate.org>

cool.. thx

i read a python book while ago, and trying to refamiliaze myself with it 
recently, your solution looks so cool..

apreciate it..
have to spend some time what u did =)

thanks


From maxm at mxm.dk  Wed Sep  4 06:12:30 2002
From: maxm at mxm.dk (Max M)
Date: Wed, 04 Sep 2002 12:12:30 +0200
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com> <oukd9.7$4E3.806@news.ecrc.de> <slrnanbm0h.1rs.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <3D75DC8E.1060501@mxm.dk>

Gerhard H?ring wrote:
> Markus Jais wrote:

>>so maybe you should learn both PHP and Python (and Ruby, of course :-)
> 
> I'd skip PHP. But then, I'm biased. And proud of it :-P



Well PHP will not teach you a lot as a programmer, but it can have good 
commercial, practical and pragmatic value.

It is often used in the wild and it is nice to have a general feel for it.

regards Max M



From nanotech at europa.com  Thu Sep 12 16:29:21 2002
From: nanotech at europa.com (czrpb)
Date: Thu, 12 Sep 2002 13:29:21 -0700 (PDT)
Subject: Help with regular expression using findall and .*?
Message-ID: <Pine.BSO.4.21.0209121328450.19471-100000@thetics.europa.com>

Could anyone help out with rewriting (still using regular expressions)
the following so that it does not cause an exception:

import re

s1=('macro\n'+'a'*200+'\norcam\n')*10
s2=('macro\n'+'a'*20000+'\norcam\n')*10

p=re.compile(r'macro.*?orcam',re.DOTALL)

for x in re.findall(p,s1):
    print x

for x in re.findall(p,s2):
    print x

thanks!! Quentin Crain


-------------------------------------------------------------------------
I think we are miscommunicating. I actually expected some confusion with
my last response and I am not sure if there actually was some, but I feel
like it. So, my message is going to assume I did not make my points, but
perhaps you did understand me. If so, then that means *I* did not
understand *YOUR* response - in which case you will need to rephrase.






From jnana4 at DELETEhotmailCAPS.com  Tue Sep 24 00:32:37 2002
From: jnana4 at DELETEhotmailCAPS.com (jano)
Date: Tue, 24 Sep 2002 04:32:37 GMT
Subject: cgi post problem
Message-ID: <FVRj9.192186$Jo.71059@rwcrnsc53>

Hi,

I am trying to post a form to a cgi script.  The form is as follows:

<form action="http://someurl.com/cgi-bin/updatexml.py" method="POST">
    <input type="text" name="/Product/Name" size="20" />
    <input type="text" name="/Product/PartNumber" size="20" />
    <input type="text" name="/Product/Inventory" size="20" />
    <input type="text" name="/Product/Cost" size="20" />
    <input type="text" name="/Product/WholesalePrice" size="20" />
    <input type="submit" value="Update Records" />
</form>

At the other end, the updatexml.py file currently just echoes back the form
data, as follows (with some wrapper code):

#!/usr/local/bin/python

import sys, traceback, cgi

excText = ""


def main():
    global excText
    try:
        form = cgi.FieldStorage()
 keys = form.keys()

        print "Content-type: text/html\n\n"
        print("""<?xml version='1.0' encoding='UTF-8'?>
        <!DOCTYPE html PUBLIC '-//W3C//DTD XHTML 1.1//EN'
'http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd'>
        <html xmlns='http://www.w3.org/1999/xhtml' xml:lang='en'>
            <head>
                <title>CGI exception</title>
                <link type='text/css' rel='stylesheet' href='style.css' />
                <script language=\"javascript\" type=\"text/js\"
href=\"scripts/Netscape4Fix.js\"></script>
            </head>
            <body>

        """)
        print "<p>** ",keys," **</p>"
        print "</body></html>"


    except:
        print "Content-type: text/html\n\n"
        import time
        errtime = '--- '+ time.ctime(time.time()) +' ---\n'
        errlog = open('cgi_errlog', 'a')
        errlog.write(errtime)
        traceback.print_exc(None, errlog)
        errlog.close()
        print("""<?xml version='1.0' encoding='UTF-8'?>
        <!DOCTYPE html PUBLIC '-//W3C//DTD XHTML 1.1//EN'
'http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd'>
        <html xmlns='http://www.w3.org/1999/xhtml' xml:lang='en'>
            <head>
                <title>CGI exception</title>
                <link type='text/css' rel='stylesheet' href='style.css' />
                <script language=\"javascript\" type=\"text/js\"
href=\"scripts/Netscape4.js\"></script>
            </head>
            <body>
            <div id=\"main\">
        """)
        print "<p class='error'>Sorry, a problem was encountered running the
cgi script: </p><br />\n"
 if excText != "":
  print "<p class='citation'>", excText, "</p>"
        print "</div></body></html>"


if __name__ == "__main__":
    main()

==================

The output of the script is always an empty [].  Does anybody know what the
problem might be?  In a previous thread, the problem was that the user had
<input id=""...> instead of using name ="", but that is not the case here.

Thanks for any help.  I have checked usenet and the python faq and used
google, to no avail.

jano




From marklists at mceahern.com  Thu Sep 19 14:00:52 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 19 Sep 2002 13:00:52 -0500
Subject: finding last match in a file
In-Reply-To: <slrnaojlin.8re.dag@omally.animagicnet.no>
Message-ID: <JHEOKEOOLIGLDHCMAHMOOEPKDPAA.marklists@mceahern.com>

> I have a number of very large files where I have to find the last line
> which matches a certain experssion.  Currently I'm simply opening the
> file and looping through it noteing each time a line matches and at the
> end of the loop I see which was the last match.  However this is very
> slow and inefficient.  Is there some way to read through the file
> backwards and simply find first match starting from the end of the file.
> The files are really too big to read the entire thing into memory.

Well this is worth considering:


http://groups.google.com/groups?selm=000101bd9b36%2405590020%2451a22299%40ti
m

// m

-




From aglyportREMOvethispart at nospam.yahoo.com  Thu Sep  5 18:13:49 2002
From: aglyportREMOvethispart at nospam.yahoo.com (Anton Graph)
Date: Thu, 05 Sep 2002 15:13:49 -0700
Subject: Memory fault - core dumped   while compiling python 2.2.1 on
   SCO
References: <3D757EFA.9050401@nospam.yahoo.com> <yu99sn0qqw1r.fsf@europa.research.att.com> <m3bs7etem9.fsf@mira.informatik.hu-berlin.de> <3D765A05.6030004@nospam.yahoo.com> <j48z2h61j9.fsf@informatik.hu-berlin.de> <3D768753.8030302@nospam.yahoo.com> <m3n0qx8j3w.fsf@mira.informatik.hu-berlin.de> <3D76A139.1030304@nospam.yahoo.com> <al69pn$sg$0@216.39.172.122> <3D76BC56.7020503@nospam.yahoo.com> <al6doq$au2$0@216.39.172.122>
Message-ID: <3D77D71D.3060305@nospam.yahoo.com>

Bengt Richter wrote:
> On Wed, 04 Sep 2002 19:07:18 -0700, Anton Graph <aglyportREMOvethispart at nospam.yahoo.com> wrote:
> 
> 
>>Bengt Richter wrote:
>>
>>>On Wed, 04 Sep 2002 17:11:37 -0700, Anton Graph <aglyportREMOvethispart at nospam.yahoo.com> wrote:
>>
> [...]
> 
>>>How about
>>>    cc -v
>>>?
>>>
>>>what does
>>>
>>>    which cc
>>>
>>>give you? And
>>>
>>>    which cc | xargs -ld
>>>?
>>
>>
>>
>>$ which cc
>>/bin/cc
>>$ /bin/cc -v
>>Usage: cc [ options ] files ...
>>$ which cc | xargs -ld
>>xargs: d is not a valid number of lines, must be a positive integer
>>
> 
> Oops, sorry, that should have been
> 
>    which cc | xargs ls -ld
> 
> Regards,
> Bengt Richter

which cc | xargs ls -ld
lrwxrwxrwx   1 root     sys           31 Oct 26  1999 /bin/cc -> 
/opt/K/SCO/unixds/5.1.1A/bin/cc

cc -V
SCO UNIX Development System  Release 5.1.1A 27Jul98
Usage: cc [ options ] files ...



From brian at sweetapp.com  Tue Sep  3 16:43:20 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Tue, 03 Sep 2002 13:43:20 -0700
Subject: PyListObject & C Modules
In-Reply-To: <m3bs7esua1.fsf@mira.informatik.hu-berlin.de>
Message-ID: <006701c2538a$89f12630$df7e4e18@brianspiv1700>

Martin, are you sure that you don't have a ref counting bug in your
example? Since PyList_GetItem borrows a reference and PyList_SetItem
steels one, you need to do an incref in between. Here is my modified
version (not tested or even compiled):

PyObject*
get_first_elem(PyObject *unused, PyObject *args)
{
  PyObject *list;
  PyObject *first;
  PyObject *result;
  if(!PyArg_ParseTuple("O", &list))
    return NULL;

  if(!PyList_Check(list)) {
    PyErr_SetString(PyExc_TypeError, "get_first_elem expects a list");
    return NULL;
  }

  if(PyList_Size(list) == 0){
    PyErr_SetString(PyExc_ValueError, "empty lists not allowed");
    return NULL;
  }
 
  first = PyList_GetItem(list, 0);
+ if (first == NULL) return NULL;
  result = PyList_New(1);
  if (!result) return NULL;
+ Py_ADDREF(first)
- PyList_SetItem(result, 0, first);
+ if (PyList_SetItem(result, 0, first) != 0) {
+   Py_DECREF(first)
+   Py_DECREF(result);
+   result = NULL;
+ }
  return result;
}

Cheers,
Brian




From ianb at colorstudy.com  Tue Sep 17 15:22:55 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 17 Sep 2002 14:22:55 -0500
Subject: CGI question
In-Reply-To: <slrnaoerk7.mgo.jboy@otaku.freeshell.org>
References: <slrnaoerk7.mgo.jboy@otaku.freeshell.org>
Message-ID: <1032290575.8110.159.camel@dsl-65-184-205-5.telocity.com>

On Tue, 2002-09-17 at 13:04, John D. Boy wrote:
> I am working on a simple CGI script to use as a framework for a website
> of mine. Basically, I want it to work as follows: When somebody accesses
> the site, a CGI (called e.g. index.py) is run. It takes an html-template
> and fills it with content from another file, called e.g. index.dat, and
> returns the completed html-document. When I want to present another
> page, I want the url to look like this: http://my.domain.name/?page. In
> that case, index.py would fill the same template with the content of a
> file called e.g. page.dat. 

BTW, I wouldn't recommend doing it this way.  If you have a cgi script
at, say, /page.py, you can access URLs like /page.py/path/to/something
-- "/path/to/something" will be stored in os.environ['PATH_INFO'].  

Then your URLs will look like real URLs, and not like queries.  And for
mod_rewrite people, if you want to serve the root of your domain off
that script, I believe this will work:

RewriteEngine On
RewriteRule ^/(.*)$ /path/to/page.py/$1

(I don't think you can use Alias in this case, but I might be wrong)

  Ian





From ataylor at ucsd.edu  Fri Sep 27 18:18:13 2002
From: ataylor at ucsd.edu (Adam Taylor)
Date: 27 Sep 2002 15:18:13 -0700
Subject: What does Python fix?
Message-ID: <e15b6fa3.0209271418.14cd5642@posting.google.com>

> Syntax is skin deep.  I'm not part of the "Lisp's problem is Lots of Insipid
> Stupid Parentheses" crowd myself, so I never expected that Dylan's choice to
> use algebraic syntax would do it good in the end; to the contrary, it
> alienated old-time Lisp purists, who were Dylan's best hope for
> wider-than-it-got adoption.
>
> Any language even *inspired* by Lisp is doomed to eternal obscurity.  The
> Lisp'ers have been gnashing their teeth over "why?!" for decades.  It's no
> longer my job to explain it, since I abandoned Lisp years ago (I couldn't
> understand the concept of nested parentheses <wink>).

I'm curious to read a bit more about why Tim Peters (and presumably
others) think that Lisp-inspired languages (even non-prefix ones) are
doomed to eternal obscurity.  Would anyone care to comment and/or give
me some pointers to commentary on the subject?  It sounds like Tim
doesn't want to spend too much time getting into it (which is entirely
fair), but maybe others can 'channel' for him, or just point me
towards already-written opinions on the matter?

Thanks,
Adam

-----
[Tim]
>> And Dylan's subsequent meteoric rise to world domination proved
that
>> the "Lisp's problem is Lots of Insipid Stupid Parentheses" crowd
was
>> right all along <wink>.

[Steven Majewski]
> But the syntax that won in Dylan was the algebraic C-like one, NOT
> the S-expr syntax.

Yup.

> How do you explain that ?

As my sig said,

>> the-reason-lisp-excels-at-getting-ignored-isn't-skin-deep-ly y'rs -
tim

Syntax is skin deep.  I'm not part of the "Lisp's problem is Lots of
Insipid
Stupid Parentheses" crowd myself, so I never expected that Dylan's
choice to
use algebraic syntax would do it good in the end; to the contrary, it
alienated old-time Lisp purists, who were Dylan's best hope for
wider-than-it-got adoption.

Any language even *inspired* by Lisp is doomed to eternal obscurity. 
The
Lisp'ers have been gnashing their teeth over "why?!" for decades. 
It's no
longer my job to explain it, since I abandoned Lisp years ago (I
couldn't
understand the concept of nested parentheses <wink>).

leaving-lisp-angst-to-the-lispers-ly y'rs  - tim


From martin at v.loewis.de  Wed Sep  4 03:06:24 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 04 Sep 2002 09:06:24 +0200
Subject: simple xml pickle
References: <mailman.1031104505.31870.python-list@python.org>
Message-ID: <m3fzwqterz.fsf@mira.informatik.hu-berlin.de>

Michael Stenner <mstenner at phy.duke.edu> writes:

> Does anyone know of anything else? 

PyXML offers xml.marshal.generic, and xml.marshal.wddx (which supports
the WDDX DTD). They require xml.sax, but are otherwise "mostly"
standalone (meaning that they could be ported to pure Python 2.x with
little effort - if you do that, please consider contributing your
patches).

HTH,
Martin



From roy at panix.com  Fri Sep  6 21:01:50 2002
From: roy at panix.com (Roy Smith)
Date: Fri, 06 Sep 2002 21:01:50 -0400
Subject: Larry Wall's comment on python...
References: <mailman.1031339711.632.python-list@python.org> <pc5bla.off.ln@boundary.tundraware.com>
Message-ID: <roy-780F0D.21015006092002@reader1.panix.com>

Tim Daneliuk <tundra at tundraware.com> wrote:
> Well, that's true, but for all of its similarity to line noise, perl has
> been an enormous contribution to our profession/community. I would never want
> to diminish Wall's stature, therefore.

There is no doubt that Perl was a major breakthrough.  It sure beats the 
horrible sed/awk/grep/shell disasters we used to write.  The fact that 
it's showing its age and better things have come along (building on 
Perl's success and learning from its mistakes) does not in any way 
diminish the magnitude of Larry's contribution.  Not to mention rn and 
patch (and probably some other stuff I'm forgetting at the moment).

> I find almost all compiled languages (even C) intrusive
> when writing code for embedded systems.

Back in the days when a pdp-11/45 was the hot platform to run Unix on, I 
used to do a certain amount of M-6800 programming (note: two zeros, not 
three).  I found the best way to write code was to write it in a 
(self-imposed) simplistic subset of C, and then hand-compile it into 
assembler.  This let me write the basic control logic in a reasonably 
expressive language, but kept me close to the bare machine.  It was 
almost like flowcharting in pseudocode.

Writing for a machine which only has 256 bytes of RAM gives you a 
somewhat different outlook on programming :-)


From Oschler at earthlink.net  Sun Sep  1 13:17:56 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sun, 1 Sep 2002 13:17:56 -0400
Subject: Add methods to a class at runtime?
References: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com> <slrnan4dl3.18h.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <YYrc9.31010$vY2.601601@e3500-atl2.usenetserver.com>

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote in message
news:slrnan4dl3.18h.gerhard.haering at lilith.my-fqdn.de...
> >>> class Foo: pass
> ...
> >>> foo = Foo()
> >>> foo.bar()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> AttributeError: Foo instance has no attribute 'bar'
> >>> def bar(self): print "bar"
> ...
> >>> Foo.bar = bar
> >>> foo.bar()
> bar
> >>>

Gerhard,

Thanks but I need a little bit more detail.  Suppose I want to do it from
within code?  For example, I create a function called
AddBarToFoo(FooObject), that given a object of type Foo (FooObject) adds the
Bar method to it dynamically, what would be the syntax then?  Would I have
to put the method declaration code, which you show being typed in to the
interpreter, in a string and then eval/exec it?  Is there a more convenient
way than that?

thx






From Oschler at earthlink.net  Tue Sep 10 15:31:07 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Tue, 10 Sep 2002 19:31:07 GMT
Subject: Reset a generator function?
Message-ID: <%Lrf9.30107$w51.7961593@news2.news.adelphia.net>

How can I reset a generator function so that it begins yielding values as it
did when first called?

thx





From syver-en+usenet at online.no  Sun Sep  8 12:45:19 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Sun, 08 Sep 2002 16:45:19 GMT
Subject: Mimicing intepreter import behaviour in embedded application
Message-ID: <uu1l0bfco.fsf@online.no>

I have embedded python in a C++ application. Now I want python code running
in this application to reach modules in the same way as code running
in the standard interpreter. So I set up a test comparing the sys.path
variable under python and my application with the intpreter embedded.

Most of the two sys.path's are equal except:
Regular python:                   Embedded:
 ('d:\\devtools\\python22\\DLLs', 'D:\\devtools\\Python22\\Lib'),
 ('d:\\devtools\\python22\\lib', 'D:\\devtools\\Python22\\DLLs'),
 ('d:\\devtools\\python22\\lib\\lib-tk',
  'D:\\devtools\\Python22\\Lib\\lib-tk'),
This line is caused by the differing locations of the .exe files
 ('d:\\devtools\\python22', 'h:\\mydocuments\\Kode\\C++\\Applications\\pyexe'),
 ('d:\\devtools\\python22\\lib\\site-packages\\Numeric',
  'D:\\devtools\\Python22'),
Here I miss the .pth file that I have for PIL.
 ('d:\\devtools\\python22\\lib\\site-packages\\PIL',
  'D:\\devtools\\Python22\\lib\\site-packages\\Numeric')]

Is there a way to more closely mimic "the real interpreter"?


-- 

Vennlig hilsen 

Syver Enstad


From martin at v.loewis.de  Mon Sep  2 01:59:20 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 02 Sep 2002 07:59:20 +0200
Subject: Problem compiling Secret Labs RexExp module in pyamazon package
References: <xy9b9.33935$Hd4.7555862@e3500-atl1.usenetserver.com>
Message-ID: <m3y9al6ifb.fsf@mira.informatik.hu-berlin.de>

"Robert Oschler" <Oschler at earthlink.net> writes:

> I'm not looking for anyone to give me a source code correction.  What I
> would like is to know how to do one of the following:  contact the makers of
> xml.dom or Secret Labs and report this?  Learn if this is a known problem
> due to some Python configuration issue that I need to follow up on.  Etc.
> In other words what I should generally do to hunt this down and fix it?

Posting to this group is a good starting point. I'm surprised you are
getting such an error. If I do

Python 2.1.3 (#2, Aug 27 2002, 09:38:37)
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> import re
>>> re.compile("&#(/d+);")
<SRE_Pattern object at 0x8188e28>

it works fine for me, on Python 2.1.3. Are you saying that this fails
for you? Could you try Python 2.1.3?

Regards,
Martin


From tdelaney at avaya.com  Fri Sep 20 00:22:24 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 20 Sep 2002 14:22:24 +1000
Subject: Destructor never called ???
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5BD@natasha.auslabs.avaya.com>

> From: Tim Roberts [mailto:timr at probo.com]
> 
> Well, then, what was all the fuss about deterministic 
> desctructors in the
> newsgroup last year or so?  Has this changed, or am I just 
> remembering this entirely wrong?

I'm not getting into this again ... the people (myself included) advocating
*non*-reliance on timely finalisation are perfectly aware of CPython's
*current* behaviour, which does (in *most* cases) have timely finalisation.

Tim Delaney



From mhammond at skippinet.com.au  Tue Sep 17 23:28:40 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 18 Sep 2002 03:28:40 GMT
Subject: Calling Python from C# - please help
References: <6de16ac9.0209130508.5d1b56c6@posting.google.com> <j4elbyt3yw.fsf@informatik.hu-berlin.de> <ur8fyxau0.fsf@online.no> <0iCg9.2350$Ee4.6516@news-server.bigpond.net.au> <3D82FA5B.3050003@mxm.dk> <uptvgvn1h.fsf@online.no> <3%Dh9.15217$Ee4.45118@news-server.bigpond.net.au> <3d877db8_5@news.bluewin.ch>
Message-ID: <IpSh9.17251$Ee4.52376@news-server.bigpond.net.au>

Samuele Pedroni wrote:

> How were you going to support eval and exec?

The compiler itself is written in CPython.  The intention was to get the 
compiler good enough to compile itself - then we will have a .NET 
compiler able to be built into the Python runtime.  exec and eval would 
leverage this.

> For the moment the unload granularity of .NET is AppDomains only and only
> upon explicit request, and types/modules/assemblies are not garbage
> collected but simply live as much as the AppDomain,

App Domains can be dynamically created tho, which may help.

> that's OK for static languages, or environments with a clear run/design time
> distinction (one can setup an AppDomain and tear it down for each
> run-cycle), or for running scripts in isolation but not if one wants IL/JIT
> speed for a enviroment that allows for redefinitions and eval, ...
> 
> In Jython we create Java bytecode and dynamically load classes for all the
> code, also eval code because in Java classes are elegible for garbage
> collection.
> 
> In .NET it seems one needs a pure interpreter that does not compile to IL
> for exec and eval support, and that seems what JScript does for jsc compiled
> code containing eval, that means the script is compiled but the evaluated
> code is only interpreted,

Im not really convinced of that.  Reflection::Emit has IL generation 
capabilities purely for "dynamic" code - the example used in their docs 
is that regex engines could compile down to IL on-the-fly.

> otherwise a long running process using eval could go out of memory.
> 
> That's something people wanting to write a feature complete impl of Pyhton
> for .NET should consider or wait MS to implement a more
> dynamic-environment-friendly unload policy.

Yes, I see your point, but AppDomains are really quite powerful.  I 
believe there are no restrictions between cross domain calls.  Managing 
the lifetime of the dynamically created domains may be an issue though.

Thanks for the comments,

Mark.



From shalehperry at attbi.com  Fri Sep  6 11:30:11 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 6 Sep 2002 08:30:11 -0700
Subject: Getting a proccesses ID
In-Reply-To: <Qu3e9.188987$_91.214523@rwcrnsc51.ops.asp.att.net>
References: <Qu3e9.188987$_91.214523@rwcrnsc51.ops.asp.att.net>
Message-ID: <200209060830.11844.shalehperry@attbi.com>

On Friday 06 September 2002 08:05, Micah Mayo wrote:
> Hi,
>
> I'm writing a little tool for SysAdmins on FreeBSD - I'm trying to make
> it as portable as possible(as far as the unices go), and have come upon
> a problem - one of the things this tool does is restarts various deamons
> on the system, in order to do that I have to send a killsignal to the
> pid. The problem is getting the process id - right now I've got a hack
> using ps, grep, and awk, which I doubt will translate well to other
> flavors of *nix because they all seem to have their own method of
> delivering info via ps. Can anyone tell me a cross platform way of
> grabbing a particular processes ID
>
> Here's what I'm doing now, by the way:
>
> pid = os.popen("ps -x | grep sshd | grep -v grep | awk '{ print $1 }'")
> pid = pid.atoi()

>>> import os
>>> os.getpid()
29421

(-:



From skip at pobox.com  Thu Sep 19 15:30:54 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 19 Sep 2002 14:30:54 -0500
Subject: Larry Wall's comment on python...
In-Reply-To: <psovg51sylq.fsf@bandit.curl.com>
References: <wkr8g7rn4c.fsf@yahoo.com>
        <bd751532.0209082142.54236307@posting.google.com>
        <allrrs$6hm$2@solaris.cc.vt.edu>
        <ambpgq$dre$1@nntp9.atl.mindspring.net>
        <mailman.1032437557.24458.python-list@python.org>
        <psovg51sylq.fsf@bandit.curl.com>
Message-ID: <15754.9710.139758.17475@12-248-11-90.client.attbi.com>

    Christopher> For instance:

    Christopher>    {paragraph Hello {bold Bold} World!}

    Christopher> Although it might be annoying to have to type the curly
    Christopher> braces, I really don't know what the alternative would
    Christopher> be....  I don't think there is a way to do this kind of
    Christopher> thing using Pythonic syntax.

(Danger Will Robinson!  I thought curl was a URL fetcher.  I never heard of
curl the language.  The above fragment is the only curl I've ever seen.
That said...)

Why wouldn't

    paragraph:
        Hello
        bold:
            Bold
        World!

work?  :-)

-- 
Skip Montanaro - skip at pobox.com
The need for gutters to be cleaned is directly proportional to how hard it
happens to be raining at the moment.



From dwelch91 at nospam.attbi.com  Sat Sep 21 10:33:00 2002
From: dwelch91 at nospam.attbi.com (dwelch)
Date: Sat, 21 Sep 2002 14:33:00 GMT
Subject: PyQt setup on Linux Mandrake 8.2
References: <3D8B4898.20808@nospam.attbi.com> <ddc19db7.0209210259.7beb52d4@posting.google.com>
Message-ID: <3D8BC2F4.8080107@nospam.attbi.com>

Armin Steinhoff wrote:
> dwelch <dwelch91 at nospam.attbi.com> wrote in message news:<3D8B4898.20808 at nospam.attbi.com>...
> 
>>I would like to setup PyQt 3.4 on my Mandrake 8.2 box. I am having 
>>difficulty, and I am worried that if I do the wrong thing, KDE will go 
>>belly-up on me since it is Qt based.
>>
>>First, question is: why isn't QTDIR defined by default? 
>>
> 
> PyQt works with Qt2.x.x and Qt 3.x.x ... so YOU have to choose the
> correct directory.
> 
> 
>>Is QTDIR a 
>>runtime setting or a development setting?
>>
> 
> Both ..
> 
> 
>>So I set QTDIR to /usr/lib/qt2 (that seems to be where the libqt2 RPMs 
>>files are). Next question is: Since I have libqt2 already on my system, 
>>how do I add the Qt development pieces without screwing up Qt/KDE? Is 
>>/usr/lib/qt2 even the correct location?
>>
> 
> Yes ... if you want to use only the latest version of Qt 2.x.x
>  
> 
>>When I run SIP 3.4 build.py, it complains about tmake not being 
>>available. Is tmake included with Qt2 development files?
>>
> 
> It's available at the Trolltech homepage or its mirrors.
> 
> 
>>SIP doesn't 
>>seem to have any docs, so I'm not sure how to proceed.
>>
> 
> Nomally you haven't to care about SIP ..
> 
>  
> 
>>Also, my RPM manager says that I have libqt2-devel (2.3.1-29mdk) 
>>installed. No tmake listed in the files, though. And, PyQt-devel 
>>(2.5-1mdk) is also listed as installed, but there aren't any .py files 
>>included, only .sip files (I found this curious).
>>
> 
> Corious ... there must be examples included.
>  
> I would suggest to dowload the latest version of Qt from the Trolltech
> homepage, just for a simple and clean recompilation.
> 
> And do the same with SIP and PyQt from the Riverbank homepage ...
> 
> It seems to be better to forget that outdated stuff provided by the 
> Mandrake distribution ...
> 
> Armin
> 
> 
>>Installing software on Linux makes me feel like such an idiot!
>>
>>Any help or pointers would be much appreciated.
>>
>>Don
>>

OK, I found and installed tmake. Now, SIP can't find some of the Qt 
pieces it needs. So, I still don't understand the situation with Qt. Is 
it really possible to install and run my applications that I write with 
say Qt 3.0.5, when KDE on my system is using Qt 2.3.1? How does/would 
KDE know which version to use? How will my apps know what version to 
use? I am assuming there can only be one QTDIR active on a system at a 
time...

Thanks for the help,

Don




From somebody at hotmail.com  Wed Sep  4 08:23:25 2002
From: somebody at hotmail.com (stephan)
Date: Wed, 04 Sep 2002 14:23:25 +0200
Subject: newsgroup for zope
Message-ID: <al4tru$o3j$03$1@news.t-online.com>

Hi perhaps a stupid question,

but does there exist a newsgrup for zope centric items?

bye

 Stephane


From jb at cascade-sys.com  Thu Sep 26 00:34:22 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Wed, 25 Sep 2002 21:34:22 -0700
Subject: Getting started
References: <uojinmcfapjf45@corp.supernews.com> <AFik9.51323$b%3.25248@fe03> <mailman.1032979596.18965.python-list@python.org> <3D920950.323B25B3@alcyone.com> <Ziqk9.167772$pX1.5989755@news2.tin.it>
Message-ID: <3D928E4E.3090209@cascade-sys.com>

Alex Martelli wrote:

>Erik Max Francis wrote:
>  
>
>>This is simply a case of differing terminology.
>>

I agree the whole is ultimately a difference of terminology but the 
choice is not arbitrary.  I've tried to search out and establish the 
true meaning of the term in an intellectually honest fashion-- I 
certainly was NOT simply making up and promoting some arbitrary 
definition of my own.

>Bingo -- and I think that if one works through all the references I
>gave one will see Cardelli cover all the major issues, though he does
>not do so in any single one of them and his exact terminology may
>differ from mine, yours, James's, "Some authors'" 
>

I'll stipulate to this.  

>(Some authors 
>refer to languages with latent types as weakly typed or dynamically typed 
>languages.) 
>

About 1/4 of the google references I found used this alternative.  Now I 
better understand why.  However, it's better established than the 
oddball third alternative you've been struggling to defend.  

>I don't think it's all that pragmatically useful to have
>a black/white definition for properties, in real life -- it may be
>good maths (or at least, much-simpler maths) 
>

I thought we were talking Computer Science, or at least Computer 
Programming, which is somewhat more along the lines of maths rather than 
real life.  Some properties are intrinsically boolean, others are not. 
 "Type Safety", "Strongly typed" or "Weakly typed" have definitions that 
make them boolean properties.  


>It may be important to know the extent and character of the deviation,
>of course.  For example, [...]
>

Absolutely true.

The "degree of safeness" or the "list of things that make a language 
weakly typed," or the "quality of a language's type or object model" are 
interesting non-boolean properties.

But this is not "weak typing".  

>On the other side of the fence, it's clearly trivial to design
>a language which can have no type problems -- 
>

True.

>just give the language
>just one type, period.  It's even easy for the compiler to ensure
>that you are indeed using that single type (as it's the only one
>you have) wherever it's required (i.e., everywhere).  It's of no
>bloody practical use whatsoever, of course.
>

False.  Neither sufficient nor necessary.

E.g., Unix's "BC" arbitrary precision arithmetic calculator had but a 
single data type and yet it was extremely useful.

>Eg., in another
>post James claimed that standard C has no objects -- 
>[...]
>Clearly, James meant "objects as in OO", 
>


You are correct in what you say above and, particularly that the meaning 
of my alliterative use of "objects" was clear in context.  My point was 
that many references including Cardelli's "Type Systems" specifically 
cite C as an example of a "weakly typed" language.  Since C lacks 
"objects" (in the OOP sense) then "weakly typed" clearly does not have 
anything to do with the relative quality of a language's object model 
(which was your position at the time).  Since it was your assertion 
about OO which I was refuting, my abbreviation C "objects" clearly meant 
C's OO features.

Ha!  And I was afraid you were going to accuse me of saying "the 
original C language was well defined".  ;o)  Which I did.  =8^O.  But it 
was completely tangent to the main argument.

Warmest Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From tundra at tundraware.com  Tue Sep  3 17:30:03 2002
From: tundra at tundraware.com (Tim Daneliuk)
Date: Tue, 03 Sep 2002 21:30:03 GMT
Subject: [ANN]  'waccess' 2.0 web log analyzer released
Message-ID: <ki93la.0t2.ln@boundary.tundraware.com>

Version 2.0 of the 'waccess' web log analyzer is released at:

       http://www.tundraware.com/Software/waccess/

This version has a bunch of new features and options.  This should
be it for quite a while barring bugfixes.  The FreeBSD port has
also been submitted.


Summary:
--------

'waccess' is a quick way to check a web access log for particular
"hits" or accesses.  'waccess' will search the log looking for and
displaying all the access records which contain any of the search keys
you name on the command line.  By default, a summary of its findings -
including total number of records processed, number of accesses per
search key, and number of unique IPs accessing each search key - is
printed.

A rich set of options let you customize 'waccess' to do things the way
you like them including setting case sensitivity, ignoring selected
address spaces in the analysis, and selecting which log file to
process.

'waccess' is not a replacement for a serious web log analysis tool,
but is surprisingly useful for "quickie" lookups and day-to-day
systems managment.

'waccess' is written in Python and needs a reasonably current version
of the langugage on your machine.

'waccess' was written for and tested on FreeBSD 4.6R and is set up to
process Apache access logs.  However, the code is clearly structured
and should port to other systems and web servers pretty trivially.

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com



From max at alcyone.com  Tue Sep 24 00:10:28 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 23 Sep 2002 21:10:28 -0700
Subject: confusion about opening files
References: <mailman.1032836791.15157.python-list@python.org>
Message-ID: <3D8FE5B4.3FF00F5F@alcyone.com>

Mark McEahern wrote:

> I read somewhere that 'file' was preferred to the use of 'open'
> because of
> the symmetry to the type:
> 
>   >>> f = file("junk.txt")
>   >>> type(f)
>   <type 'file'>
> 
> Anyway, that's how I've trained myself to spell 'open'.

It probably is, but only works in more recent versions of Python.  At
any rate, it wasn't what the original poster was asking about, so I
didn't bother addressing it.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ The average dog is a nicer person than the average person.
\__/ Andrew A. Rooney
    Bosskey.net: Counter-Strike / http://www.bosskey.net/cs/
 A personal guide to Counter-Strike.


From revyakin at yahoo.com  Mon Sep 30 16:32:18 2002
From: revyakin at yahoo.com (revyakin)
Date: 30 Sep 2002 13:32:18 -0700
Subject: xml.dom.minidom parser error
Message-ID: <fa06e058.0209301232.58fdc0fd@posting.google.com>

Hi all,

We recently upgraded from OS X to 10.2, and from python 2.1 to python
2.2.
After that my xml parser stopped working: when I attempt to parse an
xml string I get the error shown below.
I don't know whether it matters, but we also installed PIL recently,
and there's a reference to PIL in the error log as well.
Any ideas what this could mean? Writing a parser by myself does not
sound exciting.

Andrey

>>> q=xml.dom.minidom.parseString('<body />')
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/BinaryCache/python/python-3.root~193/usr/lib/python2.2/xml/dom/minidom.py",
line 965, in parseString
  File "/BinaryCache/python/python-3.root~193/usr/lib/python2.2/xml/dom/minidom.py",
line 951, in _doparse
  File "/BinaryCache/python/python-3.root~193/usr/lib/python2.2/xml/dom/pulldom.py",
line 340, in parseString
  File "/usr/lib/python2.2/site-packages/PIL/__init__.py", line 93, in
make_parser
    # ;-)
xml.sax._exceptions.SAXReaderNotAvailable: No parsers found


From aleax at aleax.it  Wed Sep 18 16:56:17 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 18 Sep 2002 20:56:17 GMT
Subject: Copy with __slots__
References: <mailman.1032358481.7625.python-list@python.org> <3K1i9.110352$ub2.2335897@news1.tin.it> <mailman.1032368496.5318.python-list@python.org>
Message-ID: <RL5i9.129147$pX1.4626313@news2.tin.it>

Christian Tismer wrote:
        ...
> You can get this to the single slot-level speed by doing
> the slot grabbing phase of above once, after the class

It's an important speed-up, yes -- but still quite a bit
slower than dealing with a single slot.

> speed can be found by not caching the __slots__ but their
> get/set wrappers and then just calling these with the new
> instance...

I've done that too (as a function, but would be trivial to
wrap as a custom metaclass):

class acceled(withslots):
    def __copy__(self):
        result = self.__class__.__new__(self.__class__)
        for get, set in self._accel:
            set(result, get(self))

def makeaccel(aclass):
    slot_names_seen = {}
    slot_gets_sets = []
    for base in aclass.__mro__:
        for s in getattr(base, '__slots__', []):
            if s in slot_names_seen: continue
            slot_names_seen[s] = True
            slot = getattr(base, s)
            slot_gets_sets.append((slot.__get__,slot.__set__))
    aclass._accel = slot_gets_sets

makeaccel(acceled)

[alex at lancelot ~]$ python -O slorno.py
1.10 noslots
0.67 withslots
0.37 acceled
[alex at lancelot ~]$ python -O slorno.py
1.14 noslots
0.69 withslots
0.37 acceled
[alex at lancelot ~]$ python -O slorno.py
1.10 noslots
0.68 withslots
0.36 acceled

Not the 0.28 or so that I got by "unwrapping the loop" (i.e., just
coding out the assignment for each slot!), but still, an important
speedup, as I said.  Thanks on behalf of the OP!

It's not clear from the above what proportion of the gain is due
to just doing some of the work in advance, what to extracting
the getter and setter, what simply to avoidance of double work
that class withslots did because it copied attribute s1 twice.  As
the OP clearly needs practice in measuring Python runtimes (he
guessed that the 0.28 blazing-fast verson would be SLOWER
than the original 1.10 one, and didn't even THINK of doing a
measurement to double check his intuition...!!!), I guess I'll leave
this homework to him as a useful exercise:-).


Alex



From clspence at one.net  Wed Sep 11 16:53:09 2002
From: clspence at one.net (Chris Spencer)
Date: 11 Sep 2002 15:53:09 -0500
Subject: How do I capture just a single keystroke?
Message-ID: <pqavnu0vqp374e4155esnctk70on886ff7@4ax.com>

I'd like to implement a "Press any key to continue..." type function, but it
seems like all of Python's input mechanisms REQUIRE the enter key to be hit.

If this is documented clearly somewhere, then you have my apologies for wasting
bandwidth.

Chris.



From gua81 at XXXyahoo.com  Thu Sep 26 04:24:35 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Thu, 26 Sep 2002 18:24:35 +1000
Subject: Getting last char in string.
References: <amue3q$h59$1@lust.ihug.co.nz> <amueqq$94qgr$1@ID-71831.news.dfncis.de>
Message-ID: <amug77$j0a$1@lust.ihug.co.nz>

Worked like magic
Thanks dude!


"James Kew" <james.kew at btinternet.com> wrote in message
news:amueqq$94qgr$1 at ID-71831.news.dfncis.de...
> "CheapSkate" <gua81 at XXXyahoo.com> wrote in message
> news:amue3q$h59$1 at lust.ihug.co.nz...
> > Hi, im writing a simple method that takes in a string input and check if
> the
> > last charcter is an '/'
> > then it will do A else do B
> >
> > my Q now is how to get the last char???
>
> if stuff[-1:] == '/':
>
> or
>
> if stuff.endswith('/'):
>
> --
> James Kew
> james.kew at btinternet.com
>
>
>




From johnroth at ameritech.net  Thu Sep 26 10:15:38 2002
From: johnroth at ameritech.net (John Roth)
Date: Thu, 26 Sep 2002 10:15:38 -0400
Subject: object changing itself to another object
References: <amu0lv$6ac$1@otis.netspace.net.au>
Message-ID: <up65bpq7uaph5a@news.supernews.com>

"Glen Murphy" <abuse at nospam.riot.com.au> wrote in message
news:amu0lv$6ac$1 at otis.netspace.net.au...
> I'm creating a program that takes user input, and passes it to an
object -
> however, I want that object itself, and not the parent to change where
that
> input goes. In the highly simplified example below, I would like the
output
> to be "Obj1","Obj2","Obj1" - but not suprisingly, I just get Obj1,
Obj1,
> Obj1
>
>
> ### code begin  ###
>
> class Obj1:
>     def key1(self):
>         # do obj1 specific stuff
>         self = Obj2()
>         print "Obj1"
>
> class Obj2:
>     def key1(self):
>         # do obj1 specific stuff
>         self = Obj1()
>         print "Obj2"
>
> a = Obj1()
>
> # simulate user keypresses
> a.key1()
> a.key1()
> a.key1()
>
> ### code end ###
>
> I know I could achieve the result I want by doing horribly complicated
trees
> of ifs and such, but I'd like my code to be nice and expandable (in
this
> example, to easily add new Objs). I've looked through the Python FAQ,
Google
> Groups and Various O'Reilly books, but I don't really know the
terminology
> for what I'm looking for, so I haven't found anything so far, so does
anyone
> have any pointers?
>
> Any help would be greatly appreciated,
> Cheers,
> Glen
>
> --
> Glen Murphy, http://glenmurphy.com/

It's a standard pattern, check out the Patterns book. Unfortunately, my
copy is in a storage locker on the other end of the continent, so I
can't
give you the name off the top of my head.

What it requires (I believe) is a level of indirection. The top level
is essentially a Facade that calls the real object. You can then change
the real object to anything you want as long as it understands the
method calls the Facade object makes.

John Roth

John Roth
>
>




From thoa0025 at mail.usyd.edu.au  Thu Sep 12 05:58:29 2002
From: thoa0025 at mail.usyd.edu.au (:B nerdy)
Date: Thu, 12 Sep 2002 09:58:29 GMT
Subject: [xml.dom.minidom] why cant it parse '&'
Message-ID: <9zZf9.24873$_l1.66951@news-server.bigpond.net.au>

im using the minidom.parseString() to parse the following text

<sxml>
<title>Congo Mombo - Guitar Gable & The Musical Kings</title>
<composer>Various</composer>
<genre>blues</genre>
<trackNo>12</trackNo>
<copyright>Westwood</copyright>
<dateInserted>2002-03-11</dateInserted>
</sxml>

when it comes to the '&' it throws an exception. is there any way around
this problem??
cheers asap




From richie at entrian.com  Tue Sep 17 05:48:15 2002
From: richie at entrian.com (Richie Hindle)
Date: 17 Sep 2002 02:48:15 -0700
Subject: Windows AutoRunning Python Hello -- feedback requested
References: <Hmuh9.70$Ha6.5066815@newssvr14.news.prodigy.com>
Message-ID: <8a6ba1da.0209170148.23bca1e8@posting.google.com>

Dennis,

> the EXE displays "hello" in a browser window and shows
> its running directory as evidence that os.py is functioning.

I like it.  I think using a web browser as a user interface is an
excellent way to go for some applications.  Every week someone here
asks "What's the best GUI toolkit for Python?" and no-one ever says
"Netscape" or "Internet Explorer", but given a system like yours plus
a framework for building an HTML user interface (based on one of the
many web application frameworks for Python) I think this system is
just as valid as Tk, wxWindows, Qt or whichever other 'real' toolkit
you care to name.

Does your system have mechanisms for serving up forms and accepting
data back from the user, or is it simply a toolkit for distributing
and running browser-based code?  My gut feeling is that the latter
would be better - sufficiently generic that any Python web application
framework could be built into it.  Zope-based Windows applications,
anyone?

-- 
Richie Hindle
richie at entrian.com


From mcherm at destiny.com  Tue Sep 17 16:09:33 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Tue, 17 Sep 2002 16:09:33 -0400
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com>	<mailman.1032107006.6411.python-list@python.org>	<am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net>	<j4znuio0kr.fsf@informatik.hu-berlin.de>	<mailman.1032204706.14737.python-list@python.org>	<m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de>	<1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de>
Message-ID: <am8259$ce0$1@main.gmane.org>

Ian Bicking <ianb at colorstudy.com> writes:
>It has been simplified, but I don't think that changes anything with
>regard to getting it into standard Python -- the same barriers still
>exist.

Martin v. Loewis wrote:
> You are making these barriers up in your mind. I don't know why you
> are doing that, though.

Okay, let's see if we can settle this... because I think it has real 
potential.

I am going to pose a hypothetical situation... please don't argue about 
whether these conditions are true now, or whether they will be true, but 
take these as assumptions:

  (1) Some version of Stackless consists of a quite small non-complicated
      patch onto C-Python.

  (2) The default behavior of this Stackless is to NOT allow stack
      switching... it must be enabled explicitly to be used.

  (3) Someone (probably NOT Christian) is willing to take on the job of
      merging it into the core and maintaining it afterward.

  (4) Stackless works on some architechures (including at least X86,
      Sparc, PPC and Cygwin). But it does NOT work on some other
      architectures. So (rather like the thread package) it is not
      as portable as Python itself.

  (5) Good reasons exist why Stackless CANNOT be implemented in Jython,
      in Python for .NET, or other Python implementations -- it's a
      C-Python-only thing.

The question is this:  Would the advantages of making Stackless part of 
standard Python outweigh the disadvantages of non-portability and 
C-Python specificity?

If the answer is "NO", (and I can see good reasons why it would be) then 
no one need bother worrying about integrating Stackless into the 
standard. Let it live a separate existance!

If the answer is "YES" (and I can see other good reasons why THAT might 
be) then integration MIGHT be possible, but there's no reason why it has 
to be, or needs to be hurried. It can wait until someone steps steps up 
to make (3) take place.

And I agree with Christian's sentiment:

> Please let me stay what I am: The one who writes Stackless
> Python, supports and maintains it. I really do not take
> responsibility for being standard or non-standard. 

... leave Christian to do what he's doing well, and let someone else 
worry about possible integration.

-- Michael Chermside






From mcfletch at rogers.com  Sat Sep  7 06:38:43 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Sat, 07 Sep 2002 06:38:43 -0400
Subject: Python from C/SDL
References: <3D793B29.3198AB0E@ihug.co.nz> <006101c25626$56c9b1c0$0101010a@local>
Message-ID: <3D79D733.1070405@rogers.com>

I'll just deal with the OpenGL stuff.  Basically, PyOpenGL-based apps 
should be able to work like this:

c draw():
	context = new MyContext;
	context.set_current();
	err = Py*_Call*( size );

	python:
		def render( dimensions ):
			do_your_rendering_here()
			return

c:
	swap_buffers();
	context.unset_current();

So, you'd likely need to find, somewhere in pymol, a function to call 
which is _after_ the Python context setup and set_current() which you 
can call.  Depending on the architecture of pymol, that's likely to 
require at least sub-classing, and possibly some re-writing, as most 
rendering libs go out of their way to be sure they're using the "right" 
context before any attempt to render, and most apps assume they'll be 
using a context they set up.  You'll likely need to pass size of the 
window from your code, as some renderers assume they know the current 
size of the rendering context (which they won't if they aren't getting 
their resize events).

For example, in OpenGLContext, you'd subclass a Context and override 
setCurrent and unsetCurrent to do nothing (or more professionally, you'd 
create C "extending" functions that set the appropriate context for your 
app, which would allow the Python system to have its own contexts 
alongside yours if you wished).  In other frameworks those calls aren't 
actually split out into easily overridable methods (it's only required 
if you're targetting multiple GUI libs), so you might have to re-write 
anywhere you see "set current" methods called.    Would suggest 
contacting the creator of PyMol for suggestions where to apply your lever.

For a "full" job of creating a PyOpenGL embedded system, you'd also 
provide initialisation/creation functions to allow Python code to create 
a context within your app (e.g. setting display modes).

There are similar questions asked once in a while, but not particularly 
with pymol AFAIK.  "How do I embed PyOpenGL apps" is the basic template. 
  Suppose I should create a tutorial for it, but I've never done it 
directly myself (though conceptually every context in OpenGLContext is 
"embedded" in the GUI library it uses, since those are all C libs, it's 
just that I do the "calling into Python" from the libs own Python wrappers).

HTH,
Mike
BTW: out of town in a few minutes, so won't be available to clarify 
anything 'til Monday, sorry.

Chris Gonnerman wrote:
> ----- Original Message ----- 
> From: "Gib Bogle" <bogle at ihug.co.nz>
> 
> 
>>Hi,
>>
>>I am a complete beginner with Python.  My interest in it is focussed on
>>a great piece of open-source software called pymol, a system for
>>visualization of molecules that uses Python and OpenGL.  I'd like to get
>>access to pymol's graphical functions from my own program, which is
>>built from C, SDL and OpenGL.
> 
...
>>I will not want to use the GUI at all.  SDL would
>>create the surface(s), and handle the buffer swapping.  My goal is to be
>>able to send a request to the pymol API asking for a frustum view to be
>>rendered on the provided surface.
>>
>>Sorry if my question is one that gets asked and answered frequently.  If
>>there is a place where this has been dealt with please direct me.
> 
> 
> Never heard this one before...
> 
> Good luck!
> 
> 
> Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
> http://newcenturycomputers.net
> 
> 

-- 
_______________________________________
   Mike C. Fletcher
   Designer, VR Plumber, Coder
   http://members.rogers.com/mcfletch/





From Oschler at earthlink.net  Wed Sep 11 18:19:20 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Wed, 11 Sep 2002 22:19:20 GMT
Subject: SuSE 7.3 Yast2-able Python 2.2.1?
Message-ID: <IjPf9.1729$Lg2.215939@news2.news.adelphia.net>

Anybody seen a "yast2" friendly version of Python for SuSE Linux 7.3 Pro?
If so please give me the URL or ftp site.  I tried using the KRUD rpm's but
the OpenSLL missing dependencies issue (libcrypto.so.2 and libssl.so.2 not
found) is kicking my ass.  I ended up forcing the issue with
'rpm -ivh --nodeps' using the i386/Linux Python2 rpm I found, but the
missing dependencies are causing my favorite Python IDE (WingIDE) to croak.
I have the feeling that the lack of yast2's help, especially when it comes
to SuSEConfig, is somehow causing other apps not to be able to find the
version of OpenSLL that definitely _is_ installed on my system (0.9.6)

Any help would be appreciated.

thx





From vmilitaru at sympatico.ca  Thu Sep 12 23:11:22 2002
From: vmilitaru at sympatico.ca (vio)
Date: Thu, 12 Sep 2002 22:11:22 -0500
Subject: [mod_python] ANNOUNCE: Mod_Python donated to ASF
In-Reply-To: <Pine.BSF.4.32.0209121812020.92256-100000@eden.ispol.com>; from grisha@modpython.org on Thu, Sep 12, 2002 at 06:25:45PM -0400
References: <Pine.BSF.4.32.0209121812020.92256-100000@eden.ispol.com>
Message-ID: <20020912221122.A1410@eye.neb>

Congrats to Gregory for this 'official' recognition for the great work he did
with mod_python. And thank you to the apache board for recognising the need
for an 'official' python effort in such an open way. Let's face it, there is no
better place for an apache module to live but close to apache (httpd)'s place.
Certainly the most intuitive place from a user's perspective. And this official
blessing brings the additional advantage to cut through the confusion: there
are not 10 'official' python' flavored apaches for a newcomer to choose from.
(enough headaches just installing the beast and learning it)

Regards,
Vio

* Gregory (Grisha) Trubetskoy <grisha at modpython.org> [020912 20:57]:
> 
> It is my pleasure to announce that Mod_Python has been donated to the
> Apache Software Foundation, and is now a subproject of the httpd server
> project (see http://httpd.apache.org/).
> 
> I am grateful to ASF for accepting this donation and committing resources
> to further the support of Mod_Python. I believe that this action will
> advance the development of Mod_Python, resulting in an ultimately better
> and more popular tool for Python developers. I also believe it will serve
> to better position Python as a language of choice for web development, a
> need that has been expressed by many.
> 
> There are no implications to the current Mod_Python users - the license is
> the same with the sole difference in that the copyright belongs to ASF
> now.
> 
> As a consequense of the donation, the CVS repository is now hosted on
> cvs.apache.org.  Do not use the SourceForge repository anymore, it will
> soon be removed.
> 
> There will also be website and mailing changes, but the details are still
> being finalized and will be announced when ready.
> 
> Regards,
> 
> Grisha Trubetskoy
> 
> 
> _______________________________________________
> Mod_python mailing list
> Mod_python at modpython.org
> http://www.modpython.org/mailman/listinfo/mod_python



From martin.klaffenboeck at gmx.at  Fri Sep  6 09:30:48 2002
From: martin.klaffenboeck at gmx.at (Martin Klaffenboeck)
Date: Fri, 06 Sep 2002 15:30:48 +0200
Subject: pygtk
Message-ID: <pan.2002.09.06.13.30.48.241597.8751@ns.kleinerdrache.org>

Hello,
  I'm new here in Python, because a few people told me that the
implementation of gtk is better in python than in Perl.  So now I try do
do object orientated programming here...

Is this a good group to discuss pygtk things or is this just for python?
(I think I have questions in both.)

So now this message is also to say 'Hello' to everybody in this room.

Maybe you'll here much from me. ;-)

Martin


From gerhard.haering at opus-gmbh.net  Wed Sep  4 13:40:45 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 4 Sep 2002 17:40:45 GMT
Subject: Bind threads to addresses -- Windows & urllib?
References: <mailman.1031157680.4128.python-list@python.org>
Message-ID: <slrnancha3.24s.gerhard.haering@haering.opus-gmbh.net>

Nick Arnett wrote:
> I can't see an obvious easy way to bind threads to IP addresses on a
> Win2K machine.  I have a multi-threaded Web spider, which uses urllib to
> retrieve pages, and would like to bind each thread to a different address
> on the machine, if possible.

I'm not sure what you mean. This is a multithreaded HTTP client app, right?
Do you have multiple network connections and want to use more than one?

If yes, then that's an issue of configuring the routing on your computer
appropriately (if that is at all possible, I doubt it's easy).

I'd also be surprised if you could saturate a 100 MBit link with a HTTP
spider.

-- Gerhard


From anton at vredegoor.doge.nl  Fri Sep 27 17:12:05 2002
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Fri, 27 Sep 2002 23:12:05 +0200
Subject: watching mutables?
References: <mailman.1033137396.14482.python-list@python.org> <an211j$oi$1@news.hccnet.nl> <Xns9296C58C9ADFCcliechtigmxnet@62.2.16.82> <k7l7qpvf.fsf@ion-tof.com> <Xns9296D3AB53777cliechtigmxnet@62.2.16.82> <mailman.1033154856.20758.python-list@python.org> <an2doh$ajo3p$1@ID-59885.news.dfncis.de>
Message-ID: <an2hnt$nh2$1@news.hccnet.nl>

On Fri, 27 Sep 2002 22:06:41 +0200, "Thomas Heller"
<theller at python.net> wrote:

>Here's what seems to be available:
>http://oomadness.tuxfamily.org/downloads/
>and this is obviously where the EventObject is contained in:
>http://oomadness.tuxfamily.org/downloads/EditObj-0.3.tar.gz

Very interesting module. I might add that windows doesn't seem to like
the .cfg file so this should be moved to some other place before
installing with 'python setup.py install'.

But...

I specifically asked for something that doesn't touch the original
object. In eventobj.py's docstring there is the following warning:

<warning>
Caution : As event management is performed by changing the class of
the instance, you use eventobj with critical objects at your own
risk... !
</warning>

So, with a lot of thanks for the results sofar ...

Is it possible?

Anton.


From tim.one at comcast.net  Thu Sep 26 19:08:10 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 26 Sep 2002 19:08:10 -0400
Subject: Who knows somefunction?
In-Reply-To: <3D9382AF.5070800@aon.at>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEONBGAB.tim.one@comcast.net>

[Gregor Lingl]
> Is there a function <somefunction> in Python for which
>
> something is <somefunction>(id(something))
>
> returns invariably true?

[Tim]
> I'm not sure you're asking the question you intended to ask, but str()
> is one answer to the question you actually asked.

[Gregor]
> I don't think so:
>  >>> something = [1,2,3]
>  >>> something is str(id(something))
> 0
>
> How did you understand my question?
> (Errors should never pass silently)

Seemingly in a way other than you asked it <wink>.

Marc-Andre Lemburg's mxTools extension adds such a function to the sys
module (sys.makeref(id)).  It's impossible for him to do this safely, so
believe his warnings when you read them.

conjuring-things-into-existence-by-naming-them-is-a-trick-
    best-left-to-gods-ly y'rs  - tim




From max at alcyone.com  Tue Sep 24 20:51:50 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 24 Sep 2002 17:51:50 -0700
Subject: Strange python and cgi!
References: <amquos$rcu$1@lust.ihug.co.nz> <3D9101AC.842F0513@alcyone.com> <amr0vu$sq5$1@lust.ihug.co.nz>
Message-ID: <3D9108A6.F59EC367@alcyone.com>

CheapSkate wrote:

> what kind of error am I expecting here?
> .... errrrrrr I don't know how to access the web server error log.

It is likely due to some environmental difference between how you run
the script from your shell and how your Web server runs the script. 
Other than that, there is no way to know without actually checking the
logs.  We can't check them for you; you're going to have to do that
yourself.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ As far as I'm concerned, being any gender is a drag.
\__/ Patti Smith
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.


From loewis at informatik.hu-berlin.de  Mon Sep 16 05:59:57 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 16 Sep 2002 11:59:57 +0200
Subject: python performance
References: <3D84E66F.6020701@linux.ie> <zm7h9.51182$AR1.2015178@bin2.nnrp.aus1.giganews.com> <IThh9.4238$cP3.10401@news.iol.ie>
Message-ID: <j4r8fumezm.fsf@informatik.hu-berlin.de>

Padraig Brady <Padraig at Linux.ie> writes:

> > Calling functions adds overhead.  Running code within functions
> > reduces overhead for object access.
> 
> Ah OK, this because it has to search a smaller namespace?

No. Inside a function, local variables are accessed with integer
indices, instead of by name. This is possible because there is static
knowledge about the local variables in a function.

> Wouldn't compiling overcome this? 

Not if you want to preserve Python semantics; the compiler would need
to emit code that performs the same actions that the interpreter
currently does.

Regards,
Martin


From yaipa at aol.com  Fri Sep 13 14:05:16 2002
From: yaipa at aol.com (yaipa)
Date: 13 Sep 2002 11:05:16 -0700
Subject: Is Python 2.1 for Solairs 8 broke on SunFreeware?
Message-ID: <8d148763.0209131005.65de04a8@posting.google.com>

Problem:
I've downloaded Python 2.1 for Sparc Solairs 8 from Sunfreeware and run the
pkgadd to install it.  All seems fine until I try running a script that 
works just fine with ActiveState's build of Python 2.1 for Solaris.  When
I run Python 2.1 from Sunfreeware, Python throws an import error on the 
'time' module.

% /usr/local/bin/python CronTar.py -c cronTar.cfg
  Traceback (most recent call last):
    File "/export/fruid/bin/CronTar.py", line 26, in ?
      import time
  ImportError: ld.so.1: /usr/local/bin/python: fatal: 
    libgcc_s.so.1: open failed: No such file or directory
%
Am I required to have gcc installed also?

Our deployment model strictly uses the Solaris pkgadd command, 
so I require either Python 2.1 or 2.2 in a standard Solaris 
package format. 

Any ideas?


From syver-en+usenet at online.no  Mon Sep 16 05:11:16 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Mon, 16 Sep 2002 09:11:16 GMT
Subject: Book for win32?
References: <am430k$9l3$03$1@news.t-online.com>
Message-ID: <uhegq710h.fsf@online.no>

Thomas Guettler <zopestoller at thomas-guettler.de> writes:

> Hi!
> 
> I am interested in book about Python on win32.
> 
> I found one from o'reilly: "Python on win32"
> 
> Are there better books or alternatives?
> 
> I am *not* looking for a python introduction, I am
> programming more than one year with python, but up to
> now mostly on unix.
> 
> Now I am intersted in win32 stuff:
>   - Opening a file-dialog
>   - COM
>   - Accessing the registry
>   - Maybe GUI Programming

Then you can't go wrong with the o'reilly book. It is a most excellent
tome for precisely the subjects that you outline there. 

Be aware that the book is written for Python 1.5 so a couple of
things, like the need to convert unicode strings passed in by COM
objects are outdated. As you already have been programming python for
over year this shouldn't be a problem.
 

-- 

Vennlig hilsen 

Syver Enstad


From DennisR at dair.com  Wed Sep 18 11:10:17 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Wed, 18 Sep 2002 15:10:17 GMT
Subject: Python comment stripper
References: <16Nh9.248$Wc6.17087796@newssvr13.news.prodigy.com> <3d8807b1$1@news.sentex.net> <Eg0i9.561$Y33.34967690@newssvr13.news.prodigy.com>
Message-ID: <tH0i9.566$Ze3.35623104@newssvr13.news.prodigy.com>

> > Why not just use .pyc files
> > all the time and be done with it?

Unstated assumption: .pyc files are often larger than corresponding .py
files.  With comments stripped, the .py file would definitely be smaller.

Otoh, .pyc files are easier to work with in some ways and where the size
contribution of libraries is not an issue, .pyc files are great in this
application.
--

Dennis Reinhardt

http://www.dair.com




From gmuller at worldonline.nl  Sun Sep 22 12:15:57 2002
From: gmuller at worldonline.nl (GerritM)
Date: Sun, 22 Sep 2002 18:15:57 +0200
Subject: Smalltalk vs. Python and COM was Re: Byte Magazine and Smalltalk
References: <MXFh9.23885$jF4.2037258@twister.southeast.rr.com> <3d87a686@tobjects.> <rd8fougqbgdivimn7vuest3ugoj57lvrr6@4ax.com> <Xns928CA588EF17Dtrippyparcplacecom@64.164.98.7> <3d87dd7c$1@tobjects.> <g9Vh9.313655$kp.1041470@rwcrnsc52.ops.asp.att.net> <Xns928D58DF09B6trippyparcplacecom@64.164.98.6> <20020918.115128.1605894428.1527@cybermesa.com> <amkpk0$or2$1@fe2.cs.interbusiness.it>
Message-ID: <amkqfs$ldg$1@reader1.tiscali.nl>

"Pino Gargiulo" <jk3380 at naida.org> schreef in bericht
news:amkpk0$or2$1 at fe2.cs.interbusiness.it...
<...snip...>
> For this project I was considering Smalltalk and Python. The first so I
> could be
> some real stuff done and Python because apparently it permits quite easy
> access
> the the mysterious (to me) m$ APIs. I really would want to stay away from
> VC++,
> and I'm looking for some quite turnaround environment since I'll be
learning
> COM on the job.
>
> Anybody care to give some advice? I played around with DST but never
looked
> much at the existance of a COM wrapper/interface.
>
Long long ago I worked in a large scale Objective-C project. Objective-C
inherited many of its good characteristics from Smalltalk. Now I am using
Python quite a lot. It has again a lot of the good (dynamic)
characteristics, with on top an excellent set of predefined modules. One of
them are the win32 modules by Mark Hammond, documented on line of course,
but also described extensively in "Python programming on Win32" by Mark
Hammond and Andy Robinson, ISBN 1-56592-621-8. I am using the ActiveState
distribution, which is a complete distribution including this win32 stuff.

In short highly recommended!

regards Gerrit

--
www.extra.research.philips.com/natlab/sysarch/





From maurice at akst.tn.tudelft.nl  Mon Sep  9 13:17:16 2002
From: maurice at akst.tn.tudelft.nl (Maurice van de Rijzen)
Date: Mon, 09 Sep 2002 19:17:16 +0200
Subject: Binary input into NumPy.
Message-ID: <alil2o$mmk$1@news.tudelft.nl>

Dear all,
I just downloaded Numerical Python and I think it will be very usefull.
However I have one major broblem:
I don't know how to import binary data from a file on disk into an array!
Can onyone help me with this problem?

Thanks



From 2002 at weholt.org  Mon Sep 23 05:05:55 2002
From: 2002 at weholt.org (Thomas Weholt)
Date: Mon, 23 Sep 2002 09:05:55 GMT
Subject: Suggesting a filename on a HTTP-request by sending someting in the header of the reply
References: <Pusi9.20169$0p1.363761@news2.ulv.nextra.no> <r5zi9.247091$2L.9053878@e3500-atl2.usenetserver.com>
Message-ID: <TPAj9.21311$0p1.377644@news2.ulv.nextra.no>

> Content-Disposition: file; filename=%(filename)s
>
> is the proper way to use the http header to show the "filename" of what is
> being downloaded.
>
>
> J Lowery

Thanks!! It works just fine now.

Best regards,
Thomas Weholt




From aleax at aleax.it  Sun Sep 15 03:49:09 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 15 Sep 2002 07:49:09 GMT
Subject: html template system?
References: <btug9.688$Ee4.1090@news-server.bigpond.net.au> <0ro7ouskb8ctv6f846gb8cfpqtl5vs1avc@4ax.com> <mailman.1032059537.4651.python-list@python.org>
Message-ID: <VXWg9.96529$ub2.1935606@news1.tin.it>

Aldo Cortesi wrote:

> ":B nerdy" <thoa0025 at mail.usyd.edu.au> wrote:
> 
>>what are good templating systems avaliable for python?
        ...
> Cheetah - http://www.cheetahtemplate.org/
> YAPTU - http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52305

Speaking as the author of YAPTU -- YAPTU is a neat little toy, but Cheetah
is a truly powerful system, particularly if your needs coincide with those
it was designed for (e.g., make it feasible for non-programmers to edit the
templates).  I'd suggest looking into Cheetah (or other powerful systems --
Cheetah looks very good to me, but I don't come close to knowing them
all!-), maybe unless you happen to be in a peculiar niche (e.g., if you
must bundle your templating system with a small app, and you don't need too
much in terms of raw power, YAPTU's tiny size may be a plus).


Alex



From ods at strana.ru  Mon Sep  2 09:04:03 2002
From: ods at strana.ru (Denis S. Otkidach)
Date: Mon, 2 Sep 2002 17:04:03 +0400 (MSD)
Subject: Must have Python modules?
In-Reply-To: <FCIc9.30407$FJ1.5585669@e3500-atl1.usenetserver.com>
Message-ID: <Pine.LNX.4.44L.0209021702300.17049-100000@ods.pp.ru>

On Mon, 2 Sep 2002, Robert Oschler wrote:

RO> development!)  I want to know what the other favorites are

mx, PIL, DocumentTemplate, TAL, MySQLdb

-- 
Denis S. Otkidach
http://www.python.ru/      [ru]
http://diveinto.python.ru/ [ru]




From amitp at Xenon.Stanford.EDU  Thu Sep  5 12:19:35 2002
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 5 Sep 2002 16:19:35 GMT
Subject: colloquial names for types?
References: <yu99bs7q50yp.fsf@europa.research.att.com> <mailman.1030295978.22776.python-list@python.org> <cf701bae.0208260905.69d0aca0@posting.google.com>
Message-ID: <al806n$j2n$1@usenet.Stanford.EDU>

 Inyeol Lee <inyeol_lee at yahoo.com> wrote:
| holger krekel <pyth at devel.trillke.net> wrote in message news:<mailman.1030295978.22776.python-list at python.org>...
| 
| > 
| > I'd try using
| > 
| >     something.__class__.__name__
| 
| Python 2.2.1 (#1, Apr 10 2002, 18:25:16) 
| [GCC 2.95.3 20010315 (release)] on sunos5
| 
| >>> "".__class__.__name__
| 'str'
| >>> [].__class__.__name__
| 'list'
| >>> {}.__class__.__name__
| 'dict'
| >>> 3.__class__.__name__
| SyntaxError: invalid syntax
| >>> dir(3)
| [ ... '__class__', ... ]
| 
| I'm confused. Why 3.__class__ raises exception?

It's a syntax error because  3.  is a valid (float) literal.  So it
gets treated like (3.)__class__, and there's no dot there.

You can either write (3).__class__ or 3..__class__, but the latter is
a float.

  - Amit

-- 
--
Amit J Patel, Computer Science Department, Stanford University
http://www-cs-students.stanford.edu/~amitp/


From dsavitsk at e-coli.net  Mon Sep 16 14:17:42 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Mon, 16 Sep 2002 18:17:42 GMT
Subject: COM object registration
Message-ID: <afph9.9468$yt3.4669302@newssrv26.news.prodigy.com>

if the module below is run from the command line, it registers and works
fine.  if, however, it is imported and the register() method is executed it
does not work (giving a 'localserver' error). (the registration process
works, but it can't then be used via Dispatch) seemingly the name of the
class (here testServer) can't be found.

even though it doesn't work locally, however, it will still work as a DCOM
server, that is, if it is registered from the command line on another
computer it works.

is this expected behavior? is it a mangled name of some sort? is there a
better way to register a COM obect via code.

# -------   sample code   -------

import pythoncom

class testServer:

    _reg_clsid_ = '{C627B7F9-E078-4872-B06D-7DB33C3BC7A4}'
    _reg_clsctx_ = pythoncom.CLSCTX_LOCAL_SERVER
    _reg_progid_ = 'test.Server'
    _public_methods_ = ['GET_VALUE']

    def GET_VALUE(self): return 0

def register():
    import win32com.server.register
    win32com.server.register.UseCommandLine(testServer)

if __name__ == '__main__': __register()

# ------- end sample code -------




From joconnor at cybermesa.com  Tue Sep 17 04:17:09 2002
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 17 Sep 2002 15:17:09 +0700
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com>
Message-ID: <20020917.151708.502278611.1586@cybermesa.com>

In article <96904b50.0209171305.6d65f574 at posting.google.com>, "Lee Gray"
<lpgray at uop.com> wrote:

> 
> Can someone point me to a good Python GUI app to demonstrate (other than
> IDLE)?

Bioreason makes some products in Python.  You can see some screenshots of
their Python/Gtk products here: http://www.bioreason.com/lead.html and
http://www.bioreason.com/drug.html

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.r4h.org/r4hsoftware


From aglyportREMOvethispart at nospam.yahoo.com  Wed Sep  4 15:08:03 2002
From: aglyportREMOvethispart at nospam.yahoo.com (Anton Graph)
Date: Wed, 04 Sep 2002 12:08:03 -0700
Subject: Memory fault - core dumped   while compiling 2.2.1 on SCO
References: <3D757EFA.9050401@nospam.yahoo.com> <yu99sn0qqw1r.fsf@europa.research.att.com> <m3bs7etem9.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3D765A13.7090103@nospam.yahoo.com>

Martin v. Loewis wrote:
> Andrew Koenig <ark at research.att.com> writes:
> 
> 
>>Looks like the same symptom I had.
>>
>>If you are using binutils 2.13, try going back to 2.12.1 and
>>see if that fixes the problem.
> 
> 
> Alternatively, use the system C compiler - Python does not support gcc
> on SCO, apparently (or else it would not try to pass -belf).
> 
> It is probably easy to add gcc support - just adjust the OpenUnix
> fragments in configure.in to SCO.
> 
> Regards,
> Martin

Ok. I run gmake distclean
and configure --without-gcc
(output from configure confirmed that it would not use gcc).

Compiling:

cc -G -KPIC -Ki486 -belf -Wl,-Bexport 
build/temp.sco_sv-3.2-i386-2.2/arraymodule.o -L/usr/local/lib -o 
build/lib.sco_sv-3.2-i386-2.2/array.so
/bin/sh: 1577 Memory fault - core dumped
gmake: *** [sharedmods] Error 139

Did not even take nearly as long for it to coredump. With optionless 
configure build was going at least for a half an hour and this one
bailed out pretty much right away...



From erick_papadakis at REMOVEFORSPAM.yahoo.com  Tue Sep 17 11:32:10 2002
From: erick_papadakis at REMOVEFORSPAM.yahoo.com (Erick)
Date: Wed, 18 Sep 2002 00:32:10 +0900
Subject: can wxpython do this?
References: <737aefee.0209151702.3f735252@posting.google.com> <spcaou804m5oo6gah31oqaeuqsuj60h0ip@4ax.com> <3d854b44@news.sentex.net>
Message-ID: <am7hv3$623$1@news01be.so-net.ne.jp>

    > A better option, although I haven't tried it myself, would be
    > to use py2exe to build an executable, then package that up


thanks Peter, but I was talking about something cross-platform. As py2exe's
page says:

     py2exe is a distutils extension to convert python scripts into
    executable windows programs, which are able to run without
    requiring a python installation.


do you know if it works across platforms?

thanks/erick




From m2 at plusseven.com  Thu Sep 19 10:46:27 2002
From: m2 at plusseven.com (Alex Polite)
Date: Thu, 19 Sep 2002 16:46:27 +0200
Subject: debugging "Too many open files"
In-Reply-To: <m3y99zj8kr.fsf@mira.informatik.hu-berlin.de>
References: <mailman.1032378623.20448.python-list@python.org> <m3y99zj8kr.fsf@mira.informatik.hu-berlin.de>
Message-ID: <20020919144627.GJ2134@matijek.plusseven.com>

On Wed, Sep 18, 2002 at 11:23:32PM +0200, Martin v. Loewis wrote:
> Alex Polite <m2 at plusseven.com> writes:
> 
> > I'm getting the "Too many open files" error.
> 
> What "the" error? An exception? A message in the system log? An SMS
> on your mobile phone?  

Well first there's a loud voice in my head telling me to eat Christian
babies and right after that my python program says 
"OSError: [Errno 24] Too many open files" 
or sometimes 
"IOError: [Errno 24] Too many open files: '/proc/mounts'" and exits.


> I recommend you list /proc/<pid of python program>/fd/
Thanks. This tells me that some of my processes have a LOT of files
open, most of them pointing to /dev/hdc. Is this normal?


-- 
Alex Polite
http://plusseven.com/gpg/



From pete at shinners.org  Wed Sep 18 00:55:56 2002
From: pete at shinners.org (Pete Shinners)
Date: Wed, 18 Sep 2002 04:55:56 GMT
Subject: How to make a C extension module backwards compatible?
References: <mailman.1032319522.8608.python-list@python.org>
Message-ID: <wHTh9.11552$L12.2846@sccrnsc02>

Andrew McNamara wrote:
> Is there any community wisdom on how to make a C extension module that
> implements a new-style class (type?) backwards compatible (albeit,
> at the cost of some functionality)?
> 
> Put another way, how *should* I make a class implemented in C behave
> like a new-style class when installed with python 2.2 and above, and
> like an old style class when installed with older pythons?
> 
> Should I just riddle the code with #ifdef's, or is there a more elegant
> way?

i've gone the #ifdef route, and it's a bit of a pain, but manageable. note 
that this was all completely undocumented when i put it together. i mainly 
just copied the builtin types included with python (with my own #ifdef 
stuff going in). here's related excerpts from the code that do what i want. 
basically my object is "RectType", and there is an object named "Rect" to 
construct them. the object is either a function or the actual RectType 
depending on the python version used. unfortunately i believe this is still 
currently undocumented, so good luck.




static PyTypeObject PyRect_Type = {
	...
#if PYTHON_API_VERSION >= 1011
	Py_TPFLAGS_DEFAULT|Py_TPFLAGS_CHECKTYPES|Py_TPFLAGS_BASETYPE,
#else
	0,
#endif
	doc_Rect_MODULE,    /* Documentation string */
#if PYTHON_API_VERSION >= 1011
	...
	rect_new,
#endif
};



#if PYTHON_API_VERSION >= 1011 /*this is the python-2.2 constructor*/
static PyObject* rect_new(PyTypeObject *type, PyObject *args,
		PyObject *kwds)
{
	GAME_Rect *argrect, temp;
	if(!(argrect = GameRect_FromObject(args, &temp)))
		return RAISE(PyExc_TypeError, "badargs");

	return PyRect_New4(argrect->x, argrect->y);
}
#endif



static PyMethodDef rect__builtins__[] =
{
#if PYTHON_API_VERSION < 1011
	{ "Rect", RectInit, 1, doc_Rect },
#endif
	{NULL, NULL}
};



void initrect(void)
{
	PyObject *module, *dict, *apiobj;
	PyType_Init(PyRect_Type);

	module = Py_InitModule3("rect", rect__builtins__, rectangle_doc);
	dict = PyModule_GetDict(module);

	PyDict_SetItemString(dict, "RectType", (PyObject *)&PyRect_Type);
#if PYTHON_API_VERSION >= 1011 /*this is the python-2.2 constructor*/
	PyDict_SetItemString(dict, "Rect", (PyObject *)&PyRect_Type);
#endif
}



From syver-en+usenet at online.no  Tue Sep 17 08:00:25 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Tue, 17 Sep 2002 12:00:25 GMT
Subject: Using emacs gud/pdb python debugger under windows
References: <uvg5axb9e.fsf@online.no> <x53cse2dqw.fsf@tupik.goethe.zz> <r8fxwzdw.fsf@online.no> <x5adml216y.fsf@tupik.goethe.zz> <87elbuo46g.fsf@tleepslib.sk.tsukuba.ac.jp> <87n0qiul6j.fsf@computer.localdomain> <x5lm623vtf.fsf@tupik.goethe.zz> <ufzwam04i.fsf@online.no> <x5y9a22b95.fsf@tupik.goethe.zz> <usn09x6br.fsf@online.no> <u1y7t6zld.fsf@hotmail.com>
Message-ID: <3cs86d38.fsf_-_@online.no>

Felix <aprogrammer at hotmail.com> writes:

> Syver Enstad <syver-en+usenet at online.no> writes:
> 
> I am just trying to find out how/when I can
> use GNU emacs for debugging my python programs, because I generally
> prefer not to leave Emacs. I am downloading the latest emacs sources from
> cvs from savannah in order to check out if things work differently with
> > that version.
> 
> u just need to fix a few things to make it recogonize the win32 path.
> here's how, open gud.el and search for the variable
> 'gud-pdb-marker-regexp',
> 
> then modify it to something like: 
> (defvar gud-pdb-marker-regexp
>   "^>
> \\([-a-zA-Z0-9_/.:\\]*\\|<string>\\)(\\([0-9]+\\))\\([a-zA-Z0-9_]*\\|\\?\\)()\\(->[^\n]*\\)?\n")
> 
> this works well for me.

Ha, ha it works, it works! (Dancing around the room) I want to hug you
Felix! I didn't think of that, obviously. So it was just pdb failing
to recognize the output from the windows version of python.

I am a happy man. Thank you very much Felix.

The version above had as line break in it. This version is without the
linebreak in the re (don't know if it matters.

  (defvar gud-pdb-marker-regexp "^> \\([-a-zA-Z0-9_/.:\\]*\\|<string>\\)(\\([0-9]+\\))\\([a-zA-Z0-9_]*\\|\\?\\)()\\(->[^\n]*\\)?\n")

-- 

Vennlig hilsen 

Syver Enstad


From loewis at informatik.hu-berlin.de  Thu Sep 19 03:43:14 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 19 Sep 2002 09:43:14 +0200
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <mailman.1032329366.6813.python-list@python.org> <j4ptvbaapm.fsf@informatik.hu-berlin.de> <3D890FB0.4030408@something.invalid> <m3admeo4u1.fsf@mira.informatik.hu-berlin.de> <ambso1$4g8vh$1@ID-59885.news.dfncis.de>
Message-ID: <j4fzw6tofh.fsf@informatik.hu-berlin.de>

Thomas Heller <theller at python.net> writes:

> > That is not true. It is more reliable to inspect the source code.
> > In general, it is very easy to do so: if there are no callbacks to
> > Python, the module is stackless-safe.
> 
> So win32all probably wxPython are incompatible with stackless?

No. If a module has callbacks, it is thread safe if it does not put
data on the stack whose addresses are used outside the function, and
it does not use setjmp, and it does not use exception handling.

Notice that the ifs are not if-and-only-ifs; there might be cases
where a module puts a global object on the stack and is still
compatible with stackless (e.g. if the function that does that does
not callback).

Reportedly, wxPython is compatible with stackless (by experimentation,
not by inspection).

Regards,
Martin



From amk at amk.ca  Sun Sep  1 17:07:24 2002
From: amk at amk.ca (A.M. Kuchling)
Date: Sun, 01 Sep 2002 21:07:24 -0000
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de>
Message-ID: <un50ccotaak86b@corp.supernews.com>

In article <3d725881.345921 at news.t-online.de>, Gerson Kurz wrote:
> in the match or find function. So basically, couldn't one come up with
> a *human readable* syntax for re, and compile that instead? Python

Maybe Ka-Ping Yee's rxb?  http://web.lfw.org/python/rxb15.py

The problem with a new syntax is that no one else would be using it, so
you'd still need to learn the existing syntax for use with grep, vi, Perl,
&c.  (It wouldn't surprise me if Perl 6's revised regexes run into this very
difficulty and don't gain much adoption.)

--amk


From dwelch91 at nospam.attbi.com  Sun Sep 29 13:51:32 2002
From: dwelch91 at nospam.attbi.com (djw)
Date: Sun, 29 Sep 2002 17:51:32 GMT
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com> <3d9739c7$0$19818$91cee783@newsreader01.highway.telekom.at>
Message-ID: <3D973DA4.4040600@nospam.attbi.com>

Jurie Horneman wrote:
> "Jurie Horneman" <jhorneman at pobox.com> wrote in message
> news:9ac02e81.0209290839.4f4d6475 at posting.google.com...
> 
>>rant
> 
> 
> My thanks to both of the people who have ignored my ranting and tried to
> help me. I believe I did what you suggested, but it didn't work, hence my
> frustration.
> 
> Right after posting my original problem, I completely removed and
> reinstalled Python. And hey, now it works, with solution 4, ie packages as
> they are described in the documentation. I don't know what I did to my
> installation to make it do that, but my problem appears to be solved.
> 
> For what it's worth - I've already been called on this once in a private
> email: although my post was meant to illustrate my frustration and criticism
> of the Python docs, it was also meant to be at least partly
> self-deprecatory. I should have known better: not because nobody out there
> gets that, but because text, and especially text sent out to millions of
> people, just is not a good place for irony - especially when one is asking
> for help. So if you felt offended: sorry, that wasn't what I intended.
> 
> However, I do stand by my criticisms of the Python docs.
> 
> Jurie.
> 
> 
> 

Although I believe that the Python docs are a bit on the brief side 
sometimes, I don't think this topic has this problem... if you look at 
the package docs at:

http://python.org/doc/current/tut/node8.html#SECTION008400000000000000000

... it has a nicely done example of a directory layout and how the 
__init__.py files fit into the picture. What I would fault the docs with 
is the fact that (having gone through this fustration myself) this 
information is not easily discoverable. How was I supposed to know that 
I needed a "package" in order to import a module from a subdirectory? 
I'm sure many would say, go buy a Python book and look there... but I 
did that, and the books aren't any better in my opinion in this area. 
This topic of importing modules from various directories has been my 
biggest fustration with the language... definitely a "wart" in my 
opinion. Don't get me wrong, discovering Python-and-All-Its-Splendor has 
changed my whole approach to software development (and made it fun 
again), but there are still a few things that could use some polish.

Regards,

Don



From list-python at ccraig.org  Tue Sep  3 01:04:45 2002
From: list-python at ccraig.org (Christopher A. Craig)
Date: 03 Sep 2002 01:04:45 -0400
Subject: Why it is so dramatical?
In-Reply-To: <3D73E035.2040300@nospam.attbi.com>
References: <200209021107.25761.b.maryniuk@forbis.lt>
	<mailman.1030961043.19190.python-list@python.org>
	<3D73E035.2040300@nospam.attbi.com>
Message-ID: <t1wptvv3bpu.fsf@kermit.wreck.org>

djw <dwelch91 at nospam.attbi.com> writes:

> Stupid question:
> 
> If ""join() is 100-1000 X faster, why doesn't the Python interperter
> translate s = s1 + s2 + ... into s = "".join( [ s1, s2, ... ] ) 
> automatically?

You won't see that kind of dramatic results if you did that.  In fact,
my testing shows it would be slower.  

s = "".join(s1, s2) is slower than s=s1+s2.  The reason you see that
kind of speed up is that list.append(s1) is a ton faster than s+=s1,
and "".join() is moved outside the loop.  Compare the following
(presume 'a' contains some string):

method a)
s=""
for x in xrange(10000):
   s += a

method b)
l = []
for x in xrange(10000):
   l.append(a)
s = "".join(l)

method c)
s=""
for x in xrange(10000):
   s = "".join([s, a])


method a creates 10,000 individual string objects
method b creates 1 list object (with a bunch of appends) and 1 string
    object
method c creates 10,000 list objects (each of two elements s and a) and
    10,000 string objects.

On my system, as I would expect, method a takes about .40 seconds,
method b takes .048, and method c takes .51.

-- 
Christopher A. Craig <list-python at ccraig.org>
"Imagination is more important than knowledge" Albert Einstein



From max at alcyone.com  Wed Sep 25 01:46:16 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 24 Sep 2002 22:46:16 -0700
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <%8ci9.138477$5r1.5541622@bin5.nnrp.aus1.giganews.com> <amltd3$3k$1@glue.ucr.edu> <3D911FE1.1050801@something.invalid>
Message-ID: <3D914DA8.A62C0659@alcyone.com>

Greg Ewing wrote:

> Personally, I rather like
> 
>     a if b else c
> 
> but Guido doesn't like that either, unfortunately.
> 
> Perhaps I'll slip it into Pyrex when no-one's looking...:-)

I went ahead and added one in empy, since it caused no conflicts;
@(expression) can be used for embedding expressions, and so @(test ?
thenExpression : elseExpression) can be used for embedding conditional
expressions.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Music washes away from the soul the dust of everyday life.
\__/ Berthold Auerbach
    7 Sisters Productions / http://www.7sisters.com/
 Web design for the future.


From marten.hedman at btk.utu.fi  Fri Sep  6 08:45:04 2002
From: marten.hedman at btk.utu.fi (Marten Hedman)
Date: Fri, 06 Sep 2002 15:45:04 +0300
Subject: Win32 question: Problem with NetLocalGroupDelMembers
Message-ID: <3D78A350.4010006@btk.utu.fi>

Hi,

Im writing a script for administrating local user groups on a remote 
WinNT computer, using Python 2.2.1 and the Win32 extensions, build 148. 
Getting information about users in a group and adding users to a group 
works fine, but when I try to delete users from a group I get an 
exception, 'TypeError: The object can not be converted to a Unicode 
object'. An interactive example session looks like this:

PythonWin 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32.
Portions Copyright 1994-2001 Mark Hammond (mhammond at skippinet.com.au) - 
see 'Help/About PythonWin' for further copyright information.
 >>> import win32net
 >>> server = '\\\\LAVENDER'
 >>> level = 3
 >>> uname = 'BTK\\testperson'
 >>> group = 'testgroup'
 >>> u={}
 >>> u['domainandname'] = uname
 >>> u=[u]
 >>> u
[{'domainandname': 'BTK\\testperson'}]
 >>> win32net.NetLocalGroupAddMembers(server, group, level, u)
 >>> mem, tot, res = win32net.NetLocalGroupGetMembers(server, group, level)
 >>> mem
[{'domainandname': u'BTK\\testperson'}]
 >>> win32net.NetLocalGroupDelMembers(server, group, mem)
Traceback (most recent call last):
   File "<interactive input>", line 1, in ?
TypeError: The object can not be converted to a Unicode object
 >>> win32net.NetLocalGroupDelMembers(server, group, u)
Traceback (most recent call last):
   File "<interactive input>", line 1, in ?
TypeError: The object can not be converted to a Unicode object
 >>>

Have I misunderstood something, or have I stumbled upon a bug?
Any help would be apprciated.

Yours,

Marten Hedman
Systems Administrator
Turku Centre for Biotechnology
Turku, Finland



From ifls at rz.tu-clausthal.de  Tue Sep 24 07:25:04 2002
From: ifls at rz.tu-clausthal.de (Lutz Schroeer)
Date: 24 Sep 2002 11:25:04 GMT
Subject: Sessionhandling
References: <c5b3549a.0209230650.34a30fca@posting.google.com> <f72dac0d.0209231211.5614b398@posting.google.com> <amp3h2$1ot7$1@ariadne.rz.tu-clausthal.de> <bvo56-tf1.ln1@junker.stroeder.com>
Message-ID: <ampi2g$28ci$1@ariadne.rz.tu-clausthal.de>

Michael Str?der <michael at stroeder.com> wrote:
: Most users do *not* configure their browsers anyway. Therefore it's 
: almost useless to use HTTP_ACCEPT_LANGUAGE or similar env vars.

Well, at least you can try. My solution to the orignal poster's problem
would be to detect the browser's language configuration and additionally
offer the possibility to change the language and store it in a (maybe
persistent) cookie.

Subsequently those with correct browser configuration will get their
preferred language and the other 90% will have the opportunity to use the
switch. 

Personally I'm always pleased if a website opens in German because sometimes
I'm not aware that a page offers a different language than English. 
Nevertheless I also like to switch back to English on some pages because the
translations are sometimes a bit "creative" :-)


-- 



From newgene at bigfoot.com  Mon Sep  9 17:34:55 2002
From: newgene at bigfoot.com (newgene)
Date: 9 Sep 2002 14:34:55 -0700
Subject: Save session in python interactive console
Message-ID: <a50631f6.0209091334.7d59476c@posting.google.com>

Hi, group,
    After I was working at python interactive console(activepython
v2.2 at Win2000 platform), I am wondering if possible to save the
current session, so that I can load it again and continue the work
next time. Just like R or Matlab, etc. Although python is a
programming language, one of the best features I like is its
interactive console. By writting my own module, I can make it as my
personal working console. If saving session is possible, it will be
much more convinient.
    The pickle module provided by python can save any object,but how
can I save all objects in current session?

Thanks in advance


CW


From whisper at oz.net  Tue Sep 17 23:08:21 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 17 Sep 2002 20:08:21 -0700
Subject: Yet another Tkinter and threads question
In-Reply-To: <4936aa67.0209171842.12cacf2a@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEIEKDFAAA.whisper@oz.net>

> After a short perusal of the group archives I haven't found anything
> answering the following, so here goes:
>
> I am polishing up the GUI for a client that uses FTP to submit MVS
> jobs to an IBM S/390, and am adding message boxes for exceptional
> circumstances i.e., user mistyped password, connection timeout, etc.
> To prevent locking up the GUI, all submissions spawn a thread, which
> calls a wrapper function around ftplib that does the right
> site-specific hanky-panky to make everything go smoothly. In the
> thread function, the code goes something like this:
>
> try:
>     submit_job()
> except socket.error:
>     showerror("Socket has barfed")
> except ftplib.error_perm:
>     showerror("You mistyped your password")
>
> The call to showerror locks the program. Any thoughts? Older posts on
> other threading+GUI questions had mentioned that Tk was not
> thread-safe; has it become so, or is this still a tar-pit?
>
> Nick
>
> A.M.D.G.

Yes, very simple. Never ever make a gui call in more then one thread. Make a
single gui thread and have it process messages from the other threads to
report/display problems such as you show. This is pretty much true for every
gui out there, not just Tkinter. Tk is not thread safe, nor wxWindows, nor
Win32. There was talk about making Tk thread safe in some future
incarnation, but it's a big job and I don't know if it's even started yet.

HTH...

Dave LeBlanc
Seattle, WA USA




From kara.klosterman at attbi.com  Thu Sep 26 19:43:33 2002
From: kara.klosterman at attbi.com (kara.klosterman)
Date: Thu, 26 Sep 2002 23:43:33 GMT
Subject: Help! This is driving me insane!
Message-ID: <FYMk9.12763$wH.1166@sccrnsc01>

Hello,
I created a very simple class that totals some data
but I am not getting the output from it that I would
expect at all.
Here is the class:
http://www.geocities.com/dbasch/insane.txt

I would expect to get:
[0, 0]
[0, 0]
[[0, 0]]

but I am getting:
[0, 0]
[0, 0]
[[1, 1]]

I dont understand how the 'grouptotals' list is
getting the incremented values prior to the
'grouptotalcounter' function being called.
furthermore, I can place a 'print grouptotals' right
after the errorlog append and it isnt the incremented
values! Please tell me this isnt a normal thing!
Thanks,
Derek Basch





From erwin at andreasen.com  Mon Sep 16 15:58:41 2002
From: erwin at andreasen.com (Erwin S. Andreasen)
Date: 16 Sep 2002 21:58:41 +0200
Subject: Py_CompileString and line numbers?
References: <slrnaoc065.9ki.sa@gamegirl.hogia.net>
Message-ID: <87hegpg0zy.fsf@andreasen.org>

Sebastian Andersson <sa at hogia.net> writes:

> I'm parsing a config file and from there I get embedded python code
> which I send to Py_CompileString, with the config file's filename given
> as filename argument to Py_CompileString. Later, when errors are found
> in the python code, the wrong line numbers are reported. Is there
> a way to tell Py_CompileString an offset for the first line that
> I give to it?

The alternative to Martin's solution is a hack -- simply prepend the
right amount of newlines before your code so the line numbers will
match up.

-- 
===============================================================
<erwin at andreasen.org>                           Herlev, Denmark     
<URL:http://www.andreasen.org/>                             <*>   
===============================================================



From sonnenbu at informatik.uni-bonn.de  Wed Sep 25 16:09:19 2002
From: sonnenbu at informatik.uni-bonn.de (Frank Sonnenburg)
Date: Wed, 25 Sep 2002 22:09:19 +0200
Subject: Newbie References Question
References: <3d91c72f$0$8558$79c14f64@nan-newsreader-02.noos.net>
Message-ID: <amt4t7$f8a$1@f1node01.rhrz.uni-bonn.de>

"Guy Rabiller" <grabiller at 3dvf.net> schrieb im Newsbeitrag
news:3d91c72f$0$8558$79c14f64 at nan-newsreader-02.noos.net...
> Hi,
>
> let say I have:
> i1 = 1
> i2 = 2
> and
> p = [i1,i2]
>
> How can I have:
> p=[*i1,*i2]
> rather than
> p=[**i1,**i2] as it is curently ?
>
> ( Sorry for this nasty hybrid syntax )
>
> Textualy, how if I want that p[0] refere to the 'i1' container, and not to
> the refererence it contains ?
>
> What I want is that if now I set:
> i1 = 4
> that automaticaly:
> p -> [4,2]
> and not keeping [1,2]

My proposal: THINK OO!

In my opinion, this is an excellent example to use the full power of object
oriented programming. Of course, this is a lot of overhead and maybe you
just wanted a very short approach. But since you told, you are a newbie,
maybe you are not quite used to this style of programming. Just try out and
experiment a little.

===============================================================

# Module point.py


class Point:
        def __init__(self, x=0, y=0):
                self._x = x
                self._y = y

        def __repr__(self):
                return 'Point (%d,%d)' % (self._x, self._y)

        def short_repr(self):
                return '(%d,%d)' % (self._x, self._y)

        def set(self, x, y):
                self._x = x
                self._y = y

class Figure:
        def __init__(self, *points):
                self._points = []
                for p in points:
                        if not isinstance(p, Point):
                                raise TypeError
                        self._points.append(p)

        def __repr__(self):
                s = 'Figure'
                # Derived class instance?
                classname = self.__class__.__name__
                if classname != 'Figure':
                        s = s + ' (%s)' % classname
                if self._points:
                        s = s + ', def. by'
                        for p in self._points:
                                s = s + ' '+p.short_repr()
                else:
                        s = s + ', no points defined'

                return s


        #
        # Now it's your turn ;-)
        #

        # def set, get, remove, ......
        #       ...


class Quad(Figure):
        def __init__(self, p0=Point(0,0), p1=Point(0,1), p2=Point(1,0),
p3=Point(1,1)):
                Figure.__init__(self, p0, p1, p2, p3)

        def set_point(self, index, point):
                if type(index) != type(0) or index < 0 or index > 3:
                        raise ValueError
                if not isinstance(point, Point):
                        raise TypeError

                self._points[index] = point

===============================================================

>>> from point import *
>>> q = Quad()
>>> q
Figure (Quad), def. by (0,0) (0,1) (1,0) (1,1)
>>> q.set_point(1,Point(555,888))
>>> q
Figure (Quad), def. by (0,0) (555,888) (1,0) (1,1)

>>> p = Point(1,2)
>>> p.set(4,2)
>>> p
Point (4,2)

===============================================================


Frank




From dwelch at nospam.vcd.hp.com  Thu Sep 26 12:29:04 2002
From: dwelch at nospam.vcd.hp.com (djw)
Date: Thu, 26 Sep 2002 09:29:04 -0700
Subject: os.makedirs: difference between nt4 and w2k
References: <u3crxdpmx.fsf@dnv.com>
Message-ID: <amvcib$inf$1@news.vcd.hp.com>

Jens Bloch Helmers wrote:
> Suppose unc_path is a UNC path refering to a non existing
> directory on another computer.
> os.makedirs( unc_path ) throws an exception if this path
> is on a windows 2000 machine but creates the directory if
> it is on a win-nt machine.
> 
> Is this a feature of win2k, or could I change the setup
> of that win2k computer to accept the creation of unc_path?
> Could this be done from Python?
> (I have administrator privileges on both computers)
> 
> Jens Helmers

I recently found that os.makedirs( <unc_path> ) does not work properly 
on win2k (haven't tried winnt). So, instead, I created a loop that walks 
the path one level at a time and calls win32file.CreateDirectory() for 
each directory level that os.path.isdir() returns false (or, you can 
simply catch the exception code rc==183 if the directory already exists, 
and continue).  I used split('\\') to generate the list of directories 
to create (element 2 ends up being the server name, element 3 the share 
name, and the slice [4:] is the remainder of the directories). Of 
course, this ends up being very Windows specific, but that doesn't sound 
like an issue for you.

So, _roughly_:

path = '\\\\server\\share\\dir1\\dir2\\'
p = path.split( '\\' )
trg = "".join( [ '\\\\', p[2], '\\', p[3] )  # \\server\share
subdirs = p[4:]  # [dir1, dir2, ...]

for subdir in subdirs:
	trg = os.path.join( trg, subdir )
	
	if not os.path.isdir( trg ):
		try:
			win32api.CreateDirectory( trg, None )
		except win32api.error, (rc, fnerr, msg ):
			if rc == 3 or rc == 53:
				print msg # File not found error
				# Do something intelligent here...
			else:
				print msg # Some other error

Hope this helps.

Don



From dbasch at yahoo.com  Thu Sep 26 19:56:43 2002
From: dbasch at yahoo.com (Derek Basch)
Date: Thu, 26 Sep 2002 16:56:43 -0700 (PDT)
Subject: This is driving me insane!
Message-ID: <20020926235643.37002.qmail@web20806.mail.yahoo.com>

Hello,
I created a very simple class that totals some data
but I am not getting the output from it that I would
expect at all. 
Here is the class:
http://www.geocities.com/dbasch/insane.txt

I would expect to get:
[0, 0]
[0, 0]
[[0, 0]]

but I am getting:
[0, 0]
[0, 0]
[[1, 1]]

I dont understand how the 'grouptotals' list is
getting the incremented values prior to the
'grouptotalcounter' function being called.
furthermore, I can place a 'print grouptotals' right
after the errorlog append and it isnt the incremented
values! Please tell me this isnt a normal thing!
Thanks,
Derek Basch


__________________________________________________
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
http://sbc.yahoo.com



From daniel.dittmar at sap.com  Thu Sep 26 06:25:09 2002
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Thu, 26 Sep 2002 12:25:09 +0200
Subject: Getting last char in string.
References: <mailman.1033027776.19152.python-list@python.org> <amumk5$p52$1@lust.ihug.co.nz>
Message-ID: <amuna6$erv$1@news1.wdf.sap-ag.de>

CheapSkate wrote:
> then what does
> s[-1:] do?

>>> s = ''
>>> s [-1]
Traceback (innermost last):
  File "<stdin>", line 1, in ?
IndexError: string index out of range
>>> s [-1:]
''

s [-1] returns a single element, s [-1:] returns a sequence. But in the
current Python definition (implementation?), single characters are strings
of length 1.

Daniel





From pedronis at bluewin.ch  Wed Sep  4 15:15:48 2002
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Wed, 4 Sep 2002 21:15:48 +0200
Subject: jython memory usage
References: <hVm5ydAY2hd9Eww$@jessikat.demon.co.uk><3d76366a$1_5@news.bluewin.ch> <rnLbuKAv5jd9Ewzz@jessikat.demon.co.uk><3d76494d$1_5@news.bluewin.ch> <YR13DVAUNld9EwXH@jessikat.demon.co.uk>
Message-ID: <3d765ce6_4@news.bluewin.ch>

Robin Becker <robin at jessikat.fsnet.co.uk> wrote in message
YR13DVAUNld9EwXH at jessikat.demon.co.uk...
> In article <3d76494d$1_5 at news.bluewin.ch>, Samuele Pedroni
> <pedronis at bluewin.ch> writes
> .....
> >> >
> >> >
> >> well actually we're running with the default jython.bat which doesn't
> >> seem to specify a memory, 1Gig is probably the default or maybe I ran
> >> out of page space. The python process never gets over 90Mb grrrhhhh.
> >
> >then maybe I'm puzzled,
> >
> >have you checked process memory usage or is just that you get
> >an "out of memory" and are estimating based on your total memory.
> >
> >Because the normal jython.bat uses the JVM default which is
> >to scream "out of memory" after 64Mb of object memory has
> >been allocated, see -Xmx here
> >
> >http://java.sun.com/j2se/1.4/docs/tooldocs/windows/java.html#options
> >
> >regards.
> ......
> we're watching it on the task monitor/performance. No other process
> running, this particular thing just eats up the world.

then the JVM is buggy (very strange btw) because is should not use more
than -Xmx memory and then say "out of memory" and the default should be 64Mb
but not 1Gb for sure.

[To prove the point consider this example program and two runs:
(under w2k btw).

* outofmem.java *

public class outofmem {

 public static void main(String args[]) {
   java.util.ArrayList list = new java.util.ArrayList();
   int count = 0;
   while(true) {
    list.add(new int[1024/4*1024*5]); // ~5MB
    System.out.println(++count);
   }
 }

}

* runs *

D:\exp\java-out-of-memory>\usr\jdk1.4.0\bin\javac outofmem.java

D:\exp\java-out-of-memory>\usr\jdk1.4.0\bin\java -version
java version "1.4.0"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.0-b92)
Java HotSpot(TM) Client VM (build 1.4.0-b92, mixed mode)

D:\exp\java-out-of-memory>\usr\jdk1.4.0\bin\java outofmem
1
2
3
4
5
6
7
8
9
10
11
Exception in thread "main" java.lang.OutOfMemoryError

D:\exp\java-out-of-memory>\usr\jdk1.4.0\bin\java -Xmx80m outofmem
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Exception in thread "main" java.lang.OutOfMemoryError

]

Are you spawning processes through os.system or something? otherwise you are
not telling some piece of the
picture.

regards.




From tundra at tundraware.com  Fri Sep  6 17:10:07 2002
From: tundra at tundraware.com (Tim Daneliuk)
Date: 06 Sep 2002 21:10:07 GMT
Subject: Larry Wall's comment on python...
References: <mailman.1031339711.632.python-list@python.org>
Message-ID: <pc5bla.off.ln@boundary.tundraware.com>

> The very idea of Larry Wall lecturing on code readability is hilarious.
> 
> -Dave
> 
>

Well, that's true, but for all of its similarity to line noise, perl has
been an enormous contribution to our profession/community. I would never want
to diminish Wall's stature, therefore.

As someone deeply steeped in Algol-derived languages (C, PL/M, PL/I, Java ...),
I find Python's indentation approach refreshing on two (pardon the expression), levels:

      - It makes the compiler/interpreter do the work instead of me -  Why should I have
        type {} and ; when the language can figure it out for itself.

      - It causes me to factor code better.  Python encourages a style that
        obviates (some of) the need for the kind of syntactic structures found in older
        block-structured languages.


Wall's comment is correct if you insist on Algol-style block
structuring.  But Python encourages you to factor in a way that makes
such block structure different and perhaps less necessary.  In
particular, the ability to combine functional, procedural, and object
constructs in a seamless way gives you the ability to do a lot of
structuring and scoping *implicitly* w/o the need for explicit
indentation or block delimiting.  Python not only "Fits Your Brain", it
changes how your brain processes and reduces problems.  *This*,
in my view is why the absence of block delimiters is not a big deal.

On a personal note, I've had an interesting year - I've been coding in nothing
but Python (Unix & Win32) and *assembler* (embedded PIC processors).  Talk about
a contrast!  Yet in both cases, the languages "Fit My Brain" for the tasks at
hand.  For instance, I find almost all compiled languages (even C) intrusive
when writing code for embedded systems.  Assembler is so much more expressive
there because *it directly reflects the state of the hardware being controlled.*
Similarly, I love Python for most other (higher level) tasks because *it painlessly
reflects the state of the computational problem being solved.*  Now, I only wish
for an excellent Python optimizing compiler with a retargetable code generator ...
*that* would be Bliss (and I don't mean the language ;)

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com



From rsrchstr at msn.com  Sun Sep 29 21:43:56 2002
From: rsrchstr at msn.com (mike henley)
Date: Mon, 30 Sep 2002 02:43:56 +0100
Subject: idea for a much needed application
References: <576c1752.0209291326.56dcdf31@posting.google.com> <an87db$bjtvs$1@ID-157863.news.dfncis.de> <an8924$bujmm$1@ID-157863.news.dfncis.de>
Message-ID: <an8a95$bhqlk$1@ID-157863.news.dfncis.de>

"mike henley" <rsrchstr at msn.com> wrote in message
news:an8924$bujmm$1 at ID-157863.news.dfncis.de...
> Just brainstorming on one possible way to do it; please advise if you
have:
>
> what would such a script/application do?
>
> it would start up with my computer, simply insert itself in my start-up
> folder or whatever
> it would just wait and detect when i go online
> then it when i do get online it would detect what my IP is
> and when if i'm disconnected then it will detect that too
> and when if i'm reconnected it would detect that too and detect my IP
> or it can just simply do this at regular intervals

Duh! an IM/jabber client would do this already.

>
> then it would use jabber/xml-rp or whatever to make a presence for
> me/my-webapplication on IM networks
> using a username or account name that i can give to others
>
> so that this way the webapplication will have "presence"
> so others will know when the webapplication is online
>
> and then if they IM it it'll reply with my IP so it appears in the IM as a
> hyperlink that they can click to launch their default browser to a url
> consisting of http://myIP
>
> and then perhaps if they wish to be automatically notified of the
> webapplication's presence they can insert their username on a form on the
> webapplication when they get their which will add it to its buddylist
>
> or even within the first IM they get from it there can be a link that if
> they wish to click on will detect and add their username automatically (as
> an option)
>
> or they reply with a message such as "add me" which will do that
>
> and either way they can be added, if they wish, to the webapplications
> buddylist so next time it goes online it can send them an IM automatically
>
> of course it should enable them to remove themselves from the buddylist
> later on if they wish
>
> sounds very much like a jabber bot
> perhaps i can get an open source one and extend or customize it

sounds much like a Jabber bot [and therefore mutli-IM-network (AIM, MSN,
Yahoo, ICQ)] that would detect what my (dynamic) IP is whenever i go online,
and then send it to others in it's buddylist if they wish so or reply to
them with it if they ask it, of course they can add it to their username,
and allow others to add/remove themselves to/from its buddylist, and have it
send them an IM automatically with the IP

>
>
> got any better ideas or there's something wrong in what i have written?
> please tell me, i have read about jabber yet.
>
>
>
>
> "mike henley" <rsrchstr at msn.com> wrote in message
> news:an87db$bjtvs$1 at ID-157863.news.dfncis.de...
> > i guess i'll rephrase; what i wish for is an "application-to-person"
(P2A)
> > IM-based "presence", so that a web-application that i host on a personal
> > webserver, on my own machine, can let others become aware of its
> "presence"
> > easily and be easy to connect to with a simple
> > click-to-launch-browser-and-go-to-my-ip. This should be easy to do. I
just
> > don't have experience with Jabber or xml-rpc, or other potential useful
> > technologies.
>
>




From zack at cs.unibo.it  Mon Sep  2 08:57:50 2002
From: zack at cs.unibo.it (Stefano Zacchiroli)
Date: Mon, 2 Sep 2002 14:57:50 +0200
Subject: portability problem with struct.{pack,unpack}
References: <slrnamv2tp.3fa.zack@lordsoth.takhisis.org> <r7nqka.tl5.ln@127.0.0.1>
Message-ID: <slrnan6o2e.5d0.zack@lordsoth.takhisis.org>

In article <r7nqka.tl5.ln at 127.0.0.1>, Radovan Garabik wrote:
> are you aware of xtalk?
> http://packages.debian.org/xtalk

yes, I'm aware of it, but this doesn't solve the portability problem of
struct.{pack,unpack} :-)))

-- 
Stefano Zacchiroli - undergraduate student of CS @ Univ. Bologna, Italy
zack at cs.unibo.it | ICQ# 33538863 | http://www.cs.unibo.it/~zacchiro
"I know you believe you understood what you think I said, but I am not
sure you realize that what you heard is not what I meant!" -- G.Romney


From lists at gregfortune.com  Thu Sep 12 00:18:23 2002
From: lists at gregfortune.com (Greg Fortune)
Date: Wed, 11 Sep 2002 21:18:23 -0700
Subject: "str.contains(part)" or alternatives?
References: <alo8jr$l04$1@news.online.de> <eWNf9.165073$l_4.88540@atlpnn01.usenetserver.com> <aloav7$mdc$1@news.online.de>
Message-ID: <3d8013e8$0$309$39cecf19@nnrp1.twtelecom.net>

<snip>> 
>> I think the current proposal is to move to
>> 
>>     if ll in s:
>> 
>> for this construct. Happy?
> 
> I would like .contains more, but at least the proposal is _much_ better
> than "if s.find(part) != -1". :-)
> 
> "if part in s" disturbs the pattern "if item in sequence" somewhat but
> that has already been weakened by "if key in dict" (that I like,
> nonetheless).
> 
> Stefan


Why not simply use .count(part) > 0 ?  I didn't catch the first part of the 
conversation, but if I'm looking for a contains, I just do.

if(search_me.count(substr) > 0):
        print 'found'
else:
        print 'not found'


Greg Fortune
Fortune Solutions


From ivan.brkanac at hl.hinet.hr  Mon Sep 16 06:59:22 2002
From: ivan.brkanac at hl.hinet.hr (AT&T)
Date: Mon, 16 Sep 2002 12:59:22 +0200
Subject: Access to CDDB or FreDB from python ?
Message-ID: <3d85b80f_3@news3.prserv.net>

I am using python 2.2. on win32 and I wanted to access CDDB from python
and there is CDDB python module on source forge. But it is not working
at least on Win32 and python 2.2 also while I was looking freedb
documentation they stated
that to calculate exactly DiscID in win32 you should not use MCI what is
method off this library.
I would like to ask if somebody have some experience in accessing cddb from
python and if so if he can share some code or some link. I managed to access
it thru one Ocx control, but this is not really in python spirit since it
works only on win32

Ivan




From rjones at ekit-inc.com  Sat Sep  7 01:48:10 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Sat, 7 Sep 2002 15:48:10 +1000
Subject: Available data types in Gadfly 1.0?
In-Reply-To: <albquv$qp4$1@main.gmane.org>
References: <albquv$qp4$1@main.gmane.org>
Message-ID: <200209071548.10254.rjones@ekit-inc.com>

On Sat, 7 Sep 2002 1:14 pm, Jeff Kowalczyk wrote:
> I'm experimenting with gadfly, and I'm looking for some documentation on
> what data types are available for columns in Gadfly 1.0? so far, I only
> know integer, float, varchar to work. Is there a date type? Boolean?

Gadfly will happily store any marshallable data (ie. builtin types like 
strings, lists, None, tuples, ...) in varchar columns. See

     http://gadfly.sourceforge.net/gadfly.html#id7

As far as the "don't count on that being true" ... I don't know of any plans 
to make it not be true.


   Richard




From chris.gonnerman at newcenturycomputers.net  Fri Sep 20 08:12:19 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Fri, 20 Sep 2002 07:12:19 -0500
Subject: sorry for this stupid question about struct module
References: <3d8a4ae7$0$280$7a628cd7@news.club-internet.fr> <zaAi9.117236$ub2.2498874@news1.tin.it>
Message-ID: <004801c2609e$f8b66f00$0101010a@local>

----- Original Message ----- 
From: "Alex Martelli" <aleax at aleax.it>


> Alan Marchand wrote:
> 
> > Hi all,
> > 
> > I tryed this :
> > 
> > e:\temp>python.exe
> > Python 2.2.1 ( #34, Apr 9 2002, ...................)
> > ...........
> >>>> import struct
> >>>> x = struct.pack( '>I', 10 )
> >>>> print repr( x )
> > '\x00\x00\x00\n'                       # why not !
> > 
> > How can I obtain the result a expect : '\x00\x00\x00\0A' ?

<< hacked out lucid explanation of repr >>

> -- but why is repr's
> behavior problematic to you?

It isn't, I'd wager.  I suspect Mr. Marchand doesn't realize
that '\n' == '\0x0A'.  His entire original post indicated that
he needed that exact value to send to a printer, which is what
makes me suspect he didn't know he actually *had* that value.

I also suspect he does know by now.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From jeremy at alum.mit.edu  Thu Sep  5 15:16:33 2002
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 5 Sep 2002 12:16:33 -0700
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com> <eppstein-56D20B.14244204092002@news.service.uci.edu>
Message-ID: <b0f083db.0209051116.142b1c25@posting.google.com>

David Eppstein <eppstein at ics.uci.edu> wrote in message news:<eppstein-56D20B.14244204092002 at news.service.uci.edu>...
> In article <m7ncnugbkm0qghf66qcub6e37b64fesdkq at 4ax.com>,
>  Gon?alo Rodrigues <op73418 at mail.telepac.pt> wrote:
> 
> > Anyway, I've been directed to PEP 227, cheerfully obeyed, and from what
> > I have understood having read-only nested scopes was more of a design
> > decision than something else. And while I can live happily with the
> > decision (some would even say happier) there is something about the "You
> > can look at it but not touch it" attitude that I dislike ;-)
> 
> I'm sure there were good reasons for this decision, but it leads to ugly 
> workarounds like making each nonlocal variable into a list and assigning 
> to x[0] instead of x...

I agree with you in principle.  I think that nested scopes would be
cleaner if they allowed rebinding names in enclosing scopes.  But it's
not a limitation that has ever caused me grief.  There are a variety
of workarounds.  The use of a list to represent a mutable variable is
one, but replacing the nested function with a class works, too.

Here's what I had to say in PEP 227 about the current state of
affairs:

    There are technical issues that make it difficult to support
    rebinding of names in enclosing scopes, but the primary reason
    that it is not allowed in the current proposal is that Guido is
    opposed to it.  His motivation: it is difficult to support,
    because it would require a new mechanism that would allow the
    programmer to specify that an assignment in a block is supposed to
    rebind the name in an enclosing block; presumably a keyword or
    special syntax (x := 3) would make this possible.  Given that this
    would encourage the use of local variables to hold state that is
    better stored in a class instance, it's not worth adding new
    syntax to make this possible (in Guido's opinion).

    The proposed rules allow programmers to achieve the effect of
    rebinding, albeit awkwardly.  The name that will be effectively
    rebound by enclosed functions is bound to a container object.  In
    place of assignment, the program uses modification of the
    container to achieve the desired effect:

    def bank_account(initial_balance):
        balance = [initial_balance]
        def deposit(amount):
            balance[0] = balance[0] + amount
            return balance
        def withdraw(amount):
            balance[0] = balance[0] - amount
            return balance
        return deposit, withdraw

    Support for rebinding in nested scopes would make this code
    clearer.  A class that defines deposit() and withdraw() methods
    and the balance as an instance variable would be clearer still.
    Since classes seem to achieve the same effect in a more
    straightforward manner, they are preferred.

Jeremy


From Padraig at Linux.ie  Thu Sep 19 08:53:37 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Thu, 19 Sep 2002 12:53:37 GMT
Subject: RegEx editors
References: <amaq0d$4aton$1@ID-144898.news.dfncis.de>
Message-ID: <lNji9.4581$cP3.11126@news.iol.ie>

Eric Arnold wrote:
> Does anybody know of any good graphical regex editors out there?

http://laurent.riesterer.free.fr/regexp/
http://regexplorer.sourceforge.net/
http://txt2regex.sourceforge.net/
http://sedsed.sourceforge.net/ (this is a python script)

P?draig.



From pyth at devel.trillke.net  Sun Sep  1 15:54:06 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 1 Sep 2002 21:54:06 +0200
Subject: Python IPC options
In-Reply-To: <ddc19db7.0209010836.50242309@posting.google.com>; from a-steinhoff@web.de on Sun, Sep 01, 2002 at 09:36:05AM -0700
References: <1030829572.943495@irys.nyx.net> <ddc19db7.0209010836.50242309@posting.google.com>
Message-ID: <20020901215406.E3893@prim.han.de>

Armin Steinhoff wrote:
> fburton at nyx10.nyx.net (Francis Burton) wrote in message news:<1030829572.943495 at irys.nyx.net>...
> [ clipp ..]
> > It would be nice if Python could provide the same kind of
> > communication mechanisms. Unfortunately, while Glish is
> > freely available, it is heavily C++ biased and runs only on
> > some Unix systems. In theory, something similar could be
> > implemented on a much wider range of platforms using
> > sockets and Python.
> > 
> > Can anyone tell me: has an interprocess communication
> > mechanism specifically with a bus architecture been
> > implemented in/for Python?
> 
> Corba 2.2 -> http://www.fnorb.org 

CORBA 2.0 AFAIK.

> Python remote objects -> http://www.sf.net/projects/pyro
> SOAP -> http://www.pythonware.com/products/soap/
> SPREAD -> http://www.python.org/other/spread/

this one has a bus-like architecture (uses IP-Multicast if possible). 

> Send/Receive/Reply for QNX6 message passing -> http://www.sf.net/projects/pyqnx

Can you summarize this one? Couldn't find details on your project home page :-)

regards,

    holger



From bruce_deletethis_dawson at cygnus-software.com  Tue Sep 24 01:14:36 2002
From: bruce_deletethis_dawson at cygnus-software.com (Bruce Dawson)
Date: Tue, 24 Sep 2002 05:14:36 GMT
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <%8ci9.138477$5r1.5541622@bin5.nnrp.aus1.giganews.com> <amltd3$3k$1@glue.ucr.edu> <3D8EB212.1000104@cygnus-software.com> <3D8EB7A9.F7CDE16@alcyone.com>
Message-ID: <3D8FF49E.3000102@cygnus-software.com>

Good point about the difference, and thanks for the pointer/reminder 
about operator.truth. It's still good to have bool though - if only for 
the compactness and readability of it.

Erik Max Francis wrote:

> Bruce Dawson wrote:
> 
> 
>>I find the (a and [b] or [c])[0] construct to obscure to be useful -
>>the
>>space it saves is wasted by the comment I have to put in to explain
>>it.
...
> 
> I use this technique frequently enough, but this isn't really what a
> conditional expression does.  A conditional expression only _evaluates_
> one of its consequents, depending on the value of the test expression. 
> The (a, b)[operator.truth(p)] evaluates both.
> 
> This is fine in many cases, of course -- such as when you're selecting
> between two strings -- but it isn't the full power of a true conditional
> expression.
> 
> 



From tim.one at comcast.net  Fri Sep 20 15:19:33 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 20 Sep 2002 15:19:33 -0400
Subject: Destructor never called ???
In-Reply-To: <OGBi9.117782$ub2.2505532@news1.tin.it>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEIKBFAB.tim.one@comcast.net>

[Tim]
> It's only the cases where Python can't possibly guess a safe order in
> which to call destructors where both __del__ and collection are
> inhibited.

[Alex Martelli]
> I think it's a bit more than that: Python *COULD" "possibly guess a
> safe order" when there is only exactly ONE object with a __del__ that
> is part of the cycle -- namely, destroy that one object first.  It
> doesn't seem to even TRY (note: I'm not saying it SHOULD -- but
> "can't possibly" might be a bit too strong as a description of what
> is actually going on).

I agree.  Way back when, the idea was to compute the maximal strongly
connected components (SCCs) of the trash reachability graph.  The derived
graph (viewing each SCC as a node in its own right) is necessarily a DAG (no
cycles), and cleanup would then proceed in an order consistent with some
topological sort of the DAG nodes (no cycles implies a topsort is always
possible).  It was noted at the time that if the set of objects
corresponding to a single SCC "supernode" contained at most one object with
a __del__, then a safe teardown order for those objects could be known (just
as you suggest).

In practice, computing SCCs is a PITA, and the simpler scheme actually
implemented does all of that *except* for dealing with would have been SCC
DAG nodes corresponding to a set of objects exactly one of which has a
__del__, had SCC DAG nodes been constructed.  Refcount semantics
automatically tear down everything else in what would have been SCC DAG node
order (btw, for those who haven't thought about it, note that "a refcount"
is exactly the same as "a predecessor count" in a topological sort
algorithm:  they have a lot in common).

Since you and I are approximately the only two people on the newsgroup whose
heads haven't exploded from trying to understand exactly what the hell I'm
talking about there <wink>, I think it's a Positive Good that Neil skipped
all the SCC complications.  "Cycles and destructors:  don't even think about
mixing them" is much easier to remember, and has never done me any harm in
real life.




From tg5027 at citlink.net  Tue Sep 17 21:40:34 2002
From: tg5027 at citlink.net (terry)
Date: Tue, 17 Sep 2002 21:40:34 -0400
Subject: Python GUI app to impress the boss?
Message-ID: <E17rTMz-0006XA-00@mail.python.org>

>>?I'm willing to bet, though, that Python will be around for longer
>> than VBScript.

Cameron,

I don't like to disagree with an icon, but I think that you are missing 
a deficiency in Python that will ultimatly limit it's usefullness in 
the vast majority of the software that is written in the world.

As long as Python refuses to admit that money is the focus of most of 
the world's software, Python will be relagated to the same status as 
FORTRAN, BASIC, and etc - except COBOL and VB.

Is there any way to get the Python world to allow for the processing of 
monetary applications without some hokey 60's constructs that involve 
in-line conversions or anticipations of calculation characteristics?

Regards,

terry



From mhammond at skippinet.com.au  Tue Sep 17 07:04:31 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 17 Sep 2002 11:04:31 GMT
Subject: Calling Python from C# - please help
References: <6de16ac9.0209130508.5d1b56c6@posting.google.com> <j4elbyt3yw.fsf@informatik.hu-berlin.de> <ur8fyxau0.fsf@online.no> <0iCg9.2350$Ee4.6516@news-server.bigpond.net.au> <3D82FA5B.3050003@mxm.dk> <uptvgvn1h.fsf@online.no>
Message-ID: <3%Dh9.15217$Ee4.45118@news-server.bigpond.net.au>

Syver Enstad wrote:
> Max M <maxm at mxm.dk> writes:
> 
> 
>>Mark Hammond wrote:
>>
>>
>>>I have no examples of this, but I should.  I'm co-authoring a .NET
>>>book which is nearly out :) -
>>
>>
>>That sounds great. I enjoyed the win32 book, and still use it as a
>>reference.
>>
>>Are you by any chance still developing Pyhton .NET, or has it
>>completely stopped?
>>
>>regards Max M
> 
> 
> Now that J# (Java .NET) has been released, I was wondering about the
> possibility of compiling Jython with the MS J# compiler to create a
> Jython .NET. Has anyone considered the possibility?

I'd love to see some Jython technology used for .NET.  At the time 
Python.NET started, there was very strict NDA, so discussing the project 
with jython-dev, or even python-dev, was taboo.  Greg Stein and I made a 
pragmatic decision to go with the p2c tool as a framework over jython as 
we alrady had experience with p2c and didnt have the time-lines to 
understand jython before embarking on our pilot code.

But see my previous post - time (generally represented as money <wink>) 
is the commodity most scarce.

Mark.



From johnroth at ameritech.net  Tue Sep  3 08:47:40 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 3 Sep 2002 08:47:40 -0400
Subject: Would you like a Snobol pattern matcher? (Re: Making regex suck less)
References: <3d725881.345921@news.t-online.de> <un7n06qrl1gm93@news.supernews.com> <3D740DFC.1040009@something.invalid>
Message-ID: <un9bkdrqeh305c@news.supernews.com>

"Greg Ewing" <see_reply_address at something.invalid> wrote in message
news:3D740DFC.1040009 at something.invalid...
> John Roth wrote:
>
> > http://www.perl.com/pub/a/2002/06/04/apo5.html
> >
> > http://www.perl.com/pub/a/2002/08/22/exegesis5.html
>
>
> I just had a brief look at this, and the underlying
> ideas seem to be a lot like the way Snobol patterns
> work.
>
> Maybe it's time for me to resurrect the Snobol-style
> pattern matching module that I started on a while
> back and never got around to releasing.
>
> Would anyone be interested in this? Its
> syntax is similar to that of Plex REs, except that
> the primitives are Snobol-like, and it uses a
> backtracking matching algorithm that's much more
> powerful than a DFA (you can write entire parsers
> in it, for example).

I don't know much about Snobol, unfortunately.
I think my biggest issue here is that we shouldn't
reinvent the wheel unless there is a good reason.
In other words, Larry is taking Perl in a specific
direction. Assuming we want to make major
changes to regex, is there any _good_ reason
for doing something conceptually different and
consequently adding to the cacophony?

John Roth


> --
> Greg Ewing, Computer Science Dept,
> University of Canterbury,
> Christchurch, New Zealand
> http://www.cosc.canterbury.ac.nz/~greg
>




From markus.vonehr at ipm.fhg.de  Tue Sep 10 10:42:12 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Tue, 10 Sep 2002 16:42:12 +0200
Subject: import Error SWIG extension
Message-ID: <3D7E04C4.A68ADF6E@ipm.fhg.de>

Hi again,

I succeeded compiling my extension (done with SWIG) but trying to 
import it I have the following error:

import mvelib
Traceback (most recent call last):
  File "<pyshell#0>", line 1, in ?
    import mvelib
SystemError: _PyImport_FixupExtension: module mvelib not loaded


I tried compiling the elemlist example from Alex Martelli (it's
made without SWIG) and it works.

I copied both of the extensions in my python home dir, so the
problem has to be in the code?
I had to rename the init_mvelib function to initmvelib that
disutils have been able to compile it.
Is it possible that the error is in the init function?

Anybody knows the cause of the error?

Markus


From fredrik at pythonware.com  Wed Sep  4 03:14:12 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 04 Sep 2002 07:14:12 GMT
Subject: MD4 and DES for unicode strings (samba passwd file manipulation)
References: <uoit1mxubi.fsf@bashful.cdf.toronto.edu>
Message-ID: <8pid9.5889$e5.1033526@newsb.telia.net>

Arcady Genkin wrote:

> And yet I can't create the right crypted strings:
>
>   >>> from Crypto.Hash import MD4
>   >>>
>   >>> m = MD4.new()
>   >>> m.update( u'111222333' )
>   >>> m.hexdigest().upper()
>   'A672E5EDFDAD90A15228E0AEC2AA9161'
>   >>> # SAMBA-encoded password "111222333":
>   >>> samba
>   'B66B458E27BD714F99A05EE3C479FBF1'
>
> Perhaps the problem lies with Unicode encoding?  Samba's man page (see
> the link below) says that the string is obtained by taking MD4 hash on
> the password, represented as a string of 16-bit little-endian unicode
> characters.

if you want a string of 16-bit little-endian unicode characters,
make sure you have one:

    m = MD4.new()
    m.update(password.encode("utf-16-le"))
    m.hexdigest().upper()

(I don't have the MD4 library, so I cannot verify this.  But if
it works, it works for both 8-bit strings with ASCII text, and
Unicode strings)

</F>




From kokohh at hotmail.com  Thu Sep 12 23:06:43 2002
From: kokohh at hotmail.com (koko)
Date: Fri, 13 Sep 2002 03:06:43 GMT
Subject: mime types 
Message-ID: <7Dcg9.8153$yt3.4264497@newssrv26.news.prodigy.com>

i am trying to retreve urls.
if it is mime text file, save the file.
if not, pass

e.g. url=http://www.uchicago.edu
mimetypes.guess_type(url)

sometimes it can give text/html. But a lot of them is text/html, but it only
gives (None, None)
Is there any accurate ways to determine the mine type?




From uwe.schmitt at procoders.net  Tue Sep 10 17:33:47 2002
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 10 Sep 2002 21:33:47 GMT
Subject: How do i print to a printer using python?
References: <6afb9d2c.0209041044.43c5f5f1@posting.google.com> <al73gj$lbt$1@wanadoo.fr> <6afb9d2c.0209091641.1d9b652e@posting.google.com>
Message-ID: <allofr$lgscc$2@hades.rz.uni-sb.de>

Naveed Iqbal <iqbal at colorado.edu> wrote:
> Thank you all for your help. I am really a novice when it comes to
> python. Will someone tell me how to install the win32ui module in
> windows?
> thanka
> naveed

I suceeded without win32all: I generated a postscript file
and copied it to the 'file' "prn:" using shutil.copyfile(). 

Greetings, Uwe.


-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           


From tjabo.kloppenburg at unix-ag.org  Tue Sep  3 15:31:40 2002
From: tjabo.kloppenburg at unix-ag.org (Tjabo Kloppenburg)
Date: Tue, 3 Sep 2002 21:31:40 +0200
Subject: Help with mod_python
In-Reply-To: <GO6c9.22678$Lj6.623043@wagner.videotron.net>
References: <GO6c9.22678$Lj6.623043@wagner.videotron.net>
Message-ID: <20020903213140.5a43330f.tjabo.kloppenburg@unix-ag.org>

hi "Joseph Youssef",

JY> Hello, I've been trying to get python to work on apache, so
JF> I downloaded mod_python...

I have the same problem on my debian linux box (woody). I changed apache.py to print sys.path (the variable with the paths where python searches modules) -- the directory of the webserver was missing!?

After this "experience" I tried the "PythonHandler mod_python.publisher" examples -- they run perfectly.

Next I decided to use PHP for the project... :-)

tk.

--
Tjabo Kloppenburg
  --=- In 111.14 Tagen ist Weihnachten! -=--

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020903/2668f99a/attachment.sig>

From jcc at ibdosnorte.com  Tue Sep 24 03:53:43 2002
From: jcc at ibdosnorte.com (=?ISO-8859-1?Q?Juan_Carlos_CORU=D1A?=)
Date: 24 Sep 2002 00:53:43 -0700
Subject: COM: Returning IDispatch or NULL
References: <a230569c.0209230700.14116a1d@posting.google.com> <Z5Oj9.12374$6g7.37565@news-server.bigpond.net.au>
Message-ID: <a230569c.0209232353.2710adc8@posting.google.com>

Mark Hammond <mhammond at skippinet.com.au> wrote in message news:<Z5Oj9.12374$6g7.37565 at news-server.bigpond.net.au>...
> Juan Carlos CORU?A wrote:
> > Hello,
> > 
> > I'm trying to develop a COM Server in python, that has a method that
> > returns an IDispatch object but sometimes it must return nothing. See
> > the following example programmed in C/SIDE (Microsoft Navision
> > Financials ERP):
> > 
> > --->
> > otherCOMServer := COMServer.returnOtherCOMServer();
> > IF ISCLEAR(otherCOMServer) THEN MESSAGE('otherCOMServer is clear');
> > <----
> > 
> > Definition of ISCLEAR (extracted from MS Navision help):
> >   The return value is TRUE if:
> >     -the automation variable has not been created. See CREATE.
> >     -the automation variable has been cleared. See CLEAR.
> >     -the automation variable has been assigned NULL by an automation
> > server.
> > 
> > I thing that in this case I must assign NULL to the otherCOMServer
> > variable in order to keep the otherCOMServer variable CLEARED.
> > What must I return from "COMServer.returnOtherCOMServer()" method to
> > assign NULL?
> > I have tried:
> >   return None
> >   return wrap(None)
> >   return 1==0
> >   return
> > 
> > but without success. The environment responds "data type not
> > supported".
> 
> None is the way to do this.  What responds "data type not supported"? 
> Have you registered your object for debugging and seen the output? 
> There may be a traceback printed.
> 
> Mark.

After executing "return None" Navision responds "This data type is not
supported by C/SIDE".

The debugging output is correct:

OutMessage created <NAVISIONSPREADLib.OutMessage instance at
0x02406260>
Object with win32trace dispatcher created
(object=<NAVISIONSPREADLib.OutMessage instance at 0x02406260>)
Returning wrapped <PyIDispatch at 0x24fde14 with obj at 0x2406468>
in _Invoke_ with 1000 1024 1 (<PyIDispatch at 0x24fde14 with obj at
0x1caad68>,)
in _Invoke_ with 1002 1024 1 (1000,)
returning None

The invoke with 1002 is the call to the method that returns NULL or an
object after 1000 ms of timeout.


From timr at probo.com  Thu Sep 26 23:43:22 2002
From: timr at probo.com (Tim Roberts)
Date: Thu, 26 Sep 2002 20:43:22 -0700
Subject: Can't suspend a thread?
References: <QZ4j9.10630$Lg2.2393215@news2.news.adelphia.net> <amognf$r6g$1@panix1.panix.com> <juj2puko63pmq5k4j4tm2eonoadvpup50d@4ax.com> <uIpk9.11787$Lg2.3232725@news2.news.adelphia.net>
Message-ID: <rqk7pu8d8nl3op3pln8bpqtrrhontqg07t@4ax.com>

"Robert Oschler" <Oschler at earthlink.net> wrote:
>
>"Tim Roberts" <timr at probo.com> wrote:
>>
>> It is worth pointing out that this "restriction" is not a Python thing; it
>> is an operating system thing.
>
>You mean a Linux/Python restriction?  Windows can suspend threads so I
>assume that's what you mean.  Note I'm not saying that you can't suspend
>threads on Linux (POSIX), I don't know if that's possible or not, I just
>know you can on Windows.

Hmm, you're quite right.  I didn't realize you could suspend another thread
in Win32.  The documentation does caution against using it in anything
other than a debugger, but it certainly exists.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From sonnenbu at informatik.uni-bonn.de  Wed Sep 18 10:09:02 2002
From: sonnenbu at informatik.uni-bonn.de (Frank Sonnenburg)
Date: Wed, 18 Sep 2002 16:09:02 +0200
Subject: 'Real' global (write var to different namespace)
References: <29242dfa.0209180310.94c915d@posting.google.com>
Message-ID: <ama153$qta$1@f1node01.rhrz.uni-bonn.de>

"Maarten van Hulsentop" <maarten at tccn.info> schrieb im Newsbeitrag
news:29242dfa.0209180310.94c915d at posting.google.com...
> Hi y'all
>
> I'm sorry if it's a FAQ but i honestly can't find to solution to my
> problem; i've created a 'store' (a class) to store vars using the
> (really cool) pickle classes. Now i can store a var doing:
> "mystore['varname'] = myvar" and read it again by "myvar =
> mystore['varname']". My question is if it's possible to create a
> function like 'mystore.RestoreVars()' that sets ALL the stored vars in
> the _calling_ namespace at once.
>
> Hope it's clear enough, i'm still a python newbie so excuse me if i
> used the wrong terms or if i didn't explain it enough.
>
> Thanks
>
> Maarten van Hulsentop

Maybe you will use this class prototype:

======================================================

class store:
  "Storage class for arbitrary data"

  def writeto(self, obj):
    "Make contents of storage instance visible in obj"

    obj.__dict__.update(self.__dict__)

======================================================

You have to pass the calling namespace explicitly, maybe you will see this
as a disadvantage, but it fits to python philosophy: explicit is better than
implicit.

======================================================


Examples:

1 - Instantiating of storage

>>> st = store()
>>> st.storage_var = 42

2 - Make visible in main scope

>>> import sys
>>> st.writeto(sys.modules['__main__'])
>>> dir()
['__builtins__', '__doc__', '__name__', 'st', 'storage_var', 'store', 'sys']
>>> storage_var
42

3 - Make visible in some object

>>> class c:
...     pass
...
>>> o = c()
>>> st.writeto(o)
>>> dir(o)
['storage_var']
>>> o.storage_var
42

4 - Usage in instance method

>>> class cc:
...     def f(self):
...             st.writeto(self)
...
>>> oo = cc()
>>> oo.f()
>>> oo.storage_var
42


======================================================

Hope it is nearly what you desired -- Frank






From eddie at holyrood.ed.ac.uk  Thu Sep 12 12:56:36 2002
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Thu, 12 Sep 2002 16:56:36 +0000 (UTC)
Subject: Reading all text in telnet connection
References: <4378fa6f.0209120722.227fe930@posting.google.com>
Message-ID: <alqh04$3tq$1@scotsman.ed.ac.uk>

mnations at airmail.net (Marc) writes:

>Hi:

>Having a small problem trying to capture all the text in a telnet
>connection. I've used all the options of eager, lazy, etc., and can't
>seem to find the right combo.

>What I'm doing is intermittently writing data and then reading all of
>the response. Unfortunately, all of the response might not immediately
>appear. I started off using read_until, but there's nothing really
>consistent that I can read to. So I need the ability to simply read
>everything current in the buffer until EOF.

I suspect you don't mean EOF, that only occurs when the connection is closed.
There's really only 2 ways to know when output is finished:
1) Find a string to match on
2) keep reading until enough time has passed that you're fairly sure there
   will be no more output

Because (2) is both slow and generally flaky option (1) is by far the best.
However you say that there is nothing to match on, if that's really true you
have to resort to hacks like waiting for time to pass.

Because (1) is by far the best it is worthwhile making the effort to find a
way of using it.  If the problem is that there are several alternatives that
might match you can use telnet.expect().  If there really is no sensible thing
such as a prompt to match on it may be possible to match on your own echoed
input if it's unique (this is especially true if you can send comments, I use
this in speaking to routers where the prompt keeps changing - I send #gurgle
and then look for this comimg back).

There is one more possibility but I don't think it's appropriate from your
description.  If what you are doing is sending a few pre-determined commands
collecting all the output and closing the connection then you MAY be able to
get away with sending all the commands in advance (hoping the far end buffers
them up) and use telnet.read_all() to wait for all the results (this depends
on the far end closing the connection as a consequence of the commands you
send).

Essentialy, if there is nothing to match on you cannot reliably automate it.

Eddie


From nhodgson at bigpond.net.au  Thu Sep 26 17:34:44 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Thu, 26 Sep 2002 21:34:44 GMT
Subject: Pythonwin (win32all build 148) on Python 2.2.1 on Win2K, bogus syntax e
References: <u47k9.46401$V7.11091779@twister.socal.rr.com> <CJ8k9.16479$6g7.52569@news-server.bigpond.net.au> <sHJk9.66556$U7.23747920@twister.socal.rr.com>
Message-ID: <U3Lk9.24576$6g7.75712@news-server.bigpond.net.au>

Aaron:

> Well, I did that, and that seems to be the problem. Is there a Pythonwin
> setting I can set to make it do standard line endings?

   PythonWin itself does standard CRLF line endings. There is a script in
the Python distribution Python22\Tools\Scripts\lfcr.py for fixing files. The
script may be located elsewhere in older versions.

   Neil





From jon+usenet at unequivocal.co.uk  Mon Sep 23 23:19:50 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Tue, 24 Sep 2002 03:19:50 -0000
Subject: Python and cgi and <a> problem!
References: <ammn6e$lj7$1@lust.ihug.co.nz> <Xns9292EA042C259cliechtigmxnet@62.2.16.82> <slrnaov5q3.nrp.jon+usenet@snowy.squish.net> <Xns929330F5A61B0cliechtigmxnet@62.2.16.82>
Message-ID: <slrnaovmel.dma.jon+usenet@snowy.squish.net>

In article <Xns929330F5A61B0cliechtigmxnet at 62.2.16.82>, Chris Liechti wrote:
>>> print '<A href="%s">%s</A>' % (filename, filename)
>> 
>> I hope none of your filenames have '"', ">" or "&" in.
> 
> yeah i avoid characters that are special to the shell in my filenames - i 
> even don't use " " very often. <wink>
> 
> sure there are improvements, but its a start.

I'm not sure recommending completely broken and insecure code is ever
a good thing. No offence intended.


From max at alcyone.com  Tue Sep 24 18:39:11 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 24 Sep 2002 15:39:11 -0700
Subject: List/Tuple bug or feature?
References: <8765wv6o0d.fsf@nospam.eml.cc>
Message-ID: <3D90E98F.95541F66@alcyone.com>

Mahesh Padmanabhan wrote:

> If this is a concatenation feature, how can I prevent something like
> this happening in the future ?

It's a standard feature, and doesn't have anything to do with tuples or
lists:

>>> 'a' 'b'
'ab'

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Once the people begin to reason, all is lost.
\__/ Voltaire
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From aahz at pythoncraft.com  Fri Sep 20 00:33:32 2002
From: aahz at pythoncraft.com (Aahz)
Date: 20 Sep 2002 00:33:32 -0400
Subject: Destructor never called ???
References: <76710511.0209111450.40999dd8@posting.google.com> <tp70ou0r0j11as6enhpsaqul4oqa3mlu6t@4ax.com> <alp8n3$bsd$1@panix1.panix.com> <4e8lou018f9uosl5p9mb0tt6691mfivamj@4ax.com>
Message-ID: <ame8es$ppc$1@panix2.panix.com>

In article <4e8lou018f9uosl5p9mb0tt6691mfivamj at 4ax.com>,
Tim Roberts  <timr at probo.com> wrote:
>aahz at pythoncraft.com (Aahz) wrote:
>>In article <tp70ou0r0j11as6enhpsaqul4oqa3mlu6t at 4ax.com>,
>>Tim Roberts  <timr at probo.com> wrote:
>>>
>>>However, Python is a garbage-collected system.  Unlike in COM, an object is
>>>NOT destroyed as soon as its reference count goes to 0.  Your object will
>>>only be destroyed if the garbage collector happens to run, and decides that
>>>it might be a convenient time to actually free up the space your object
>>>occupies.
>>
>>Wrong.  CPython uses reference counting backed by GC.
>
>Well, then, what was all the fuss about deterministic desctructors in the
>newsgroup last year or so?  Has this changed, or am I just remembering this
>entirely wrong?

There are several different issues.  First of all, __del__ does not get
called if an object is destroyed by GC (GC gets used to clean up cyclic
garbage).  Secondly, Jython does not use refcounting; it's pure GC, just
like Java.  Therefore, if you want to write portable applications, you
should not rely on refcounting.

Finally, there's a stylistic issue in programming.  Some people believe
that the "explicit is better than implicit" design goal requires
cleaning up after oneself.  Actually, it's unfair to say that it's
"just" a stylistic issue -- it's certainly possible that calling out to
libraries requires that one issue explicit close() calls.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From huaiyu at gauss.almadan.ibm.com  Fri Sep 13 18:32:52 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Fri, 13 Sep 2002 22:32:52 +0000 (UTC)
Subject: Restricting methods in derived classes
References: <mailman.1031932520.1147.python-list@python.org>
Message-ID: <slrnao4psk.7ej.huaiyu@gauss.almadan.ibm.com>

Mark McEahern <marklists at mceahern.com> wrote:
>Huaiyu,
>
>I feel compelled to gently ask, "What's the point of this rigamarole?"

Maybe to show how to make a solution larger than the problem itself?  :-)
But seriously, your first solution actually contains all I needed.  Here's
what I adopted in the end:

def undefined(self, *args, **kwargs):
	raise NotImplementedError

_common = ['__class__', '__defined__', '__dict__', '__doc__',
			'__getattribute__', '__init__', '__new__']

def restrict(foo):
	for a in dir(foo):
		if a not in _common + foo.__defined__:
			setattr(foo, a, undefined)

class foo(dict):
	def __init__(self, *args):	print "foo", args
	def put(self, x):		pass
	__defined__ = ['put', 'get', 'keys', 'items', 'values']

restrict(foo)

if __name__ == '__main__':
	f = foo()
	print f.keys
	print f.keys()
	f['a'] = 'b'

Maybe there is a simple way to wrap up restrict() in a metaclass, so that
instead of restrict(foo) we can say class foo(dict, restricted): ...
I'm still not so used to metaclasses to figure that out, however.

Huaiyu


From martin at v.loewis.de  Wed Sep  4 17:39:03 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 04 Sep 2002 23:39:03 +0200
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com> <yu99u1l5bkra.fsf@europa.research.att.com> <al5t94$vse$0@216.39.172.122>
Message-ID: <m3bs7da0zs.fsf@mira.informatik.hu-berlin.de>

bokr at oz.net (Bengt Richter) writes:

> OTOH, if "everything is an object," then the namespace of
> an enclosing scope could also be an object, and as such
> could support rebinding of names therein ;-)

It is an object - the frame object. There is just no easy way to refer
to that object.

Regards,
Martin


From akakakhel at attbi.com  Sat Sep 14 18:43:52 2002
From: akakakhel at attbi.com (Ali K)
Date: Sat, 14 Sep 2002 22:43:52 GMT
Subject: IDLE
References: <wpbf9.264320$aA.45691@sccrnsc02> <alnb0g$2tl8$1@norfair.nerim.net>
Message-ID: <IYOg9.480794$UU1.77370@sccrnsc03>

OS: Win98
Version 2.2
double clickidle.pyw
yes




From gustav at morpheus.demon.co.uk  Sat Sep 21 15:42:07 2002
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: 21 Sep 2002 20:42:07 +0100
Subject: Python 2.2.1 compatible with Tcl/Tk 8.4?
References: <yu9965x05vkr.fsf@europa.research.att.com> <j4ptv89zwk.fsf@informatik.hu-berlin.de>
Message-ID: <u1kjuo34.fsf@morpheus.demon.co.uk>

loewis at informatik.hu-berlin.de (Martin v. L?wis) writes:

> Andrew Koenig <ark at research.att.com> writes:
> 
> > Now that Tcl/Tk 8.4 is out, can anyone tell me whether Python 2.2.1
> > will work with it?  
> 
> It appears to work fine here, and I can't see a reason why it would be
> incompatible. Notice that the setup.py of 2.2.1 is already prepared to
> search for Tcl/Tk 8.4 on its own.

Can ActivePython be persuaded to use TCL 8.4? I notice it has its own
internal TCL 8.3 DLLs etc. If I install TCL 8.4, will Python notice?

Hmm, experimenting, it looks like it doesn't. So I must have to tell
Python to use the installed TCL 8.4. How?

Thanks,
Paul.


From lpgray at uop.com  Wed Sep 18 15:50:00 2002
From: lpgray at uop.com (Lee Gray)
Date: 18 Sep 2002 12:50:00 -0700
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com>
Message-ID: <96904b50.0209181149.429e8533@posting.google.com>

lpgray at uop.com (Lee Gray) wrote in message news:<96904b50.0209171305.6d65f574 at posting.google.com>...
> I'm a fairly new programmer with experience in ASP/VBScript/DHTML.
[...] 
> Can someone point me to a good Python GUI app to demonstrate (other
> than IDLE)?

Thanks for all the suggestions, folks!

I've collected lots of links to a variety of very nice screenshots,
and I've been playing with the wxWindows demo and Boa Constructor. Boa
was a bit disappointing, not as an app, but because of its current
alpha status - but it's got great potential and I'll definitely look
forward to future releases. I'll check out PythonCard too.

Regarding .Net AND Python, that's certainly a possibility down the
road, but unfortunately I don't have the time to devote to learning
both at the moment, and if we definitely go .Net, no one (besides me)
is going to see any value in adding Python.

Thanks again... and feel free to send more ideas!
Lee


From b.maryniuk at forbis.lt  Tue Sep  3 03:52:31 2002
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Tue, 3 Sep 2002 09:52:31 +0200
Subject: Why it is so dramatical?
In-Reply-To: <20020902.104707.1789376348.1554@cybermesa.com>
References: <mailman.1030957743.4021.python-list@python.org> <mailman.1030962662.4274.python-list@python.org> <20020902.104707.1789376348.1554@cybermesa.com>
Message-ID: <200209030952.31569.b.maryniuk@forbis.lt>

On Monday 02 September 2002 05:47, Jay O'Connor wrote:
> As a result Perl is much faster at some kinds of operations than
> Python, and vice-versa.

Yeah, actually, but I just would like to find some optimal Python solution :)

-- 
Regards, Bogdan

We did it for smallpox, we'll also win over on ISO 8859-1 ... ;-)
	-- Markus Kuhn after eradicating one more ISO 8859-1 file from his disk




From brobbins333 at shaw.ca  Mon Sep 30 12:53:13 2002
From: brobbins333 at shaw.ca (brobbins333 at shaw.ca)
Date: Mon, 30 Sep 2002 16:53:13 GMT
Subject: rank-0 arrays?
Message-ID: <3d988160.79509684@news>

In Numeric Python, what are rank-0 arrays?
Must not arrays have at least 1 dimension?

BR


From stefanus.arief at ai.astra.co.id  Mon Sep 23 23:26:44 2002
From: stefanus.arief at ai.astra.co.id (Arief)
Date: Tue, 24 Sep 2002 10:26:44 +0700
Subject: confusion about opening files
In-Reply-To: <3D8FD2F3.81F3831C@alcyone.com>
Message-ID: <PLELIKCBPOPJEENEJHFBMEAGECAA.stefanus.arief@ai.astra.co.id>

>-----Original Message-----
>> O. K. Now I am confused.  I thought that when I imported the os
>> module, i
>> would have to call its functions like this: os.open()    Why am I just
>> using
>> the function name?
>
>Because there are two different open functions here.  One is in the
>builtins, one is in the os module.  They are not the same:
>
>>>> import os
>>>> print open.__doc__
>file(name[, mode[, buffering]]) -> file object
>
>Open a file.  The mode can be 'r', 'w' or 'a' for reading (default),
>writing or appending.  The file will be created if it doesn't exist
>when opened for writing or appending; it will be truncated when
>opened for writing.  Add a 'b' to the mode for binary files.
>Add a '+' to the mode to allow simultaneous reading and writing.
>If the buffering argument is given, 0 means unbuffered, 1 means line
>buffered, and larger numbers specify the buffer size.
>Note:  open() is an alias for file().
>
>>>> print os.open.__doc__
>open(filename, flag [, mode=0777]) -> fd
>Open a file (for low level IO).
>
>The os.open is a low-level API and is rarely going to be the one you
>want.

The os.open is os syscall (system call). It is used to open file, fifo,
device, etc in UNIX or Linux environment (I don't know about MS Windows).
Because it is an os syscall, so any further operation should use appropriate
os syscall too, for example:

import os
fname = '/tmp/a.txt'
fd = os.open(fname)	# open file, get file descriptor used by OS
os.write(fd, 'this is only example\n')	# further operation - write file
...
os.read(fd, 1024)		# further operation - read file 1024 bytes
...
os.close(fd)	# closing file

The built in function: "open" uses c function "fopen". It is a buffered
version of os syscall "open" implementation. You should use appropriate
function for further operation too, for example:

fname = '/tmp/a.txt'
f = open(fname, 'w')
f.write('this is only example\n')
f.close()

It is "not guaranteed" that as soon as you complete the f.write call, the
file is written with data you've feed. Because it is a buffered operation.
So to assure the OS to write your text immediately, you should do this:

fname = '/tmp/a.txt'
f = open(fname, 'w')
f.write('this is only example\n')
f.flush()		# flush buffered data for previous write operation
f.close()


I hope it would help you. Any correction is welcomed ...

-----
Arief




From phncontact.libero.it at   Fri Sep 20 03:28:52 2002
From: phncontact.libero.it at  (Benjamin)
Date: Fri, 20 Sep 2002 07:28:52 GMT
Subject: general thoughts
Message-ID: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com>

hi everybody

first: only read this post if you have enough time
(real meaning: please don't flame me just because
i don't ask any concrete questions "wasting" your time :)

i have an unusual problem with programming python,
nothing tecnical, something which goes "deeper".
you may call it inspiration, see me as an artist
who lacks of inspiration (i know there are many
people who think programming is an art).
some weeks ago i began learning python as my first
(i hope some others will follow) programming language,
i've been going through some tutorials, but they begin
to bore me. the only interesting thing (from a beginners
point of view, of course) i programmed is a small
database script which receives user input and stores
things in a txt file. that was real fun, writing code,
debugging, rewriting, analyzing, trying to figure out
the problem, in one word: coding. but now i don't know
what to programm. sounds weird, doesn't it? :) what
should i program?? at the moment i have windowsXP
installed, but i hope to switch soon to some linux
distribution or freeBSD. maybe my problem gets better
then.
i've been learning python not for becoming a programmer
(i study modern languages at the university), but for having
fun. my personal motto is "i want to know how it works".
i am always trying to deepend my computer knowledge.
maybe i should learn more about computer security and
write security stuff, or cryptography (hey i could try
to write the ultimative windows security patch that fills
all security holes in m$'s os & co. and get rich ^_^). it
certainly wouldn't be boring. i don't think i will ever write
programs for living, as a job. when i finished highschool
many people thought i will study computer science because i
were (and still am) into computers - i actually did quite
the opposite. i think if it's a job, it's something i MUST
do - and therefore not interesting anymore -, and not something
i CAN do if i want.
well, i hope i didn't bore you :) i would really like if you
share your opinions about those things with me/us.

bye, benjamin


From imbosol at vt.edu  Fri Sep 20 15:10:47 2002
From: imbosol at vt.edu (Carl Banks)
Date: Fri, 20 Sep 2002 15:10:47 -0400
Subject: average?
References: <mailman.1032541648.27654.python-list@python.org>
Message-ID: <amfscs$paa$1@solaris.cc.vt.edu>

Mark McEahern wrote:
> I know I can average a sequence like this:
> 
>  def average(seq):
>    return reduce(operator.add(seq) / len(seq))
> 
> Yes, I see that will raise ZeroDivisionError if seq == [].  Aside from that,

Appropriate, I'd say, since the average of a zero-length sequence is
undefined.


> 1.  Is there anything else dumb about the above?

Well, yes; it seems to me you meant this:

    def average(seq):
        return reduce(operator.add,seq)/len(seq)

What I can't figure out is how you got from this to what you
typed. :-)


> 2.  Am I missing a builtin cousin of min, max that would do this for me?

In pure Python, I don't think so.  I would be surprised if the above
function (when typed correctly) isn't the fastest way in pure Python.


-- 
CARL BANKS
http://www.aerojockey.com


From jggramlich at yahoo.com  Tue Sep  3 14:40:49 2002
From: jggramlich at yahoo.com (Joshua Gramlich)
Date: 3 Sep 2002 11:40:49 -0700
Subject: PyListObject & C Modules
Message-ID: <bdade4c9.0209031040.4cf4b948@posting.google.com>

Hello,

I have a problem where I need to pass a Python list to a C module, and
have a differnent list passed from the C module back to Python.

I am trying to find examples of this type of activity.  I've found the
documentation at www.python.org (the C api), but I'm a bit green (new)
and cannot read the List Objects page with any, well, usefulness. 
Thus, I cannot figure out exactly if this can or cannot be done and
what the syntax of usage of PyListObject should be...

Little help?


From mark at freelance-developer.com  Wed Sep 18 18:59:35 2002
From: mark at freelance-developer.com (Mark Nenadov)
Date: Wed, 18 Sep 2002 18:59:35 -0400
Subject: linux python ideas
References: <3D88E578.6010001@diespammerdieuselesspython.com>
Message-ID: <20020918.185929.653468858.1365@x1-6-00-d0-09-7e-c7-3e.wido3.on.cogeco.ca>

In article <3D88E578.6010001 at diespammerdieuselesspython.com>, "Rob
Andrews" <rob at diespammerdieuselesspython.com> wrote:


> I'll be giving a presentation on Python to my local LUG
> (http://lugoj.org) in a few weeks, and would like to point out at least
> a few things that would be of particular interest to linux users.  The
> idea is not "how to program in Python", which would take more than one
> quick presentation, but a one-off demonstration of ways in which Python
> is particularly useful for such a group.  Any suggestions would be
> appreciated, since my notes are pretty generic so far.

Since you aren't trying to teach the language, but rather provoke some
interest in Python, it may be a good idea to just show a few specific
libraries and problem domains in which Python can excel.

Here are some topics that might suffice:

1) Overview of open source projects that use Python
2) About GUI libraries available for Python
3) Zope
4) ReportLAB (a Python library for PDF generation)
5) Jython

~Mark (http://www.freelance-developer.com)


From jb at cascade-sys.com  Tue Sep 17 02:32:01 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 16 Sep 2002 23:32:01 -0700
Subject: How to represent sets
References: <alvd8p$1sq$1@news.hccnet.nl> <mailman.1032075137.18192.python-list@python.org> <7xn0qjacx6.fsf@ruckus.brouhaha.com>
Message-ID: <3D86CC61.3080909@cascade-sys.com>

Paul Rubin wrote:

>"James J. Besemer" <jb at cascade-sys.com> writes:
>  
>
>>For dense sets isomorphic with integers, the long integer
>>implementation probably will be superior.
>>    
>>
>
>Using immutable long integers to represent mutable sets doesn't make
>much sense.  Better to use arrays like the ones implemented by the
>array module.
>

First off, by "superior" I meant superior to a dictionary implementation.

Second, I don't see why arrays would be that much better than long 
integers.  Depends on the relative efficiency of arrays vs. long integers.

Perhaps I give the Python implementers too much credit, but I also 
expect that the implementation of long integers would be about as good 
as it gets in Python with arbitrary bit vectors.

I may be overly optimistic on this last point but even if the 
implementation of long integers is mediocre, I'd expect it to be better 
than arrays.

Given the generality of having to handle a variety of sub-types, off 
hand, I'd expect arrays to have a little more built-in overhead. 
 Furthermore, the array module has no methods for doing arithmetic on 
the entire array.  Thus to AND or OR two arrays together, you need to 
write an explicit loop or map().  I can't believe it would be as fast as 
the builtin operations for long integers.  And the work to make the 
result array the proper length would have to be spelled out in long 
hand, while it's atomic with long integers.

You seem to attach some special significance to the fact that long 
integers are "immutable" while arrays are not.  However, from the little 
bit that I've studied, the implementation is smart enough to reuse the 
erstwhile 'immutable' values of intermediate results from expressions, 
thus eliminating a pair of create/destroy operations.  I believe they 
also optimize special cases like

    bits1 |= bits2

and reuse the 'immutable' representation of the value of bits1 if the 
reference count is 1.  If not, then it should do.  In the more general 
cases, both implementations (arrays or long ints) would have to create a 
new destination result and probably free an old one.  

Maybe you're right but I just don't see why an array implementation of a 
bit set would be faster than a long integer implementation.

Do you have any benchmarks to support your argument?

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From BPettersen at NAREX.com  Mon Sep 16 17:26:05 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 16 Sep 2002 15:26:05 -0600
Subject: file mode attributes...
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201253349@admin56.narex.com>

> From: Martin v. L?wis [mailto:loewis at informatik.hu-berlin.de] 
> 
> "Bjorn Pettersen" <BPettersen at NAREX.com> writes:
> 
> > I was wondering if this was the best way of implmenting these 
> > functions (windows)?
> 
> Yes.

Thanks. Do you think there would be any interest in adding these kinds of utility functions to e.g. the os.path module?

[snip]

> Notice that, on WNT/2k/XP, you might also need to modify the 
> access control list to give users permissions, especially if 
> the files are on NTFS. Access and modification of the ACL is 
> not supported in standard Python.

win32all isn't standard on Windows <wink>? Seems like it would be pretty straightforward (although tedious) with win32security.ACL() etc. Thankfully I don't have to deal with that for this project (I just need to make sure the current process doesn't delete the files accidentally :-)

-- bjorn



From pyth at devel.trillke.net  Fri Sep 20 18:02:46 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 21 Sep 2002 00:02:46 +0200
Subject: Minimal debug/rep functionality
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8BF0A5BA@natasha.auslabs.avaya.com>; from tdelaney@avaya.com on Fri, Sep 20, 2002 at 11:18:35AM +1000
References: <B43D149A9AB2D411971300B0D03D7E8BF0A5BA@natasha.auslabs.avaya.com>
Message-ID: <20020921000246.D30315@prim.han.de>

Delaney, Timothy wrote:
> > From: holger krekel [mailto:pyth at devel.trillke.net]
> > 
> > Delaney, Timothy wrote:
> > > > From: Peter Hansen [mailto:peter at engcorp.com]
> > > > 
> > > As I said, I *expect* to be able to release (there's no 
> > commercial reason
> > > not to, and we have done so with such things in the past).
> > 
> > Maybe we are turning circles [1] but why don't you do it, then? 
> > 
> > Maybe you need to invest some time in obfuscating your
> > source code and you don't want to do it? 
> > 
> > Why do you *expect* others to solve this problem for you?
> 
> Holger, what the hell are you talking about?

Oh, i am very sorry.  I completly mixed up discussions 
and for some reason posted to the complete wrong thread :-(
I thought this still was the thread on how to protect a
commercial release of a python program ....

It was late in the night, i had a sloppy network connection
and all that. so please take my excuse,

    holger



From gerhard.haering at gmx.de  Sun Sep  1 15:13:56 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sun, 1 Sep 2002 21:13:56 +0200
Subject: Making regex suck less
In-Reply-To: <3d725881.345921@news.t-online.de>
References: <3d725881.345921@news.t-online.de>
Message-ID: <20020901191356.GA3759@lilith.ghaering.test>

* Gerson Kurz <gerson.kurz at t-online.de> [2002-09-01 18:31 +0000]:
> [...] Anyway, that got me thinking on why do we have to deal with
> regular expressions like r"((?:a|b)*)", when in most cases the code
> will look something like this:
> 
> r = re.compile("<some cryptic re-string here>")
> ...
> r.match(this) or r.find(that)

If you only use the RE once, you can use the module-level functions ;-)

> which means the real time is not spent in the compile() function, but
> in the match or find function. So basically, couldn't one come up with
> a *human readable* syntax for re, and compile that instead?

That's equally powerful? Most probably not.

> Also, I think it would already be an improvement if the syntax
> provided for clear and easy-to-understand special cases, like
> 
> re.compile("anything that starts with 'abc'")

s.startswith("abc")
s.lower().startswith("abc")

> and if you cannot find something in the special cases for you, you can
> always go back to 
> 
> re.compile("<some cryptinc re-string here>")
> 
> After all, *everyone* starting with re thinks the syntax is cryptic
> and mind-boggling, and only if you get yourself into the "re mindset",
> you understand things like r"\s*\w+\s*=\s*['\"].*?['\"]" instantly. If
> we had an easier syntax, more people would be using re ;) 
> 
> Is the idea utterly foolish? 

I don't really know. IMO if you have very simple string-searching, then
you can probably get away with the string methods, and if you have very
complex stuff, then you'll probably be better of with a parser generator
(like SimpleParse, which is very readable, IMO).

I don't find regular expressions that unreadably, especially when I
consider that I'd have to write many lines of error-prone Python code
instead. Stuff like this is just too convenient:

# working around zxDateTime limitations:
if JYTHON:
    import re

    ISO_DATE_RE = re.compile(r"(\d\d\d\d)-(\d\d)-(\d\d)")
    def DateFrom(s):
        match = ISO_DATE_RE.match(s)
        if match is None:
            raise ValueError
        return DateTime(*map(int, match.groups()))

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))



From syver-en+usenet at online.no  Tue Sep 10 13:05:18 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Tue, 10 Sep 2002 17:05:18 GMT
Subject: embedded Python sample
References: <2a0f4cc4.0209100520.1d3a723b@posting.google.com>
Message-ID: <ur8g16aiw.fsf@online.no>

monnetb at yahoo.com (bruno Monnet) writes:

> Hi,
> 
> I'm trying to call a python script from my c++ code. All I need is to
> invoke the script and check the return status.
> 
> does anyone has a simple example ? All I found until now is very
> complex and ugly code :-((

I think something like this will do (untested)

void runPythonFile(char* myfile)
{
FILE* fileptr = fopen(myfile, 'r'); 
Py_Initialize();
int result = PyRun_SimpleFile(fileptr, myfile);
Py_Finalize();
return result;
}

You should get some ideas by reading the Extending/Embedding section
in the python manual, and then read the relevant sections the Python/C api.




-- 

Vennlig hilsen 

Syver Enstad


From gerhard.haering at gmx.de  Fri Sep 13 12:11:21 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 13 Sep 2002 16:11:21 GMT
Subject: function pointer
References: <ffcedbc9.0209110215.644c1788@posting.google.com> <aln5ce$5kt$1@news.netpower.no> <slrnanu7i1.22k.gerhard.haering@haering.opus-gmbh.net> <mailman.1031895797.20045.python-list@python.org>
Message-ID: <slrnao43h8.r4.gerhard.haering@lilith.my-fqdn.de>

James J. Besemer wrote in comp.lang.python:
> Gerhard H?ring wrote:
> 
>>Broken quoting - yuck! Your newsreader is broken. Please repair it (using
>>third party software) or make Usenet a better place by throwing away that
>>crapware.
>>
> 
> I believe this error occured when the line was wrapped by the
> python-list mailer, not necessarily the submitter's newsreader or
> email program.

No, it's one of these Outlook Express bugs that are known for years at
Microsoft, but never got fixed. It happens when text of the referenced
message gets quoted and Outlook Express does its wrapping of text to
the set width later. It's most annoying to those who have to read that
broken quoting style. There are other bugs that are known equally
long, but at least they don't annoy anyone besides those that deserve
it, like:

begin  loveletterforyou.vbs
hahaha
end

No matter where OE or Outlook find this in a message, they interpret
it as an attachment, ignoring the rest of the message. This has also
been known for years, but instead of just fixing their email parsing,
they wrote a laughable MSDN article that recommends to not use
"begin  " in messages.

>>If we want to keep the terminology right, Python doesn't have variables. It
>>has names and values. 
> 
> The fact that The Tutorial and other Python doc uses the term "variable" 
> makes it confusing for newcomers.

Sorry, I probably was too harsh about this. It's just that using the
terms 'name' and 'value' instead of variable made it a lot simpler for
me to understand the concepts of the Python data model. In fact, I
think I got them from the Python Language Reference, which explains
the Python data model using the terms:

- objects, values and types
- identifiers (names)

http://www.python.org/doc/current/ref/objects.html
http://www.python.org/doc/current/ref/atom-identifiers.html

The reason why I objected to the term variable was that (at least for
me) it has the connotation of strong typing, where a variable can only
hold values or references to objects of (subclasses of) the declard
type. Dismissing the use of 'variable' to describe Python behaviour
was premature, though. It certainly has its uses, it's just that it is
IMO less accurate than the terms from the Python Language Reference.

-- Gerhard


From christophe.delord at free.fr  Mon Sep  9 13:35:42 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Mon, 9 Sep 2002 19:35:42 +0200
Subject: random.shuffle question
References: <3k4f9.7500$yt3.3831288@newssrv26.news.prodigy.com>
Message-ID: <20020909193542.08c6937a.christophe.delord@free.fr>

It works but not as you may expect.

l[4:8] is another list made with l[4] to l[7] and changing l[4:8] does not change l. For example :

>>> l = [1,2,3,4,5,6,7,8,9]
>>> l2 = l[4:8]
>>> l2[2] = 99
>>> l2
[5, 6, 99, 8]
>>> l
[1, 2, 3, 4, 5, 6, 7, 8, 9]

So when you shuffle l[4:8] you shuffle a copy of a sublist of l. In the previous example you can consider that l2 is the argument given to shuffle.

One solution is to shuffle the sublist and copy the sublist in the list:

>>> l = [1,2,3,4,5,6,7,8,9]
>>> l2 = l[4:8]
>>> random.shuffle(l2)
>>> l[4:8] = l2
>>> l
[1, 2, 3, 4, 6, 7, 8, 5, 9]


But this may be inefficient on larger lists.


Christophe.


On Mon, 09 Sep 2002 16:51:11 GMT
"dsavitsk" <dsavitsk at e-coli.net> wrote:

> why does this not work to shuffle part of a list?
> 
> >>> l = [1,2,3,4,5,6,7,8,9]
> >>> random.shuffle(l[4:8])
> >>> print l
> [1, 2, 3, 4, 5, 6, 7, 8, 9]
> 
> and more importantly, is there a better way to do this?
> 
> -d
> 
> 


-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From mertz at gnosis.cx  Mon Sep 16 09:52:54 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Mon, 16 Sep 2002 09:52:54 -0400
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de>
 <pUdh9.9377$yt3.4604908@newssrv26.news.prodigy.com>
Message-ID: <2Ieh9kKkXMhU092yn@gnosis.cx>         

"dsavitsk" <dsavitsk at e-coli.net> wrote previously:
|If your code is that valuable, run it on your server remote from end
|users as a trade secret, or force them to agree to a license and protect
|it w/ copyright.

I would reformulate the issue slightly:  If you BELIEVE your code is
that valuable, seek psychiatric help.

Yours, Lulu...

--
 mertz@   _/_/_/_/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY:_/_/_/_/ v i
gnosis  _/_/                    Postmodern Enterprises         _/_/  s r
.cx    _/_/  MAKERS OF CHAOS....                              _/_/   i u
      _/_/_/_/_/ LOOK FOR IT IN A NEIGHBORHOOD NEAR YOU_/_/_/_/_/    g s




From wesc at deirdre.org  Thu Sep  5 14:40:07 2002
From: wesc at deirdre.org (Wesley J. Chun)
Date: 5 Sep 2002 11:40:07 -0700
Subject: ANN: BayPIGgies mtg Wed 9/11 7:30pm
Message-ID: <fcd9edc1.0209051040.5a06b606@posting.google.com>

BayPIGgies: Silicon Valley-San Francisco Bay Area Python Users Group

Agenda:   Using a New MVC Architecture to Create a
          Pythonic XML-based Web Application Framework
When:     September 11, 2002 @ 7:30pm
Where:    Stanford University, Palo Alto, CA
Speakers: Paul McGavin, Donovan Preston, Sam Penrose from InterSight

We will discuss how we are applying a new Python-based MVC
architecture to create an efficient, Pythonic web application
framework that separates presentation templates from the Python
source code with DOMTemplate. The framework componentizes behavior
into reusable objects with DOMWidgets and DOMHandlers.

After creating web applications in three separate Python frameworks
-- Webware for Python, Zope and Apache/Python cgi -- InterSight
decided to create its own MVC framework, based on some of the ideas
from IBM and Lutris Enhydra's Barracuda project, to power their
automated publishing applications. Their applications enable scalable
content: the ability to produce, from a single set of managed data
and images, many different marketing communication pieces: press-
ready Adobe InDesign and QuarkXPress files, web pages, order sheets,
sample stickers, price tags and more.

We are actively seeking speakers for BayPIGgies, esp. October 9th!
If you would like to give a talk at one of our meetings (any Python-
related topic), please contact us!

more info including directions:    http://www.baypiggies.net

hope to see some of you on Wednesday!

-wesley

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

"Core Python Programming", Prentice Hall PTR, ? 2001
    http://starship.python.net/crew/wesc/cpp/

Silicon Valley-San Francisco Bay Area Python Users Group (BayPIGgies)
    http://www.baypiggies.net

wesley.j.chun :: wesc at deirdre.org
cyberweb.consulting : henderson, nv : cyberweb at rocketmail.com
http://roadkill.com/~wesc/cyberweb/


From duncan at NOSPAMrcp.co.uk  Mon Sep  2 11:19:44 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Mon, 2 Sep 2002 15:19:44 +0000 (UTC)
Subject: Newbie's unsurmountable pb with IO stream trick
References: <aba30b75.0209020633.26ba4c70@posting.google.com>
Message-ID: <Xns927DA446C54D0duncanrcpcouk@127.0.0.1>

spendius at muchomail.com (Spendius) wrote in 
news:aba30b75.0209020633.26ba4c70 at posting.google.com:

> So I have a challenge for you guys who are enough acquainted
> with Python: how could I perform the same with this language ?

What operating system are you running on? I believe there are ways to do 
this on unix or on windows, but they are totally different.

Do you really need to access your database through this frontend program? 
Your life might be much easier if you connected to the database directly.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From noah at noah.org  Mon Sep  9 16:08:00 2002
From: noah at noah.org (Noah)
Date: 9 Sep 2002 13:08:00 -0700
Subject: Integer overflow in fcntl on some Linux platforms. What's going on?
Message-ID: <c9d82136.0209091208.5987de97@posting.google.com>

Hello,

Can anyone help me figure out what the problem is here?
On some Linux platforms when I try to execute the following code:

    s = struct.pack("HHHH", 24, 80, 0, 0)
    x = fcntl.ioctl(sys.stdout.fileno(), termios.TIOCSWINSZ, s)

I get this exception:

    'signed integer is greater than maximum'

It seems that Linux does not like the integer from sys.stdout.fileno(), but 
this has worked fine on all the BSD machines and one Linux 2.2.20 machine 
I tested it on. So far only RedHat 7.1 has this problem.
Or maybe it does not like the constant termios.TIOCSWINSZ (set win size), yet
it has no problem with the similar constant termios.TIOCGWINSZ (get win size).

The following is a test script that demonstrates this problem. 
It logs all steps to a file called LOG. 
It crashes after STEP 5 inside getwinsize() at STEP getwinsize() 2. 
This line is the ioctl() call. This is the results I get in my LOG file:
        STEP 1: pid: 274968
        STEP 2: stdout.fileno(): 1
        STEP 3: stdout.fileno() hex: 1
        STEP 4: termios.TIOCSWINSZ: 2148037735
        STEP getwinsize() 1.
        STEP getwinsize() 2.
        STEP getwinsize() 3.
        STEP 5: getwinsize() #1: (54, 80, 0, 0)
        STEP setwinsize() 1.
        STEP setwinsize() 2.
        EXCEPTION: signed integer is greater than maximum

Any help would be appreciated.

Yours,
Noah

#!/usr/bin/env python
import os, sys, termios, fcntl
import struct

def setwinsize():
        # Assume ws_xpixel and ws_ypixel are zero.
        log ('STEP setwinsize() 1.')
        s = struct.pack("HHHH", 24, 80, 0, 0)
        log ('STEP setwinsize() 2.')
        x = fcntl.ioctl(sys.stdout.fileno(), termios.TIOCSWINSZ, s)
        log ('STEP setwinsize() 3.')

def getwinsize():
        log ('STEP getwinsize() 1.')
        s = struct.pack("HHHH", 0, 0, 0, 0)
        log ('STEP getwinsize() 2.')
        x = fcntl.ioctl(sys.stdout.fileno(), termios.TIOCGWINSZ, s)
        log ('STEP getwinsize() 3.')
        return struct.unpack("HHHH", x)

def log (s):
        fout = open ('LOG','a+')
        fout.write(s)
        fout.write('\n')
        fout.close()

pid = os.getpid()

log ('STEP 1: pid: ' + str (pid) )
log ('STEP 2: stdout.fileno(): ' + str(sys.stdout.fileno()) )
log ('STEP 3: stdout.fileno() hex: %x' % sys.stdout.fileno() )
log ('STEP 4: termios.TIOCSWINSZ: ' + str(termios.TIOCSWINSZ) )
try:
        log ('STEP 5: getwinsize() #1: ' + str (getwinsize()))
        setwinsize()
        log ('STEP 6: getwinsize() #2: ' + str (getwinsize()))
except Exception, e:
        log ('EXCEPTION: ' + str(e))
        sys.exit(0)

print
print 'Log file written. It is called "LOG".'
print
print open('LOG').read()


From loewis at informatik.hu-berlin.de  Fri Sep 27 07:46:10 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 27 Sep 2002 13:46:10 +0200
Subject: py2exe Error creating NT Service
References: <an1epm$rc5$03$1@news.t-online.com>
Message-ID: <j4ptuz4pvh.fsf@informatik.hu-berlin.de>

"Harald Schneider" <h_schneider at marketmix.com> writes:

>     exec compile(open(self.script).read(), self.script, "exec") in
> mod.__dict__
>   File "<string>", line 101
> 
> SyntaxError: invalid syntax
[...]
> Any ideas?

I would print self.script, and look at line 101 of the file named by
self.script.

Regards,
Martin



From joconnor at cybermesa.com  Thu Sep 26 02:09:41 2002
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Thu, 26 Sep 2002 13:09:41 +0700
Subject: How do you pronounce "__" (double underscore)?
References: <525eaba.0209260956.72bf6786@posting.google.com> <3D934C5C.4000603@mxm.dk>
Message-ID: <20020926.130938.1789366143.1524@cybermesa.com>

In article <3D934C5C.4000603 at mxm.dk>, "Max M" <maxm at mxm.dk> wrote:

> Pat Notz wrote:
>> @ = "at"
>> ! = "bang"
>> . = "dot"
>> * = "star"
>> 
>> Is there a short (one or two syllable) pronounciation for "__" (double
>> underscore).  "Underscore underscore" and "double underscore" are kind
>> of long.  Somthing like "splat"?  Anyone?
> 
> _ = 'under'
> 
> __ = 'under-under'


How about "underder"??

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.r4h.org/r4hsoftware


From marklists at mceahern.com  Thu Sep 26 09:25:35 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 26 Sep 2002 08:25:35 -0500
Subject: Is there a way to find the name of the method currently being executed?
In-Reply-To: <3D9278D0.9060307@arielpartners.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOIELKEBAA.marklists@mceahern.com>

import sys

def func1():
  print "my name is %s" % myname()

def myname():
  steps_back = 1
  return sys._getframe(steps_back).f_code.co_name

func1()



From jeff at sasmor.com  Sun Sep  8 16:58:48 2002
From: jeff at sasmor.com (Jeff Sasmor)
Date: Sun, 08 Sep 2002 20:58:48 GMT
Subject: Zope Status?
References: <Xns9281221A07167bosahvnetscapenet@66.185.95.104>
Message-ID: <cSOe9.3024$io.1100@nwrddc03.gnilink.net>

Speaking from what I understand about this subject,  it's not
simple for the Zope Corp to merely plug-in a new version of
Python and make a release.  Disregarding any issues about
changes they might want to make that would take advantage of
new features of Python 2.2 (and up), the underlying instrumentality
of Zope, the ZODB (Zope Object DataBase) as it exists in  Zope 2.5
and so on, will not run under Python versions > 2.1 due to
incompatibilities of the C extension classes with the changes made
(in Python 2.2) regarding type-class unification.

Obviously, there's already a version of ZODB for Python 2.2, but
I think it's not something that one cannot just 'drop in' to a current
version of Zope.

That's just one thing. Also, Zope 3 is a clean-slate redesign; that is,
it is being rearchitected with no initial regard for existing Zope
applications.  I suppose opinions go either way on this, but overall
it is probably a Good Thing.  I recall reading that there will eventually
be backwards compatibility, perhaps with some interface layer, but
I suspect that the initial releases will be for very advanced developers.

BTW, I don't speak as a member of the Zope team, just someone who's
been lurking on all the dev lists for Zope  3 and the ZODB as I am
developing an app based on the ZODB and need to track what these
folks are doing.

Apologies in advance to Zope Corp for any inaccuracies.
jms
--
#--------------------------------
Jeff Sasmor
jeff at sasmor.com
"Bo" <bosahvremove at netscape.net> wrote in message
news:Xns9281221A07167bosahvnetscapenet at 66.185.95.104...
> I apologise if this isn't the proper forum to do this in.
>
> 1) Can anyone tell me roughly if\when Zope is moving to python 2.2?
>
> 2) Also any idea when they might release Zope 3?
>
> 3) Anyone close to the Zope team who is willing to shed some light on the
> near future for things Zope?
>
>
> Great thanks in advance.
>
> Bo




From see_reply_address at something.invalid  Sun Sep 22 21:44:42 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Mon, 23 Sep 2002 13:44:42 +1200
Subject: Python about file I/O
References: <amgnnl$5qc9n$1@ID-49758.news.dfncis.de> <3d8da3df@nntp.server.uni-frankfurt.de>
Message-ID: <3D8E720A.90501@something.invalid>

Michael Janssen wrote:

> existence-check is:
> os.access(file,os.F_OK)


But it's generally considered preferable to just go
ahead and try to open the file, and catch the exception
if it fails. Some reasons for preferring this approach
include:

   * It will catch all possible reasons for failing
     to open the file, not just lack of existence or
     permissions.

   * It's not prone to something happening in between
     testing and opening which causes the file to
     no longer be openable.

   * It's more efficient -- only one file system
     operation instead of two.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From sholden at holdenweb.com  Sat Sep  7 14:33:07 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 7 Sep 2002 14:33:07 -0400
Subject: encryption (passwords)
References: <3d75d115$0$221$4d4ebb8e@news.nl.uu.net> <slrnanblej.1rs.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <aFre9.16609$l_4.13261@atlpnn01.usenetserver.com>

"Gerhard H?ring" <gerhard.haering at opus-gmbh.net> wrote in ...
> Gumuz wrote:
> > I am quite a newbie regarding encryption stuff. There are some issues
> > boggling my head and I suppose these come from the fact that I have no
> > clue how to apply encryption in a real-life situation.
> >
> > I am using network-sockets to communicate between two(or more) python
> > applications. I want to encrypt these messages, or at least a part of it
> > e.g. password and such.
>
> Ok. I'd recommend to use SSL instead of inventing your own crypo layer.
I'd
> probably go with PyOpenSSL.
>
I'd agree that implementing your own encryption technique is unlikely to be
secure. Use the work that's already been done.

> > My limited understanding of encryption tells me that I need a 'key' to
> > decrypt stuff.
>
> There are two kinds of encryption:
> - symmetric encryption
> - assymemtric encryption
>
> Symmetric works by both ends having the same key. The sender encrypts the
> message with the key, the receiver decrypts it using the same key. The
> problem is that you have to distribute the keys first, and have a secure
> way of distributing the keys.
>
In other words, a second (secure) channel is required to comunicate the key
to be used on the first channel. Fortunately teh second secure channel can
be the same channel but using public-key (asymmetric) encryption.

> Asymmetric encryption works by splitting the key up into a private and a
> public key. The private key can only be used for decrypting, the public
key
> only for encrypting messages.
>
Not strictly true, it depends on the requirements. Aysmmetric encryption can
be used to provide three different features:

1. Privacy. In this case the sender encrypts the message with the
recipient's public key. Since it can now only be decrypted by the holder of
the corresponding private key, the assumption is that only the intended
recipient will be able to decrypt it. This is the case Gerhard describes.

2. Authentication. Here the sender encrypts the message with their own
private key. Anyone with access to the sender's public key can decrypt it,
and since the decryption process gave a legible message it is reasonable to
assume that it really was encrypted by the holder of the corresponding
private key.

3. Integrity. Here the sender generates a message digest (some function
computed from the content of the message, ideaaly using a recognized
cryptographic hashing function). The digest is of fixed length, and with a
"good" hashing function it won't be possible to generate arbitrary message
digests with a particular digest value. The sender then encrypts the digest
value with their private key, producing what is normally calles a "digital
signature". The recipient decrypts the digest with the sender's public key,
and independently verifies (by generating another message digest locally)
that the message they received has the same digest value as that inserted by
the sender. If this is not the case then either the content or the digest
has been tampered with, and the message content should not be trusted.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                        pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From eric.brunel at pragmadev.com  Thu Sep  5 04:17:31 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Thu, 5 Sep 2002 10:17:31 +0200
Subject: How do i print to a printer using python?
References: <6afb9d2c.0209041044.43c5f5f1@posting.google.com>
Message-ID: <al73gj$lbt$1@wanadoo.fr>

Naveed Iqbal wrote:
> I have to print to a standard printer. I take the data from the
> keyboard and print it using the printer. Please help me....my job
> depends on this!!!!

Things are quite different on Unix and Windows. Just a sample (untested) 
program to print "Hello world!" to the default printer on both platforms:

----Unix----------------------
import os
p = os.popen('lpr', 'w')
p.write("Hello world!\n")
p.close()
------------------------------

----Windows-------------------
import win32ui

dc = win32ui.CreateDC()
dc.CreatePrinterDC()
dc.SetMapMode(4)        # This is UI_MM_LOENGLISH
## With this map mode, 12 points is 12*100/72 units = 16
font = win32ui.CreateFont({'name' : 'Arial', 'height' : 16})
dc.SelectObject(font)

dc.StartDoc("HelloWorld")
dc.StartPage()
dc.TextOut(40, 40, "Hello world!")
dc.EndPage()
dc.EndDoc()
------------------------------

And now you wish you were on Unix ;-)!

For more information on the Windows version, see the documenattion for 
win32ui @ 
http://aspn.activestate.com//ASPN/Python/Reference/Products/ActivePython/PythonWin/modules.html 
and Micro$oft MFC documenation @ 
http://msdn.microsoft.com/library/default.asp

Good luck!

PS: I'm actually a little biased towards Unix... In fact, if you have the 
printer directly connected to your PC, opening LPT1: and just writing the 
text to it should work. But if it's a network printer, the only solution is 
the code above...

We also can make things a little more complicated on Unix by using 
PostScript. But it's still shorter:

import os
p = os.popen('lpr', 'w')
p.write('%!PS\n')
p.write('/Helvetica findfont 12 scalefont setfont\n')
p.write('/cm { 72 mul 2.54 div } def\n')
p.write('1 cm 28 cm moveto\n')
p.write('(Hello world!) show\n')
p.write('showpage\n')
p.close()
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From sholden at holdenweb.com  Wed Sep 18 12:25:22 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 18 Sep 2002 12:25:22 -0400
Subject: Loading modules from several directories
References: <a1247c5b.0209171610.243f3c02@posting.google.com> <33803989.0209172247.2e6afc1f@posting.google.com> <a1247c5b.0209180607.843526e@posting.google.com>
Message-ID: <6Q1i9.1149$bh1.989@fe03>

"Eugene Pervago" <eugene at ulka.com> wrote ...

[...]
> If I put the project_a path first, it loads its modules but can't
> import the /onedir/foo modules. It seems package can only be in one
> place, once Python finds it, it just stops looking elsewhere.
>

What would you rather have it do? Load them both? Find every possible
package and load the oine with the most recent modification date/time? Load
the longest one? Load the one whose checksum is closest to your birthdate?

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From whisper at oz.net  Mon Sep  9 19:28:09 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 9 Sep 2002 16:28:09 -0700
Subject: a re question
In-Reply-To: <pan.2002.09.09.16.10.38.482129.20940@presidency.com>
Message-ID: <GCEDKONBLEFPPADDJCOEIECEEOAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Rajarshi Guha
> Sent: Monday, September 09, 2002 13:11
> To: python-list at python.org
> Subject: a re question
>
>
> Hi,
>   I have a file with lines of the format:
>
> 001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23
> 001 Xya FGh   143557789   7 100 09/05/2002 20:23:23
>
> I am trying to extract the 9 digit field and the single digit field
> immediatley after that.
>
> When I use Visual Regexp to try out the regexp
>
> (\d{9,} {3,}\d)
>
> it highlights the 2 fields exactly.
>
> But when I use the following Python code I get None:
>
> >> s='001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23'
> >> p = re.compile(r'(\d{9,} {3,}\d)')
> >> print p.match(s)
> >> None
>
> Could anybody point out where I'm going wrong?
>
> Thanks,

Yes - match always starts looking for a match at the beginning of the line -
try p.search instead. See python22/doc/lib/matching-searching.html

I would also suggest using \s{3, } instead of what you have for clarity in
skipping the spaces between the 9 digit number and the single digit number.

You might find that putting parens around each element you want to extract a
convenient shortcut; then if you do elts = p.search(s), elts.group(1) will
be the 9 digit number and elts.group(2) will be the single digit. The
pattern would then look like:

r'(\d{9, })\s{3, }(\d)'

you could also do num9, num1 = elts.group(1,2) if you're in a hurry ;)

HTH,

Dave LeBlanc
Seattle, WA USA




From josh_py at sautez.com  Tue Sep 17 13:15:19 2002
From: josh_py at sautez.com (josh_py at sautez.com)
Date: Tue, 17 Sep 2002 10:15:19 -0700
Subject: using python as cgi on apache unix
References: <am6fm301477@enews2.newsguy.com> <orHh9.167$UJ1.17889587@newssvr14.news.prodigy.com>
Message-ID: <am7nv0075d@enews2.newsguy.com>

"Dennis Reinhardt" <DennisR at dair.com> wrote in message
news:orHh9.167$UJ1.17889587 at newssvr14.news.prodigy.com...
> > ... want to move this to my unix server.  Unfortunately, no matter what
I
> > do I can not get apache on my linux box to execute any python scripts.
I
> > get the following errors in apache:
> > Premature end of script headers: /var/www/html/cgi-bin/test.py
>
> What do you get when you type "test.py" at the shell prompt when logged
into
> the cgi directory on the RH box at the command line?  Do you get:
> ---
> ---
> Content-Type: text/html
> ---
> hello
>
> where "---" is taken as a blank line?  I agree with Gerhard that the
initial
> print "\n\n" is wrong.  But your comment trying the cgi.test suggests that
> something else is going on.  The above test may sniff out that something
> else.
>
>
> Dennis Reinhardt
>
> http://www.dair.com
>
>

I do get the correct results.  I also removed the blank lines and it still
didnt' work.

I have now moved the script to yet another machine and now it works.  I must
have Apache configured in such a way on the one machine that it is confused.
It knows the .py are executible and such but does not execute them
correctly.  There is a problem with how apache is seeing the file and
executing the program.  I will have to review the httpd.conf file in both
cases to try to figure it out.

Thanks for the suggestion and pointing out my errors with the print
statements.

Josh




From muk_pr at yahoo.com  Wed Sep 11 18:50:24 2002
From: muk_pr at yahoo.com (vb)
Date: 11 Sep 2002 15:50:24 -0700
Subject: Destructor never called ???
Message-ID: <76710511.0209111450.40999dd8@posting.google.com>

Hi, 
  I have the following code my problem is when I call any methods of  
       TCVector_Object the destructor never executes.

I mean

v = module.CreateTCVector()
del v;
//destructor runs. message shown .OK

but 
in

v = module.CreateTCVector()
v.Append(123);
del v;//or closing python interpreter

destructor never gets called. Any ideas?

BTW are there any extending&embedding docs (except ASPN & Python docs)
I realy have hard times in that ref. counting stuff and
is my way the official way of adding methods to a new type ?

Thanks.



headers.............

staticforward PyTypeObject TCVector_Type;
staticforward PyMethodDef TCVector_methods[];

struct TCVector_Object{
    PyObject_HEAD
    vector<int> List;
};

static PyObject*
new_TCVector(PyObject* self, PyObject* args)
{
    TCVector_Object *obj;

    if (!PyArg_ParseTuple(args,""))
        return NULL;

    obj = PyObject_New(TCVector_Object, &TCVector_Type);
    return (PyObject*)obj;
}

static void
delete_TCVector(PyObject* obj)
{
    ShowMessage("Deallocating");
    PyObject_Del(obj);
}

static PyObject *
TCVector_getattr(PyObject *obj, char *name)
{   //Methods will be looked here.
    AnsiString attrname = name;
    if (attrname == "Count")
    {
        return Py_BuildValue("i",((TCVector_Object*)obj).size());
    }
    else
        return Py_FindMethod(TCVector_methods, obj, name);
}

static PyObject *
TCVector_Append(PyObject* self, PyObject* args)
{
  int i;
  if (!PyArg_ParseTuple(args,"i",&i))
  {
    return NULL;
  }
  ((TCVector_Object*)self).push_back(i);
  Py_INCREF(Py_None);
  return Py_None;
}

static PyObject *
TCVector_Clear(PyObject* self, PyObject* args)
{
  ((TCVector_Object*)self).clear();
  Py_INCREF(Py_None);
  return Py_None;
}

static PyTypeObject TCVector_Type = {
    PyObject_HEAD_INIT(&PyType_Type)
    0,
    "TCVector",
    sizeof(TCVector_Object),
    0,
    delete_TCVector, /*tp_dealloc*/
    0,               /*tp_print*/
    TCVector_getattr,          /*tp_getattr*/
    0,          /*tp_setattr*/
    0,          /*tp_compare*/
    0,          /*tp_repr*/
    0,          /*tp_as_number*/
    0,          /*tp_as_sequence*/
    0,          /*tp_as_mapping*/
    0,          /*tp_hash */
};

static PyMethodDef TCVector_methods[] = {
    {"Append", (PyCFunction)TCVector_Append, METH_VARARGS,""},
    {"Clear", (PyCFunction)TCVector_Clear, METH_VARARGS,""},
    {NULL, NULL, 0, NULL}           /* sentinel */
};


static PyMethodDef samplemodule_methods[] = {
    {"CreateTCVector", new_TCVector, METH_VARARGS,
     "Create a new TCVector object."},
    {NULL, NULL, 0, NULL}
};

void __export __stdcall initsamplemodule()
{
    Py_InitModule("samplemodule", borland_tools_methods);
}


From agenkin at cdf.toronto.edu  Fri Sep 13 15:37:40 2002
From: agenkin at cdf.toronto.edu (Arcady Genkin)
Date: Fri, 13 Sep 2002 19:37:40 GMT
Subject: Generating Win95 (lanman) password hashes with Python
Message-ID: <uo7khpsmt7.fsf@bashful.cdf.toronto.edu>

I need to generate Win95 password hashes from a Python script.  The
PyCrypto module (http://www.amk.ca/python/writing/pycrypt/) provides
DES class:

http://www.amk.ca/python/writing/pycrypt/pycrypt.html#SECTION000400000000000000000

that might be able to do what I need.  The hash that I need is the
following:

   LAN Manager Password Hash

   This is a 32-bit hexadecimal sequence that represents the password
   Windows 95 and 98 clients will use. It is derived by encrypting the
   string KGS!@#$% with a 56-bit DES algorithm using the user's
   password (forced to 14 bytes and converted to capital letters)
   twice repeated as the key. If there is currently no password for
   this user, the first 11 characters of the hash will consist of the
   sequence NO PASSWORD followed by X characters for the
   remainder. Anyone can access the share with no password. On the
   other hand, if the password has been disabled, it will consist of
   32 X characters. Samba will not grant access to a user without a
   password unless the null passwords option has been set.

My first problem is that I don't understand how to get two 8-byte
strings from the 14-character password.  I'm guessing that the
password is padded with something.

The second problem is the "mode" parameter that I need to pass to the
DES.new() function.  Any idea what I might need?  The dir() on DES
gives the following available modes:

'MODE_CBC', 'MODE_CFB', 'MODE_CTR', 'MODE_ECB', 'MODE_OFB', 'MODE_PGP'

Any help would be appreciated.
-- 
Arcady Genkin : CDF sysadmin : http://cdf.toronto.edu/~agenkin/contact.html

guilt is the cause of more disauders
than history's most obscene marorders                       (E.E. Cummings)


From pecora at anvil.nrl.navy.mil  Sun Sep 29 09:05:02 2002
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Sun, 29 Sep 2002 09:05:02 -0400
Subject: [Mac] Editor on Macintosh?
References: <1fj7za2.1erbqps1uvzpj8N%news@agapow.net> <B9BC89EF.12189%jonathan@onegoodidea.com>
Message-ID: <290920020905027300%pecora@anvil.nrl.navy.mil>

>From the looks of it this is OS X only, right?  -- Lou

In article <B9BC89EF.12189%jonathan at onegoodidea.com>, Jonathan Hogg
<jonathan at onegoodidea.com> wrote:

> I'm quite happy with BBEdit.
> 
>     <http://www.bbedit.com/>
> 
> One feature I'm growing particularly fond of is the new worksheet windows.
> These are like MPW worksheets if you're familiar with MPW (Ah! thems were
> the days). Basically you get an editor window that has a live shell
> underneath it. By hitting the ENTER key instead of return, the last line
> typed (or currently selected text) is executed and the stdout inserted at
> the cursor point in the window.
> 
> This isn't so useful for very interactive stuff and I wouldn't use it as a
> substitute for a terminal window, but for recording and monitoring the
> output of programs it's fantastic. Basically I open a new worksheet window
> and type something like this at the top:
> 
>     cd /to/project/directory
>     ./start-program --non-interactive --with-lots-of-debugging-output
> 
> Then I just select the lines, hit ENTER, and watch the output fill the
> window. Then I can scroll up and down and use BBEdit's grep searching
> facilities on the output as it's running.
> 
> Yes, I'm sure Emacs can do all of this, but I don't want a terminal-based
> editor pretending to be a GUI application.
> 
> Oh, another thing. BBEdit has a command-line tool now for firing up the
> editor allowing you to type:
> 
>     % bbedit foo.py bar.py
> 
> and have them open up in BBEdit.
> 
> Jonathan (no stock options, just a happy user)


From claird at starbase.neosoft.com  Thu Sep  5 14:49:49 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 5 Sep 2002 13:49:49 -0500
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Sep  4)
References: <448A0F008E0AFE11.19B9EF86BEE2B191.107D5020B234F743@lp.airnews.net> <al5tvu$1mtjjr$1@ID-71831.news.dfncis.de> <al5ue4$sj2$1@peabody.colorado.edu> <al5ugn$sj2$2@peabody.colorado.edu>
Message-ID: <D4DFCC30D8BE1773.7E547BD5FB183F60.A4EA187BF6686FA3@lp.airnews.net>

In article <al5ugn$sj2$2 at peabody.colorado.edu>,
Fernando =?ISO-8859-1?Q?P=E9rez?=  <fperez528 at yahoo.com> wrote:
>Fernando P?rez wrote:
>
>> James Kew wrote:
>> 
>>> I can't persuade Google to find this message-id; any chance of a link?
>> 
>> google_groups('comp.lang.python anansispaceworks.com')[0]
>
>Oops. Make that 
>
>google_groups('comp.lang.python anansispaceworks.com scientific')[0]
			.
			.
			.
Short form:  look for Hancock at <URL: http://
groups.google.com/groups?th=8acf74098f129cf3 >.

And accept my apology for failing to resolve all the
references properly before sending out the "Python-URL!".
-- 

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


From Maric.MICHAUD at cirec.com  Tue Sep 17 08:56:57 2002
From: Maric.MICHAUD at cirec.com (Maric MICHAUD)
Date: Tue, 17 Sep 2002 14:56:57 +0200
Subject: PyQT License
Message-ID: <27E327AB8201B54495D54105FBBBB1485D205A@SRVMSG.GERLAND2.local>

I did not really understand terms of Qt and PyQt license.
Is it GPL for both Win32/UNIX platforms ?



From marklists at mceahern.com  Fri Sep 27 18:17:18 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 27 Sep 2002 17:17:18 -0500
Subject: watching mutables?
In-Reply-To: <JHEOKEOOLIGLDHCMAHMOEECOECAA.marklists@mceahern.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOIECPECAA.marklists@mceahern.com>

[I apologize for not providing context, but I've already deleted my copy of
the specific message I'm replying to.]

The reason a metaclass solution is not possible without modifying the
original code has to do with how the metaclass is determined:

  http://www.python.org/2.2.1/descrintro.html#metaclasses

  1.  First, Python looks for '__metaclass__' within the class' dict.
  2.  Python uses the metaclass of any base class.
  3.  Python looks for a global __metaclass__.
  4.  Python uses the classic metaclass types.ClassType.

With builtin lists and dicts, Python never gets past step 2, so you have to
intervene forcefully with step 1 if you want to override the metaclass of a
builtin class, which means you have to first subclass it:

  class WatchedDict(dict):

    __metaclass__ = Watched

// m




From claird at starbase.neosoft.com  Wed Sep 18 08:34:38 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 18 Sep 2002 07:34:38 -0500
Subject: Facilities for accounting arithmetic (was: Python GUI app to impress the boss?)
References: <mailman.1032311543.27487.python-list@python.org>
Message-ID: <A169FAA5F4815882.8AAD59FBFA694D4B.34D4691F49E031FC@lp.airnews.net>

In article <mailman.1032311543.27487.python-list at python.org>,
terry  <tg5027 at citlink.net> wrote:
>
>>>?I'm willing to bet, though, that Python will be around for longer
>>> than VBScript.
>
>Cameron,
>
>I don't like to disagree with an icon, but I think that you are missing 
>a deficiency in Python that will ultimatly limit it's usefullness in 
>the vast majority of the software that is written in the world.
>
>As long as Python refuses to admit that money is the focus of most of 
>the world's software, Python will be relagated to the same status as 
>FORTRAN, BASIC, and etc - except COBOL and VB.
>
>Is there any way to get the Python world to allow for the processing of 
>monetary applications without some hokey 60's constructs that involve 
>in-line conversions or anticipations of calculation characteristics?
			.
			.
			.
If it helps, I have a soft spot for OS/400.  I spent the
early '80s writing BCD arithmetic packages in various
assemblers.

Terry, who needs disagreement more urgently than an icon?

You and others have raised this point before.  I haven't
taken it as seriously as I should, just because I know
that Microsoft and Sun (among others) are utterly unsympa-
thetic to it.  I need to ponder that ... anomaly (?) more.
Thanks for making an issue of it.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From stephen.boulet at motorola.com  Fri Sep 27 17:47:09 2002
From: stephen.boulet at motorola.com (Stephen Boulet)
Date: Fri, 27 Sep 2002 16:47:09 -0500
Subject: Better terminal for windows?
References: <3D94B037.2020202@motorola.com> <mailman.1033158756.25404.python-list@python.org>
Message-ID: <3D94D1DD.40407@motorola.com>

Gerhard H?ring wrote:
> * Stephen Boulet <stephen.boulet at motorola.com> [2002-09-27 14:23 -0500]:
> 
>>I'm using python 2.2.1 on winnt.
>>
>>If I want to run python interactively (and perhaps using ipython), is 
>>there something better than the "cmd" terminal?
> 
> 
> Well, you could install MSYS from the mingw project, which is sort of a
> Cygwin-lite :-)

Thanks. I'll have a look at it.

>>I'm used to bash on linux at home, but I'd like to avoid installing 
>>cigwin for now. Thanks.
> 
> 
> cmd.exe isn't *that* bad once you activate tab completion in the
> registry. Have you done that?

Nope. How do you do that? I couldn't figure it out just looking through 
regedit.

> 
> 4NT is also great, but it's shareware.
> 
> -- Gerhard
> 

I've come to the conclustion that the python shell in idle really isn't 
all that bad. Maybe I'll just stick with that.

-- Stephen



From jepler at unpythonic.net  Wed Sep 11 22:40:13 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Wed, 11 Sep 2002 21:40:13 -0500
Subject: iterators and generators, is this the Python Way???
In-Reply-To: <3D7FF921.2040305@one.net>
References: <3D7FF921.2040305@one.net>
Message-ID: <20020911214013.A1795@unpythonic.net>

On Wed, Sep 11, 2002 at 10:17:05PM -0400, Michael Schneider wrote:
> Hello All.
> 
> I just updated my python to 2.2 and noticed that generators, and 
> iterators were in there.
> 
> I often parse through many files (one at a time), and wanted the speed 
> of read,  and
> the convience of   for x in myGen:

If this beats 'for line in open("myfile"): ...' by much, there's a problem .. 
can you post some benchmark numbers?  Anyway, I don't think this can outperform the simpler
	for x in file.readlines(): ...
since yours adds another layer of Python code in the middle...

I think that files can be iterated over in 2.2, I only have 2.3 on hand:

$ python2.3 -c 'import sys; print list(sys.stdin)' < /etc/resolv.conf 
['search unpythonic.net\n', 'search localnet\n', 'nameserver 206.222.212.218\n']

f.xreadlines() or iter(f) should both operate in the same way, by using
readlines with a size hint, then returning successive lines from the
returned list, refilling the list as necessary.  At the time, this
benchmarked "just as fast as" slurping the whole file with a single
readlines(), but with a reasonable memory footprint even for gigabyte files.
(At least when none of the lines are insanely long)

For benchmarks made at the time, start near
    http://mail.python.org/pipermail/python-dev/2001-January/011269.html

Tim Peters quoted Neel Krishnaswami:
| Quick performance summary of the current solutions:
|
| Slowest: for line in fileinput.input('foo'):     # Time 100
|        : while 1: line = file.readline()         # Time 75
|        : for line in LinesOf(open('foo')):       # Time 25
| Fastest: for line in file.readlines():           # Time 10
|          while 1: lines = file.readlines(hint)   # Time 10
|          for line in xreadlines(file):           # Time 10
|
| The difference in speed between the slowest and fastest is about
| a factor of 10.

Jeff



From trash at terastat.com  Tue Sep 17 13:23:57 2002
From: trash at terastat.com (bap)
Date: Tue, 17 Sep 2002 17:23:57 GMT
Subject: PIL v.1.1.3 and grabscreen
Message-ID: <NyJh9.344277$_91.442717@rwcrnsc51.ops.asp.att.net>

I'd like to use the grabscreen function from Lundh's site
(http://effbot.org/downloads) however it won't install or run (I use Python
2.2, it may be compiled for 2.1). The function is now included (according to
the documentation) in PIL 1.1.3 -- but the PIL site only has binary versions
of 1.1.2, for v1.1.3 the site refers to Lundh's site -- where again the
installation routine doesn't work (asks for a directory but will accept no
input). So -- does anyone know of any working sources of either grabscreen
or a binary distributable for windows/activestate python 2.2 of PIL 1.1.3?




From gerhard.haering at gmx.de  Wed Sep  4 17:45:22 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 4 Sep 2002 21:45:22 GMT
Subject: Database BSD on Mac?
References: <3144be3c280c062e5e71a98ab191f664.98737@mygate.mailgate.org> <mailman.1031095985.2997.python-list@python.org> <_7vd9.57894$ub2.928923@news1.tin.it>
Message-ID: <slrnancvnh.107.gerhard.haering@lilith.my-fqdn.de>

Max wrote in comp.lang.python:
>> > I was looking on BSD but I'm not sure it is ported to all platforms.
>> > Wich db should I utilize?
>>
>> Do you really need one? Or do either pickle or shelve suffice?
> 
> I read shelve can be used for small amount of data... will it work with 5mb?

Certainly. And probably for _lots_ of more data, if your problem fits
an on-disk hashtable like shelve.

-- Gerhard


From peter at engcorp.com  Thu Sep 19 19:40:37 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 19 Sep 2002 19:40:37 -0400
Subject: Minimal debug/rep functionality
References: <J%ai9.893$1D.574@nwrddc01.gnilink.net> <3d894106@news.sentex.net> <PQpi9.16355$t%6.7733@nwrddc02.gnilink.net>
Message-ID: <3d8a606a@news.sentex.net>

Mark wrote:
> Peter Hansen wrote:
> 
>>Sorry, I couldn't find a solution that was larger than six lines:
>>
>> >>> def trace(frame, event, arg):
>>...   if event == 'line':
>>...      print '%s: %s' % (frame.f_code.co_filename, frame.f_lineno)
>>...   return trace
>>...
>> >>> import sys
>> >>> sys.settrace(trace)
>>
> 
> This is exactly what I needed.  Now, can you explain to me how I hook in a 
> "local" trace function to take over once I execute the function I'm 
> interested in tracing?  I'd like to use the "event" argument to take note 
> of calls that are 1 level deep (ie in my foo function, any calls made) and 
> turn off tracing of these until they return back to the level of foo.

I guess you'd need a global to track the level, or implement the trace
"method" as an object with a __call__ method so you can track state as
well as receive the tracing calls.

As for local tracing... the above function actually does that already.
As the docs note, the global function (installed by sys.settrace)
returns a reference to a local trace function when event is "call".
I'm just using the same function for both purposes as a cheat. :)

> I presume I could implement this with a "if event == 'call':" condition, but 
> I can't seem to get a global trace function to hook up to a local trace 
> function ... it does one or two things right, then goes crazy.

Post sample code if you're having trouble... but I recommend playing
a lot at the interactive prompt before doing so.  You learn lots more
that way.

-Peter



From jpm at papercut.org  Mon Sep 30 15:04:24 2002
From: jpm at papercut.org (Joao Prado Maia)
Date: Mon, 30 Sep 2002 15:04:24 -0400 (EDT)
Subject: Question about __getattribute__/__getattr__
In-Reply-To: <JHEOKEOOLIGLDHCMAHMOIEJDECAA.marklists@mceahern.com>
Message-ID: <Pine.LNX.4.44.0209301503100.12980-100000@mercury.sabren.com>

On Mon, 30 Sep 2002, Mark McEahern wrote:

> >         if cached_value == None:
> 
> This is an aside, but comparison to None should be via identity rather than
> equality; e.g.,
> 
>   if x is None:
> 
> or:
> 
>   if x is not None:
> 

Yeah, I always forget that when switching from PHP to Python. Anyway, my 
main question was how to get the list of arguments passed to a 
method/function. Any ideas ? :)

Cheers,
Joao




From jason at tishler.net  Mon Sep 16 12:42:59 2002
From: jason at tishler.net (Jason Tishler)
Date: Mon, 16 Sep 2002 12:42:59 -0400
Subject: Accessing Windows Registry under cygwin
In-Reply-To: <f5b1y8298w2.fsf@cogsci.ed.ac.uk>
References: <f5b1y8298w2.fsf@cogsci.ed.ac.uk>
Message-ID: <20020916164259.GG424@tishler.net>

Henry,

Sorry for the sluggish response time, but I just returned from
vacation...]

On Mon, Sep 09, 2002 at 09:59:41PM +0100, Henry S. Thompson wrote:
> On 11 September 2001 Norman Vine said "I have managed to get the
> '_winreg' module working with Cygwin < _winreg = Python access to the
> Windows Registry >" [1].  Assiduous searching does not reveal that
> this work ever surfaced -- where might I find it?

See the following:

    http://sourceforge.net/tracker/?group_id=5470&atid=305470&func=detail&aid=554807

Are you willing to help?

Thanks,
Jason



From peter at engcorp.com  Mon Sep  2 02:27:31 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 02 Sep 2002 02:27:31 -0400
Subject: curses: AttributeError: initscr
References: <un5dhs4f7qlte8@corp.supernews.com> <un5iq3ht257s22@corp.supernews.com> <mailman.1030934343.21056.python-list@python.org>
Message-ID: <3D7304D3.28D48684@engcorp.com>

Gerhard H?ring wrote:
> 
> * Cesar <g0w1s0 at yahoo.ca> [2002-09-01 22:22 -0400]:
> > in the same directory where my script live, there was an old file named
> > curses.py. Then the instruction : "import curses" imported the bad one...
> 
> You're not the first one to have this kind of problem.

Too true. :)

> A way to debug such problems is to start Python with the option "-v" or
> even "-vv", which helps you track down where modules are really loaded
> from.

The first thing I generally do, since I'm inevitably testing something
like this out at the interactive prompt, is to as for "module.__file__"
and see where the file I'm importing is really coming from.  I've
been surprised on more than one occasion.

-Peter


From jpm at papercut.org  Mon Sep 30 11:30:55 2002
From: jpm at papercut.org (Joao Prado Maia)
Date: Mon, 30 Sep 2002 11:30:55 -0400 (EDT)
Subject: Question about __getattribute__/__getattr__
Message-ID: <Pine.LNX.4.44.0209301035390.1837-100000@mercury.sabren.com>

Hi,

I'm trying to do the following:

class Real_Class:
    def real_method(self, name):
        return name

class Cache:
    r = None

    def __init__(self):
        self.r = Real_Class()

    def __getattr__(self, attr):
        cached_value = self.get_cache(attr)
        if cached_value == None:
            # need to eventually save the returned value in the cache
            return getattr(self.r, attr)
        else:
            return cached_value

    def get_cache(self, attr):
        return None

if __name__ == '__main__':
    t = Cache()
    print t.real_method('spam')


Now, the code works correctly - it calls the appropriate method name with 
the argument 'name'. However, I would like to cache the return value of 
the method call for a specific match of method name / method arguments. 
That is, I would cache:

t.real_method('spam')

And:

t.real_method('eggs')

in two different cache files.

It all comes down to the question: how do I know the full list of 
arguments passed to a function/method ? This way I'll be able to md5() the 
full string and have a different cache file for each method call.

Any ideas ?

Cheers,
Joao




From loewis at informatik.hu-berlin.de  Wed Sep  4 05:03:10 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 04 Sep 2002 11:03:10 +0200
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com>
Message-ID: <j4n0qydt4h.fsf@informatik.hu-berlin.de>

Benjamin <phncontact.libero.it@> writes:

> is it worth to go deeper into python?

That depends on what you are trying to do. Python can help in doing
lots of things, since it has good libraries for many application areas.

If you want to discuss further, please tell us what kinds of things
you want to do. 

You might just start using it, as well - take a small problem of yours
that you always want to get solved, and see how easy it is to solve in
Python. Don't hesitate to ask here if you have specific questions.

Regards,
Martin



From DeepBleu at DeepBleu.org  Sat Sep 21 10:48:56 2002
From: DeepBleu at DeepBleu.org (DeepBleu)
Date: Sat, 21 Sep 2002 14:48:56 GMT
Subject: pyrhon and oracle
References: <5b8834c2.0209201412.50e77861@posting.google.com> <pan.2002.09.20.23.29.54.67293.2172@verizon.net> <B__i9.505088$q53.17034835@twister.austin.rr.com>
Message-ID: <sF%i9.286967$eK6.8391777@twister.austin.rr.com>

"Lee Gray" <lee100 at sport.rr.com> wrote in message
news:B__i9.505088$q53.17034835 at twister.austin.rr.com...
> Satheesh Babu wrote:
>
> > cx_Oracle might be the *easiest* to use. I use DCOracle while working
> > with Zope, but for standalone development, cx_Oracle is tops.
> >
> > Download the driver from
> > http://www.computronix.com/utilities.shtml
> >
> > go to Python DB-API specs
> > http://www.python.org/topics/database/DatabaseAPI-2.0.html
> >
> > Here's an example using DB-API
> > http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/81189
> >
> > vsb
> > http://vsbabu.org/
> >
>
> I'm just looking into Python, so these articles don't mean much to me
yet...
> Does all this mean that Python doesn't talk to an Oracle DB "out of the
> box"?  For Windows, can Python use ADO?
>
In all programming languages you have to add components/packages to do
certain things.  In the case of Python you have to add a package that allows
you to use a database.  There are many packages made for Python for this
purpose.
Also, Python can use ADO in Windows.  Please search for Python and ADO in
Google.  There are articles that describe how to do that.




From chris.gonnerman at newcenturycomputers.net  Mon Sep 30 01:13:33 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Mon, 30 Sep 2002 00:13:33 -0500
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <an7bvd$burft$1@ID-125932.news.dfncis.de> <an7p7t$cpg$0@216.39.172.122> <mailman.1033338759.6775.python-list@python.org> <enf8na.86m.ln@ix.netcom.com>
Message-ID: <006701c26840$20e0e340$ba01010a@local>

----- Original Message ----- 
From: "Dennis Lee Bieber" <wlfraed at ix.netcom.com>


> Chris Gonnerman fed this fish to the penguins on Sunday 29 September 
> 2002 03:33 pm:
> 
> 
> > You ask how it is a different problem.  Simple.  A human,
> > training in decimal math in grade school, knows that
> > 
> >     1/3 = 0.3333333...
> > 
> > and would *expect* it to round to 0.33 (still in two 
> > decimal
> 
>         So, in 2 decimal place arithmetic
> 
>         a = 1.00 / 3.00
>         b = a * 3.00
> 
> gives
>         b = 0.99
> 
> which is NOT equal to 1... But this is okay because it's done 
> is "decimal arithmetic"? <G>

It's not "OK"... it's EXPECTED.

>         The telling phrase is "training in DECIMAL math"... 
> Most computers don't work in decimal math, so the 
> user /should/ be ready to learn the ins&outs of binary 
> floating point...

Incorrect.  The *user* expects the dang computer to produce
the same results as for a manual calculation.  In other words,
truly serious businesspeople expect to be able to hand-check
the numbers.  The occasional (but far too common) binary float
anomalies are unacceptable in that environment.

> 
>         For every example of "what is wrong with binary 
> floating point" someone else can find a similar case when 
> using decimal math.

True, but the decimal anomalies are *expected* and the binary
anomalies are *surprising*...  I've been a programmer for 
a long time, working in everything from a scientific lab to 
a retail store.  I still don't understand the ins and outs,
as you put it, of binary floating point.

> Even VB's "currency" type (which looks to be a scaled 64-bit 
> integer) carries 4 decimal places -- which allows for 
> "proper" results when displaying final results in two decimal 
> places with rounding.
> 
>         1/3 => 0.3333   internally
>         3*0.3333 => 0.9999
>         1.00 for display after rounding.

I hate to give anything to Microsoft, but they seem to have a
good, workable, not-often-surprising idea here.

I still prefer the FixedPoint.py module's programmer-controlled
scaling.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From python473 at yahoo.com  Fri Sep 27 11:10:11 2002
From: python473 at yahoo.com (John Howard)
Date: 27 Sep 2002 08:10:11 -0700
Subject: file open verification
Message-ID: <9eabe547.0209270710.17ca112d@posting.google.com>

Two questions - somewhat related:

Situation - I have several files on an apache server that are being
accessed by
python programs that are receiving data from html forms.

q1: How can the status of a file be checked?  If two forms have been
submitted
and the data is being added to a common file, how do I know that data
from both forms are being written correctly to the common file?

q2: Related to above - really more of a cgi question - does the apache
server keep all these file accesses separate some way?


From sismex01 at hebmex.com  Fri Sep 20 09:25:44 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Fri, 20 Sep 2002 08:25:44 -0500
Subject: strong/weak - dynamic/static [Was: Getting started]
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233024FD928@mail.hebmex.com>

> 
> >So, you see, in C (a strong, static typed language) I obtained
> >a char from a floating-point number; not the string representation,
> >just the first byte.
> 
> C is a statically typed language, but to call it "strongly" typed
> is a bit of a stretch. C is at best, middle of the road in type
> strength.
> 
> C//
> 

You are of course right; I ommited some (intended) quotes
around the "strong" in "(a strong, static typed language),
so it would reflect my true intention.

:-)

-gus




Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From never at mind.info  Wed Sep  4 19:09:50 2002
From: never at mind.info (G. Willoughby)
Date: Thu, 5 Sep 2002 00:09:50 +0100
Subject: How do i print to a printer using python?
References: <6afb9d2c.0209041044.43c5f5f1@posting.google.com>
Message-ID: <al63r8$3bo$1@news7.svr.pol.co.uk>

I would really like to know how to do this too, basically i want to print to
a standard bubblejet print (non postscript) from a Tkinter Text widget, any
ideas?

--G. Willoughby




From shalehperry at attbi.com  Wed Sep 11 18:05:38 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 11 Sep 2002 15:05:38 -0700
Subject: problem with lists
In-Reply-To: <20020911215423.GA30455@habana.easygolucky.de>
References: <20020911215423.GA30455@habana.easygolucky.de>
Message-ID: <200209111505.38492.shalehperry@attbi.com>

On Wednesday 11 September 2002 14:54, Manuel Hendel wrote:
> for line in lines:
>         line = string.strip(line)[1:-1]
>         fields = string.split(line, "|")
>
>         domains = []
>
>         if not domains.index(fields[3:4]):
>             domains.append(fields[3:4])
>
> print domains
>
> This brings the following error message:
> :!./pop3create.py pop3data
>
> [No write since last change]
> Traceback (most recent call last):
> File "./pop3create.py", line 22, in ?
> if domains.index(fields[3:4]):
> ValueError: list.index(x): x not in list
>
> Can someone explain this to me?
>
> Thanks,
>     Manuel

You want to use :

if fields[3:4] not in domains:
    domains.append(fields[3:4])

.index() only works if the item is actually in the list.



From gerhard.haering at gmx.de  Fri Sep  6 03:48:36 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Fri, 6 Sep 2002 09:48:36 +0200
Subject: Timothy Rue
In-Reply-To: <898662b7d85d2c966c0bd45d563d5d2d@xganon.com>
References: <898662b7d85d2c966c0bd45d563d5d2d@xganon.com>
Message-ID: <20020906074835.GB848@lilith.ghaering.test>

* xganon <remailer at xganon.com> [2002-09-06 02:31 -0500]:
> Does Timothy Rue still Troll here?  If not, how did you get rid of
> him?

Easy:

gerhard at lilith:~/src/conf > grep -B2 threeseas News/score
Score: =-9999
Expires: 2/20/2003
        From: "Timothy Rue"
        <threeseas at earthlink\.net
gerhard at lilith:~/src/conf > grep threeseas .muttrc_scoring
# "Timothy Rue" <threeseas at earthlink.net>
score '~f threeseas' =0

Good thing I have my personal config files in CVS, so I have these
settings on all the computers/OSes I work on :)

> We need him out of our group now,

Are _you_ trying to troll now?

-- Gerhard



From member at dbforums.com  Tue Sep 17 08:26:27 2002
From: member at dbforums.com (MPS)
Date: Tue, 17 Sep 2002 12:26:27 +0000
Subject: Modify AD key fields programmatically via ADSI or VB Script
Message-ID: <1824301.1032265587@dbforums.com>

I am new to ADSI and was wondering if anybody knows a way to modify AD
key fields (i.e./ DN) programmatically via ADSI or VB Script? Here is
the exact scenario:

We are attempting to synchronize the population in our new Active
Directory deployment to the active individuals identified within our
various database systems of record.  To accomplish that, we generate the
following ldap string describing the DN for every individual daily:

TOM SMITH,ou=ACCOUNTING,ou=people,dc=biz,dc=url,dc=com

along with values for numerous attributes.  As noted above, the DN is
comprised of the concatenate of FIRST NAME, MIDDLE NAME, and LAST NAME
as well as that individual's PRIMARY DEPARTMENT (in this case
ACCOUNTING).

All of these are implemented into AD via a LDIF generated by a third
party synchronization product (Simplesync).

If any of these four elements (FIRST NAME, MIDDLE NAME, LAST NAME,
DEPARTMENT) change (as they will from time to time), the result in AD is
to add a "new" individual and "delete" to old.  This is undesirable.

Is there a way to programmatically detect that any of the above data
elements are changing prior to executing the bulk synchronization and
MODIFY the DN in AD in place or any other solution that will move this
person to a different DN without losing the rest of their existing AD
information.

Thanks a million for any help - MPS

--
Posted via http://dbforums.com



From mhammond at skippinet.com.au  Fri Sep 20 21:19:47 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 21 Sep 2002 01:19:47 GMT
Subject: Python COM and variable number of arguments
References: <mailman.1032529407.31552.python-list@python.org> <sn04lhyb.fsf@boeing.com>
Message-ID: <TOPi9.1071$6g7.2034@news-server.bigpond.net.au>

Greg Green wrote:
> In this case I was using the makepy generated wrapper. I have
> M. Hammonds book and used the method of ensuring I was using the
> early-bound object.
> 
> So am I out of luck even if I have the wrapper?

You could also try passing with pythoncom.Missing or pythoncom.Emtpy in 
place of the positional arguments.  You may need to experiment to see 
what one works - it depends on the COM object.

Mark.



From david.abrahams at rcn.com  Sun Sep 15 11:20:59 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Sun, 15 Sep 2002 11:20:59 -0400
Subject: Extending a C++ App
References: <996eade8.0209120912.3a653d7@posting.google.com> <n0qnhvfn.fsf@boeing.com> <996eade8.0209122352.45721938@posting.google.com> <altqiu$fjr$1@bob.news.rcn.net> <slrnao96ee.52u.Gareth.McCaughan@g.local>
Message-ID: <am28vl$3bt$1@bob.news.rcn.net>

"Gareth McCaughan" <Gareth.McCaughan at pobox.com> wrote in message
news:slrnao96ee.52u.Gareth.McCaughan at g.local...
> David Abrahams wrote:
> >
> > "Bwuce_Wee" <bwucew at yahoo.com> wrote in message
> > news:996eade8.0209122352.45721938 at posting.google.com...
> >
> > >     I'm busy investigating Boost.Python...
> >
> > In that case, get Boost.Python v2 using the instructions at
> >
> > http://mail.python.org/pipermail/c++-sig/2002-May/001100.html
> >
> > It's nearing release, and v1 isn't long for this world.
>
> If you recommend new users to start with v2 rather than v1,
> then it's already time for a (pre-)release, no?

Almost. We're just working out the final administrative issues which will
allow us to do that.

--
-----------------------------------------------------------
           David Abrahams * Boost Consulting
dave at boost-consulting.com * http://www.boost-consulting.com





From a-steinhoff at web.de  Thu Sep  5 05:39:48 2002
From: a-steinhoff at web.de (Armin Steinhoff)
Date: 5 Sep 2002 02:39:48 -0700
Subject: Installing Python 2.2.1 on WinNT
References: <mOSb9.38460$C8.132490@nnrp1.uunet.ca>
Message-ID: <ddc19db7.0209050139.34c12ed1@posting.google.com>

"Phil Rittenhouse" <phil.rittenhouse at dspfactory.com> wrote in message news:<mOSb9.38460$C8.132490 at nnrp1.uunet.ca>...
> Hi all,
> 
> I've run into a (actually two) problem installing Python 2.2.1 on Windows NT
> 4.0.
> At the end of the install, it throws up an error "Could not load the DLL
> library $%#%^$.
> The specified module could not be found."

I have recently installed Python 2.2.1 on a NT 4.0 machine with SP6.

The install process worked flawless. 
Is your dowloaded install file OK ??

- Armin


From martin at v.loewis.de  Tue Sep 17 02:05:35 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 17 Sep 2002 08:05:35 +0200
Subject: Large file support and stat vs stat64 in extension modules
References: <pan.2002.09.10.16.03.39.756929.4320@remove.Stanford.EDU> <pan.2002.09.16.22.58.21.620527.4504@remove.Stanford.EDU>
Message-ID: <m3znuhb174.fsf@mira.informatik.hu-berlin.de>

Ben Escoto <ben at remove.Stanford.EDU> writes:

> define_macros=[("_LARGEFILE_SOURCE", 1), ("_FILE_OFFSET_BITS", 64)]
> 
> when declaring the ext_modules does something similar and seems to work.
> 
>     Again, this is a bit ad hoc, but seems to work, and I don't know of a
> better way.

If Python was configured for LFS, then those defines appear in
pyconfig.h. So including Python.h before any system headers is the
recommend, and better, way.

Regards,
Martin



From DennisR at dair.com  Wed Sep 18 13:00:06 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Wed, 18 Sep 2002 17:00:06 GMT
Subject: Python comment stripper
References: <16Nh9.248$Wc6.17087796@newssvr13.news.prodigy.com> <3d8807b1$1@news.sentex.net> <Eg0i9.561$Y33.34967690@newssvr13.news.prodigy.com> <S41i9.127467$pX1.4558233@news2.tin.it>
Message-ID: <pi2i9.615$3s2.32637168@newssvr13.news.prodigy.com>

> Makes sense.  Let's see, what about a one-(logical)-liner (warning,
> untested!):
Alex,

Thanks for the starter code.  This will be a good Python learning program
for me.  However, I may defer implementing source comment extraction until
after first release.

> What are you using now for compression?
I am using WinRar 3.0 (www.rarlab.com).  As an exe inside a zip file, my
hello example takes 402 Kb.  The user must extract and run the EXE manually.
As an EXE inside a self-extracting RAR format EXE, the same example is 361
Kb.  The latter is smaller and easier for recipient. (ref:
www.dair.com/pyhello.zip and www.dair.com/pyhello.exe).

> Module gzip is probably a good idea to help you reduce size, too.
At the time extraction is going on, it is impractical for Python to be doing
the extraction -- there is the Python DLL (v2.0: 653 Kb and v2.2: 821 Kb)
for starters.  I am aiming for a small footprint.  The 361 Kb download has
Python22.dll inside and that DLL  is used when the hello program runs. :

Thanks, Dennis
 --

Dennis Reinhardt

http://www.dair.com

>




From tjreedy at udel.edu  Thu Sep 12 12:45:17 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 12 Sep 2002 16:45:17 GMT
Subject: wrapping yield ?
References: <alqehl$9qv$1@nntp0.reith.bbc.co.uk>
Message-ID: <xw3g9.338518$On.14183798@bin3.nnrp.aus1.giganews.com>

"Michael Sparks" <Michael.Sparks at rd.bbc.co.uk> wrote in message
news:alqehl$9qv$1 at nntp0.reith.bbc.co.uk...
> Hi,
>
> I'd like to be able to wrap "yield"  in a function - is that
> possible, and if not (my eading of stuff  I've seen so far
> says not) are there any plans to make it possible?
>
> Whilst it may seem odd, I'd like to be able to do this:
>
> def my_generator(arg):
>     dostuff....
>     while(1):
>        waitForSomethingCheckingPeriodically()
>        processit

You have to put in a yield x *somewhere* in order to yield anything to
the caller.
Hard to see why this is a problem.  Something like

x = wait....
yield process(x)

TJR

You are




From amichail at cse.unsw.edu.au  Thu Sep 12 02:51:23 2002
From: amichail at cse.unsw.edu.au (Amir Michail)
Date: 11 Sep 2002 23:51:23 -0700
Subject: ignoring dtd file in sax parser
Message-ID: <1010a26.0209112251.3efd57f4@posting.google.com>

Hi,

I'm trying to parse XML files that specify
a dtd.  However, I would like to ignore the dtd,
and in fact, the dtd file is not even present.

I tried:

        parser.setFeature(feature_namespaces, 0)
        parser.setFeature(feature_validation, 0)

        parser.setContentHandler(ph)

        try:
            parser.parse(f)
    #        print "NO PARSE ERRORS"
        except xml.sax._exceptions.SAXParseException:
            print "XML parse errors in " + fn


However, it still complains about not finding the dtd file.

Amir


From lbrannma at cablespeed.com  Sun Sep 22 20:52:57 2002
From: lbrannma at cablespeed.com (Lance)
Date: Sun, 22 Sep 2002 17:52:57 -0700
Subject: sys.ps2 assignment has no effect
References: <uos1np8t53k396@corp.supernews.com> <FQqj9.149901$pX1.5369568@news2.tin.it>
Message-ID: <uospb1lclehc93@corp.supernews.com>

Thanks Alex.. it's with IDLE. That explains things.

Lance

"Alex Martelli" <aleax at aleax.it> wrote in message
news:FQqj9.149901$pX1.5369568 at news2.tin.it...
> Lance wrote:
>
> > I'm running the latest Python on Windows XP. The default installed
> > secondary string appears to be blank spaces rather than an ellipse. The
> > following has no effect:
> >
> > import sys
> > sys.ps2 = '... '
> >
> > Any suggestions?
>
> Is this with the text-mode interactive Python (which would
> surprise me, as I've run the latest Python on XP w/o such
> problems), or with IDLE or other such IDE?  IDLE does not
> use PS2, afaik (it always uses spaces instead) and other IDEs
> might do likewise.
>
>
> Alex
>




From gerhard.haering at opus-gmbh.net  Wed Sep 11 03:57:22 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 11 Sep 2002 07:57:22 GMT
Subject: RELEASED email package version 2.3
References: <mailman.1031715254.9581.python-list@python.org>
Message-ID: <slrnanttkg.1mk.gerhard.haering@haering.opus-gmbh.net>

Barry A. Warsaw wrote:
> I've just released the (standalone) email package version 2.3.

Is there any summary on changes, possible incompatibilities and deprecated
features in comparison with the email package in Python 2.2 available?

-- Gerhard


From cliechti at gmx.net  Tue Sep 17 15:15:03 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 17 Sep 2002 21:15:03 +0200
Subject: More on Protecting Source Code
References: <GCEDKONBLEFPPADDJCOEAEDDFAAA.whisper@oz.net> <mailman.1032260602.11981.python-list@python.org>
Message-ID: <Xns928CD8319CD25cliechtigmxnet@62.2.16.82>

Kerim Borchaev <warkid at storm.ru> wrote in
news:mailman.1032260602.11981.python-list at python.org: 

> Hello David,
> 
> Tuesday, September 17, 2002, 7:26:06 AM, you wrote:
> 
> DL> I suppose one solution is to modify the Python interpreter with
> different DL> op-codes and that ought to make it somewhat painful for
> the average hacker. DL> A better solution is to make a .pyc file
> approximately as hard as a binary DL> .exe file to decompile - however
> that could be done. 
> 
> Just a thought.
> Keeping your python modules encripted(e.g. with rotor module?) you can
> reimplement __import__ builtin to properly handle importing these
> modules. 

that sounds nice at first, but how would you code the custom import? if one 
has the readable code for that one using it to read the encrypted files is 
a no-issue. and if you decrypt it the plain text sources are in memory and 
a simple snapshot of the applications memory reveals the source.

a simple way to semi obfuscate is to use the freeze tool which embedds the 
python bytecodes in a C array. as you also link against a python 
interpreter there are no version issues and distributing gets easy.

if somebody wants to change the opcodes for this python interpreter - go 
on, that should be easy and the python decompilers won't work out of the 
box.

if somebody believes that there should be better security there might be 
ways, but i think that you need expert knowledge in encryption, the OS it 
runs on, attack methods of crackers etc. to achieve a much higher level of 
security. probably too much effort just to eventualy loose...
just distributing bytecode (forzen or not) should be save enough for most 
commercial purposes. i don't think that java is any more secure than python 
in this respect.

chris
-- 
Chris <cliechti at gmx.net>



From anton at vredegoor.doge.nl  Mon Sep 30 12:48:03 2002
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Mon, 30 Sep 2002 18:48:03 +0200
Subject: watching mutables?
References: <an1p3c$jrd$1@news.hccnet.nl> <an880d$ncp$0@216.39.172.122>
Message-ID: <an9vhe$c30$1@news.hccnet.nl>

On 30 Sep 2002 01:05:17 GMT, bokr at oz.net (Bengt Richter) wrote:

<good explanation deleted>

I have slightly adjusted your code to result in the code below, this
works if imported by Idle like this:

Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help
>>> from watcher import Watcher
>>> w = Watcher('m',locals())
>>> m = 1
"m" is bound to new object: 1


So this seems to be ok. However if the script below is executed
itself, nothing happens. Also importing the module in another script
and then executing a test script gives no output. The watch function
isn't called at all. I guess the tracer has to know more about the
other arguments in:

    def watch(self, frame, event, arg):

Could this please be elucidated.

Thanks for your insightful remarks and code.

Anton.


#watcher.py
from sys import settrace, modules
from marshal import dumps

class Watcher:

    def __init__(self, name, namespace):
            self.name = name
            self.ns = namespace
            obj = namespace.get(name)
            self.objid = id(obj)
            self.objs = dumps(obj)
            settrace(self.watch)

    def watch(self, frame, event, arg):
            obj = self.ns.get(self.name)
            objid = id(obj)
            if objid != self.objid:
                self.objid = objid
                what = 'new'
            else:
                what = 'same but mutated'
            objs = dumps(obj)
            if self.objs != objs:
                print '"%s" is bound to %s object: %s' % (self.name,
what, `obj`)
                self.objs = objs
            elif what == 'new':
                what = 'new but equivalent'
                print '"%s" is bound to %s object: %s' % (self.name,
what, `obj`)
            return self.watch

    def finished(self):
            settrace(None)

def test():
    # huh ?
    w = Watcher('m', locals())
    m = 1

if __name__=='__main__':
    test()



From thorsten at thorstenkampe.de  Mon Sep 23 17:03:21 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Mon, 23 Sep 2002 23:03:21 +0200
Subject: Three dumb questions (ordered by dumbness descending)
Message-ID: <amnvit$7kqdt$1@ID-77524.news.dfncis.de>

Okay, here they are:

1. Why is 'zip(zip(x)) != x' (should be the same because it's the 
transposed)


2. Peter Norvig mentions in "Python for Lisp Programmers" some 
"don'ts": "[x] + y" and "x[1:]". Are there more things to avoid 
(especially in a loop)?


3. random.shuffle: the documentation says: "Note that for even rather 
small len(x), the total number of permutations of x is larger than the 
period of most random number generators; this implies that most 
permutations of a long sequence can never be generated." I translate 
this to "don't shuffle lists with more than 15 elements if you want a 
randomly ordered list".

Alex Martelli says the same: "For example, Python users who are 
calling random.shuffle on a sequence of substantial length have a need 
for an underlying random generator with a *HUGE* period, as a 
necessary although not sufficient condition towards ensuring that all 
permutations of the shuffled sequence get generated with equal 
likelihood.

Wichman-Hill has a period of less than 7,000 billions (7e12),
insufficient to account even for the permutations of a sequence
whose length is just 16 (16! > 2e13)."

Otherwise in the "Python Cookbook" (Selecting Random Elements from a 
List Without Repetition) the shuffle version is called "the winner".

So: can I use random.shuffle or do I have to write my own - slower - 
version to get a random list?


Thorsten


From sam at localhost.localdomain  Wed Sep 11 19:05:22 2002
From: sam at localhost.localdomain (newbie)
Date: Thu, 12 Sep 2002 00:05:22 +0100
Subject: I must be an idiot, please pity me!
References: <VAOf9.2572$Pa7.155732@newsfep1-gui.server.ntli.net> <3D801757.1020901@nospam.free.fr>
Message-ID: <oYPf9.2894$XE6.174556@newsfep1-win.server.ntli.net>

On Thu, 12 Sep 2002 01:25:59 +0100, laotseu wrote:


> Creating a db connection in each function for sure not be very
> efficient.But what's wrong with passing the database connection to the
> functions in mod1, mod2, and modXXX ???

That connection has to be passed to the other functions (where needed of
course) in mod1. Say, mod1.functiona may use functions f1, f2 and f3. My
gut feeling of having to pass that connection to all of these said "this
can't be right". This is what I've done but it sure doesn't feel right!


> Anyway, you can make it simpler, like this :

Yay, now for the good stuff!

 
> in each module,
> - create a global db_connection variable - add an init() function, that
> takes a db_connection as argument and assign it to the module's global
> db_connection variable - and don't forget to call this function before
> any other.

Yes, I can follow that. Once I saw that "global" was more "local" I've
never really considered using it.


> Another, more OO way to do this would be to rewrite your modules as
> classes, each getting a db_connection argument in it's __init__(), and
> then, in the main script, create the needed objects.

So the resulting object would be the same as what I return with now?

> HTH
> Laotseu

Thank you.
Sammy


From shalehperry at attbi.com  Tue Sep 24 11:15:15 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 24 Sep 2002 08:15:15 -0700
Subject: confused with lambda~
In-Reply-To: <20020924084839.30815.qmail@web21307.mail.yahoo.com>
References: <20020924084839.30815.qmail@web21307.mail.yahoo.com>
Message-ID: <200209240815.15208.shalehperry@attbi.com>

On Tuesday 24 September 2002 01:48, black wrote:
> that describe in reference documentation is too short, I couldnt even
> understand what it trying to say. any explanation please ?
>

lambda is way to define a function without giving it a name.

lambda x,y: x + y

defines a function which takes two parameters -- x and y -- and returns their 
sum.

do_sum = lambda x,y: x + y

do_sum(2,3)

yields 5.  As you can see we store the function as do_sum.  So let's take this 
a step further.

numbers = range(1,21) # numbers 1 - 20
sum = reduce(lambda x,y: x + y, numbers)

sum will be the sum of all numbers 1 - 20.  This is equivalent to the 
following code:

sum = 0
for n in numbers:
    sum = sum + do_sum(sum, n)

The idea is lambda let's you write little one off functions when you need them 
rather than have to define one somewhere and likely only use it once.

In a lambda on expressions are valid so no assignment or print calls can 
happen.  If you need that you must define a function.

Hope this helps.



From gerhard.haering at gmx.de  Tue Sep 10 18:33:31 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 10 Sep 2002 22:33:31 GMT
Subject: test (ignore)
References: <3d7e2957$1_2@nopics.sjc>
Message-ID: <slrnansspr.6im.gerhard.haering@lilith.my-fqdn.de>

Adonis wrote in comp.lang.python:
> just a test, please ignore.

What exactly is wrong with alt.test or one of the other .test groups?

-- Gerhard


From david.abrahams at rcn.com  Fri Sep 27 19:08:20 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Fri, 27 Sep 2002 19:08:20 -0400
Subject: Problem w/ boost_python_debug.dll
References: <cb95025b.0209142123.82c016e@posting.google.com>
Message-ID: <an2q6m$grl$1@bob.news.rcn.net>

[Boost.Python questions should go to the C++-sig
(http://www.python.org/sigs/c++-sig/); I seldom watch this list carefully.]


This is a limitation of our build system (sorry, we're working on it).
boost_python_pydebug.lib is really built as boost_python.lib, then
copied/renamed, a strategy that doesn't work out well. You have several
options:

    1. rename the debug products back to boost_python.lib/.dll, and link
with that. Keep debug and release versions in separate directories. Keep
your PATH straight so that the wrong one is never loaded by mistake

    2. Use Boost.Build to build your project. It always keeps the
debug/release/debug-python versions in separate directories, and
automatically points at the right one (which is why this doesn't cause me
all kinds of pain every day).

    3. Temporarily edit libs/python/build/Jamfile to change the name of the
library target from boost_python to boost_python_pydebug. Rebuild
everything. There will be a properly-named-and-functioning debug library in
the bin/ directory tree. Grab it and away you go.

Sorry for the inconvenience. I'm going to try to get distutils going for v2;
that should make some of these things easier.

HTH,
Dave

--
-----------------------------------------------------------
           David Abrahams * Boost Consulting
dave at boost-consulting.com * http://www.boost-consulting.com


"Richard" <kapis at hotmail.com> wrote in message
news:cb95025b.0209142123.82c016e at posting.google.com...
> I'm using Boost.Python (v1) to embed & extend a C++ application. I've
> downloaded ActiveState's distribution and built the debug python
> library and DLL.
>
> I built the release and python-debug versions of Boost.Python as is
> documented. I also define BOOST_DEBUG_PYTHON in my C++ files since I
> have the debug python libraries.
>
> Enough with the background, let's get to the problem. For debug
> builds, I link with boost_python_pydebug.lib. But when I run the
> program, it still tries to load up boost_python.dll, NOT the debug
> version! The only way I know to fix the problem is to rename
> boost_python_pydebug.dll so that it is loaded instead.
>
> Does anyone know a better fix? Any help would be appreciated.
>
> Thanks.




From gua81 at XXXyahoo.com  Thu Sep 26 03:48:37 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Thu, 26 Sep 2002 17:48:37 +1000
Subject: Getting last char in string.
Message-ID: <amue3q$h59$1@lust.ihug.co.nz>

Hi, im writing a simple method that takes in a string input and check if the
last charcter is an '/'
then it will do A else do B

my Q now is how to get the last char???

Thanks




From mnations at airmail.net  Thu Sep 12 11:22:07 2002
From: mnations at airmail.net (Marc)
Date: 12 Sep 2002 08:22:07 -0700
Subject: Reading all text in telnet connection
Message-ID: <4378fa6f.0209120722.227fe930@posting.google.com>

Hi:

Having a small problem trying to capture all the text in a telnet
connection. I've used all the options of eager, lazy, etc., and can't
seem to find the right combo.

What I'm doing is intermittently writing data and then reading all of
the response. Unfortunately, all of the response might not immediately
appear. I started off using read_until, but there's nothing really
consistent that I can read to. So I need the ability to simply read
everything current in the buffer until EOF.

What is the best way to accomplish this?

Thanks,
Marc


From gerhard.haering at gmx.de  Sat Sep 14 06:30:08 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 14 Sep 2002 10:30:08 GMT
Subject: Skipping Parts of a For Loop
References: <pan.2002.09.14.12.23.44.103401.723@thomas-guettler.de>
Message-ID: <slrnao63tf.r8.gerhard.haering@lilith.my-fqdn.de>

Thomas Guettler wrote in comp.lang.python:
> Is there a way to skip some parts in a for loop?
> [...]
> Is there a way to skip some parts?
> 
> "continue" skips one element.

"continue" skips the rest of the for loop.

> But how can I skip several?

With "continue" :-)

-- Gerhard


From aleax at aleax.it  Tue Sep 24 08:42:47 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 24 Sep 2002 12:42:47 GMT
Subject: Dynamic loading of modules
References: <PrJj9.12293$Uu.3503@news1.bredband.com> <87sn00zeke.fsf@pokey.henrik-motakef.de> <amnvgg$71f$1@maud.ifi.uio.no>
Message-ID: <b5Zj9.136116$ub2.2967284@news1.tin.it>

Johan Fredrik ?hman wrote:

> I tried this below.  But as you see, even if the Password.py has a "def
> init()" it doesn't work.
> 
> Python 2.2 (#1, Mar 26 2002, 15:46:04)
> [GCC 2.95.3 20010315 (SuSE)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
>>>> z = __import__("plugins.Password")
>>>> z.init()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> AttributeError: 'module' object has no attribute 'init'

__import__ returns the TOP-LEVEL module in the package even when you
import a dotted name.  SO, you need:

x = __import__('plugins.Password')
x = getattr(x, 'Password')

and now you should have the effect you want.

x = x.Password

would be OK too instead of the getattr call, of course, but the
latter has the advantage that you can use a variable to hold
the attribute (submodule) name:-).


Alex



From aahz at pythoncraft.com  Wed Sep 18 16:29:43 2002
From: aahz at pythoncraft.com (Aahz)
Date: 18 Sep 2002 16:29:43 -0400
Subject: Remarks to Python 2.3
References: <20020918222421.328088eb.use-net@schabi.de>
Message-ID: <amannn$p4b$1@panix1.panix.com>

Normally, I tell people posting to python-dev that they're in the wrong
place and should go to c.l.py.  However, this time, I think the reverse
suggestion is in order: you should post your questions to python-dev.

In article <20020918222421.328088eb.use-net at schabi.de>,
Markus Schaber  <use-net at schabi.de> wrote:
>
>I studied http://python.org/dev/doc/devel/whatsnew/contents.html and was
>very impressed - you manage to get python even more fascinating with
>every release.
>
>I especially like enumerate() and Bool.
>
>But I have some remarks:
>
>- Mutable sets support some in-place operations - should those, just as
>the in place operations on lists, like sort(), return None instead of
>the new list to make clear that they are in place operations and for the
>sake of consistency?
>
>- When replacing list elements using the extended slice syntax - e. G.
>in a[::2] = range(0, -2, -1) - the left side must have exactly the same
>numer of elements as the right side. So, to cite the example, a[::2] =
>range(3) is illegal for a=[0, 1, -1, 3]. 
>
>Now I could image that it makes sense to allow one single additional
>element to be added to the list in case only the step is specified and
>the operation doesn't produce a gap. So the given example above would be
>valid, giving a sequence of five elements, whereas with a=[0, 1, 2] it
>would stay illegal (as it would produce a gap). a[0:3:2]=range(3) would
>stay illegal as it explicitly specifies the length of the range,
>a[0:3:5]=range(3) could be allowed.


-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From greek_bill at yahoo.com  Tue Sep 17 16:12:37 2002
From: greek_bill at yahoo.com (Morfeas)
Date: Tue, 17 Sep 2002 21:12:37 +0100
Subject: Instantiating and initialising C++ with embedded Python
Message-ID: <K0Mh9.2591$pr5.174881@newsfep1-win.server.ntli.net>

Hello,

I'm developing a game in C++ and I've come to a point where I think I need 
to use a scripting languge (and why not python?!). 

I have a nice C++ class hierarchy that does all sorts of things. I need to 
be able to create (i.e. instantiate) and initialise a bunch of objects at 
run time. For example I need to do the following (in pseudo-code) :

if (car jumps over obstacle)
{
   award points to player;
}

I could hard-code this with C++, but I'd like the flexibility of doing this 
only for cerain game levels, etc.

Since I've only started reading on Python, can you please tell me if the 
following is right :

1. My main program is in C++, so I need to embed the python interpreter.

2. I need to allow python access to my C++ classes, therefore I need to 
extend the embedded interpreter.

I think I'm ok with doing (1), but not so sure about (2). I've been reading 
the python documentation, but they seem to focus on C rather than C++. I 
believe I need to write some python-friendly class definitions, etc.

Can someone give me some of the following :

a) a brief description of the whole process
b) some code examples
c) suggestions about useful tools

Thanks,

Morfeas


-- 
www.freecfm.com/m/morfeas


From vze3m6zd at verizon.net  Sun Sep  8 13:33:02 2002
From: vze3m6zd at verizon.net (dioscoro)
Date: Sun, 08 Sep 2002 17:33:02 GMT
Subject: Why Python?
References: <mailman.1031446691.2023.python-list@python.org>
Message-ID: <3D7B8AC2.2040008@verizon.net>

> Also, I started recently to learn PHP. I wouldn't recommend it for a 
> newbie at programming, unless he/she needs it.

you not recommend beacuse php is is more dificult? are you? you must 
mean it lacks the the power of c++ or python.




From fafhrd at datacom.kz  Fri Sep  6 01:02:30 2002
From: fafhrd at datacom.kz (Nikolay Kim)
Date: 06 Sep 2002 12:02:30 +0700
Subject: tuple problem
In-Reply-To: <1031287998.6307.44.camel@fafhrd>
References: <1031287998.6307.44.camel@fafhrd>
Message-ID: <1031288576.6308.49.camel@fafhrd>

sorry! it's my error





From mikeb at mitre.org  Tue Sep  3 16:00:16 2002
From: mikeb at mitre.org (Mike Brenner)
Date: Tue, 03 Sep 2002 16:00:16 -0400
Subject: [Q] win32com/Excel
Message-ID: <3D7514D0.902AA220@mitre.org>

Hi Robin, 

That was great, but most importantly, it was FAST!

Would it be possible to tell how you did that so quickly? Is there a reference that would tell me that x1LastCell is in constants? or that you put the activeWindow in front of ActiveCell?

Do you know of any references as to how to do it in Microsoft Project whose object model seems much less documented than Excel's is?

I know how to use Project Explorer in the Visual Basic Editor inside the Microsoft OLE applications like Microsoft Project, and I know how to get the Python list of objects using makepy. However, that does not seem to be enough to figure out many questions in Microsoft Project (and even some questions in Microsoft Excel).

What would be really nice is a document that explains what each object in the list is used for. For example, when opening a Microsoft Project file, it asks a bunch of questions like: are macros okay? should I open related files? should I save the changes I think you made to the previous file? 

I would like to learn how to cancel out those messages when I open a Microsoft Project file from Python. What book or application do you recommend I purchase? Or what file can I download to learn how to do these things?

Thanks,
Mike



>I would like to emulate this Visual Basic statement from Python, using
>Com:
>
>        Range(Range("A1"), ActiveCell.SpecialCells(xlLastCell)).Select
>
>this statement selects the active worksheet (which is, incidentally,
>not similar to UsedRange).  
>
>any clues?  
>
>thank you,
from win32com.client import Dispatch
from win32com.client import constants
xl = Dispatch("Excel.Application")
AB = xl.ActiveWorkbook
AS = AB.ActiveSheet
AW = xl.ActiveWindow
R = AS.Range('A1',AW.ActiveCell.SpecialCells(constants.xlLastCell))
print AB.Name, AS.Name, R
-- 
Robin Becker




From Chris.Barker at noaa.gov  Wed Sep 18 13:01:01 2002
From: Chris.Barker at noaa.gov (ChrisBarker)
Date: Wed, 18 Sep 2002 10:01:01 -0700
Subject: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org>
Message-ID: <amabpb$5rm$1@news.nems.noaa.gov>

terry wrote:

 > Contrast that
> with the following VB code:  Amount = Qty * UnitPrice

> The amount of validation rises exponentially 
> when the calculations in a program are implementation specific, as 
> opposed to language intrinsics.  Each calculation must be tested at the 
> limits of each of the variables involved - at each instance.

Well, sure. This is obviously a stupid way to go about...That's what 
object oriented programming is all about. What you need to do is create 
a currency class that behaves like it should, and then use it everywhere 
you want currency. Then you can do:

Amount = Qty * UnitPrice

just like VB.

> The quick reply to your observation is:  Yes, anything is possible - 
> but is forcing a language into an environment to which it is not 
> intrinsicly suited a reasonable approach?

This is not an issue of suitability of Python.

> I think Python needs to incorporate an intrinsic data type that 
> monetary

Yes, it would be4 very useful to have that data type, but it really 
doesn't make any difference whether it is intrinsic or not.


> and shouldn't be accomplished with overlays on integer or 
> character variable types.

Why not? what do you think VB is doing? All machines that VB runs on use 
binary integer and floating point arithmetic..so a currency data type 
HAS to be based on those. It's just a question of whether you write the 
"overlays" in Python, C, assembly, or whatever. And you are free to 
extend Python in those languages if you like. The only reason I can see 
to not writing your class in Python is performance, and unless you are 
doing a lot of calculations, that is probably a non-issue.

If someone writes a nice one, maybe it could even be put intot he 
standard library.

> Python needs this data type, that is not 
> hardware oriented, to become truly mainstream and be a threat to VB.

Python needs the equivalent of the MS marketing machine to be a "threat" 
to VB.

> I say this with all due respect to all the other positive attributes of 
> Python and it's flexibility to solve a myriad of other non-business 
> application problems.

-Chris



-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov



From jb at cascade-sys.com  Fri Sep  6 21:01:59 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Fri, 06 Sep 2002 18:01:59 -0700
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com>
Message-ID: <3D795007.5000700@cascade-sys.com>

Rod Stephenson wrote:

>Larry Wall Wrote:
>
>"Python is cool to look at small bits of, but I think the "outline"
>syntax breaks down with larger chunks of code. I'm with Aristotle on
>the structure of discourse--a story should have a beginning, and
>middle, and an end. So should blocks"
>
>Any comments?
>
Typically, he substitutes a cute particle of rhetoric for a meaningful 
response.

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	






From tim.one at comcast.net  Sun Sep 29 17:56:50 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 29 Sep 2002 17:56:50 -0400
Subject: Python's import: why doesn't it work?
In-Reply-To: <3d976fbc$0$23666$91cee783@newsreader02.highway.telekom.at>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEJKBHAB.tim.one@comcast.net>

[ Jurie Horneman]
> Actually, I just tried this in a DOS shell under Windows 98:
>
> python -v script.py >output.txt
>
> It lists the verbose output straight to the DOS shell, not to
> output.txt. The output is useless, I can't scroll back.
>
> Oddly,
>
> python -h >output.txt
>
> correctly sends the output to output.txt.
>
> Does anybody know if this is a bug,

Arguably so, but it's a command.com (the Win9x shell) limitation.  Your
first example printed to stdout, your second to stderr, and command.com
simply doesn't allow redirecting stderr, or setting the DOS box history to a
usable value.  cmd.exe (the WinNT/2K/XP shell) does both.

> and how I can work around it?

Programming life under command.com is impossible without tcap:

    http://www.simtel.net/pub/pd/11141.html

It's free, tiny, and works great.  It's the only Win9x "console capturing"
utility I've tried that actually works, and doesn't cause system
instability.




From thorsten at thorstenkampe.de  Mon Sep  9 20:14:35 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Tue, 10 Sep 2002 02:14:35 +0200
Subject: Python primer - comments appreciated!
Message-ID: <aljdhe$1q63hj$1@ID-77524.news.dfncis.de>

"Python, anyone?"... Yep, this is my first one. Comments on
style/readability and technique/flow appreciated!

#v+
def quotient_set(seq, func, partition='nonbool'):
    """ partition <seq> into equivalence classes

    With the 'bool' parameter 'quotient_set' always returns two lists: the
    first  containing the 'true' items and the second containing the false
    ones. """
    if partition == 'bool':
    # Yes, I know about 'filter', but this returns just the 'true
    # elements'...
    
        # put this in one line?
        true_class  = []
        false_class = []

        for item in seq:
            if func(item):
                true_class.append(item)
            else:
                false_class.append(item)
        return [true_class, false_class]
        
    else:
        # put this in one line?
        quotient_set =         []
        representative_class = []

        for item in seq:
            representative = func(item)

            # "Try: seq.index(item)" is IMO rather ugly in a loop. But
            # it's even better than "if item in seq: seq.index(item)",
            # which goes twice through seq
            try:
                quotient_set_index = representative_class.index(representative)
            except ValueError:
                quotient_set.append([item])
                representative_class.append(representative)            
            else:
                # This somehow "inelegant", because all I want to do is:
                # "seq[index] = seq[index].append(item)" Suggestions?
                equivalence_class = quotient_set[quotient_set_index]
                equivalence_class.append(item)
                quotient_set[quotient_set_index] = equivalence_class
                
        return quotient_set
#v-

To get a feeling for what 'quotient_set' does, try:
>>> f = lambda x: x % 3
>>> quotient_set([1, 2, 3, 4, 5, 6, 7, 8, 9], f)
[[1, 4, 7], [2, 5, 8], [3, 6, 9]]

>>> quotient_set([1, 2, 3, 4, 5, 6, 7, 8, 9], f, 'bool')
[[1, 2, 4, 5, 7, 8], [3, 6, 9]]

#v+
def set(seq):
    """ make <seq> a true set by removing duplicates """
    # Shortest way to do a 'null function'? 'None' doesn't work...
    f = lambda x: x  # identity function
    
    return [item[0] for item in quotient_set(seq, f)]    
#v-


From donn at u.washington.edu  Tue Sep  3 12:37:57 2002
From: donn at u.washington.edu (Donn Cave)
Date: 3 Sep 2002 16:37:57 GMT
Subject: Synchronous capture of stdout and stderr
References: <akoe0c$ejn$1@news.tudelft.nl> <IFTb9.588$Dr5.1262@news-server.bigpond.net.au> <akqrc3$a31$1@news.tudelft.nl>
Message-ID: <al2oh5$21o2$1@nntp6.u.washington.edu>

Quoth Konrad Wojas <usenet at wojas.vvtp.tudelft.nl>:
| Neil Hodgson wrote:
| > Konrad Wojas wrote:
| > 
| >> I'm working on a build tool that captures the stdout and stderr of an
| >> application and mails the output if a compile fails. I'm now using
| >> popen2.popen3 and the select module, but I don't receive the two streams
| >> in the same order as the application sends them.
| > 
| >     A common problem here is that the application buffers the streams
| > when writing to a pipe but not when writing to a terminal. Using a pty
| > instead may convince the child application to run unbuffered. Is there a
| > flag (like Python's -u) to run the application in unbuffered mode?
|
| No, the applications are make and gcc :)
|
| I guess I'll have to use popen4 and don't distinguish between stdout and 
| stderr. Maybe I could try writing a LD_PRELOAD wrapper some time.

I wouldn't expect make and gcc to buffer output like that.  Personally,
a single output stream sounds better to me anyway in this case, because
I don't recognize any great semantic distinction between standard and
error outputs in a make job.

But in answer to your immediate question, I would suggest two things -
use the UNIX file descriptor, file.fileno(), and the posix module read
and write functions, to avoid unwanted buffering, and secondly, use
the select.select() function to find output that's ready to read.

	Donn Cave, donn at u.washington.edu


From jeske at chat.net  Sun Sep 15 02:26:08 2002
From: jeske at chat.net (David Jeske)
Date: Sat, 14 Sep 2002 23:26:08 -0700
Subject: html template system?
In-Reply-To: <btug9.688$Ee4.1090@news-server.bigpond.net.au>
References: <btug9.688$Ee4.1090@news-server.bigpond.net.au>
Message-ID: <20020914232608.B20812@mozart.chat.net>

There is am extremely fast template system called Clearsilver which
has some similarities to Cheetah (Python) or Velocity (Java), except
that it's a C-library and very language neutral. A Python binding is
available:

http://www.clearsilver.net/

On Fri, Sep 13, 2002 at 11:24:55PM +0000, :B nerdy wrote:
> what are good templating systems avaliable for python?
> 
> ive been at http://zebra.sourceforge.org and the documentation is kinda sus.
> zope DocumentTemplate is included with zope. but i want a smaller footprint
> system.
> 
> what are my options? cheers
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

-- 
David Jeske (N9LCA) + http://www.chat.net/~jeske/ + jeske at chat.net



From jb at cascade-sys.com  Mon Sep  9 20:38:52 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 09 Sep 2002 17:38:52 -0700
Subject: a re question
References: <pan.2002.09.09.16.10.38.482129.20940@presidency.com>
Message-ID: <3D7D3F1C.8000804@cascade-sys.com>

Rajarshi Guha wrote:

>Hi,
>  I have a file with lines of the format:
>
>001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23
>001 Xya FGh   143557789   7 100 09/05/2002 20:23:23
>
>I am trying to extract the 9 digit field and the single digit field
>immediatley after that.
>
Regex is great but on the surface it appears to be overkill for your 
application.

I would like to suggest some alternatives not using regex.

(A)  IF all the fields are fixed width (up to and including the fields 
of interest, but not necessarily the ones  following) then you can 
extract sub fields by simple indexing into the string.

E.g., assuming a single space or TAB for a separator and that variable 
'line' contains one of the above data lines, then something like

    line[14:23]

would extract the larger numeric field.  (I may have counted wrong -- 
you may have to debug that fragment before using it.)

(B) If the fields are variable width (as your regex suggests) BUT always 
separated by spaces or tabs, you can simply split the line into fields:

    fields = line.split()

and then,

    fields[4] and fields[5]

would contain the nonwhite space contents of your desired numeric 
fields.  The split function (in the string module) takes an optional 
argument to specify separators (e.g., commas) other than whitespace.

I expect these alternatives would be faster than regex, though I have 
not measured to make sure.

If I'm mistaken and the fields are all run together, without whitespace 
separators, then you're stuck with regex.  However, then your existing 
expressions likely need more work to work right in that case.

"There's more than one way to do it!"

Regards

--jb


-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	






From BPettersen at NAREX.com  Fri Sep 13 15:29:14 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 13 Sep 2002 13:29:14 -0600
Subject: Embed extended Python
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192012532DC@admin56.narex.com>

> From: Mathieu Tremblay [mailto:mtremblay at golemlabs.com] 
> 
> Hello, 
>    I am quite new to Python and I'd like to embed extended python. 
> The problem is that the documentation I can find on the 
> internet is always related to C standard types and not to C++ objects.

[snip]

It isn't clear to me exactly what you're having problems with so I'll
just give you the general overview of what you need to do... First you
need to initialize Python and your extension modules. In this case I'm
using SIP to wrap C++ (hence the call to initlibsip), and my wrapped
module is libnrxc.dll).

    PyInitializer::PyInitializer() {
        // Initialize Python
        Py_Initialize();
        
        // initialize SIP (the wrapper generator) library
        initlibsip();

        // initialize the Narex library
        initlibnrxc();

        PyObject* mainmod = PyImport_AddModule("__main__");
        m_namespace = PyModule_GetDict(mainmod);
        Py_INCREF(m_namespace);
        
        // make sure the nrx module is loaded
        try {
            PyObject* m_result = 
                PyRun_String(
                    "import nrx\n",    // has to end in \n
                    Py_file_input, 
                    m_namespace, 
                    m_namespace);

            if (m_result == NULL) {
                // getTraceback() returns the traceback as a
std::string.
                throw new NException(getTraceback().c_str(), N_EX_NULL);
            } else {
                Py_DECREF(m_result);
            }
        } catch(NException *e) {
            std::cout << e->getMessageText();
        }
    }
    
After all that you should be able to call into your C++ module:

    PyObject* res = PyRun_String("nrx.CWheel()", Py_eval_input,
m_namespace, m_namespace);    
    
res is now your C++ instance wrapped in a PyObject. To get back the C++
object, you'll have to call a function from your wrapping library that
will do that for you, e.g. for SIP:

    int error;
    CWheel = (CWheel*)sipConvertToCpp(res, sipClass_CWheel, &error);
    
I haven't used swig in a while, so I don't know what the swig call is,
but you get the idea...

When you want to go the other way (have c++ object, need PyObject),
there is a similar call:

    PyObject* val = sipMapCppToSelf(&myWheel, sipClass_CWheel);

hth,
-- bjorn



From mstenner at phy.duke.edu  Thu Sep 26 10:41:28 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Thu, 26 Sep 2002 10:41:28 -0400
Subject: off topic -- spam assassin
In-Reply-To: <200209261407.g8QE76Wo028975@ratthing-b246.strakt.com>; from lac@strakt.com on Thu, Sep 26, 2002 at 04:07:06PM +0200
References: <200209261407.g8QE76Wo028975@ratthing-b246.strakt.com>
Message-ID: <20020926104128.A12812@phy.duke.edu>

On Thu, Sep 26, 2002 at 04:07:06PM +0200, Laura Creighton wrote:
> On the advice of several of you here, I installed spam assassin, and
> had a pleasant six months of nearly spam free mail.  Thank you, each
> and every one of you who recommmended it.  Alas, the spammers have
> got smarter, and I am now getting 25+ pieces of spam that aren't
> caught by spam assassin every day again. :-(  What's new for our
> side of the arms race?  

Have you upgraded to the most recent version of spamassassin.  My
observations are similar to yours, but staying on the current version
of SA works nicely for me.  (although I don't get as much spam as you,
and maybe your spammers are smarter than mine)

I think the SA heyday is over.  Once it left obscurity, I think the
spammers started using it to tailor their messages.  This is an
inevitability and only means that we'll need to stay on top of the
most recent version.

					-Michael

-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From michaels at one.net  Fri Sep 20 08:07:34 2002
From: michaels at one.net (Michael Schneider)
Date: Fri, 20 Sep 2002 08:07:34 -0400
Subject: Python sequences reference - problem?
References: <mailman.1032443760.5578.python-list@python.org> <YZli9.114624$ub2.2439313@news1.tin.it>
Message-ID: <3D8B0F86.9000600@one.net>

I also like python.

When doing pure math  I have been using J for prototyping algorithms.

J is basically APL without the special symbols .


Good paper with many links for APL and J:

ftp://watserv1.uwaterloo.ca/languages/apl/software-library/index.html

http://www.jsoftware.com/



These are expecially good for heavy linear algerbra,

Have fun,
Mike

Alex Martelli wrote:

>Michal Kaukic wrote:
>
>>Hi,
>>   
>>   we like Python - I use it (with Numeric and SciPy modules)
>>for teaching of Numerical analysis and my colleague is doing
>>some research in discrete optimization in Python. And he often
>>surprises me with non-conformal use of language :-)
>>
>>   We are mathematicians, not programmers. I suppose, he believes
>>that Python "thinks" like Mathematician. Yesterday, he was struggling
>>
>
>I think the concept of "MODIFYING an object" is pretty far from
>most areas of maths I've dabbled in (I'm an engineer, not a
>mathematician, but do find maths fascinating).  It seems to me
>that the only real way to get a language that WILL "think like
>a mathematician" is to take your pick among languages that have
>no concept of "modifying an object" - which nowadays basically
>means, Functional Programming languages.  Personally I find
>Haskell the most elegant one -- clean syntax (uses whitespace
>much like Python), clean semantics (lazy evaluation the default
>everywhere).  Last I checked, the best Haskell compilers were
>able to produce very decent code for many problems -- performance
>will in most cases be better than Python's.  If you want even
>MORE performance, I suspect some less-puritanical but still FP
>language (such as O'Caml) can probably deliver it.
>
>If for some reason you don't want to get into FP, then I think
>you can forget everything about "thinking like a mathematician".
>If data modification IS allowed, then your chosen language needs 
>to choose either of two possibilities:
>
>-- COPY-semantics: an assignment such as
>        x = y
>   inherently copies y, so that any future change to y has no
>   more effect on x
>
>-- REFERENCE-semantics: an assignment such as
>        x = y
>   makes x a synonym of y, so that both refer to the same
>   object: any future change to that object affects any
>   synonym for the object equally
>
>Many languages choose a MIX of these two kinds of behavior,
>but that's the worst of both worlds, as it complicates things
>a lot.  Generally, OLD languages (such as Fortran) relied on
>copy semantics (references or synonyms were not considered,
>and when they could happen, as in argument-passing, they were
>prohibited by the language rules, although those rules were
>rarely enforced).  Modern languages by and large have moved
>to reference semantics, which is generally much cleaner even
>though in some cases performance can suffer (optimization is
>the main reason some languages complicate things by mixing
>reference and copy semantics).  "Generally" is important: LISP,
>one of the oldest languages around (second-oldest after
>Fortran, among those still widely used today), relied on
>reference semantics from the start.  The resulting elegance
>has much to do with Lisp's acceptance in the mathematical
>community, IMHO.
>
>Python uses reference semantics.  When you want a copy, you
>ask for a copy -- it's as simple as this.  If you don't ask
>for a copy, you don't get a copy... you get a reference to
>the original.  There is no complication: it's _always_ this
>way.  (OK, OK, at the margin there are doubtful cases: in
>particular, a SLICE is a copy-request for all built-in types,
>but a reference, sharing data with the original array, when
>you use the Numeric package... Numeric had to move _away_
>from copy and onto reference in this case to avoid accidental
>copies of huge array slices... take this as the exception
>that proves the rule, OK?-).
>
>Python does have some data types that are immutable, which
>helps move it a little bit closer to FP -- numbers, strings
>and tuples are immutable.  (Even Java has strings as
>immutables -- SOME of FP's niceties do start to move into
>the mainstream, albeit slowly:-).  But the main style is still
>very much NOT FP, rather relying on modifiable data.
>
>>with the code shown below (somewhat simplified):
>>    
>>
>>L_L=[[1,2,3], [2,3,4], [3,4,5]]
>>
>>def do_process(LL):
>>    n=len(LL[0])
>>    rec = [0]*n
>>    Res=[]
>>    
>>    for row in LL:
>>        rec[0] = row[0]
>>        # further code - modifying rec
>>        # lots of conditional processing and such...
>>        # ...
>>        Res += [rec]
>>                
>>    return Res
>>
>>---------------------------------------------
>>
>>After calling do_process(L_L), the "expected" result should be
>>
>>   [[1, 0, 0], [2, 0, 0], [3, 0, 0]]
>>
>>but the true result is
>>
>>   [[3, 0, 0], [3, 0, 0], [3, 0, 0]].
>>
>>
>>   Yes, this is fully in accordance with how the Python language should
>>behave:
>>       Res += [rec] inserts references to list object rec,
>>       which are further modified... (he should use copy(rec) instead).
>>
>
>Right in all respects, yes.
>
> 
>
>>   But there is nothing to make this behaviour clearly VISIBLE in code.
>>
>
>*EVERY* assignment in Python uses reference semantics.  Every occurrence
>of an assignment, therefore, makes reference behavior "clearly VISIBLE".
>
>Presumably fish don't find water "clearly VISIBLE", being used to that,
>much as it is for us and air.  When something "is *EVERYWHERE*", how
>can it be "clearly VISIBLE" at the same time?-)
>
>
>>If I work with pointers in C/C++ I know and see they are pointers.
>>
>
>That's because C uses copy semantics, and uses pointer to emulate
>reference semantics when needed.  That's just the reverse of Python,
>which uses reference semantics, and explicit calls to copy to
>emulate copy semantics when needed.
>
>C++ lets you define "references" which IMPLY reference semantics
>rather than copy semantics -- and you don't SEE what they are at
>the point of use (only at the point of definition).  Flaming about
>that here, one way or another, won't be very useful:-).  But there
>are LOTS of precedents: Pascal (!!!) lets you declare arguments
>to a procedure as either reference or copy -- then you don't see
>the difference at the point of use; Visual Basic emulated that, too
>(the default has changed in recent releases of VB, but that's
>-another- flame:-); languages with sophisticated macro systems,
>such as Dylan or relatively-recent LISPs, let you do all that and
>worse.
>
>But C and Python are SIMPLE languages: one semantics is pervasive
>(reference for Python, copy for C), you take explicit measures
>when you want the OTHER semantics, and those measures are visible
>at the point of use (well, _mostly_ -- C does complicate things
>a bit by having arrays "decay to" pointers, Python has the
>"slice aren't copies in Numeric" issue -- but that's definitely
>at the margin).
>
>>You can say - we also know that rec is list object and so be careful
>>with it.  Yes, but consider the complex code where the similar constructs
>>are very easy to overlook. And debugging of such code can be frustrating.
>>
>
>The issue has nothing to do with rec being a list object, really.
>You may see the issue as having to do with the fact that rec is
>(any kind of) *MUTABLE* -- indeed, a language without mutable data
>DOES shield you from these issues.  If you tried to mutate non-
>mutable data, as rec[0] = .... does, you'd get an exception at
>runtime (if every kind data was immutable, of course, the language
>wouldn't even have any CONSTRUCTS to express mutation).
>
>
>>   My colleague was in state of despair and made thoughtful remarks
>>about FORTRAN and Python similarity. He believes that Python is corrupting
>>his (computer) memory...
>>
>
>Fortran (up to F77 -- I lost touch with it afterwards) had reference 
>semantics in all parameter passing to subprograms (with aliasing forbidden 
>in rather complex ways -- very few compilers ever enforced that, letting
>the hard-to-debug cases for the programmers:-)), copy semantics for all 
>assignments.  I can't see any parallel between that and Python's
>behavior, not even in the "anti-parallel" sense of Python // C.
>
>
>>   So what is the point? I wrote this message in hope that there are
>>more people with similar experience. My question is - how to explain
>>to novice non-programmer users (maybe mathematically "infected")
>>the quirks of Python sequence objects? Which methodology to use
>>
>
>First of all, somebody with reasonable math bent will prefer a
>more general explanation -- so, forget sequence objects, because
>that's just one specific cases.  ALL object-mutation works the
>same way, whether it be of a sequence or non-sequence.
>
>>in programs so we can clearly see we work with "pointers"?
>>
>
>You're in water.  Water is *everywhere*.  When you think there
>is nothing there, there's water.  Just get used to it.  Trying
>to muddy up the water in an attempt to make it more visibile
>would be counter-productive.
>
>>Or maybe someone can propose the changes to the Language to overcome
>>this (psychological) barrier? I feel this as a serious obstacle
>>in using Python (my students experienced it too).
>>
>
>It appears, though you never state it, that the behavior you
>would consider "natural" would have to do with copy semantics.
>I find that weird, absolutely weird -- I've seen it happen to
>old died-in-the-wool programmers without solid mathematical
>bases, because for 20 or 30 years of their lives they had only
>seen copy semantics and could no longer conceive of any other,
>but that's exactly the reverse from the "mathematical bent"
>you keep mentioning.
>
>Accepting your mention of "mathematical bent" as relevant, I
>repeat my suggestion that you look into functional programming.
>FP is seriously cool, IF all the programmers involved have a
>solid mathematical outlook on life, the universe and everything.
>
>If for whatever reason you want to keep using languages based
>on the idea of data modification, you won't find a simpler or
>more regular one than Python.  And it would be absurd to pick
>old-enough languages that use copy semantics because some
>previous experience (NOT with maths itself, surely?!) makes
>you feel that's natural and divergences from it should be
>made "very VISIBLE" -- besides Python, you'd be cutting yourself
>off from Java, Eiffel, C# ... reference semantics IS the wave
>of the last decade-plus, and probably of the next one too
>(unless FP comes into its own, but I've sort of stopped holding
>my breath for THAT -- there just isn't enough of an audience
>with a pervasive-enough mathematical turn of mind:-).
>
>Just remembed that EVERYTHING works by reference, and what
>could possibly be the problem...?
>
>
>Alex
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020920/40ca22ad/attachment.html>

From mstenner at phy.duke.edu  Tue Sep 10 08:39:24 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Tue, 10 Sep 2002 08:39:24 -0400
Subject: UNIX processes
In-Reply-To: <200209101233.05796.b.maryniuk@forbis.lt>; from b.maryniuk@forbis.lt on Tue, Sep 10, 2002 at 12:33:05PM +0200
References: <200209101233.05796.b.maryniuk@forbis.lt>
Message-ID: <20020910083924.A4265@phy.duke.edu>

On Tue, Sep 10, 2002 at 12:33:05PM +0200, Bo M. Maryniuck wrote:
> Hi, all.
> Somebody know how to list all the processes in RIGHT way on Linux/UNIX?
> I need to find the zombies, stopped processes etc.

The problem is that almost every different UNIX(like) system has a
different interface to the kernel/process table.  You have a couple of
choices:

  1) use the program "ps", which does the right thing for any given
     os/kernel and presents a fairly standardized report, or

  2) pick the os(s) that you care about, and write your own code to do
     it directly.  On linux, that can be done pretty nicely through the
     /proc/ filesystem ("man proc").

					-Michael
-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From djc at object-craft.com.au  Mon Sep 23 11:39:35 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 24 Sep 2002 01:39:35 +1000
Subject: M$SQL connections from linux
References: <1032528860.599940@news.etel.ru> <memo.20020923162224.544B@a.a>
Message-ID: <m3lm5sd8aw.fsf@ferret.object-craft.com.au>

>>>>> "Mark" == Mark Charsley <mark.charsley at REMOVE_THIS.radioscape.com> writes:

Mark> In article <1032528860.599940 at news.etel.ru>, black at flamingo.ru
Mark> (Vasiliy A.
Mark> Chernoivan) wrote:

>> Hello all Is there some module for python which performs the task?
>> Is it possible at allto access data under M$SQL from linux machine?

Mark> Not used it myself, but mxODBC should do the job:

Mark> http://www.egenix.com/files/python/mxODBC.html

Or this:

        http://object-craft.com.au/projects/sybase/

Compile with the latest FreeTDS from CVS (not 0.53)

        http://object-craft.com.au/projects/sybase/sybase/node5.html

You may have a small problem with 0.35pre1 and FreeTDS in which case
you should try this version of Sybase.py until I make a new release
(in a week or so):

        http://object-craft.com.au/pipermail/python-sybase/2002-September/000059.html

- Dave

-- 
http://www.object-craft.com.au


From jb at yahoo.de  Mon Sep  9 16:01:32 2002
From: jb at yahoo.de (JB)
Date: Mon, 09 Sep 2002 22:01:32 +0200
Subject: Python 2.3
Message-ID: <3d7cfc19_5@news.newsgroups.com>

Will rational numbers be implemented in Python 2.3? And when 
is the first beta due?

TIA,
-- 
Janos Blazi


-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From aleax at aleax.it  Wed Sep 25 17:57:13 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 21:57:13 GMT
Subject: Getting started
References: <uojinmcfapjf45@corp.supernews.com> <AFik9.51323$b%3.25248@fe03> <mailman.1032979596.18965.python-list@python.org> <3D920950.323B25B3@alcyone.com>
Message-ID: <Ziqk9.167772$pX1.5989755@news2.tin.it>

Erik Max Francis wrote:

> "James J. Besemer" wrote:
> 
>> However "lame" or "weak" you may consider this feature to be, it does
>> NOT constitute "weak typing" per the formal definition.  This is
>> simple
>> polymorphism, as you seem to acknowledge, same in Perl as in Python.
> 
> The problem here is that there is no official "formal definition."  Some
> use "weak typing" to mean any type system where the typing can be foiled
> by casts; some use it to mean any system where operations, not the
> objects themselves, determine the types of the objects.  (I favor the
> latter definition, since it carries more information about the
> language.)  This is simply a case of differing terminology.

Bingo -- and I think that if one works through all the references I
gave one will see Cardelli cover all the major issues, though he does
not do so in any single one of them and his exact terminology may
differ from mine, yours, James's, "Some authors'" (e.g. Scheme's R4R :

"""
Scheme has latent as opposed to manifest types. Types are associated with 
values (also called objects) rather than with variables. (Some authors 
refer to languages with latent types as weakly typed or dynamically typed 
languages.) Other languages with latent types are APL, Snobol, and other 
dialects of Lisp. Languages with manifest types (sometimes referred to as 
strongly typed or statically typed languages) include Algol 60, Pascal, and 
C. 
"""

and so on.  I don't think it's all that pragmatically useful to have
a black/white definition for properties, in real life -- it may be
good maths (or at least, much-simpler maths) but "practicality beats
purity".  In geometry, a 2D shape either is or isn't a circle: there
is no "being more circular" or "being less circular" -- it's black
and white.  In real life, we don't really care in most application
for a 1-micron deviation on a shape of 1 meter's diameter -- we can
and do consider it "pretty much a circle" for all practical purposes.
One whose deviation is 10 microns is going to be "less circular"
but still OK for most purposes, etc, etc.

It may be important to know the extent and character of the deviation,
of course.  For example, Java or Eiffel are compile-time type-safe
*EXCEPT* for some covariance issues -- in arrays for Java, in
procedure parameters for Eiffel.  C++ has no covariance issues,
but does have explicit casts, unions, etc.  Fortran has no covariance
nor cast issues, but does give compliant compilers a license to NOT
enforce all checks that would be needed to ensure all language rules
are complied with -- IF you follow all the language rules you're OK,
BUT the fact that your program compiles correctly with a compliant
compiler does NOT give you certainty that the rules are indeed
followed -- in Fortran IV and Fortran 77, procedure parameters,
equivalence statements, and mismatch of common statements for the
same common between separately compiled subprograms, are all points
to check (some compilers can check _some_ rules -- all are indeed
enforceable, at runtime at worst, but Fortran's typical emphasis on
speed means they often aren't).

On the other side of the fence, it's clearly trivial to design
a language which can have no type problems -- just give the language
just one type, period.  It's even easy for the compiler to ensure
that you are indeed using that single type (as it's the only one
you have) wherever it's required (i.e., everywhere).  It's of no
bloody practical use whatsoever, of course.


In practice, we want practically useful definitions of things --
"weak typing" is one, but clearly not the only one.  Eg., in another
post James claimed that standard C has no objects -- something
that will no doubt break Kernighan and Ritchie's hearts, since
they write:
"""
An object is a manipulatable region of storage; an lvalue is 
an expression referring to an object
"""
and so on.  [the issue IS important -- an rvalue *NEED NOT*
"refer to an object" in C, while an lvalue *DOES*].

Clearly, James meant "objects as in OO", but that's not
what I wrote and to which he was vehemently responding -- no
matter, he clearly has his own definition of what "an object"
IS, no matter what one of a language's inventor claims an
object is *in the language he invented*...


Our discipline hijacks many terms of everyday use -- e.g.,
to describe something that is done, or is to be done, in a
generic way, I can't safely call it a process nor a task
nor a routine nor a procedure nor a job nor ... all have
specific technical meanings in some language and/or operating
system, and/or in the literature.

It's easier in Italian -- because most such terms are used
in the English original, therefore most Italian equivalents
are free for their ordinary use.  Similarly in the late
Roman Republic some Latin scholars apparently claimed that
Greek was the only language suited to philosophy -- and
didn't note that Latin *with* embedded Greek phrases was in
fact much better, because technical terms appeared as such
and didn't hijack the everyday, ordinary significance
of the word or phrase.


Alex



From eugene at ulka.com  Wed Sep 18 18:38:55 2002
From: eugene at ulka.com (Eugene Pervago)
Date: 18 Sep 2002 15:38:55 -0700
Subject: Loading modules from several directories
References: <a1247c5b.0209171610.243f3c02@posting.google.com> <33803989.0209172247.2e6afc1f@posting.google.com> <a1247c5b.0209180607.843526e@posting.google.com> <6Q1i9.1149$bh1.989@fe03>
Message-ID: <a1247c5b.0209181438.3a1c0c9a@posting.google.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message news:<6Q1i9.1149$bh1.989 at fe03>...
> "Eugene Pervago" <eugene at ulka.com> wrote ...
> 
> [...]
> > If I put the project_a path first, it loads its modules but can't
> > import the /onedir/foo modules. It seems package can only be in one
> > place, once Python finds it, it just stops looking elsewhere.
> >
> 
> What would you rather have it do? Load them both? Find every possible
> package and load the oine with the most recent modification date/time? Load
> the longest one? Load the one whose checksum is closest to your birthdate?

They do not intersect. Actually the structure is something like this:
/onedir/foo
   dbutils.py
   utils.py
   auth.py

/anotherdir/foo - Nothing
/anotherdir/foo/projectA
    User.py
    Product.py

In Unix I would do it using symlinks, not possible on Windows. I can
think of two solutions:
 
1) Copy /onedir/foo/*.py to anotherdir/foo/ for each project 
   (Doesn't seem like a good solution, though easily doable since the
common
    files are under version control)
2) Instead of foo.projectA have a foo_projectA or fprojectA to avoid
problems
   with nested packages.

Neither of those two is particularly attractive to me. Any
suggestions?

Thanks,
Eugene Pervago


From ajs at ix.netcom.com  Tue Sep 17 11:04:49 2002
From: ajs at ix.netcom.com (Arthur)
Date: Tue, 17 Sep 2002 11:04:49 -0400
Subject: More on Protecting Source Code
Message-ID: <001c01c25e5b$92a4e470$9865fea9@arthur>

Janes writes:
<snip>
>It is precisely the promise of commercial reward that makes America the
>incredible fount of invention and creativity that it is.
<snip>

You sound like a neo-something  - like myself.

To me the open source movement has offered the opportunity to maintain a
subversive stance in the context of an essentially "conservative" world
view.  Sort of wholesome subversion. Makes the folks looking to make $ from
their on intellectual property work a bit harder - if they are competing
with folks willing to throw theirs into the ring for free. No question it
has *driven* innovation.


That -

And there is certainly the opportunity for abuse of processes that are
themselves reasonably sound - Enron and friends in some instances, Disney
(for example, and in a totally different sense - and in some ways for me,
closer to home), I would argue, in some others.

The worst thing "we" could do is leave the high ground screaming on these
kinds of situations to the folks with an essentially different world view..

Art







From sholden at holdenweb.com  Fri Sep 13 15:23:51 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 13 Sep 2002 15:23:51 -0400
Subject: Problem with HTTPS request
References: <uo2fveepf88175@corp.supernews.com> <b0f083db.0209130558.2aa60398@posting.google.com> <uo45ara0l62cdc@corp.supernews.com> <slrnao4dkn.288.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <mYqg9.8349$WV.1529@fe03>

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote ...
> Christian Reyes wrote in comp.lang.python:
[...]
> >> "Christian Reyes" <creyes at ea.com> wrote in message
> > news:<uo2fveepf88175 at corp.supernews.com>...
> ^^
>
> Broken quoting.
>
> >> > [...] Note: HTTPS support is only available if the socket module was
compiled
> > with
> ^^
>
> Broken quoting again. Please throw away that OE crapware or fix it
> using Morver or whatever.
>
Gerhard:

You're showing distinct signs of having too much time on your hands :-)
Can't you fix your brain instead so it will let you ignore other people's
mailer problems?

an-unreconstructed-OE-user-myself-ly y'rs  - steve
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From gmuller at worldonline.nl  Sat Sep  7 06:44:48 2002
From: gmuller at worldonline.nl (GerritM)
Date: Sat, 7 Sep 2002 12:44:48 +0200
Subject: Smoothing a discrete set of data
References: <3csmdso5.fsf@morpheus.demon.co.uk>
Message-ID: <alcleu$c1n$1@reader1.tiscali.nl>

"Paul Moore" <gustav at morpheus.demon.co.uk> schreef in bericht
news:3csmdso5.fsf at morpheus.demon.co.uk...
<..snip...>
> I have a set of data, basically a histogram. The data is pretty
> variable, and I'd like to "smooth" it to find trends. Actually, this
> comes up a *lot* for me - some examples: sampled IO rates on a machine
> - I want to look for trends (is IO higher overnight or during the day,
> etc) or fuel consumption for my car (do I use less fuel since I had
> the service).
<..snip...>
> Can anyone help me? I can't believe that this problem has never come
> up before, but I can't find any literature on it.
>
Yes, this problem is very common. For instance in Physics it occurs very
often. In physics the approach would be to fit "lines", where the line form
depends on the particular physics.

In your example it sounds as if less theory is available (how about
operations research?), which means that you have less a priori knowledge of
the signal. Your brain picks out the evident peaks.

One approach is to filter out the high frequent noise by simply applying a
low pass filter, ie a weighted average of neighboring points.

A more complex approach would be to assume some distribution in peaks (for
instance Gaussian) and fit a set of these peaks over the data. Depending on
your expectation the peaks can be parameterized for width and amplitude for
instance.

I expect that in the end your interpretation is still required, but the job
might be eased by this kind of computer assistance.

> Thanks in advance,
> Paul Moore.

regards Gerrit

--
www.extra.research.philips.com/natlab/sysarch/




From phoebe_1 at att.net  Sat Sep 21 10:27:10 2002
From: phoebe_1 at att.net (Holden Caulfield)
Date: 21 Sep 2002 07:27:10 -0700
Subject: Solaris 64 bit compilation
References: <c2595393.0209201844.2ed62265@posting.google.com> <m3sn03k44m.fsf@mira.informatik.hu-berlin.de>
Message-ID: <c2595393.0209210627.44f672a8@posting.google.com>

Please see insert. I apoligize for not providing all the info, as I
did not have
access to the machine.

martin at v.loewis.de (Martin v. Loewis) wrote in message news:<m3sn03k44m.fsf at mira.informatik.hu-berlin.de>...
> phoebe_1 at att.net (Holden Caulfield) writes:
> 
> >  This is my first attempt at compiling Python v2.2.1 in *64bit* mode 
> > on a Solaris 5.9 system. I am using the Sun Compiler (the latest I do not
> > remember the version). I am migrating a few of my apps (all developed on
> > 64bit platforms) to Solaris, so I *need* the 64bit mode.
> 
> 
> I can't follow your conclusion - if you migrate applications which
> have been developed on 64-bit platforms, there is no inherent need to
> compile them as 64-bit binaries on Solaris 9.
> 

Ok, What I have is Python C-Extensions that I wrote for a third party
vendors
software. Unfortunately, they have decided to ship *only* 64 bit
libraries for
Solaris. So, you see I need to build python in that mode.

> > But, it is failing miserably. It compiles *fine* in 32bit mode (the default),
> > but it does me no good. As some of the extensions library that I have are
> > 64bit. Before, I start hacking I thought I would give a query here and see
> > if anyone had any success?
> 
> It compiles fine for me, with SunPRO 6U1.

What I have is :
cc: Sun WorkShop 6 update 1 C 5.2 2000/09/11
Like I said, it compiles fine for the default mode, which is creating
32bit
extensions,modules. Like I said before, I *need* to have it in 64-bit
objects,
modules,extensions etc.

> 
> > 
> > Here are my preliminary results:
> > - Flags used "-xarch=v9 -xcode=pic32".
> > The first thing I get is the bail-out in "pyport.h":
> >  Where the #LONG_BITS != (8 * sizeof(LONG)) fails!!!
> 
> Please try to report failures carefully. It does not say sizeof(LONG),
> instead, it says SIZEOF_LONG. Which gives rise to the question: What
> value does SIZEOF_LONG have, in pyconfig.h?
> 
cc -c -DNDEBUG -g -Xc xarch=v9 -xcode=pic32 -I. -I./Include
-DHAVE_CONFIG_H  -o Modules/p
ython.o Modules/python.c
"./Include/pyport.h", line 480: #error: "LONG_BIT definition appears
wrong for platform (bad gcc/glibc config?)."


> > So, I commented it out, first to proceed and see what happens. Most of them
> > compiled fine except, I got some serious errors in the:
> > fileobject.c file probably related to the LP64 model.
> > 
> > I will try to add more info, when I get around to it.
> 
> More info would be required if you want help. Start with reporting the
> error messages as the compiler produces them, instead of interpreting
> them, and reporting your interpretation.
> 

So after commenting it out, here is what I get.

cc -c -DNDEBUG -g -Xc xarch=v9 -xcode=pic32 -I. -I./Include
-DHAVE_CONFIG_H  -o Objects/f
ileobject.o Objects/fileobject.c
"Objects/fileobject.c", line 282: argument #2 is incompatible with
prototype:
        prototype: long : "/usr/include/iso/stdio_iso.h", line 209
        argument : union  {double _d, array[2] of int _l}
"Objects/fileobject.c", line 325: return value type mismatch
"Objects/fileobject.c", line 355: assignment type mismatch:
        union  {double _d, array[2] of int _l} "=" long
"Objects/fileobject.c", line 393: assignment type mismatch:
        union  {double _d, array[2] of int _l} "=" long
"Objects/fileobject.c", line 407: operands have incompatible types:
         union  {double _d, array[2] of int _l} "==" int
"Objects/fileobject.c", line 462: operands have incompatible types:
         union  {double _d, array[2] of int _l} "==" int
"Objects/fileobject.c", line 468: argument #1 is incompatible with
prototype:
        prototype: long : "./Include/intobject.h", line 37
        argument : union  {double _d, array[2] of int _l}
"Objects/fileobject.c", line 548: argument #2 is incompatible with
prototype:
        prototype: union  {double _d, array[2] of int _l} :
"/usr/include/unistd.h", line
 341
        argument : long
"Objects/fileobject.c", line 550: operands have incompatible types:
         union  {double _d, array[2] of int _l} ">=" int
"Objects/fileobject.c", line 551: assignment type mismatch:
        union  {double _d, array[2] of int _l} "=" long
"Objects/fileobject.c", line 553: operands have incompatible types:
         union  {double _d, array[2] of int _l} "<" int
"Objects/fileobject.c", line 555: operands have incompatible types:
         union  {double _d, array[2] of int _l} ">" union  {double _d,
array[2] of int _l
}
"Objects/fileobject.c", line 555: operands have incompatible types:
         union  {double _d, array[2] of int _l} ">=" int
"Objects/fileobject.c", line 556: operands have incompatible types:
         unsigned int "+" union  {double _d, array[2] of int _l}
"Objects/fileobject.c", line 556: operands have incompatible types:
         unsigned int "-" union  {double _d, array[2] of int _l}
"Objects/fileobject.c", line 1173: cannot recover from previous errors
cc: acomp failed for Objects/fileobject.c
*** Error code 2
make: Fatal error: Command failed for target `Objects/fileobject.o'

> Regards,
> Martin


From marco at reimeika.ca  Sun Sep 22 01:34:36 2002
From: marco at reimeika.ca (marco)
Date: 22 Sep 2002 01:34:36 -0400
Subject: "time" module under AIX?
References: <yibhegi7l40.fsf@coxeter.math.toronto.edu>
Message-ID: <yibelbm7fkj.fsf@coxeter.math.toronto.edu>

marco <marco at reimeika.ca> writes:

> I've just compiled python 2.2.1 under AIX and all seemed
> to go well. However, I'm having some trouble importing
> a few modules such as "time" and "math" e.g.

OK, it seems it was a permission problem. I compile/install
stuff as a non-privileged user and then change ownership.
However, it turns out that "make install" doesn't use the
"install" command consistently and sometimes just copies
stuff (without setting any permissions). In my case the
files in lib-dynload ended up being 750 and this was causing
problems (there may be others, I have to check carefully)...

Cheers,

--
marco at reimeika.ca
Gunnm: Broken Angel
http://reimeika.ca/


From bokr at oz.net  Mon Sep 16 13:58:10 2002
From: bokr at oz.net (Bengt Richter)
Date: 16 Sep 2002 17:58:10 GMT
Subject: Converting Python app to C++ completely
References: <aklm6c$r1r$1@news.swman.net.uk> <3D85D9EB.9020102@thinkware.se> <roy-2A7271.09563416092002@reader1.panix.com>
Message-ID: <am563i$4m5$0@216.39.172.122>

On Mon, 16 Sep 2002 09:56:34 -0400, Roy Smith <roy at panix.com> wrote:

>Magnus Lycka <magnus at thinkware.se> wrote:
>> A common approach is to factor out the performance critical
>> parts (once you have actually measured) and rewrite them in
>> C or C++. The profiler is a good tool for this measuring.
>
>I'm in the process of doing exactly that.  I've got an app I wrote in 
>Python which needed speeding up.  Profiling showed that 99% of the CPU 
>time was spent in one class, which did a lot of low-level character 
>processing.  Mostly what it did was thrash around creating string 
>objects and tearing them down again in a rather un-pythonic way.
>
>Seemed like a perfect opportunity to learn C++, so I re-wrote the class 
>in that language.  Somewhat surprisingly, I only got about a 2x speedup.  
>I think the reason is because now instead of spending all my time 
>allocating and deallocating Python strings, I'm doing the same with C++ 
>strings :-)
>
>I'm going to take a shot at re-writing it again using C-style arrays of 
>characters.  It'll be interesting to see how much speedup I get.  I 
>figure one of two things will happen; either the C version will be much 
>faster than the C++ version, or it won't.  Either way, I figure I will 
>have learned something about C++, and maybe about Python too :-)

Maybe you could post the culprit class and see what alternative _algorithms_
people come up with. It sounds like the badness is in your methods ;-)

Regards,
Bengt Richter


From robie at 1373.net  Mon Sep 23 22:30:21 2002
From: robie at 1373.net (robie1373)
Date: Mon, 23 Sep 2002 20:30:21 -0600
Subject: confusion about opening files
Message-ID: <3d8fcda4$0$198$75868355@news.frii.net>

I am trying to open a file,

>>> import os
>>> file = os.open("e:\\d12", "r")
Traceback (most recent call last):
  File "<pyshell#4>", line 1, in ?
    file = os.open("e:\\d12", "r")
TypeError: an integer is required

I get this error if the mode is r, r+, a+ or whatever.  I also get it
whether the file exists or not. Can someone explain what I am doing wrong?




From dsavitsk at e-coli.net  Fri Sep  6 19:34:32 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Fri, 06 Sep 2002 23:34:32 GMT
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <albdk2$2iro$1@norfair.nerim.net>
Message-ID: <cYae9.7120$yt3.3465240@newssrv26.news.prodigy.com>

"Gillou" <nospam at bigfoot.com> wrote in message
news:albdk2$2iro$1 at norfair.nerim.net...
> "Rod Stephenson" <viking_kiwi at yahoo.com> a ?crit dans le message de news:
> wkr8g7rn4c.fsf at yahoo.com...
> > Slashdot has a list of questions posed to Larry Wall (perl). When
> > asked his thoughts on other scripting languages, he makes the
> > following observation about python
> >
> > "Python is cool to look at small bits of, but I think the "outline"
> > syntax breaks down with larger chunks of code. I'm with Aristotle on
> > the structure of discourse--a story should have a beginning, and
> > middle, and an end. So should blocks"
> >
> > I'm not quite sure what he's trying to get at here - I guess that for
> > a long heavily indented chunk of code, you could lose track of the
> > overall structure, but I don't write code this way.
> >
> > Any comments?
>
> Long heavily indented chunks of code is a really bad programming style
> (whatever's the language).
> Modern python editors can find blocks, hide/show blocks, auto
indent/dedent
> blocks, and find start/end of blocks.

Further, those of us used to using, say, the treeview in Outlook [Express]
to read newsgroups, or other similarly structured interfaces, find the
hierarchy of Python indenting quite natural.

-d


> Tabnanny does a fine job on ugly indenting.
>
> def longfunction():
>     # 300 lines of code below
>     ...
>     return
> # end longFunction()
>
>




From mhammond at skippinet.com.au  Wed Sep 25 18:43:13 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 25 Sep 2002 22:43:13 GMT
Subject: PythonWin CVS problem
References: <mailman.1032936513.10390.python-list@python.org>
Message-ID: <5_qk9.20506$6g7.64886@news-server.bigpond.net.au>

David LeBlanc wrote:
> I'm using the same repository I've been using for the past 6 months or more:
> 	:pserver:anoncvs at cvs.pythonpros.com:/home/cvsroot
> 
> Just tried it again and got the same: "permission denied".
> 
> David LeBlanc
> Seattle, WA USA 
> 
> 
>>-----Original Message-----
>>From: python-list-admin at python.org
>>[mailto:python-list-admin at python.org]On Behalf Of Dilton McGowan II
>>Sent: Tuesday, September 24, 2002 20:30
>>To: python-list at python.org
>>Subject: Re: PythonWin CVS problem
>>
>>
>>David LeBlanc wrote:
>>
>>
>>>Second day it's inaccessable with "permission denied"
>>>
>>>David LeBlanc
>>>Seattle, WA USA
>>
>>Are you using cvs.pythonpros.com? It worked for me yesterday and 
>>again just 
>>now. They seem to have the latest source. Just a friendly suggestion.
>>-- 
>>http://mail.python.org/mailman/listinfo/python-list

It may be a local permissions problem.  Try checking out the tree to a 
new directory.

Mark.



From etralex at ua.pt  Mon Sep 23 16:23:28 2002
From: etralex at ua.pt (Ricardo B)
Date: Mon, 23 Sep 2002 21:23:28 +0100
Subject: Very complex Python/C++ embedding/extension question(s)
References: <3d8cfd3c$1@news.swissonline.ch>
Message-ID: <1032812577.470847@newsfront3>

On Sun, 22 Sep 2002 01:13:23 +0200, Bernhard Glueck wrote:

> Hi there
> I am developing a new multimedia/game/vr architecture, and at the moment
> i am trying to embed python as a scripting language. I have a slew of
> questions about the Python/C API..
> 
> 
Your architecture could do with some better explanation. Anyway, based on
that you can add new Actor/Type classes in run-time, I'll assume that
Actor/Type's child classes don't extend their parent's interface and that
you have a system where new classes can be dynamically registred.

Therefore, you only need a C++ to Python wrapper to the base class, and
you can use that wrapper for instances of child classes. 
The same holds true for Python to C++ wrappers.
You should provide an interface for the childclasses written in Python to
register themselves and to see what classes are avaliable, the sameway you
do with the C++ classes.

About the third question: 
Why don't you just use actor.Call("DoSomething", ...) in Python code, as
you do in C++?
You're just begging for increased confusion.


-- 
	Ricardo



From andrewm at object-craft.com.au  Tue Sep 17 23:24:36 2002
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Wed, 18 Sep 2002 13:24:36 +1000
Subject: How to make a C extension module backwards compatible?
Message-ID: <20020918032436.0DC653C21B@coffee.object-craft.com.au>

Is there any community wisdom on how to make a C extension module that
implements a new-style class (type?) backwards compatible (albeit,
at the cost of some functionality)?

Put another way, how *should* I make a class implemented in C behave
like a new-style class when installed with python 2.2 and above, and
like an old style class when installed with older pythons?

Should I just riddle the code with #ifdef's, or is there a more elegant
way?

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/



From spoermeg at voldelig.com  Tue Sep 17 19:33:37 2002
From: spoermeg at voldelig.com (Terje Johan Abrahamsen)
Date: Tue, 17 Sep 2002 18:33:37 -0500
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <mailman.1032152058.11043.python-list@python.org> <m3admizdzq.fsf@mira.informatik.hu-berlin.de>
Message-ID: <am8epu$3ml4d$2@ID-53055.news.dfncis.de>

"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m3admizdzq.fsf at mira.informatik.hu-berlin.de...
> Gerhard H?ring <gerhard.haering at gmx.de> writes:
>
> > You should perhaps reconsider if and why you need to hide your source
> > code.
>
> You should also recognize that, if your computer can execute and
> perform a certain algorithm, users can always observe simulate the
> computer operations, thus breaking any protection scheme.
>
> You'll have to give your customers a sealed computer and no permission
> to install additional software, or communicate with the internet, if
> you want to reliably prevent them from analysing your program.

I guess that will be a little overkill. I am not writing military software
or software that has uniqe algoritms. I just want to make sure that if
someone is going to steal my program to sell it themselves, that they have
to actually write the code. Not only get mine, remove the copyrights, fill
in what they want and sell it away. If Py2Exe stores the source in pretty
much open for reading, I need to use another way of protecting it. But, the
programs will be availiable for download, and will be possibly to copy. I
just want to make it a little harder.




From martin at v.loewis.de  Sat Sep 21 19:28:22 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 22 Sep 2002 01:28:22 +0200
Subject: Python 2.2.1 compatible with Tcl/Tk 8.4?
References: <yu9965x05vkr.fsf@europa.research.att.com> <j4ptv89zwk.fsf@informatik.hu-berlin.de> <u1kjuo34.fsf@morpheus.demon.co.uk>
Message-ID: <m3y99u7wix.fsf@mira.informatik.hu-berlin.de>

Paul Moore <gustav at morpheus.demon.co.uk> writes:

> Hmm, experimenting, it looks like it doesn't. So I must have to tell
> Python to use the installed TCL 8.4. How?

You need to recompile _tkinter.pyd.

Regards,
Martin


From marklists at mceahern.com  Mon Sep 23 19:06:34 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 23 Sep 2002 18:06:34 -0500
Subject: list problem
In-Reply-To: <-29615135.1032818024099.JavaMail.nobody@webmail2.brturbo.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOKENPEAAA.marklists@mceahern.com>

[jubafre at brturbo.com]
> i have:
> x=[['0020', '0000x9'], ['0030', '0000xa'], ['00B4', '0000x8'],
> ['0030', '0000xb'], ['00F0'], ['0001'], ['0003'], ['0005']]
>
> and i want:
> y=['0020', '0000x9', '0030', '0000xa', '00B4', '0000x8', '0030',
> '0000xb','00F0', '0001',
> '0003', '0005']
>
> how to do this?

You want to flatten a sequence.  Here's a recipe that does that:

  http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/118845

I'll leave it as an exercise for you to deal with the string type
appropriately.  I think the recipe will try to flatten the string (although
I have not tested to confirm that) and that's probably not what you want.

// m

-




From brueckd at tbye.com  Thu Sep 26 11:22:00 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Thu, 26 Sep 2002 08:22:00 -0700 (PDT)
Subject: Python threading?
In-Reply-To: <ikEk9.26204$sR2.463163@news4.ulv.nextra.no>
Message-ID: <Pine.LNX.4.44.0209260809490.22517-100000@ns0.tbye.com>

On Thu, 26 Sep 2002, Thomas Weholt wrote:

> This might be just a stupid question, but I'm trying to implement a
> filesharing-app, based on the BaseHTTPServer in the standard python-distro.
> This will of course be heavy IO-processing involved. Will threading still be
> a bad idea? The project is aimed at small user-groups ( I do not think one
> person will have more than say 5-10 concurrent connections at a time, most
> of them downloading files 1MB + in size),

At those low load levels, threads will be fine.

> as if threading is a very bad idea no matter what, still it's used almost
> everywhere ( I got no statistics to back up that statement ;-) ).

You have to factor in the issue of complexity too though. The threading
approach is almost always much simpler and many applications don't need to
be high performance, so threading is acceptable. That said, the asynch
approach _is_ useful to know and understand - if for no other reason than
it makes you understand the problem from a very different angle. And after
awhile it really starts to "click" in your head and make sense. State 
machines are nifty. :)

> NB! It would be nice to be able to limit the number of threads, actually the
> number of concurrent connections, on the server too, so if anybody has any
> comments on some way of doing that and still handle request gracefully

Is the client side a web browser or a custom app? Your request handler 
could return a nice "server is busy" (e.g. HTTP 503) message to a browser, 
but if you want to explicitly cap the number of threads you'll need to 
subclass the server. It's been awhile since I've looked at SocketServer, 
but I think it has a check_request or verify_request or something method 
that you could override to send back an HTTP 503 if the server is 
overloaded (already has N threads running) and then return false (as in, 
"don't process this request any further).

-Dave




From phncontact.libero.it at   Wed Sep  4 03:59:23 2002
From: phncontact.libero.it at  (Benjamin)
Date: Wed, 04 Sep 2002 07:59:23 GMT
Subject: Servizi Internet
References: <mailman.1031056925.7383.python-list@python.org>
Message-ID: <f9fbnu8dghga0tbpvf7nsv64jo3csrlsm1@4ax.com>

by the way, how many people understand italian in
this nesgroup? ^_^


From danb_83 at yahoo.com  Sun Sep 22 00:36:04 2002
From: danb_83 at yahoo.com (Dan Bishop)
Date: 21 Sep 2002 21:36:04 -0700
Subject: Math optimization
References: <6T5j9.145247$pX1.5195840@news2.tin.it> <mailman.1032645989.8709.python-list@python.org>
Message-ID: <ad052e5c.0209212036.47bd0590@posting.google.com>

jepler at unpythonic.net wrote in message news:<mailman.1032645989.8709.python-list at python.org>...
> Did you bother benchmarking?  I'd be surprised to see a huge difference,
> since the overhead of the bytecode loop should be much much greater than
> the cost of a single division operation.

There is a very slight difference.  On my PC,

a / 2.0 takes 824 ns
a * 0.5 takes 784 ns


From mhammond at skippinet.com.au  Tue Sep 17 07:14:57 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 17 Sep 2002 11:14:57 GMT
Subject: win32security.LogonUser not working
References: <x6Pg9.8439$yt3.4473830@newssrv26.news.prodigy.com> <L9Pg9.8440$yt3.4474038@newssrv26.news.prodigy.com> <ZT2h9.8834$yt3.4540499@newssrv26.news.prodigy.com>
Message-ID: <3D870F4B.7000904@skippinet.com.au>

{posted and mailed]

dsavitsk wrote:
> I have gone so far as to assign every possible right in dir(ntsecuritycon),
> but i am still getting the same error. I have also assigned nearly everyone
> and every thing on my network the right to act as the os. any ideas?
> 
> -d
> 
> 
> "dsavitsk" <dsavitsk at e-coli.net> wrote in message
> news:L9Pg9.8440$yt3.4474038 at newssrv26.news.prodigy.com...
> 
>>should have mentioned
>>python:   2.1.3
>>win32all: 145
>>win2k:    sp2
>>
>>
>>"dsavitsk" <dsavitsk at e-coli.net> wrote in message
>>news:x6Pg9.8439$yt3.4473830 at newssrv26.news.prodigy.com...
>>
>>>I am trying to run code as another user. I copied the code below from
>>
>>ASPN,
>>
>>>so it would seem that someone has gotten this to work. I keep getting
>>
> the
> 
>>>error
>>>
>>>pywintypes.api_error (1314, 'LogonUser', 'A required privilege is not
>>
> held

This is a *privilege* issue, not a *permissions* issue.

This same issue happens when trying to reboot a server - from the 
RebootServer.py script I posted recently:

import win32security
import win32api
import sys
import time
from ntsecuritycon import *

def AdjustPrivilege(priv, enable = 1):
     # Get the process token.
     flags = TOKEN_ADJUST_PRIVILEGES | TOKEN_QUERY
     htoken = 
win32security.OpenProcessToken(win32api.GetCurrentProcess(), flags)
     # Get the ID for the system shutdown privilege.
     id = win32security.LookupPrivilegeValue(None, priv)
     # Now obtain the privilege for this process.
     # Create a list of the privileges to be added.
     if enable:
         newPrivileges = [(id, SE_PRIVILEGE_ENABLED)]
     else:
         newPrivileges = [(id, 0)]
     # and make the adjustment.
     win32security.AdjustTokenPrivileges(htoken, 0, newPrivileges)

Look for C/C++ samples using the LogonUser function in the way you are 
trying to, and I am confident you will find similar calls to the above 
in that sample code.

Let us know if you sort it out - or better still, send me a sample .py 
file that uses it which I can include in win32all :)

Mark.



From hst at empolis.co.uk  Tue Sep 17 05:53:23 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Tue, 17 Sep 2002 10:53:23 +0100
Subject: Regular expression help needed
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB04F41A@hendrix.empolisuk.com>

Torkil Grindstein wrote:
> Sent: 17 September 2002 10:04
> To: python-list at python.org
> Subject: Regular expression help needed
> 
> 
> Hi.
> 
> It's been a while since I played around with regexps, and I
> realize that I really need some help here.
> 
> Mission: I have a string containing a whole WML document. I want
> to extract data from following occurences:
> 
> <meta name="michael" content="owen"/>
> 
> That is, I want to extract the content of any occurences of
> the meta "michael" key. There may be several occurences in one
> document.
> <meta name="michael" content="owen"/>
> <meta name="michael" content="jackson"/>
> will give me the results "owen" and "jackson".
> 
> Of course, it is possible that the document has written the
> tags in uppercase (META Name, META NAME, etc), but I could
> lowercase the whole document string prior to searching. (The
> content is case insensitive for my purposes.)
> 
> I would really be glad if someone out there took this challenge..:)
> 
> Cheers,
> Torkil
> -- 

Superficially, this is easy:

>>> s='<meta name="michael" content="owen"/><meta name="michael" content="jackson"/>'
>>> r=re.compile('<meta\s+name="michael"\s+content="([^"]*)"')
>>> r.findall(s)
['owen', 'jackson']

but I can't really recommend it as it contains too many implicit assumptions:

1) There are no comments or processing instructions that could contain similar data
2) The attributes always come in the order specified i.e. name first, content second, any others following
3) The attributes are always double quoted and there are no spaces surrounding the =

All of these can be got round by using REs, but as the data is in (presumably well-formed) XML an easier solution would be to parse the XML using SAX and simply examine the attributes of each element named "meta" or "META".

Harvey

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.



From tjabo.kloppenburg at unix-ag.org  Wed Sep 18 22:41:52 2002
From: tjabo.kloppenburg at unix-ag.org (Tjabo Kloppenburg)
Date: Thu, 19 Sep 2002 04:41:52 +0200
Subject: Regex question
In-Reply-To: <amapuf$45t8t$1@ID-144898.news.dfncis.de>
References: <amapuf$45t8t$1@ID-144898.news.dfncis.de>
Message-ID: <20020919044152.50f6bab2.tjabo.kloppenburg@unix-ag.org>

hi,
"Eric Arnold" <umop at psyon.org> wrote:

> The string I'm capturing is as follows:
> Name:AAA,BBB,CCC

why regexp?

just do:
splitted1 = s.split(":",2)     # --> splitted[0] == Name
events = splitted1.split(",")  # = list of events

it is not that cool, but it runs without brainsucking experiments... :)

tk.
--
Tjabo Kloppenburg
  --=- In  95.82 Tagen ist Weihnachten! -=--

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020919/b8ed832d/attachment.sig>

From megabytemonster at hotmail.com  Tue Sep  3 16:29:23 2002
From: megabytemonster at hotmail.com (Stuart)
Date: 3 Sep 2002 13:29:23 -0700
Subject: winsound.Beep() frequencies
Message-ID: <3ab719f7.0209031229.33667461@posting.google.com>

While looking at the python documentation, I noticed that there are no
frequency/pitch tables for using with the winsound module.  Does
anyone know whether this documentation exists, and/or what the
corresponding frequencies to pitches are?

Thanks,
Stuart


From neel at mediapulse.com  Thu Sep 12 18:37:20 2002
From: neel at mediapulse.com (Michael C. Neel)
Date: Thu, 12 Sep 2002 18:37:20 -0400
Subject: [mod_python] ANNOUNCE: Mod_Python donated to ASF
Message-ID: <C0FC22C08B82074A88B50061764157772C63C6@johnson.mediapulse.net>

Let me be the first to say, in the fashion that the Internet has taught
us to celebrate a great victory or other action....

Woot.

Mike

--
Michael C. Neel
There are only 10 types of people in the world;
those who understand binary and those who don't.


-----Original Message-----
From: Gregory (Grisha) Trubetskoy [mailto:grisha at modpython.org] 
Sent: Thursday, September 12, 2002 6:26 PM
To: mod_python at modpython.org; python-list at python.org;
announce at apache.org; dev at httpd.apache.org
Cc: jim at apache.org
Subject: [mod_python] ANNOUNCE: Mod_Python donated to ASF



It is my pleasure to announce that Mod_Python has been donated to the
Apache Software Foundation, and is now a subproject of the httpd server
project (see http://httpd.apache.org/).

I am grateful to ASF for accepting this donation and committing
resources
to further the support of Mod_Python. I believe that this action will
advance the development of Mod_Python, resulting in an ultimately better
and more popular tool for Python developers. I also believe it will
serve
to better position Python as a language of choice for web development, a
need that has been expressed by many.

There are no implications to the current Mod_Python users - the license
is
the same with the sole difference in that the copyright belongs to ASF
now.

As a consequense of the donation, the CVS repository is now hosted on
cvs.apache.org.  Do not use the SourceForge repository anymore, it will
soon be removed.

There will also be website and mailing changes, but the details are
still
being finalized and will be announced when ready.

Regards,

Grisha Trubetskoy


_______________________________________________
Mod_python mailing list
Mod_python at modpython.org
http://www.modpython.org/mailman/listinfo/mod_python



From tchur at optushome.com.au  Thu Sep 12 17:04:22 2002
From: tchur at optushome.com.au (Tim Churches)
Date: Fri, 13 Sep 2002 07:04:22 +1000
Subject: Detecting Russian and Ukrainian character sets
Message-ID: <3D810156.67A62B78@optushome.com.au>

Here are two questions for Russian and Ukrainian Python users:

1) I understand that a common problem when processing text data 
collected from various sources in Russia and the Ukraine is 
the mixture of character sets which are used - MS-DOS, Windows,
Linux, Unix and mac machines may all use one (or more) of a number
of character sets to encode strings, and when such data are 
supplied in text files, there is usually no indication of
which character set was used. Is this correct?  
http://czyborra.com/charsets/cyrillic.html has a listing of
known Cyrillic character sets.

2) Are there any Python routines available for automatically
deducing which character set was used to encode a particular
text file (or a particular string)? There is a module for
Perl called Lingua:RU:Charset which seems to address this problem
(see
http://www.freebsd.org/cgi/url.cgi?ports/russian/p5-Lingua-RU-Charset/pkg-descr
)
at least for Russian encodings.

Tim C



From roy at panix.com  Sun Sep 22 21:08:23 2002
From: roy at panix.com (Roy Smith)
Date: Sun, 22 Sep 2002 21:08:23 -0400
Subject: more on weak typing..
References: <uospj56t55o858@corp.supernews.com>
Message-ID: <roy-50C656.21082322092002@reader2.panix.com>

"Lance" <lbrannma at cablespeed.com> wrote:
> So Python is both dynamically typed (because it doesn't use explicit
> datatype declarations) and strongly typed (because once a variable has a
> datatype, it actually matters).

A favorite interview question of mine is to ask the applicant to 
classify a whole bunch of attributes as more likely to belong to a 
compiled language, an interpreted language, or neither.  I'm not so much 
interested in the actual answers as I am to see that the applicant knows 
what I'm talking about.

One of those attributes is "strongly typed".  I once had somebody tell 
me that compiled languages are strongly typed because they support 
longer variable names (i.e. it takes more keystrokes to type in the 
program).


From sholden at holdenweb.com  Tue Sep 10 18:57:14 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 10 Sep 2002 18:57:14 -0400
Subject: Pickle problem
References: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220D10@hendrix.empolisuk.com> <mailman.1031675293.21444.python-list@python.org>
Message-ID: <IOuf9.123572$l_4.80128@atlpnn01.usenetserver.com>

"Michael Chermside" <mcherm at destiny.com> wrote ...
[good advice on getting pickled]

> PS: If what you are trying to do is save in the pickle the exact way
> that the "match" method works because you are afraid that the
> maintainers of Python may come out with a new version of Python in which
> match works differently and is no longer backward compatible, and you
> want your code to still work the same way in this hypothetical new
> version of Python, then you really *DO* need to store (2).
> Unfortunately, it's still not possible (AFAIK). Fortunately, the
> maintainers Python wouldn't do that to you.
>

Not if you make a contribution to the PSU, about which I recently heard





From mfranklin1 at gatwick.westerngeco.slb.com  Mon Sep 16 12:06:58 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Mon, 16 Sep 2002 16:06:58 +0000
Subject: getting username from uid
In-Reply-To: <pan.2002.09.16.10.05.30.511334.24338@presidency.com>
References: <pan.2002.09.16.10.05.30.511334.24338@presidency.com>
Message-ID: <200209161606.58840.mfranklin1@gatwick.westerngeco.slb.com>

On Monday 16 Sep 2002 2:05 pm, Rajarshi Guha wrote:
> Hi,
>   when I do a stat on a file I can get the UID of the owner. IS there
> anyway to get from UID to username (or directly to username for that
> matter)?

On *nix  look at the pwd standard module...

$ pydoc pwd
Python Library Documentation: module pwd

NAME
    pwd

FILE
    /usr/local/lib/python2.2/lib-dynload/pwd.so

DESCRIPTION
    This module provides access to the Unix password database.
    It is available on all Unix versions.

    Password database entries are reported as 7-tuples containing the
    following
    items from the password database (see `<pwd.h>'), in order:
    pw_name, pw_passwd, pw_uid, pw_gid, pw_gecos, pw_dir, pw_shell.
    The uid and gid items are integers, all others are strings. An
    exception is raised if the entry asked for cannot be found.

FUNCTIONS
    getpwall(...)
        getpwall() -> list_of_entries
        Return a list of all available password database entries, in arbitrary 
order.
        See pwd.__doc__ for more on password database entries.

    getpwnam(...)
        getpwnam(name) -> entry
        Return the password database entry for the given user name.
        See pwd.__doc__ for more on password database entries.

    getpwuid(...)
        getpwuid(uid) -> entry
        Return the password database entry for the given numeric user ID.
        See pwd.__doc__ for more on password database entries.

DATA
    __file__ = '/usr/local/lib/python2.2/lib-dynload/pwd.so'
    __name__ = 'pwd'



HTH
Martin
-- 
### Python Powered Signature
I started writting this email on 
Mon Sep 16 16:05:22 2002




From martin at v.loewis.de  Mon Sep 30 15:39:00 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 30 Sep 2002 21:39:00 +0200
Subject: About data structure
References: <3d98a5d8$0$496$7a628cd7@news.club-internet.fr>
Message-ID: <m3vg4nw9m3.fsf@mira.informatik.hu-berlin.de>

malan <malan at club-internet.fr> writes:

> I need to build a data structure like this :
> 
> size	4 bytes ( example 0x0000000A )
> flag	1 byte  ( example 0x87 )
> trans   8 bytes ( example 'URTVMI1' )
> F5	3 bytes ( example 0x00125C )
> 
> How can I define that structure, format it and send it via a socket ?

Your specification is incomplete, as you don't give the byte
order. Assuming it has to be big-endian, that would be

>>> import struct
>>> struct.pack(">IB8s3b", 0xA, 0x87, "URTVMI1", 0x0, 0x12, 0x5c)
'\x00\x00\x00\n\x87URTVMI1\x00\x00\x12\\'

For details, help(struct).

HTH,
Martin


From claird at starbase.neosoft.com  Thu Sep 19 08:57:40 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 19 Sep 2002 07:57:40 -0500
Subject: Need advice on choosing the technology for intranet based application
References: <3d88fd1e@news.ivm.net> <Gn7i9.1769$ik2.430@fe02>
Message-ID: <108666F172046BF0.AFF78CED7027291C.080DCC403BC7489C@lp.airnews.net>

In article <Gn7i9.1769$ik2.430 at fe02>,
Steve Holden <sholden at holdenweb.com> wrote:
>"Nenad Propadovic" <propadovic.nenad at debitel.net> wrote ...
			.
		[much thoroughly
		sensible advice]
			.
			.
>> The GUI has a part I consider complex: I have to present some statistics
>in
>> form of charts or graphs.
>
>The Python Image Library (PIL) is very good for creating graphics. You can
>create them "on the fly" in your web server and then serve back HTML
>documents with references to the created graphics.
			.
			.
			.
It strikes me that Mr. Propadovic is tackling
several new techniques and technologies at once.
Given that, I have an alternative to suggest for
"some statistics in form of charts or graphs".

There's a trick for making making simple bar
charts which requires no extra modules or pro-
gramming.  It's so simple, in fact, that it makes
barcharting easy in shell or assembler or C.

Apparently I've only documented it in Tcl <URL:
http://wiki.tcl.tk/barchart > (see references
there, though, for Python hints).  If you need
a sample dilation in Python, just say so here.

Also, Mr. Propadovic:  take Steve's advice to 
heart.  Even the CGI about which he writes need
not be as challenging as your comments suggest
you fear it to be.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From jslowery at hotmail.com  Fri Sep 20 02:24:14 2002
From: jslowery at hotmail.com (Jeremy Lowery)
Date: Fri, 20 Sep 2002 01:24:14 -0500
Subject: Suggesting a filename on a HTTP-request by sending someting in the header of the reply
References: <Pusi9.20169$0p1.363761@news2.ulv.nextra.no>
Message-ID: <r5zi9.247091$2L.9053878@e3500-atl2.usenetserver.com>

> I got a simple webserver serving files, but the files are found using
> lookup based on ids ( like http://localhost/dl?id=10, where 10 is a key in
a
> dictionary pointing to a file ), not their filename so when I send
> content-type = 'application/octet-stream' in the headers to the client I
get
> the proper "Save as ..."-dialog, but the filename is all screwed up. It
seem
> to be the "dl"-part of the url above. I've tried to read the RFC
> 2616-document and sent "Content-Disposition"-headers etc. but with no
luck.
>
> What headers do I need to send for it to get the proper filename in the
> "Save as ..."-dialog? Say for instance that the filename I wanted the
client
> to save the file as was "test.zip".

The Content-Type (along with the file extension in Internet Explorer)
dictate what the browser
will do when it recieves the file. Internet Explorer also completely
ingnores content-types, and
it tries to guess what the content-type of the document is. Depending on the
MIME type
configuration of the client, it may
prompt to download, automatically open the file, etc. The easiest (and most
common way) to to
include this functionality is to put "right click and save target as" near
the link. the
Content-Disposition will give a source of the file (a filename; very useful
when using CGI to
dynamically send files), but it has nothing to do with the content-type and
how the browser
reaacts to it.

Content-Disposition: file; filename=%(filename)s

is the proper way to use the http header to show the "filename" of what is
being downloaded.


J Lowery






From cbbrowne at acm.org  Fri Sep 20 23:39:48 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: 21 Sep 2002 03:39:48 GMT
Subject: Decimal arithmetic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amabpb$5rm$1@news.nems.noaa.gov> <3D8B1537.4010803@thinkware.se> <ad052e5c.0209201128.1a53108@posting.google.com> <amgcgf$5e1fd$1@ID-125932.news.dfncis.de> <7xn0qcfafp.fsf@ruckus.brouhaha.com>
Message-ID: <amgpm4$5si22$1@ID-125932.news.dfncis.de>

Paul Rubin <phr-n2002b at NOSPAMnightsong.com> wrote:
> Christopher Browne <cbbrowne at acm.org> writes:
>> The COBOL (or PL/I, or anyone else accustomed to BCD data types) guy
>> will justifiably look at the bizarro FP rounding as being, well,
>> bizarre.
>> 
>> There's _no_ good reason for subtracting two perfectly good decimal
>> values from one another to result in having weird artifacts appear in
>> some 17th decimal place, aside from having a willingness to admit
>> fairly deranged arithmetic.
>
> Um, put $1000.00 in an imagined bank account and have it collect 5%
> annual interest, compounded daily for 25 years.  How much money is in
> the account at the end?
>
> The "true" answer is (1+(.05/365))**(365*25) dollars, assuming 365 day
> years.
>
> The IEEE double precision answer will be extremely close to that
> amount.
>
> What happens if you do that calculation in BCD?  How would you even
> specify how to do it?  I don't think you will get nearly as much
> accuracy.

REALITY CHECK for a moment.

What do you think A Real Live bank uses to compute interest?
Financial transactions are most certainly /not/ computed based on the
quasi-infinite precision that you are imagining.

Were you expecting IEEE double precision arithmetic, applied over the
period of 25 years?

Or perhaps, instead, might they use BCD arithmetic, applied on each
date that interest is to be accrued?

If I put $1000 in a /real/ bank account, earning 5% interest,
compounded daily, paid monthly, it is more than just /slightly/ likely
that the bank in question would, using code likely written in PL/I or
COBOL, put money into the account every month, with amounts rounded
based on BCD arithmetic.

The "true answer" will most certainly be the BCD-based answer; that's
the amount of money you'd have in the account 25 years later.

If you protest that the "true" answer is supposed to be
(1+(.05/365))**(365*25) dollars, and that the bank is short-changing
you by 7 cents, the bank will quite glibly tell you to look at the
interest policy, which /won't/ be based on waiting 25 years to apply
the interest.
-- 
(reverse (concatenate 'string "gro.mca@" "enworbbc"))
http://cbbrowne.com/info/oses.html
One man's constant is another man's variable.  -- Alan Perlis


From BPettersen at NAREX.com  Mon Sep  9 14:58:48 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 9 Sep 2002 12:58:48 -0600
Subject: Embedded Python implementation
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201253160@admin56.narex.com>

> From: Evan [mailto:tagith1394 at hotmail.com] 
[snip]
> I have inserted a small amount of C to print the contents of 
> com_list, 
> right before the PyRun_String, here is the complete output (*** marks 
> beginning and end of the script)
> 
>   ***
> 112 (p) 114 (r) 105 (i) 110 (n) 116 (t) 32 ( ) 34 (") 104 (h) 101 (e) 
> 108 (l) 108 (l) 111 (o) 34 (") 10 (
> )  ***
>       File "<string>", line 2
> 
>         ^
> SyntaxError: invalid syntax

I can't tell from your output, but there are a couple of gotcha's:

 - the line terminator needs to be the same as for your
   Python installation.
 - The last character(s) in the string needs to be a line
   terminator.

hth,
-- bjorn



From steve at ferg.org  Fri Sep 13 08:32:04 2002
From: steve at ferg.org (Stephen Ferg)
Date: 13 Sep 2002 05:32:04 -0700
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <ac677656.0209111225.4a141bc5@posting.google.com> <1a3a9ffa.0209120856.1de06f57@posting.google.com> <b16e4ef7.0209121745.44e91aa@posting.google.com>
Message-ID: <b16e4ef7.0209130432.7fb3ee61@posting.google.com>

> I can't really say WHY using a class structure is desirable.  

It is coming back to me now.

As I was working on easygui.py (http://www.ferg.org/easygui/), I
noticed two things. (1) I was using global variables a lot.  (2) I had
several GUI "chunks" in one Python program, and when they were all at
the same level, the code was really unstructured.  Very hard to
understand.

Granted, a lot of my confusion was simply newbie confusion.  But I
thought that the next Tkinter development I did, I would see if
structuring the GUI component into classes would help with these
problems.  Even in the limited development I've been doing for
"Thinking in Tkinter", it seems to me that: (1) being able to refer to
"self." variables removes the need to make so many variables global. 
(2) Grouping the GUI material into classes really helps bring some
structure to the program that it didn't have before.  I suspect that
the importance of this (which may not be that obvious in small
programs) will become overwhelmingly obvious in larger and more
elaborate GUIs. One of my "to do" items is to test this hypothesis by
re-writing easygui.py using a class structure.

So right now, my hypothesis is that -- although Tkinter technology
does not *require* you to structure your GUI as class -- doing so is
probably desirable simply from the standpoint of making a large
program more manageable.


From edwardam at interlix.com  Sun Sep 22 02:00:46 2002
From: edwardam at interlix.com (Edward Muller)
Date: 21 Sep 2002 23:00:46 -0700
Subject: Python HTTPSConnection/HTTPResponse Problem
Message-ID: <9abedeb.0209212200.5c669a50@posting.google.com>

I am trying to use HTTPSConnection and HTTPResponse objects to process
payments with Authorize.net. I've search (using google) the net and
couldn't find any solid examples, so I'm not sure if I'm doing this
right.

I *know* that I am creating and sending the request correctly because
Authorize.net sends me an email with the data in as verification. The
problem is that calling conn.getresponse() (actually calling
con.sock.makefile('rb',0) also does this) ends up hanging until the
connection times out.

Authorize.net AFAIK sends only one thing back in response to the HTTPS
POST, and that is a comma seperated list of values that include the
condition of the transaction (as you can interact with the customer).

My though was that perhaps the things get really messed up because the
data passed back isn't a normal HTTP response ... so I tried playing
with the behind the scenes stuff (peeking into httplib.py) .. . and
like I said above ... calling conn.sock.makefile('rb',0) ... which
looks like what I need to do to get a file object hangs as well ...

So I dunno...

Oh and PS conn is a HTTPSConnection instance ...

Any help would be very appreciated!


From martin at v.loewis.de  Tue Sep 17 01:33:43 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 17 Sep 2002 07:33:43 +0200
Subject: Pickling slotted instances
References: <mailman.1031805375.9030.python-list@python.org> <j4fzwf349u.fsf@informatik.hu-berlin.de> <mailman.1032144560.28257.python-list@python.org> <m3it16ze98.fsf@mira.informatik.hu-berlin.de> <mailman.1032158478.7315.python-list@python.org> <j44rcqpfam.fsf@informatik.hu-berlin.de> <mailman.1032223960.7874.python-list@python.org>
Message-ID: <m3n0qhch8o.fsf@mira.informatik.hu-berlin.de>

Andrew McNamara <andrewm at object-craft.com.au> writes:

> I'd agree with this. The __slots__ notation is a bit of a wart, but then
> again, it's consistent with several other tricks.

Depending on which tricks you mean, I think they all deserve proper
syntax (static and methods, properties, super, ...), e.g.

class Foo:
  def m1(kl)[classmethod]:
    pass
  def m2()[staticmethod]:
    pass
  def m3()[get('x')]:
    return -1

> I see your point, but it seems like most of the machinery is already
> there.  It's a shame that slotted classes aren't more "complete",
> because they are the perfect answer to people who use tuples instead
> of classes for performance reasons (they are only slightly slower
> than a tuple).

A shame for whom? As I said - unless somebody contributes a solution,
there is none.

Regards,
Martin


From DennisR at dair.com  Mon Sep 16 20:07:03 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Tue, 17 Sep 2002 00:07:03 GMT
Subject: Windows AutoRunning Python Hello -- feedback requested
Message-ID: <Hmuh9.70$Ha6.5066815@newssvr14.news.prodigy.com>

For Windows platforms (only, sorry), I have packaged enough Python to import
the os.py library and its dependencies for distribution as an EXE to systems
not having Python installed.  Python is running as a CGI program on top of a
web server I wrote so that the interface is HTML, supporting IE 4+, Netscape
4.7+, and Opera (4+?).  The demo is at http://www.dair.com/hellopy.zip  When
decompressed and run, the EXE displays "hello" in a browser window and shows
its running directory as evidence that os.py is functioning.  Compressed ZIP
size is 401 Kb, expanded is 922 Kb.

1) Is this the right group to discuss/get feedback?
2) Is this functionality useful enough to others to pursue?
3) Other comments?

Thanks,  Dennis
--

Dennis Reinhardt

http://www.dair.com




From barry at zope.com  Fri Sep 13 11:45:16 2002
From: barry at zope.com (Barry A. Warsaw)
Date: 13 Sep 2002 11:45:16 -0400
Subject: RELEASED email package version 2.3
References: <mailman.1031715254.9581.python-list@python.org> <slrnanttkg.1mk.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <ba7khpzyer.fsf@anthem.wooz.org>

>>>>> "GH" == Gerhard H?ring <gerhard.haering at opus-gmbh.net> writes:

    GH> Barry A. Warsaw wrote:
    >> I've just released the (standalone) email package version 2.3.

    GH> Is there any summary on changes, possible incompatibilities
    GH> and deprecated features in comparison with the email package
    GH> in Python 2.2 available?

Yes, there's a detailed NEWS file in the package, which describes all
the changes in the history of the package.  There's lots of new stuff
and a few deprecated APIs, so it would be worthwhile reading the NEWS
file over (I'm not posting it here due to size).

The documentation is still lagging, but I intend to update that during
the Python 2.3 alpha cycle.

-Barry





From bowman at montana.com  Fri Sep 27 23:58:27 2002
From: bowman at montana.com (bowman)
Date: Fri, 27 Sep 2002 21:58:27 -0600
Subject: Better terminal for windows?
References: <3D94B037.2020202@motorola.com>
Message-ID: <7I9l9.12791$V64.3332431@newsfeed.slurp.net>

Stephen Boulet wrote:
> 
> If I want to run python interactively (and perhaps using ipython), is
> there something better than the "cmd" terminal?

There is a standalone tcsh that works rather nicely.



From never at mind.info  Wed Sep 18 07:38:36 2002
From: never at mind.info (G. Willoughby)
Date: Wed, 18 Sep 2002 12:38:36 +0100
Subject: Could someone please try and py2exe this?
Message-ID: <am9ojd$8tv$1@newsg3.svr.pol.co.uk>

I've had lots of trouble trying to 'py2exe' this app. i keep getting import
errors, can someone else please try it and see if you have any luck? If so
could you please upload it somewhere where i could get it or send me it via
email?

Its a Win32 app BTW:

http://www.thecalm.fsnet.co.uk/MapperGUI/files/MapperGUI%20v1.6%20Win32.zip

Thanks,

--G. Willoughby
email: sab(at)freeuk(dot)com




From chris.gonnerman at newcenturycomputers.net  Sun Sep 29 23:42:52 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sun, 29 Sep 2002 22:42:52 -0500
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05> <an7bvd$burft$1@ID-125932.news.dfncis.de> <an7p7t$cpg$0@216.39.172.122> <mailman.1033338759.6775.python-list@python.org> <7xd6qwqtgz.fsf@ruckus.brouhaha.com> <mailman.1033352259.30842.python-list@python.org> <7xelbcgo56.fsf@ruckus.brouhaha.com>
Message-ID: <003501c26833$75827560$ba01010a@local>

----- Original Message ----- 
From: "Paul Rubin" <phr-n2002b at NOSPAMnightsong.com>


> "Chris Gonnerman" <chris.gonnerman at newcenturycomputers.net> writes:
> > > >>> (.70 * .05) + .000000001
> > > 
> > > rounds to the right thing.
> > 
> > Cool.  Now PROVE that's right in all cases.
> 
> What do you mean by "right"?  What do you mean by "all 
> cases"?

The original subject is *business* arithmetic.  There is no
choice but to use decimal arithmetic if you want your math
results to match good-old-fashioned pen and paper math.

That defines "right."  "All cases" refers to all numeric 
domains common to *business* math.  The exact number of
decimal places and the rounding rule (0.5 up/0.4 down vs. 
"banker's rounding" which is rounding 0.5 to an even number) 
will vary from nation to nation and sometimes from industry
to industry, but the expectation of reliable rounding results
is still the same.

Sorry not to have defined the terms; if you are following 
the discussion I would expect you to know this.  If we were
talking about tracking satellites or computing the mass of 
the electron I'd defer to the experts (those who work in
those fields) regarding the best sort of computer math to 
use.  When it comes to business, though, I have experience.

As another poster noted, it really doesn't matter if *we*
think that results such as the above are "good enough," as
our bosses/customers/legislators/users control the definition
of "good enough."

Most of them can't stand to mislay a penny.

I've never worked for a boss who wasn't human (although I
have wondered sometimes if I was working for a weasel in 
human clothing).

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From sholden at holdenweb.com  Wed Sep 11 20:23:38 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 11 Sep 2002 20:23:38 -0400
Subject: [FEEDBACK] Is this script efficient...is there a better way?
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net>
Message-ID: <L9Rf9.168314$l_4.127784@atlpnn01.usenetserver.com>

"Bob X" <bobx at linuxmail.org> wrote ...
> I am a newbie at this and this is a script I have come up with. I am
> looking for pointers on any "newbie" gotchas or ways to do it better.
>
> I use this to parse a log file (30MB+) for keywords and write the lines
> that are found into another file.
>
> Any thoughts would be appreciated...  :-)
>
> Bob
>
> #!/usr/local/bin/python -w
>
> # import the needed libs
> import sys, string
>
> # make sure the command line arguments are there
> if len(sys.argv) < 3:
>      print "usage: fread.py [log file] [hit file]"
>      sys.exit(1)
>
Unless the return is important, you'll find

    sys.exit("usage: fread.py [log file] [hit file]")

more convenient.

> # open the files with some error checking
> try:
>      inFile = open(sys.argv[1],"r")
> except IOError:
>      print "Cannot open log file!\n"
>      sys.exit(1)
>
> try:
>      outFile = open(sys.argv[2],"w")
> except IOError:
>      print "Cannot open hits file!\n"
>      sys.exit(1)
>
Your error checking is exemplary. Normally I'm happy to have a Python
exception explain the error condition, but of course your explicit messages
are great.

> # build list of keywords
> kw = [ "some", "words" ]
>
> # loop through the list and print the lines to a file
> for line in inFile.readlines():
>      for badword in kw:
>          if line.find(badword) > -1:
>              result = '%s %s' % (badword, line)
>              print result            # Print the result
>              outFile.write(result)   # Write the result
>
If you only wanted to know the lines that match at least one of the words
you could follow the write() with a "break", since there's no need to
continue the loop then.

Also, the "print" looks like a debugging statement.

> # close the files
> inFile.close()
> outFile.close()
>
> # let me know when it's done
> print "Finished processing file..."
>
Highly commendable. Another approach would be to build a regular expression
from the keywords list and search for the r.e. in each line. The following
code is untested...

kw = ["some", "words"]
kwpat = "|".join(kw)
pat = re.compile(kwpat)
for line in inFile.readlines():
    if pat.search(line):
        outFile.write(line)

for example.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From oren-py-l at hishome.net  Thu Sep 12 06:06:15 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Thu, 12 Sep 2002 06:06:15 -0400
Subject: [xml.dom.minidom] why cant it parse '&'
In-Reply-To: <9zZf9.24873$_l1.66951@news-server.bigpond.net.au>
References: <9zZf9.24873$_l1.66951@news-server.bigpond.net.au>
Message-ID: <20020912100615.GB41211@hishome.net>

On Thu, Sep 12, 2002 at 09:58:29AM +0000, :B nerdy wrote:
> im using the minidom.parseString() to parse the following text
> 
> <sxml>
> <title>Congo Mombo - Guitar Gable & The Musical Kings</title>
> <composer>Various</composer>
> <genre>blues</genre>
> <trackNo>12</trackNo>
> <copyright>Westwood</copyright>
> <dateInserted>2002-03-11</dateInserted>
> </sxml>
> 
> when it comes to the '&' it throws an exception. is there any way around
> this problem??

It's not specific to Python or minidom. & and <> are reserved characters 
in SGML/HTML/XML.

&  =>  &amp;
<  =>  &lt;
>  =>  &gt;

	Oren



From thoa0025 at mail.usyd.edu.au  Sun Sep 15 05:56:38 2002
From: thoa0025 at mail.usyd.edu.au (:B nerdy)
Date: Sun, 15 Sep 2002 09:56:38 GMT
Subject: regular expression help [port]
Message-ID: <qPYg9.7377$Ee4.17838@news-server.bigpond.net.au>

Greetings, this is a regular expression for an email address
"^[a-z0-9_]+@[a-z0-9\-]{2,}\.[a-z0-9\-\.]{2,}$"

i wish to valid a variable against this string. would someone be so kind to
supply a few lines of code to do this?

cheers




From richie at entrian.com  Fri Sep 13 03:49:41 2002
From: richie at entrian.com (Richie Hindle)
Date: 13 Sep 2002 00:49:41 -0700
Subject: Smoothing a discrete set of data
References: <mailman.1031850555.28301.python-list@python.org>
Message-ID: <8a6ba1da.0209122349.454a4991@posting.google.com>

[ Explanation of moving averages and running medians ]
> So, I hope this will help answer your question.

Whether or not it answers Paul's question, it was an interesting and
well-written explanation - many thanks for taking the time to post it!

-- 
Richie Hindle
richie at entrian.com


From tjreedy at udel.edu  Sun Sep 15 18:59:36 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Sun, 15 Sep 2002 22:59:36 GMT
Subject: representing a literal single slash
References: <dd5454fb.0209141828.1ad24260@posting.google.com> <3D84E9F9.8090101@linux.ie> <3D84FEE6.90600@linux.ie>
Message-ID: <sh8h9.52633$AR1.2065934@bin2.nnrp.aus1.giganews.com>

"Padraig Brady" <padraig at linux.ie> wrote in message
news:3D84FEE6.90600 at linux.ie...
> > mystr = mystr.replace(r'\\', r'\')
>
> Emm this actually doesn't work. Why can't you:
> s=r'\'
> Is it a bug?

Ref Manual 4.1:
When an `r' or `R' prefix is present, a character following a
backslash is included in the string without change, and all
backslashes are left in the string. For example, the string literal
r"\n" consists of two characters: a backslash and a lowercase `n'.
String quotes can be escaped with a backslash, but the backslash
remains in the string; for example, r"\"" is a valid string literal
consisting of two characters: a backslash and a double quote; r"\" is
not a valid string literal (even a raw string cannot end in an odd
number of backslashes). Specifically, a raw string cannot end in a
single backslash (since the backslash would escape the following quote
character). Note also that a single backslash followed by a newline is
interpreted as those two characters as part of the string, not as a
line continuation.

Terry J. Reedy







From sismex01 at hebmex.com  Tue Sep 24 18:12:33 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Tue, 24 Sep 2002 17:12:33 -0500
Subject: List/Tuple bug or feature?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233025AB17A@mail.hebmex.com>

It's one of python's features :-)

For when you wanna write a long string, SQL for example:

Query = "SELECT what, when, where, why "
        "FROM t1, t2, t2 "
        "WHERE t1.id = t2.id_t1 and t3.id_t2 = t2.id and t1.id_t2 = t3.id "
        "AND t1.answer = 'OK'"

Kinda neat.

-gus

> -----Original Message-----
> From: Mahesh Padmanabhan [mailto:news at nospam.eml.cc]
> Sent: Tuesday, September 24, 2002 5:06 PM
> To: python-list at python.org
> Subject: List/Tuple bug or feature?
> 
> 
> 
> Hi,
> 
> I spent a lot of frustrating hours tracking this down and I am
> confused as to why I see this behavior.
> 
> >>> x = ('a' 'b')
> >>> x
> 'ab'
> 
> >>> x = ['a' 'b']
> >>> x
> ['ab']
> 
> As you might have guessed, I missed a comma between two elements in
> a fairly large list.
> 
> If this is a concatenation feature, how can I prevent something like
> this happening in the future ?
> 
> This is with Python 2.1.3 on Debian Linux.
> 
> Python 2.1.3 (#1, Sep  7 2002, 15:29:56) 
> [GCC 2.95.4 20011002 (Debian prerelease)] on linux2
> Type "copyright", "credits" or "license" for more information.
> 
> TIA,
> 
> -- 
> Mahesh Padmanabhan
> 
> -------
> 
> Replace nospam with prana to email me.
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 

Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From michaels at one.net  Wed Sep 18 07:51:31 2002
From: michaels at one.net (Michael Schneider)
Date: Wed, 18 Sep 2002 07:51:31 -0400
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <3D881B66.1080200@something.invalid> <j4wupjafun.fsf@informatik.hu-berlin.de>
Message-ID: <3D8868C3.9090005@one.net>


Martin v. L?wis wrote:

>Greg Ewing <see_reply_address at something.invalid> writes:
>
>>Last I heard, Christian was moving pieces of the C
>>stack while there could be pointers to it in use.
>>
>
>Yes, that's how stackless works, and that cannot change.
>
>>That's definitely dangerous, as he discovered when
>>it broke Tk.
>>
>
>What is the danger? A program crashes? It is dangerous to cross a
>street at rush hour; sitting in front of a computer is rarely
>dangerous.
>
>Stackless adds a few limitations as to what C extensions can do. If
>those restrictions are followed, it won't crash.
>
>>I'd be very nervous about using Stackless until
>>it stops doing that.
>>
>
>Why is that? Those crashes can only happen if you use the stackless
>modules (and only if you use them with a stackless-unsafe extension).
>If you don't import stackless, no stack is copied.
>


This is an important point that I missed.  I thought that a stackless 
python was
always stackless, and that I would need to have two version of python 
around.

1) python.exe  - normal python ( for TK)
2) pythonsl.exe - stackless python.


Please correct me if I misintrepreted your statement above, but this is 
what I got out
of it :

    "stackless python behaves the same as 'normal' python, until you 
import the
    stackless module.  At that point, it becomes stackless python.  This 
allows you
     to run 'stackless unsafe' modules with the same interpretor as 
stackless modules,
     in a predictable manner (as long as you don't import stackless, and 
an unsafe module)"


Is this the correct understanding????

If so, this is really good news!!!!!!!!!!

Thank you very much,
Mike




>
 ... good stuff cut

>
>Regards,
>Martin
>



From owen at nospam.invalid  Mon Sep 16 18:25:26 2002
From: owen at nospam.invalid (Russell E. Owen)
Date: Mon, 16 Sep 2002 15:25:26 -0700
Subject: Python on MacOS X 10.2: a warning and instructions
Message-ID: <am5log$1ip2$1@nntp6.u.washington.edu>

MacOS X 10.2 (Jaguar) breaks the fink installation of Python (due to 
moving the ncurses package).

MacOS X 10.2 does include Python (at least if you install the 
Developer's CD, not sure otherwise) but it has no readlines support 
(i.e. no command recall) and no Tkinter support.

I upgraded before I realized this, and needed python too much to wait 
for the fink binary, so I spent a day and a half figuring out how to do 
the installation (with help from the MacPython list). Instructions are 
at <http://arcserve.astro.washington.edu/InstallingPythonOnJaguar.html> 
if you are unfortunate enough to need them.

These instructions install a version that uses xfree86 (an X server) to 
display windows. However, a release version of the native aqua version 
of tcl/tk now exists (version 8.4.0) but so far the only binary 
installer is for 8.4a4, an older version. The source installer comes 
with no instructions and what I tried failed. If anybody gets Python 
with Tkinter to worth with the aqua version of tcl/tk 8.4.0 under MacOS 
10.2, I would love instructions.

-- Russell


From klausman-un250602 at tuts.net  Mon Sep  9 13:06:50 2002
From: klausman-un250602 at tuts.net (Tobias Klausmann)
Date: 9 Sep 2002 17:06:50 GMT
Subject: Python plugins
References: <alf7rc$1p06l1$1@ID-41653.news.dfncis.de> <mailman.1031495592.26635.python-list@python.org> <3D7C13B3.2070606@nospam.free.fr>
Message-ID: <alikfa$1pqvck$1@ID-41653.news.dfncis.de>

Christian Tismer <tismer at tismer.com> wrote:
> If you executed a proper __import__, all modules and funcs should
> be accessible in the normal way.
> There is no need to exec.

This is what I overlooked: the lowlevel interface to __import__

Thanks a lot for your help (to laotseu, too).

Greets,
Tobias



From tim.one at comcast.net  Fri Sep  6 23:29:52 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 06 Sep 2002 23:29:52 -0400
Subject: Larry Wall's comment on python...
In-Reply-To: <3d796a95$1@news.sentex.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCOELCBCAB.tim.one@comcast.net>

[Peter Hansen]
> Anyone know how long is the largest chunk of code Larry Wall has
> actually written in Python?

It was 23 lines longer than the largest chunk of code Guido has actually
written in Perl.  The difference is that Larry's code worked, and Guido's
still doesn't.  I suspect Larry may have taken a wrong lesson from that
<wink>.

> The fact he has to say "I think" about this issue implies he is
> guessing.  Maybe if he actually tried it he'd conclude differently.

It was just a sound bite from an interview, and blocks aren't quite the same
thing in Perl (for example, a Perl block can have its own variables, but a
Python block can't).  I bet that if Guido had designed Perl, it would still
have curlies, and I defy anyone to prove that wrong.  Heh.

leave-rationality-to-the-rationals-this-is-language-design-ly y'rs  - tim




From anton at vredegoor.doge.nl  Sun Sep 15 02:39:01 2002
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Sun, 15 Sep 2002 08:39:01 +0200
Subject: How to represent sets
References: <alvd8p$1sq$1@news.hccnet.nl> <1VIg9.16142$5r1.654088@bin5.nnrp.aus1.giganews.com>
Message-ID: <am19v9$jdk$1@news.hccnet.nl>

On Sat, 14 Sep 2002 15:50:21 GMT, "Terry Reedy" <tjreedy at udel.edu>
wrote:

>> I would like to reopen this discussion.
>
>I don't remember any pronouncement from Guido or anyone else closing
>it, except perhaps with regard to what he will personally support.

How about:

discussion.close()
discussion.open()

Anton.



From mihan at foss.kharkov.ua  Wed Sep  4 09:00:55 2002
From: mihan at foss.kharkov.ua (Mykhaylo Sorochan)
Date: Wed, 4 Sep 2002 16:00:55 +0300
Subject: forum for zope
References: <al4tru$o3j$03$1@news.t-online.com>
Message-ID: <al5054$187$1@hyppo.gu.net>

> but does there exist a newsgrup for zope centric items?
There exists a forum
www.zope-forum.org




From mcherm at destiny.com  Thu Sep 12 16:08:13 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Thu, 12 Sep 2002 16:08:13 -0400
Subject: Coding Style: Quotes
References: <bt1g9.17429$Pf7.590136@news1.west.cox.net> <20020912144710.GA78280@hishome.net>
Message-ID: <alqs76$ua$1@main.gmane.org>

Oren Tirosh wrote:
> On Thu, Sep 12, 2002 at 02:25:11PM +0000, John Waycott wrote:
> 
>>I'm curious if others have adopted any standards for choice of single vs.
>>double quotation marks to delimit strings. [...]
> 
> The default repr of a string uses single quotes unless the string 
> contains a single quote. Most of the time I use the same convention but 
> I'm not 100% consistent.
> 
> For docstrings I always use triple double-quotes even if the docstring 
> fits in a single line.
> 
> I use triple single-quotes to comment out blocks of code. This is very 
> useful since "if 0:" would require re-indentation.
> 
> 	Oren

Encouragingly, I use EXACTLY the same conventions and practices that 
Oren does, having developed them independently for the same reasons.

-- Michael Chermside






From cbrown at metservice.com  Thu Sep 12 21:14:58 2002
From: cbrown at metservice.com (Colin Brown)
Date: Fri, 13 Sep 2002 13:14:58 +1200
Subject: Project idea
Message-ID: <3d813c12$1@news.nz.asiaonline.net>

On 6/09/2002 I wrote:

>>Hi
>>
>>If you are interested in a Windows platform project for automated testing
>>then read on.
>>
>>There is an excellent freeware package "AutoIt" available from
>>http://www.hiddensoft.com/AutoIt/
>>that provides a DLL module that could be readily interfaced to Python.
>>
>>"AutoIt" is a simple tool that can simulate key presses, mouse movements
and
>>window commands (maximize, minimize, wait for, etc.) in order to automate
>>any windows based task (or even windowed DOS tasks).
>>
>>Combining these features with say the ability to produce checksums of
>>snapshotted screen frames for
>>test comparisons would provide the basis for a nice GUI regression test
rig.
>>
>>Unfortunately I do not have the time to follow up on this myself so I
thought
>>I would pass it on.
>>
>>Colin Brown
>>PyNZ

On re-examination, this is even easier because the package already comes
with an ActiveX component.
Here is an example of how simple it is to use:

Python 2.2.1
>>> import win32com.client
>>> conn=win32com.client.Dispatch('AutoIt.Control')
>>> help(conn)

<COMObject AutoItX.Control>

>>> conn.WinGetActiveTitle()
u'Python (command line)'
>>> conn.WinMinimize('Python (command line)','')

& the window minimises!

>From here on in even a beginner should be able to have some fun :)

Colin





From max at alcyone.com  Wed Sep 18 22:23:54 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 18 Sep 2002 19:23:54 -0700
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com>
Message-ID: <3D89353A.2EC1C052@alcyone.com>

"daniel w. moore" wrote:

> It's occurred to me after some brainstorming on the subject that the
> C/perl construct
> 
>   a ? b : c
> 
> can be arranged like this in python:
	...
> Has anyone put this to good use, i wonder?
> Or have I just made a daring and heroic discovery in my own particular
> idiom ;)

Sorry, but no, this isn't new.  A variation is mentioned in the FAQ and
these two variants -- using and/or combinations and lambdas -- gets
discussed in the group frequently.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ When you love somebody, you're always in trouble.
\__/ Col. Sherman Potter
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.


From c6re at sdsumus.sdstate.edu  Thu Sep 19 23:33:02 2002
From: c6re at sdsumus.sdstate.edu (Buckshot)
Date: 19 Sep 2002 20:33:02 -0700
Subject: e vs exp()?
Message-ID: <3a838998.0209191933.1f22e977@posting.google.com>

Can anyone here explain why I'm getting different results from
e**2 and exp(2)?  Which is more accurate?  

Python 2.2 (#1, Mar 27 2002, 14:56:58)
[GCC 2.95.3 20010315 (release)] on sunos5
Type "help", "copyright", "credits" or "license" for more information.
>>> from math import *
>>> e**2
7.3890560989306522
>>> exp(2)
7.3890560989306504


From jdhunter at ace.bsd.uchicago.edu  Mon Sep 23 11:13:49 2002
From: jdhunter at ace.bsd.uchicago.edu (John Hunter)
Date: Mon, 23 Sep 2002 10:13:49 -0500
Subject: numpy array assignment
In-Reply-To: <m2bs6on46w.fsf@mother.paradise.lost> (John Hunter's message of
 "Mon, 23 Sep 2002 09:58:15 -0500")
References: <m2bs6on46w.fsf@mother.paradise.lost>
Message-ID: <m2lm5skac2.fsf@mother.paradise.lost>

>>>>> "John" == John Hunter <jdhunter at ace.bsd.uchicago.edu> writes:

    John> I have a 2 dimensional array X which is NxM and a 1
    John> dimensional array v which is Nx1.  I want to assign v to the
    John> k-th column of X

This is what I've come up with so far -- if there is a better/faster
way, please let me know....

import Numeric

def putcol(k, Z, v):
    (n,m) = Z.shape
    Numeric.put(Z, k+m*Numeric.arange(n), v)

def putrow(k, Z, v):
    (n,m) = Z.shape
    Numeric.put(Z, k*m+Numeric.arange(m), v)


Z = Numeric.zeros( (3,5), Numeric.Float  )
vc = Numeric.ones( (3,1), Numeric.Float )
vr = Numeric.ones( (1,5), Numeric.Float )

putcol(3, Z, vc)
putrow(1, Z, vr)
print Z




From bruce_deletethis_dawson at cygnus-software.com  Mon Sep 23 02:18:09 2002
From: bruce_deletethis_dawson at cygnus-software.com (Bruce Dawson)
Date: Mon, 23 Sep 2002 06:18:09 GMT
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <%8ci9.138477$5r1.5541622@bin5.nnrp.aus1.giganews.com> <amltd3$3k$1@glue.ucr.edu>
Message-ID: <3D8EB212.1000104@cygnus-software.com>

thp at cs.ucr.edu wrote:

> Terry Reedy <tjreedy at udel.edu> wrote:
> 
> + "daniel w. moore" <dan65536 at hotmail.com> wrote in message
> + news:33b491f.0209181820.17b6ea70 at posting.google.com...
> +> It's occurred to me after some brainstorming on the subject that the
> +> C/perl construct
> +>
> +>   a ? b : c
> +>
> +> can be arranged like this in python:
> +>
> +>   (a and [b] or [c])[0]
> 
> + If you *know* that b will evaluate as True (perhaps because it is a
> + non-null constant, then this can be simplified to 'a and b or c'.  I
> + first used this several years ago.
> 
> But why use grotesque hacks in the first place?  "a ? b : c" is a
> well-established syntax for conditional expressions.  Why doesn't
> Python simply adopt it?
> 
> Tom Payne


I assume that there are reasons why it could clutter or make ambiguous 
the syntax - but I'm not expert.

I find the (a and [b] or [c])[0] construct to obscure to be useful - the 
space it saves is wasted by the comment I have to put in to explain it.

It occurs to me that the bool() built in function allows for what is - 
in my opinion - a cleaner solution. Easier for mortals to comprehend.

Basically, construct a two entry list of the two items and then index 
that with your expression, converted to bool. The bool function is 
essential since there was previously no generic way of getting the 
true/false value of an expression.

For example:

Python 2.2.1 (#34, Sep 14 2002, 09:40:33) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
 >>>
 >>> a = "Hello"
 >>> b = "Goodbye"
 >>> myString = ""
 >>> [a,b][bool(myString)]
'Hello'
 >>> myString = "Something"
 >>> [a,b][bool(myString)]
'Goodbye'



From lee100 at sport.rr.com  Fri Sep 20 08:14:00 2002
From: lee100 at sport.rr.com (Lee Gray)
Date: Fri, 20 Sep 2002 12:14:00 GMT
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com>
Message-ID: <ciEi9.496880$q53.16851520@twister.austin.rr.com>

Benjamin wrote:

> the problem, in one word: coding. but now i don't know
> what to programm. sounds weird, doesn't it? :) what
> should i program??

A friend and I joked for years that we were going to learn C, but just 
couldn't get into it... because we didn't have a project or problem to 
solve with it!  Although I never learned C, I finally did learn to program, 
and have often found myself in a short slump simply because I didn't have a 
project at the time.

You need a problem to solve, or as the saying goes, "an itch to scratch".  
If you have that, and you enjoy using the language, you will find that fun 
again.

Lee



From max at alcyone.com  Mon Sep  9 14:23:15 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 09 Sep 2002 11:23:15 -0700
Subject: random.shuffle question
References: <3k4f9.7500$yt3.3831288@newssrv26.news.prodigy.com>
Message-ID: <3D7CE713.93BC9FAF@alcyone.com>

dsavitsk wrote:

> why does this not work to shuffle part of a list?
> 
> >>> l = [1,2,3,4,5,6,7,8,9]
> >>> random.shuffle(l[4:8])
> >>> print l
> [1, 2, 3, 4, 5, 6, 7, 8, 9]
> 
> and more importantly, is there a better way to do this?

random.shuffle shuffles the list you hand it, and when you create a
slice of a list, that sublist is separate from the original list.  So
you're creating a new sublist, shuffling it, and then tossing it away.

You could simply chop up the list, shuffle the part that you want, and
then reassemble it:

>>> import random
>>> l = [1, 2, 3, 4, 5, 6, 7, 8, 9]
>>> lBegin = l[:4]
>>> lMiddle = l[4:8]
>>> lEnd = l[8:]
>>> random.shuffle(lMiddle)
>>> l2 = lBegin + lMiddle + lEnd
>>> l2
[1, 2, 3, 4, 6, 5, 8, 7, 9]


-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ The perfection of innocence, indeed, is madness.
\__/ Arthur Miller
    Max Pandaemonium / http://www.maxpandaemonium.com/
 A sampling of Max Pandameonium's music.


From gustav at morpheus.demon.co.uk  Tue Sep 10 15:35:43 2002
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: 10 Sep 2002 20:35:43 +0100
Subject: Smoothing a discrete set of data
References: <3csmdso5.fsf@morpheus.demon.co.uk> <Buue9.492218$m91.19280936@bin5.nnrp.aus1.giganews.com>
Message-ID: <8z2963jk.fsf@morpheus.demon.co.uk>

"Terry Reedy" <tjreedy at udel.edu> writes:

> "Paul Moore" <gustav at morpheus.demon.co.uk> wrote in message
> news:3csmdso5.fsf at morpheus.demon.co.uk...
> > I have a set of data, basically a histogram.
> 
> Actually, the two examples you give are *not* histograms.  A histogram
> is a discrete frequency graph: how many numbers fit in each bit.  Note
> that binning numbers like this them tosses away any order and other
> covariate info.  Unimodal histograms are usually smoothed by moment
> matching: mean, standard deviation, and possibly more.

Thanks for the clarification. As became increasingly obvious to me as
I read the various replies, I am not being adequately clear in my own
mind what I'm trying to do. Part of this is a result of not thinking
in the right terms (for some reason, I didn't see the question in
terms of stats, even though I was thinking of things like averaging
and regression...), and part is simply a matter of oversimplifying and
losing the key details of what I'm trying to do.

Given that this is basically statistics, it's ironic that I did maths
as my university degree, and while I didn't do a lot of stats in my
course, it was an area I had had a strong "amateur" interest in prior
to doing the degree course.

> > sampled IO rates on a machine
> > - I want to look for trends (is IO higher overnight or during the
> > day,
> 
> If you take measurements (samples) every hour, for instance, you have
> a time series.  There are many books on this subject alone.

Interesting - I'd heard of time series, but had never got far enough
into the subject to get a proper feel for what they covered...

> > etc) or fuel consumption for my car (do I use less fuel since I had
> > the service).
> 
> This is a standard question with standard methods.  If you ignore
> order and other covariates and group measurements as before and after,
> a t-test or signed rank test would be appropriate.

When you put it like that, it's so obvious!! I think I was so fixated
on making the picture look like the point I was trying to make (too
much time working with business graphics, where the graph is there to
make a point, rather than to explain the facts :-()

> > Normally, what I do with things like this is draw a graph, and try
> > to spot the trends "by eyeball". But it feels to me like I should
> > be able to write something which smooths the data out. I just
> > don't see how
> 
> As a statistician, I am a fan of eyeballing raw data (with appropriate
> caveats about testing what you think you see) in addition to numerical
> analysis.  However, each statistical procedure is aimed as answering a
> question (and many are based on some assumption about the data).  So
> you need to better formulate what you want to know.

You're right. I think that applying some of the suggestions other
posters made (low-pass filtering the graph, for example) would help to
display some of the grosser trends better (I think that was another
aspect of what I was looking for, but I see now that it *is* a
separate aspect). Once I have that "feel", I need to get more
analytical about specifics.

Thanks for your suggestions, and for all the other posters' comments
as well. This has reawakened an old interest for me, and I look
forward to dusting off some of my old books and learning more...

Paul.


From joconnor at cybermesa.com  Wed Sep 11 16:20:48 2002
From: joconnor at cybermesa.com (Fearless Freep)
Date: 11 Sep 2002 13:20:48 -0700
Subject: need Python and Zope experienced programers
References: <3d647529.27575997@news.earthlink.net> <15717.3618.449268.769068@gargle.gargle.HOWL> <mailman.1031751095.31551.python-list@python.org>
Message-ID: <7d3dc526.0209111220.35007fff@posting.google.com>

Michael Chermside <mcherm at destiny.com> wrote in message news:<mailman.1031751095.31551.python-list at python.org>...
> Skip Montanaro wrote:

> > You might want to post to the Python Jobs board:
> > 
> >     http://www.python.org/Jobs.html
> > 
> 
> Unfortunately, it's for posting job LISTINGS only, not job REQUESTS. But 
> it seems I've found something anyhow (though without Python :-( ).

For what it's worth, I'm an experienced Python programmer looking for
work (part or full time, contract or permanant)

Take care,

Jay O'Connor
joconnor at cybermesa.com


From oren-py-l at hishome.net  Wed Sep  4 07:38:26 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Wed, 4 Sep 2002 07:38:26 -0400
Subject: encryption (passwords)
In-Reply-To: <3d75d115$0$221$4d4ebb8e@news.nl.uu.net>
References: <3d75d115$0$221$4d4ebb8e@news.nl.uu.net>
Message-ID: <20020904113826.GA71214@hishome.net>

On Wed, Sep 04, 2002 at 11:29:15AM +0200, Gumuz wrote:
> I am using network-sockets to communicate between two(or more) python
> applications. I want to encrypt these messages, or at least a part of it
> e.g. password and such.
> 
> My limited understanding of encryption tells me that I need a 'key' to
> decrypt stuff. So I figure that this key will bey written(hard-coded?)
> somewhere in my Python-script and it needs to be known at both sides in
> order to encrypt and decrypt. Somehow, I can't believe this is true. I can't
> get over the idea that if my application becomes open-source anyone could
> see the key and decrypt my messages.

You can get secret communication without any prior agreement between the
parties. The Diffie-Hellman key exchange protocol lets two parties 
communicating over an insecure channel create a shared secret and a passive 
listener will not be able to find it.

Take a look at http://magaf.org/dh.py

There is a catch, though: this gives you secrecy but not authentication.
If you encrypt the data stream using this shared secret as a key you will
be communicating in perfect secrecy with someone who may or may not be who
you meant to communicate with. It could be someone pretending to be the 
other party (a man-in-the-middle attack)

So secrecy has a technical solution, but it only moves the problem 
elsewhere.  Authentication has *no* simple technical solution because it's
more of a meatspace problem than a cyberspace problem. Any solution
would depdend on lots of tricky assumptions about the users of the system, 
the potential threats and the environment in which it runs.

	Oren





From rjones at ekit-inc.com  Wed Sep 18 20:18:40 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Thu, 19 Sep 2002 10:18:40 +1000
Subject: Python equivalent to perldoc?
In-Reply-To: <ama2tm$40up1$1@ID-77524.news.dfncis.de>
References: <Olvh9.7898$Lg2.1389903@news2.news.adelphia.net> <kaPh9.8886$Lg2.1572004@news2.news.adelphia.net> <ama2tm$40up1$1@ID-77524.news.dfncis.de>
Message-ID: <200209191018.40935.rjones@ekit-inc.com>

On Thu, 19 Sep 2002 12:34 am, Thorsten Kampe wrote:
> *  Bob X
>
> > "Thorsten Kampe" <thorsten at thorstenkampe.de> wrote in message
> >
> >>> Shouldn't Python know its own libs?
> >>
> >> It does:
> >> ,---
> >>
> >> | Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on
> >> | win32 Type "help", "copyright", "credits" or "license" for more
> >> | information.
> >> |
> >> | >>> import pydoc
> >>
> >> `---
> >
> > It doesn't...you are executing from with Python...I am trying to do a ViM
> > script that executes it.
>
> So Vim doesn't doesn't know about Python's libs. No surprise.

Actually, vim's been able to import and use python libs for a long, long time. 
There's no reason why vim shouldn't be able to hook into pydoc.


    Richard




From justnotworking at yahoo.com  Thu Sep 12 18:06:51 2002
From: justnotworking at yahoo.com (mowestusa)
Date: Thu, 12 Sep 2002 17:06:51 -0500
Subject: Newbie Looking for direction
Message-ID: <uo23o95eb8tg87@corp.supernews.com>

I have never programmed in a language before.
I have developed complex databases with Access (Office 95 version) just with
the visual tools.
I would like to learn a computer language or database as a hobby that I
would use to keep track of things in my real job.
I have worked through "Non-programmer's Tutorial for Python".
I'm reading "Programming Python (First Edition, got cheap)" right now.

Most of the programs that I would like to develop would be relatively simple
databases.
I have Access 95 and Paradox 10 (with runtime).  I'm wondering if I'm going
down the right path in learning Python or if I would be better served by
just learning to use Paradox 10 well.  The one problem I have with Paradox
or Access is that they seem so slow. Even if you have a simple database it
takes a long time to start up and run.  I would also like to make things
that I could share with colleagues who have no interest in programming, but
could use some of the database tools I would develop for our line of work.

So far it seems like I'm a long way away from being able to be productive
with Python, but that maybe because my end goal is too complex.

Any advise would be appreciated.

mowestusa




From Oschler at earthlink.net  Sat Sep 21 16:52:00 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sat, 21 Sep 2002 20:52:00 GMT
Subject: Can't suspend a thread?
Message-ID: <QZ4j9.10630$Lg2.2393215@news2.news.adelphia.net>

I was hoping to have a master thread that woke up and suspended several
other threads dynamically, as needed.  From the books I have, that seems to
be impossible in Python?  Is this true?  If so, is there a workaround?

thx





From sludtke at bcm.tmc.edu  Mon Sep 30 13:12:15 2002
From: sludtke at bcm.tmc.edu (Steve Ludtke)
Date: Mon, 30 Sep 2002 12:12:15 -0500
Subject: XMLRPC calls don't work in threads, or rather, they don't play
 well with others ! (help)
References: <3D97D190.9030003@bcm.tmc.edu> <cb035744.0209300456.70c45108@posting.google.com>
Message-ID: <3D9885EF.7020703@bcm.tmc.edu>


Will Stuyvesant wrote:
> [Steve Ludtke]
> 
>>Both of the server.()  calls prevent other threads from running
> 
> 
> Um, do you really think they prevent other *threads* from running?  Or
> would it be the case that there can be only one server call being
> handled at any time at a *specific url:portnumber*.

I am currently using only 1 thread to make the XMLRPC calls. It still 
prevents other threads from running. I'm working on a script to 
demonstrate the problem.



From LogiplexSoftware at earthlink.net  Fri Sep 20 09:47:09 2002
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 20 Sep 2002 06:47:09 -0700
Subject: More on Protecting Source Code
In-Reply-To: <3d86b4a0$0$23173$afc38c87@news.optusnet.com.au>
References: <mailman.1032232999.22661.python-list@python.org> 
	<3d86b4a0$0$23173$afc38c87@news.optusnet.com.au>
Message-ID: <1032529629.7312.214.camel@localhost.localdomain>

On Mon, 2002-09-16 at 21:50, Steven wrote:

> "Security through obscurity is no security against a determined attacker"
> 
> So, someone who is willing to hacka nd hack and hack away at your code, can
> get at it, and this is true for things coded in assembler, C, Python or
> Befunge. Most people aren't determined attackers, they're simply Joe Average
> who needs to get a  job done.

And even Joe Average is able to download the cracked software from one
of the many sites that exist to distribute such software.  This is the
problem:  people think that they can protect their software from the
"average" user, and they could except that the "average" user has the
resources of the cracker community at his disposal.  

 
> An operating system is a complex beast, compiled into machine code its even
> more complex, how many people out there are using an unlicensed copy of
> Windows or Office? they didn't decompile, that would take skill, they just
> copied it bit for bit and posted the registration key.

True, but if skill is required and the software is worth it, someone
will take the time and they get no greater joy than posting their work
for others to see (and download).

> Do a very very simple risk analysis. Who is your target audience, what is
> the price of your software, will people bother going through your bytecode,
> or will they just post the registration key to a website? Will stolen
> software lead to a loss in revenue - or will those pirate users be people
> who would never have bought it in the first place? (yes, the pirate users
> are annoying, and you would feel ripped off, but OTOH you haven't lost
> money, and you've gained a bunch of users)

Very true.  Microsoft benefited greatly from this stance (not worrying
too much about unlicensed copies of Windows and other applications).  
Not enforcing a copyright is a quick way to get something to become a
"standard".  You can always enforce the copyright later, when everyone
is dependent on your software (again, see Microsoft).

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From robin at jessikat.fsnet.co.uk  Wed Sep  4 13:10:07 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 4 Sep 2002 18:10:07 +0100
Subject: jython memory usage
References: <hVm5ydAY2hd9Eww$@jessikat.demon.co.uk>
 <3d76366a$1_5@news.bluewin.ch>
Message-ID: <rnLbuKAv5jd9Ewzz@jessikat.demon.co.uk>

In article <3d76366a$1_5 at news.bluewin.ch>, Samuele Pedroni
<pedronis at bluewin.ch> writes
>From: Robin Becker <robin at jessikat.fsnet.co.uk>
>> A colleague has been running tests of reportlab under jython-2.1 several
>> of the tests use prodigious amounts of memory (one even failed). We're
>> using win2k+j2sdk1.4.0_01 as the jython engine.
>>
>> Anyone know what sorts of things cause x10 memory usage? I find one
>> particular test needs more that 1Gb and then falls over for lack of
>> memory.
>
>That means you are running jyhton through:
>
>java -mx1024m  -cp ...jython.jar org.python.util.jython ...
>
>and it uses all that memory and would need more?
>
>regards.
>
>
well actually we're running with the default jython.bat which doesn't
seem to specify a memory, 1Gig is probably the default or maybe I ran
out of page space. The python process never gets over 90Mb grrrhhhh.
-- 
Robin Becker


From klausm0762 at yahoo.de  Wed Sep 25 05:13:07 2002
From: klausm0762 at yahoo.de (Klaus Momberger)
Date: 25 Sep 2002 02:13:07 -0700
Subject: Python Compiling
References: <amfveg$5g64k$1@ID-53055.news.dfncis.de>
Message-ID: <80a8af7d.0209250113.28ae4c72@posting.google.com>

"Terje Johan Abrahamsen" <spoermeg at voldelig.com> wrote in message news:<amfveg$5g64k$1 at ID-53055.news.dfncis.de>...
.....
> 
> Is it not possible to write a complete compiler like the C compiler for
> Python? Or is it just that nobody has done it? Or must the language be
> constructed differently?
> 
Slightly off-topic, but if you are looking for a language which can
be run by an interpreter and can be compiled as well, maybe you should take a
look at Ocaml:

http://www.ocaml.org

-klaus


From marklists at mceahern.com  Thu Sep 26 11:02:17 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 26 Sep 2002 10:02:17 -0500
Subject: revenge of the python puzzle
In-Reply-To: <Pine.LNX.4.44.0209261018590.22263-100000@hydrogen.sabren.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOAEMFEBAA.marklists@mceahern.com>

[Michal Wallace]
> Oh well. :) I think this is just not a good way to make a
> python puzzle. 
> 
> I guess I could avoid all keywords with something like this:
> 
>    open("/dev/stdout","w").write("hello world")
> 
> and then try to make it as convoluted as possible, but then
> it's not really cross platform.
> 
> Was it obvious what it did? :)

One man's obvious is another man's enigma.  I thought it was a fine puzzle.

Another way to make it harder would be to scramble all characters.

// m



From sismex01 at hebmex.com  Mon Sep  9 10:50:38 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Mon, 9 Sep 2002 09:50:38 -0500 
Subject: xml.minidom is stripping out my CRLF's in attrib values!!
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062330237C578@mail.hebmex.com>

> > All line breaks must have been normalized on input to #xA 
> > as described in 2.11 End-of-Line Handling, so the rest of
> > this algorithm operates on text normalized in this way. 
> > 
> > Begin with a normalized value consisting of the empty string.
> > 
> > For each character, entity reference, or character reference in the
> > unnormalized attribute value, beginning with the first and continuing
> > to the last, do the following: 
> > 
> > For a character reference, append the referenced character to the
> > normalized value. 
> > 
> > For an entity reference, recursively apply step 3 of this algorithm
> > to the replacement text of the entity. 
> > 
> > For a white space character (#x20, #xD, #xA, #x9), append a space
> > character (#x20) to the normalized value. 
> > 
> > For another character, append the character to the normalized value.
> > 
> > If the attribute type is not CDATA, then the XML processor must
> > further process the normalized attribute value by discarding any
> > leading and trailing space (#x20) characters, and by replacing
> > sequences of space (#x20) characters by a single space (#x20)
> > character. 
> > 
> > Note that if the unnormalized attribute value contains a character
> > reference to a white space character other than space (#x20), the
> > normalized value contains the referenced character itself
> > (#xD, #xA or #x9). This contrasts with the case where the
> > unnormalized value contains a white space character (not a
> > reference), which is replaced with a space character (#x20)
> > in the normalized value and also contrasts with the case where
> > the unnormalized value contains an entity reference whose
> > replacement text contains a white space character; being
> > recursively processed, the white space character is replaced
> > with a space character (#x20) in the normalized value. 
> > 
> > All attributes for which no declaration has been read should be
> > treated by a non-validating processor as if declared CDATA. 
> > 
> 
> So the only way to get a newline into an attribute is to 
> escape it in using an entity reference.
> 

WOOHOO!!

Thanks Duncan, miniXML grows by the minute!

-gus

pd:

For those wondering, "miniXML" is a tiny pure-Python
implementation of an XML tokenizer (a generator),
a parser (a class which uses the tokenizer) a
document-object class (for in-memory access, read-write),
an event-processing class (for streaming), an object
pickling mixin class, and some other stuff tossed in
for good measure.

For those times when you don't need the big guns.

-gus
--

Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From emile at fenx.com  Wed Sep 11 10:50:12 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 11 Sep 2002 07:50:12 -0700
Subject: EDI parsing
References: <746127ef.0209110128.53665a23@posting.google.com>
Message-ID: <alnl4o$i81$1@main.gmane.org>

Onno Ebbinge:
> I'm writing an application (in Python) that is part of a bigger
> infrastructure for B2B messaging. One communication channel is an EDI
> interface to the B2B messaging system.
>
> I need to parse a (subset) of the UN EDIFACT (aka EDI) standard
> (homepage: http://www.unece.org/trade/untdid/welcome.htm). After doing
> the initial reading of the standard it looks like it is quite hard to
> make a parser for this standard (and 'constructor').

My experience is with the UCC UCS standards, so some of this may be
different in the UN standard.

Once you have the full specification of the documents you'll be
exchanging in a form that python can digest, it's not that hard to build
the parser and/or constructor.  The full specs are available to members
on CD in PDF form.  I once wrote an extremely ugly extraction utility
that preprocessed using pdf2text and then parsed the resulting mess into
something I could then use as something like

from edixn import EdiXn
for source in sourcemessages:
    tempIn = EdiXn(source)

where sourcemessages is a python list of the individual transactions
extracted from the edi envelope.  EdiXn knew about 810, 850, 852, 855,
875 and 880 transactions, allowing me to write appropriate handlers for
each type.

>
> My questions:
>
> 1) Is there a python EDI module? I can't find any :(

I didn't then either.

> 2) Is there a (python) EDI to XML converter?

That wouldn't be hard.  If there's already a DTD for edi, that would
help a lot.  ;-)

> 3) Is there anything that I can use to easily interface with EDI?
>    (preferably in lib and DLL form)

Try some of the commercial suppliers.  There's big money (both expense
and potential) in this, and that's probably why we don't find
implementations strewn about the web.

>
> If the above is not available or fails...
>
> [You have to know that I've never written a parser in Python before.
>  The last (big) parser I wrote was in C with a few years back the help
>  of lex and yacc if memory serves me right.]
>
> What is the best approach to writing an EDI parser in Python?
>

To parse the received messages, I've always found it easiest to use and
parse the raw ascii coming in.  There are commercial packages that allow
you to map and export or even map-to-map import, but as most of the
customers I've done this for were typically forced into it by demands of
the channel, and the commercial offerings rarely seemed as nimble as
simply writing a one-to-one utility taking the edi order to the database
or the resulting invoice to edi.  I have generally used a commercially
available transport package, KMart at that time being the notable
exception.  Outbound messages follow suit.

If you are writing this as a part of a portable b2b application then
you've got your hands full as you've entered the map-making utility
supplier market.  You'll need to address the entire specification, which
my hardcopy of  my deprecated 003050UCS shows to be some ~1000 _dense_
pages.  Make sure you've got the funding to do it right, as this is no
low-budget add-on.

The 'edi specification' always felt akin to the 'rs232 specification' to
me, only instead of 25 pins to muck up, they offer 100s.

Not-sure-how-this-could-help-ly y'rs,

--

Emile van Sebille
emile at fenx.com

---------






From skip at pobox.com  Wed Sep  4 14:37:07 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 4 Sep 2002 13:37:07 -0500
Subject: Why read-only nested scopes?
In-Reply-To: <j4wuq1d7mh.fsf@informatik.hu-berlin.de>
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com>
        <j4wuq1d7mh.fsf@informatik.hu-berlin.de>
Message-ID: <15734.21203.49404.785473@12-248-11-90.client.attbi.com>

    Gon?alo> So why hasn't this been extended when nested scopes were
    Gon?alo> introduced, e.g. by reusing the global directive or some other
    Gon?alo> one?

    Martin> I think the main reason is that nobody could propose an
    Martin> acceptable syntax. Reusing the global directive does not work.

For the gory details, read PEP 227:

    http://www.python.org/peps/pep-0227.html

especially the section entitled "Rebinding names in enclosing scopes".

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From joec at mill.co.uk  Mon Sep  2 12:52:00 2002
From: joec at mill.co.uk (Joe Connellan)
Date: Mon, 02 Sep 2002 17:52:00 +0100
Subject: compiling extensions on linux
References: <3D738D4E.35B6C712@mill.co.uk> <j44rd8jqu9.fsf@informatik.hu-berlin.de>
Message-ID: <3D739730.DEB80478@mill.co.uk>

it does have a initmyextension declared - Here's a simple example of my
problem - I've taken all c++ related stuff out so I can use gcc

#include "Python.h"
#include<stdio.h>

int
converter(void *from, unsigned long *to)
{
        to = (unsigned long*)from;
        printf("%d %d %d", to[0], to[1], to[2]);
        return 1;
}

static PyObject *
myextension_printHi(PyObject *self, PyObject *args)
{
        printf("Hi\n");
        Py_INCREF(Py_None);
        return Py_BuildValue("");
}

static PyMethodDef myextension_methods[] = {
        {"printHi", myextension_printHi, 1, "printHi() doc string"},
        {NULL, NULL}
};

void
initmyextension(void)
{
        Py_InitModule("myextension", myextension_methods);
}


$ gcc -shared -o myextension.so myextension.cpp
-I/usr/local/dist/Python-2.2.1/Include/ -I/usr/include/python2.2/

Python 2.2 (#1, Apr 12 2002, 15:29:57)
[GCC 2.96 20000731 (Red Hat Linux 7.2 2.96-109)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import myextension
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ImportError: dynamic module does not define init function (initmyextension)
>>>


Thanks for the help

Joe

"Martin v. L?wis" wrote:

> Joe Connellan <joec at mill.co.uk> writes:
>
> > I'm trying to port an extension of mine from windows to linux and am not
> > sure how to include the myExtension.def file in the compile
>
> You shouldn't have a .def file in the first place - I recommend that
> you use distutils instead.
>
> > eg if I use
> >
> > g++ -shared -o myextension.so
> >
> > it compiles fine but I get the following error when importing it into
> > python
> >
> > >>> import myextension
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in ?
> > ImportError: dynamic module does not define init function
> > (initmyextension)
>
> Nothing to this respect is needed on Linux. Are you sure your module
> does define a initmyextension function?
>
> It might be that it defines initmyextension__Fv instead, since you
> compile your extension as a C++ program.
>
> You should compile it as a C program, unless it really is a C++
> program, in which case you need to define the init function as extern
> "C".
>
> HTH,
> Martin



From rsrchstr at msn.com  Mon Sep 30 18:05:59 2002
From: rsrchstr at msn.com (mike henley)
Date: Mon, 30 Sep 2002 23:05:59 +0100
Subject: idea for a much needed application
References: <576c1752.0209291326.56dcdf31@posting.google.com> <pasepuc0r73mh5vlcsu1l7erl4e2b17qvi@4ax.com> <an831g$bfrqn$1@ID-157863.news.dfncis.de> <ujihpu0gvloetcsadfeg706rcs64o51mjj@4ax.com>
Message-ID: <anahsf$ccdrt$1@ID-157863.news.dfncis.de>

actually, featureprice.com offer a good service; they support perl, php,
python, tcl.

"Bart Van Hemelen" <tenthousand.NOSPAM at yahoo.com> wrote in message
news:ujihpu0gvloetcsadfeg706rcs64o51mjj at 4ax.com...
> In the dim and distant past on Mon, 30 Sep 2002 00:40:24 +0100, it was
> rumoured that "mike henley" <rsrchstr at msn.com> spake thus on the
> subject of "Re: idea for a much needed application" in comp.lang.perl:
>
> >"other than the ones already present on a web-server", exactly! i want to
be
> >ably to use whatever language, modules or applications i choose, and over
> >the year use others, freely. if i choose perl, php, or even python, ruby
or
> >tcl, then i want to be able to do that.
>
> Would be sys admin hell.
>
> Quite frankly, you'd probably be cursing about your hosting service
> being crap because this would cause so much hassle to the people who
> have to keep the servers running (patches and upgrades to software,
> plus scripts by inexperienced users going haywire).
>
> I know that if I'd was working at a firm offering such things, they'd
> have to pay me a lot since it woudl mean I'd have to be extremely good
> at a whole lot of things, which in turn would mean expensive
> subscriptions for users.
>
> --
> Bart Van Hemelen
> http://pr1nc3.com/BVH/
> --------------------------------------
> Life would be a lot easier if you could take a
> look at the source code every once in a while.




From paul at boddie.net  Wed Sep  4 04:11:51 2002
From: paul at boddie.net (Paul Boddie)
Date: 4 Sep 2002 01:11:51 -0700
Subject: https in Python?
References: <ac259413.0209031134.3778c6ac@posting.google.com> <m37ki2su6r.fsf@mira.informatik.hu-berlin.de> <slrnana6iq.t8.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <23891c90.0209040011.7d9e972@posting.google.com>

Gerhard H?ring <gerhard.haering at gmx.de> wrote in message news:<slrnana6iq.t8.gerhard.haering at lilith.my-fqdn.de>...
>
> > s_gherman at yahoo.com (Sorin Gherman) writes:
> > 
> >> Is it possible to handle https using Python's libraries? 
> 
> See my page http://www.cs.fhm.edu/~ifw00065/pyssl/index.html for
> Windows.

See also Robin Dunn's page: http://alldunn.com/python/

I've had some success with the Python 2.1 libraries, and they have the
advantage that you don't need to abandon the usual APIs to use https
with them; that seems to be the case for M2Crypto, for example.

> I wonder why ActiveState doesn't provide SSL support with their Python
> distribution. I thought they want to distinguish themselves by adding
> value ;-)

I can imagine that there's still some paranoia about shipping built-in
cryptography, even in jurisdictions which don't prohibit such exports.
Also, it's additional workload for those responsible, since there are
certain special obligations in providing up-to-date, secure, reliable
implementations of crypto packages.

Paul


From Padraig at Linux.ie  Fri Sep  6 08:46:44 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Fri, 06 Sep 2002 12:46:44 GMT
Subject: tree data structure
References: <FM0e9.2860$cP3.6165@news.iol.ie> <Xns9281891D87291duncanrcpcouk@127.0.0.1>
Message-ID: <Us1e9.2865$cP3.6131@news.iol.ie>

Duncan Booth wrote:
> Padraig Brady <Padraig at Linux.ie> wrote in 
> news:FM0e9.2860$cP3.6165 at news.iol.ie:
> 
> 
>>Hi,
>>
>>I've a graph like stucture represented
>>by a list of tuples:
>>
>>[(L0,S2), (S2,S0), (S2,S1), (S1,Q2), (S0,Q1), (S2,S3)]
>>
>>And would like to convert it to
>>a tree structure like:
>>
>>           S0 - Q1
>>         /
>>L0 - S2 - S3
>>         \
>>           S1 - Q2
>>
>>Any tips appreciated.
>>
> 
> Does this help?
> 
> data = [('L0', 'S2'), ('S2', 'S0'), ('S2', 'S1'), ('S1', 'Q2'), ('S0', 
> 'Q1'), ('S2', 'S3')]
> 
> # Produce a tuple where each node contains its name
> # and a list of its child nodes:
> #  (node, [child1, ...])
> nodes = {}
> for parent, child in data:
>     nodes.setdefault(parent, (parent, 
> []))[1].append(nodes.setdefault(child, (child, [])))
> 
> result = nodes[data[0][0]]
> 
> assert result==('L0', [('S2', [('S0', [('Q1', [])]), ('S1', [('Q2', [])]), 
> ('S3', [])])])
> 
> 

It certainly does!
Just one thing...Why use lists instead of tuples?

thanks,
P?draig.

p.s. I am completely amazed by Python. What is it now.. 4 hours
and 43 minutes and I've learnt Python enough (from the web docs)
to be able to parse a "graph format" text file and outputting XML
(albeit with a little help :-)). Fantastic!



From whisper at oz.net  Tue Sep 17 15:38:30 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 17 Sep 2002 12:38:30 -0700
Subject: Earing money with Python software (was: More on Protecting Source Code)
In-Reply-To: <4jt7ma.fse.ln@lenashee.goebel-consult.de>
Message-ID: <GCEDKONBLEFPPADDJCOEMEIBFAAA.whisper@oz.net>

In my actually having done it experience, utilities and tools are a hard
sell. The market is small and developers tend to get along without or build
rather then buy unless it's just too good to do without. Developers have a
habbit of 'sharing' good tools too - heh.

Offering commercial support would be my suggestion. The efbot could probably
give you chapter and verse on that!

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Hartmut Goebel
> Sent: Tuesday, September 17, 2002 11:51
> To: python-list at python.org
> Subject: Earing money with Python software (was: More on Protecting
> Source Code)
>
>
> Martin v. Loewis wrote:
>
> > For software, I think the economical risk is that somebody uses your
> > software without giving you the money you want - not that somebody
> > sits down and studies your code for days, to write a program that he
> > could have written himself from scratch in the same time.
>
> Which brings me to another (related) topic): How to earn money
> distributing Python software?
>
> I'm currently trying to make up my mind about exactly this problem
> area: How can I get some reward back when writing software like
> decompyle and bridgekeeper?
>
> I would be very gald if someone could share his/her experiances on
> this topic!
>
> - Sell it via micro-payment?
> - Declare it "Shareware" and hope somebody pays for it?
> - Offer commercial support?
> - Develop two different versions, a "free" one which is missing some
>   features and a commercial/registered/supported one offering more
>   sophisticated functionallity?
>
> (Any other comments are welcome, too.)
>
> Regards
> +++hartmut
> --
> | Hartmut Goebel      | h.goebel at crazy-compilers.com               |
> | crazy-compilers.com | compilers which you thought are impossible |
> --
> http://mail.python.org/mailman/listinfo/python-list




From loewis at informatik.hu-berlin.de  Mon Sep  9 03:48:12 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 09 Sep 2002 09:48:12 +0200
Subject: Difference between static method and a function as a class attribute?
References: <pqchnu40hs1dgqc4bct2shi4lbpl1oe80i@4ax.com> <j4znuvb3nm.fsf@informatik.hu-berlin.de> <nlohnusn3fi2b2ksltv1kcgpl6h4340sgv@4ax.com>
Message-ID: <j4ofb7bo3n.fsf@informatik.hu-berlin.de>

Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:

> Haaaa (a light goes on in an otherwise dark and empty skull) so the
> function is implicitely mutated into an unbound method! Wasn't aware of
> this, thought that Python did nothing at all and that is why I asked -
> Always learning new things.

Yes. Guido once phrased this as "functions become methods *on the way
out* of the class". This has surprised me at first as well, but I now
consider it quite elegant, as it allows

C.m = f

where f is a function; C.__dict__['m'] will still be the function, but
C.m will be a method (whose im_func is f).

Regards,
Martin


From see_reply_address at something.invalid  Sun Sep 22 21:28:40 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Mon, 23 Sep 2002 13:28:40 +1200
Subject: Avoiding referencing (was: Python sequences reference - problem?)
References: <mailman.1032443760.5578.python-list@python.org> <YZli9.114624$ub2.2439313@news1.tin.it> <amejvb$55qsi$1@ID-77524.news.dfncis.de>
Message-ID: <3D8E6E48.5070708@something.invalid>

Thorsten Kampe wrote:

> 2. "If you don't want referencing, call it with a copy of 'foo' like: 
> part(foo[:], [2, 3])"
> 
> Yes, but how do I know in advance that 'part' modifies 'foo'? I might 
> not be the author of 'part'.


But it ought to be clear from the documented purpose
of 'part' whether it changes the object that you pass
to it or not. It's generally expected that a function
won't make changes to its arguments unless there's a
good reason.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From jason at tishler.net  Mon Sep 16 14:00:57 2002
From: jason at tishler.net (Jason Tishler)
Date: Mon, 16 Sep 2002 14:00:57 -0400
Subject: Accessing Windows Registry under cygwin
In-Reply-To: <f5bbs6x7t0k.fsf@cogsci.ed.ac.uk>
References: <f5b1y8298w2.fsf@cogsci.ed.ac.uk>
 <20020916164259.GG424@tishler.net> <f5bbs6x7t0k.fsf@cogsci.ed.ac.uk>
Message-ID: <20020916180057.GQ424@tishler.net>

Henry,

On Mon, Sep 16, 2002 at 06:18:19PM +0100, Henry S. Thompson wrote:
> Jason Tishler <jason at tishler.net> writes:
> > Are you willing to help?
> 
> I wish I were competent to help -- I have no experience with
> programming to WIN32 APIs.

Hmm...  You are the one with the itch...

> I'll download the package and have a look.

Now that's the attitude!

> It's not entirely clear from the sourceforge log what's actually
> needed next.

Why not ask via the SourceForge Patch Manager?  I'm sure that Martin or
Jerry would appreciate any help.

BTW, if you don't care about being portable with Win32 Python, then try
using the following workarounds:

    1. /proc/registry (but is read-only)
    2. regtool (but spawns another process)

Jason



From dsavitsk at e-coli.net  Mon Sep 16 01:22:29 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Mon, 16 Sep 2002 05:22:29 GMT
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de>
Message-ID: <pUdh9.9377$yt3.4604908@newssrv26.news.prodigy.com>

It is also worth repeating the often mentioned point that this is a job for
intellectual property law. If your code is that valuable, run it on your
server remote from end users as a trade secret, or force them to agree to a
license and protect it w/ copyright.

-d


"Terje Johan Abrahamsen" <spoermeg at voldelig.com> wrote in message
news:am3ms1$2e49b$1 at ID-53055.news.dfncis.de...
> If I would like to write some code that would not be possible for others
to
> read, after I distribute the program, would Python be suitable? I know
that
> there are a few programs like Py2Exe, that creates .exe files. But, I also
> saw this:
>
> --------------------
>  A zip-compatible archive is built, containing all files from this
>   directory as well as your script, and appended to a custom python
>   interpreter supplied with py2exe.
> --------------------
>
> Doesn't sound like it is protected very well. Is there some alternative to
> Py2Exe that can do the job, or should I look for a compiled language
> instead?
>
>
>




From shifflett at nps.navy.mil  Thu Sep 12 17:32:39 2002
From: shifflett at nps.navy.mil (David Shifflett)
Date: Thu, 12 Sep 2002 14:32:39 -0700
Subject: Signals for GtkEntry widgets
Message-ID: <3D8107F7.9090408@nps.navy.mil>

I am trying to find which signal is emitted
when the user presses Enter while in a GtkEntry widget.

I finally found some online docs which listed the various signals
for the various widgets at
http://developer.gnome.org/doc/API/2.0/gtk/

Unfortunately, these are terse, mainly documenting
the API, and not describing uses/functionality/etc

IF someone could point me to a reference,
or an example, it would be most appreciated

Meanwhile I will keep hacking away until
I stumble upon something that works

Thanks in advance
David Shifflett




From blackhole at notformail.invalid.de  Tue Sep 24 08:25:53 2002
From: blackhole at notformail.invalid.de (Volker Apelt)
Date: 24 Sep 2002 14:25:53 +0200
Subject: callback + __call__ + TypeError: unbound method; on non method
Message-ID: <lg8z1rtvzi.fsf@mu.biosolveit.local>

Hi,

I have a problem with a stored function in 
a callback class object.  I try to create a class 
(Callback) which accepts a function (a callable) as argument,
stores it in an object attribute (self._m_func) 
and uses it in Callback's.__call__ function.

The function stored is _not_ a method of 
of the calling class.

I tried several types of calls to self._m_func
but I allways end up in 

TypeError: unbound method call_me() must be called with instance as first argument

I tried some other class (XXX) without a .__call__
method, which calls self._m_func and it succeeds. 

But I can't remove the .__call__ method from my 
Callback class. The method is essential for
Callback, because an instance of it is 
used by some wrapped C-API function (not shown) 
to initiate some action on the python side 
of my code.

Do you know what is wrong in the following 
example?

I'm looking for an acceptable workaround or 
fix.

--------------------------  demo code follows
#!/usr/bin/python 
#  -*- python -*- 

# python -v 2.1.1 on linux (Suse 7.3) 
import time

# the called function
def call_me():
    return time.time

# demo: this class works as expected  
class XXX:
    def __init__(self):
        #self.m_func = time.time
        self.m_func = call_me

    def callback_currTime(self):
        return self.m_func()() # becomes self.m_func() with time.time above

    def direct_currTime(self):
        return time.time()

# defekt: this class doesn't    
class Callback:
    """
    """
    _default_Func = call_me # class wide default callback function
    
    def __init__(self):
        """ """
        self._m_func = Callback._default_Func
        
    def __call__(self, case=None):
        """ """
        ## do something before calling back
        ### -- left out
        ## all attempts end in this type error
        ## TypeError: unbound method call_me() must be called \
        ##   with instance as first argument
        ## find a valid form of calling the function stored 
        ## in _m_func 
        ##   fazit: None was found
        if   case is None or case == 1:
            return self._m_func()
        elif case == 2:
            return self.Func()()
        elif case == 3:
            C = self.Func()()
            return C()
        else:
            return "ignored"
        ##

    def useFunc(self,func=None):
        """ """
        if func is None:
            self._m_func = Callback._default_Func
        else:
            self._m_func = func

    def Func(self):
        """ """
        return self._m_func
#
#####

x = XXX()

print x.direct_currTime()
print x.callback_currTime()

x = Callback()

# try all cases 
for c in xrange(5):
    try:
        print x(c)
    except TypeError, e:
        print "CASE ",c,": ",e

##  OUTPUT  of this script 
#1032868956.78
#1032868956.78
#ignored
#CASE  1 :  unbound method call_me() must be called with instance as first argument
#CASE  2 :  unbound method call_me() must be called with instance as first argument
#CASE  3 :  unbound method call_me() must be called with instance as first argument
#ignored

# Local Variables:
#  (auto-fill-mode nil)
# End:
#end of file
---------------------- end of demo code

run it to see the problems.

thanks 


Volker

-- 
Volker Apelt                   




From martin at v.loewis.de  Mon Sep 16 01:39:12 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Sep 2002 07:39:12 +0200
Subject: [xml] convert funny chars to char entites?
References: <vNah9.10101$Ee4.23987@news-server.bigpond.net.au>
Message-ID: <m3elbuze67.fsf@mira.informatik.hu-berlin.de>

":B nerdy" <thoa0025 at mail.usyd.edu.au> writes:

> ive got a string "Il Est N?" and i want to put it into a xml document
> 
> is there a quick and easy way to conver the funny characters to char
> entities so i can store them?

I recommend you properly declare the encoding of the document, e.g.

<?xml version="1.0" encoding="iso-8859-1"?>

(assuming the ? is meant to be encoded in Latin-1).

XML does not define external entities for non-ASCII characters, so you
would have to create a DTD first if you want them.

Regards,
Martin



From nospam at horvath.com  Thu Sep 26 02:14:03 2002
From: nospam at horvath.com (Bob Horvath)
Date: Thu, 26 Sep 2002 06:14:03 GMT
Subject: combining mailbox and email modules
References: <NPck9.581867$UU1.104088@sccrnsc03> <slrnap41re.1mh.matt@happy-hour.t-island.com>
Message-ID: <LAxk9.79207$gA4.27843@sccrnsc02>

Matthew Dixon Cowles wrote:
> On Wed, 25 Sep 2002 06:36:29 GMT, Bob Horvath <nospam at horvath.com>
> wrote:
> 
> 
>>I am having trouble figuring out how to get the mailbox and email
>>modules to play together so that I can use email module functions on
>>every mail in the mailbox.  Is there a way to feed a factory from
>>the email module to mailbox module?
> 
> 
> Yes, there is:
> 
> 
>>>>import mailbox
>>>>import email
>>>>i=mailbox.Maildir("Mail/testmail",email.message_from_file)
>>>>i.next()
>>>

OK, I tried that.  Looks like the email package is being picky about 
things where I can't see a problem.


I keep getting the following error:

email.Errors.HeaderParseError: Not a header, not a continuation

Which I assume means a line that doesn't start with a space or tab 
that also does not contain a ':'.

The problem is that it looks good to me.  I've played around with 
it, and if I take out the MIME gunk (It actually is a Klez virus), 
it parses ok.

Here is what happens...



Python 2.2.1 (#1, Sep  7 2002, 14:34:30)
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
 >>> import email
 >>> fp = file('large')
 >>> msg = email.message_from_file(fp)
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "/usr/lib/python2.2/email/__init__.py", line 35, in 
message_from_file
     return _Parser(_class).parse(fp)
   File "/usr/lib/python2.2/email/Parser.py", line 41, in parse
     self._parsebody(root, fp)
   File "/usr/lib/python2.2/email/Parser.py", line 134, in _parsebody
     msgobj = self.parsestr(part)
   File "/usr/lib/python2.2/email/Parser.py", line 45, in parsestr
     return self.parse(StringIO(text))
   File "/usr/lib/python2.2/email/Parser.py", line 40, in parse
     self._parseheaders(root, fp)
   File "/usr/lib/python2.2/email/Parser.py", line 81, in _parseheaders
     raise Errors.HeaderParseError(
email.Errors.HeaderParseError: Not a header, not a continuation
 >>>

... and here it the message...

 From mickey at verizon.net  Tue Aug 20 10:03:52 2002
Return-Path: <mickey at verizon.net>
Delivered-To: bob at horvath.com
Received: from localhost (bert [127.0.0.1])
         by horvath.com (Postfix) with ESMTP id 9FE612A21D
         for <bob at horvath.com>; Tue, 20 Aug 2002 10:03:51 -0500 (CDT)
Delivered-To: horvath at enteract.com
Received: from imap.enteract.com
         by localhost with IMAP (fetchmail-5.9.11)
         for bob at horvath.com (multi-drop); Tue, 20 Aug 2002 10:03:51 
-0500 (CDT)
Received: (qmail 40801 invoked by alias); 20 Aug 2002 14:58:55 -0000
Delivered-To: alias-horvath_com-bob at horvath.com
Received: (qmail 40787 invoked by uid 0); 20 Aug 2002 14:58:53 -0000
Received: from out007pub.verizon.net (HELO out007.verizon.net) 
(206.46.170.107)
   by pop3-1.enteract.com with SMTP; 20 Aug 2002 14:58:53 -0000
Received: from Urn ([160.39.139.46]) by out007.verizon.net
           (InterMail vM.5.01.05.09 201-253-122-126-109-20020611) 
with SMTP
           id <20020820145852.GWSH1953.out007.verizon.net at Urn>
           for <bob at horvath.com>; Tue, 20 Aug 2002 09:58:52 -0500
From: mickey <mickey at horvath.com>
To: bob at horvath.com
Subject: Document.cookie
MIME-Version: 1.0
Content-Type: multipart/alternative;
         boundary=B969j3A4oe4I6mIO538a14886om5n
Message-Id: <20020820145852.GWSH1953.out007.verizon.net at Urn>
Date: Tue, 20 Aug 2002 09:58:53 -0500
Status: RO
X-Status:
X-Keywords:
X-UID: 52

--B969j3A4oe4I6mIO538a14886om5n
Content-Type: text/html;
Content-Transfer-Encoding: quoted-printable

<HTML><HEAD></HEAD><BODY>
<iframe src=3Dcid:S3vx2Jnbs6R29 height=3D0 width=3D0>
</iframe>
<FONT></FONT></BODY></HTML>

--B969j3A4oe4I6mIO538a14886om5n
Content-Type: audio/x-wav;
         name=Wsr.exe
Content-Transfer-Encoding: base64
Content-ID: <S3vx2Jnbs6R29>

TVqQAAMAAAAEAAAA//8AALgAAAAAAAAAQAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAA=9
--B969j3A4oe4I6mIO538a14886om5n
--B969j3A4oe4I6mIO538a14886om5n
Content-Type: application/octet-stream;
         name=visa_nfl_espn_freq_JS[1].htm
Content-Transfer-Encoding: base64
Content-ID: <S3vx2Jnbs6R29>

PGh0bWw+DQo8Ym9keT4NCjxzY3JpcHQgbGFuZ3VhZ2U9ImphdmFzY3JpcHQiPg0KIHZhciBC
Cj==
--B969j3A4oe4I6mIO538a14886om5n--

bert%



From thorsten at thorstenkampe.de  Fri Sep 13 13:17:49 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Fri, 13 Sep 2002 19:17:49 +0200
Subject: [ANN] PySQLite 0.3.0 released
References: <mailman.1031924237.12603.python-list@python.org>
Message-ID: <alt6k1$t0jg$1@ID-77524.news.dfncis.de>

* sismex01 at hebmex.com
>> From: Bob X [mailto:bobx at linuxmail.org]
>> [Merciless Fullquote]
>
> Yeesh! You had to quote the whole message just to
> add two words?

That's nothing. Lately I've seen someone on the tutor list /quoting 
the whole message digest/ while replying to /a single message/. Now 
isn't that sweet?

Thorsten 


From jstavnitzky at atsautomation.com  Fri Sep 20 23:28:40 2002
From: jstavnitzky at atsautomation.com (Jay)
Date: 20 Sep 2002 20:28:40 -0700
Subject: Porting woes
References: <3a4b2539.0209200809.407d50d5@posting.google.com> <j4sn041tuz.fsf@informatik.hu-berlin.de>
Message-ID: <3a4b2539.0209201928.10c577b9@posting.google.com>

> It checks whether the cls argument to classic_mro really is a classic
> class. It is surprising that it calls into classic_mro. Can you tell
> which line in Py_ReadyTypes comes from? (type, list, string, ...?)

It occurs in the initialization of the type..

> It would be good to study the value of type, and tp_bases, at this
> point. It might be that some initialization of global objects has
> failed to happen, or has happened incorrectly.

Thanks for the suggestion!! It appears as though this particular
problem goes away if I extend the static initialization of the
PyTypeObject PyBaseObject_Type to explicitly set the tp_bases value
(and the other six values that are not explicitly assigned at the end
of the structure to NULL).  If I don't explicitly zero these values,
they pick up whatever happens to be left over in those memory
locations, is it possible that this is a compiler bug?? I assume that
this code compiles correctly on most other compilers, or is zeroing
out memory locations before use an operating system function??

I hope that there aren't other places in Python where structures are
not fully initialized (and I may see a bug of this type again).

Thanks,
Jay

loewis at informatik.hu-berlin.de (Martin v. L?wis wrote in message news:<j4sn041tuz.fsf at informatik.hu-berlin.de>...
> jstavnitzky at atsautomation.com (Jay) writes:
> 
> > This failure is occuring during the function classic_mro().  The
> > stack at that point is
> > classic_mro(),mro_implementation(),mro_internal(),PyType_Ready(),
> > PyTypeReady(),Py_ReadyTypes(),Py_Initialize(),Py_Main().
> > 
> > Help!! What is happening at this point, and what could be causing this
> > assertion to fail?
> 
> It checks whether the cls argument to classic_mro really is a classic
> class. It is surprising that it calls into classic_mro. Can you tell
> which line in Py_ReadyTypes comes from? (type, list, string, ...?)
> 
> None of these are classic classes, so this code should not have been
> executed. It appears that this was executed because of the test in
> 
> 		if (PyType_Check(base))
> 			parentMRO = PySequence_List(
> 				((PyTypeObject*)base)->tp_mro);
> 		else
> 			parentMRO = classic_mro(base);
> 
> i.e. base appears not to be a type. This, again, should not happen:
> base is
> 
> 		PyObject *base = PyTuple_GET_ITEM(bases, i);
> 
> which in turn comes from
> 
> 	bases = type->tp_bases;
> 
> All base types of the types initialized in Py_ReadyTypes have only
> types as bases, so it is surprising that one of them has a base which
> is not a type.
> 
> It would be good to study the value of type, and tp_bases, at this
> point. It might be that some initialization of global objects has
> failed to happen, or has happened incorrectly.
> 
> Regards,
> Martin


From benji_75 at hotmail.com  Fri Sep 13 05:46:14 2002
From: benji_75 at hotmail.com (Ben Gerblich)
Date: Fri, 13 Sep 2002 19:16:14 +0930
Subject: no RHL 7.1 RPMs for python2.1
Message-ID: <F209zvb5YI9IJsaJbqJ00000222@hotmail.com>

Why now does the RPMs for Python2 not appear on python.org?

I am having trouble with my application core dumping.

I have downloaded and installed the following on July2002:
python2-2.1-5.i386.rpm
python2-devel-2.1-5.i386.rpm
python2-tkinter-2.1-5.i386.rpm
python2-tools-2.1-5.i386.rpm
wxPython-2.3.2.1-1-Py21.i386.rpm
wxGTK-2.3.2-1.i386.rpm
wxGTK-devel-2.3.2-1.i386.rpm
using Red Hat Linux 7.1

Is there a known problem?
BenG.


_________________________________________________________________
Chat with friends online, try MSN Messenger: http://messenger.msn.com




From fgranger at alussinan.org  Fri Sep 20 14:20:15 2002
From: fgranger at alussinan.org (=?ISO-8859-1?Q?Fran=E7ois_Granger?=)
Date: Fri, 20 Sep 2002 20:20:15 +0200
Subject: python on the mac.
References: <c07d0edc.0209190734.44c7f41f@posting.google.com>
Message-ID: <1fitsez.1wgbphvp6l8kmN%fgranger@alussinan.org>

Guy <guy at lightwork.co.uk> wrote:
> I have a script which I created on a win32 machine, I now want to port
> it to a mac sys9x, python version 2.2.1.

These pages will teach you all you need....

http://python.org/doc/current/mac/using.html

-- 
YB> Phase 1: m?diter sur la question "Qu'est qu'une mauvaise herbe" ?
JLV> surtout un jugement de valeur pl?onastiquement jud?o-chr?tien
> et plus largement religieux
YB> Monsanto est leur Dieu et Roundup est son proph?te ?


From arturs at iidea.pl  Wed Sep 11 13:41:34 2002
From: arturs at iidea.pl (Artur Skura)
Date: Wed, 11 Sep 2002 17:41:34 +0000 (UTC)
Subject: New text-mode ui library
Message-ID: <slrnanv0nh.e1b.arturs@aph.waw.pdi.net>

If you looked at (c)tinter and did not like it, you may check
ntinter, a new text-mode UI library based on Python bindings
for ncurses by Bartosz Lambach. It's available from 

http://iidea.pl/~arturs/ntinter-0.1.tar.gz

(sample screenshoot: http://arturs.iidea.pl/b.png)

-- only basic objects are implemented (Dialogs, Labels,
Text Entries, Buttons), but it's quite useful.

Regards,
Artur
-- 
http://www.amnesty.org.pl


From bloke at ii.net  Sun Sep  8 07:27:52 2002
From: bloke at ii.net (Rob Hall)
Date: Sun, 8 Sep 2002 19:27:52 +0800
Subject: python program to watch a file - example.
References: <mailman.1031464092.30848.python-list@python.org> <ales54$2un$0@216.39.172.122>
Message-ID: <3d7b3437$0$12727@echo-01.iinet.net.au>

Get a fucking life Bengt!!!


Bengt Richter <bokr at oz.net> wrote in message
news:ales54$2un$0 at 216.39.172.122...
> On Sun, 8 Sep 2002 01:47:12 -0400, "Wayne R" <wayne at tbnets.com> wrote:
>
> >This is a multi-part message in MIME format.
> Why?
> >
> >------=_NextPart_000_0039_01C256D9.B8CB5320
> >Content-Type: text/plain;
> > charset="us-ascii"
> >Content-Transfer-Encoding: 7bit
> >
> >
> >I am trying to write a python program that will run as a deamon and
> >watch a file for data to arrive and then wake up and perform duties on
> >that data.   I found ways to make a python program run as a deamon but I
> >am at a loss to find anything related to watching a file for data so
> >that it can then process the data.  I didn't really want to do a sleep
> >loop or something like, I am sure there is a better way.
> >
> >Any ideas or links to more info?
> >
> Don't annoy people you want help from, for starters.
> What system are you targeting? (Outlook in the headers says you're on
windows
> generating this abominable mix of actual message and useless cruft).
>
> >------=_NextPart_000_0039_01C256D9.B8CB5320
> >Content-Type: text/html;
> WHY?!!
>
> Regards,
> Bengt Richter




From gerhard.haering at gmx.de  Sat Sep 28 22:27:27 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 29 Sep 2002 02:27:27 GMT
Subject: Fast-track to installing/implementing a SWIP module?
References: <Mktl9.13603$Lg2.3951792@news2.news.adelphia.net>
Message-ID: <slrnapcp8g.617.gerhard.haering@lilith.my-fqdn.de>

Robert Oschler wrote in comp.lang.python:
> I am a Python newbie so bear with me here.  I just downloaded the PyProlog
> package, which allows Python to access SWI-Prolog through a C-code SWIP
> module.  I just need the basics such as, proper directory to install files
> and instructions for making sure the SWIP module can find SWI-Prolog, and
> for making sure the package will run properly in Python. Is there a URL or
> doc that can get me started?  I would assume that I have to compile the 'C'
> source for the SWIP module, and the resultant object file (.so?) in a proper
> directory?  These the kinds of questions I need to get answered.  I read the
> 3 README's in the PyProlog package but nothing seemed to suffice.

If I understand correcly, you only have to have SWI-Prolog in your PATH, then
the following:

    python setup.py build

will build everything and

    python setup.py install

will install your modules.


Explanation: This function:

    def get_atom(atom):
        query = "pl -f none -g 'current_prolog_flag(%s, V), write(V), nl.' -t 'halt'"
        f = os.popen(query % atom)
        ...

in setup.py should do the magic of finding the include and library paths to
link against SWI Prolog.

-- Gerhard


From aleax at aleax.it  Thu Sep 19 02:49:43 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 19 Sep 2002 06:49:43 GMT
Subject: monetary applications (et al)
References: <mailman.1032390744.1503.python-list@python.org>
Message-ID: <bsei9.130231$pX1.4674970@news2.tin.it>

terry wrote:

> 
>>>> Michael wrote:
> for how to emulate numeric types.
>>>>
> 
> Right, I have looked at that.  As near as I can tell there is a vast
> difference between 'emulation' and 'intrinsic' - and the burden of that
> difference falls on the programmer/customer in your assessment.  The

I don't understand what you mean.  Where's the "vast difference",
and what burden falls on the "programmer/customer"?  If a type is
not built-in (e.g.: cStringIO.StringIO, array.array, mmap.mmap, to
mention a few non-intrinsic types distributed with core Python),
then you need an "import something" statement.  That's not deeply
different from, e.g., needing an "import math" before you can
compute trig functions (then you call math.sin, of course, not
just sin).  Is *THIS* what you consider "a vast difference"???  I
can't think of anything else.

Python users who need to compute trig functions have never squealed
out load about the intolerable burden of having to "import math"
to do so.  On the contrary, people with such needs have developed
a large, now widespread package called Numeric, which supplies a
zillion facilities for such computations, and is used as a base by
a large number of other, mode specialized computational modules.

This substantial minority of Python users start almost all of their
Python programs with "import Numeric" or "from Numeric import".

Doesn't seem to have done any damage to Python's standing in the
field of scientific computation -- on the contrary, decoupling the
computational parts from Python's core allows the two parts to
evolve and get released independently, thus benefiting both those
who need such advanced maths, and those who couldn't care less.

So what distinguishes the set of people who'd like a "currency"
type from those who wanted advanced numerical maths -- except
maybe that one set is more inclined to whining, the other, to doing
something about it...?


Alex



From martin at v.loewis.de  Wed Sep 18 00:45:30 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 18 Sep 2002 06:45:30 +0200
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <mailman.1032152058.11043.python-list@python.org> <m3admizdzq.fsf@mira.informatik.hu-berlin.de> <am8epu$3ml4d$2@ID-53055.news.dfncis.de>
Message-ID: <m3wupjhpn9.fsf@mira.informatik.hu-berlin.de>

"Terje Johan Abrahamsen" <spoermeg at voldelig.com> writes:

> If Py2Exe stores the source in pretty much open for reading, I need
> to use another way of protecting it.

Py2Exe does not store source code.

Regards,
Martin



From robie at 1373.net  Mon Sep 23 22:49:17 2002
From: robie at 1373.net (robie1373)
Date: Mon, 23 Sep 2002 20:49:17 -0600
Subject: confusion about opening files
References: <3d8fcda4$0$198$75868355@news.frii.net> <3D8FCFF3.FD52AFE5@alcyone.com>
Message-ID: <3d8fd215$0$192$75868355@news.frii.net>

O. K. Now I am confused.  I thought that when I imported the os module, i
would have to call its functions like this: os.open()    Why am I just using
the function name?

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3D8FCFF3.FD52AFE5 at alcyone.com...
> robie1373 wrote:
>
> ...
> >     file = os.open("e:\\d12", "r")
> > TypeError: an integer is required
> >
> > I get this error if the mode is r, r+, a+ or whatever.  I also get it
> > whether the file exists or not. Can someone explain what I am doing
> > wrong?
>
> You meant open, not os.open.
>
> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ Insight -- the titillating knack for hurting!
> \__/ Colette
>     Erik Max Francis' bookmarks / http://www.alcyone.com/max/links/
>  A highly categorized list of Web links.




From dreed at capital.edu  Thu Sep 19 17:04:41 2002
From: dreed at capital.edu (Dave Reed)
Date: Thu, 19 Sep 2002 17:04:41 -0400
Subject: monetary applications (et al)
In-Reply-To: <3D8A161E.8060906@swbell.net> (message from Greg Goodman on Thu,
	19 Sep 2002 18:06:57 GMT)
References: <mailman.1032390744.1503.python-list@python.org> <bsei9.130231$pX1.4674970@news2.tin.it> <mailman.1032443727.4727.python-list@python.org> <3D8A161E.8060906@swbell.net>
Message-ID: <200209192104.g8JL4f816159@localhost.localdomain>

> From: Greg Goodman <chironsw at swbell.net>
> Date: Thu, 19 Sep 2002 18:06:57 GMT

<snip>


> On the other hand, cents aren't really indivisible in the American 
> marketplace, where stock prices often change by fractions of a dollar, 
> as in "General Motors down 2 and 1/8".

I didn't realize the stock market was so bad that nobody looks at it
anymore. :-) US stock exchanges have all been using dollars and cents
for a while now.

<snip>

Dave



From ark at research.att.com  Fri Sep 27 19:28:49 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 27 Sep 2002 23:28:49 GMT
Subject: What does Python fix?
References: <e15b6fa3.0209271418.14cd5642@posting.google.com>
Message-ID: <yu99ptuzui4u.fsf@europa.research.att.com>

Adam> I'm curious to read a bit more about why Tim Peters (and presumably
Adam> others) think that Lisp-inspired languages (even non-prefix ones) are
Adam> doomed to eternal obscurity.  Would anyone care to comment and/or give
Adam> me some pointers to commentary on the subject?

You might start here:
http://www.ai.mit.edu/docs/articles/good-news/good-news.html

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From nospam at here  Thu Sep  5 13:59:48 2002
From: nospam at here (Ricardo)
Date: Thu, 5 Sep 2002 18:59:48 +0100
Subject: How to get os.popen subprocess PID?
References: <3D775167.6060307@pp.htv.fi>
Message-ID: <1031248567.61579.0@doris.uk.clara.net>

> I have a subprocess opened with os.popen, producing undefined length of
> data.  I want to read maybe some of this data and then get rid of the
> subprocess, not waiting until the end of data.  I have thought of
> os.kill-ing the process.  So far I haven't been able to determine how to
> get the pid of the subprocess, required by os.kill .. Any pointers?
>

Instead of using os.popen() you could import the popen2 module and use the
Popen3 or Popen4 classes.

Objects of these classes contain a 'pid' attribute.







From wham10 at mad.scientist.com  Mon Sep 16 06:57:28 2002
From: wham10 at mad.scientist.com (Srijak Rijal)
Date: Mon, 16 Sep 2002 05:57:28 -0500
Subject: (no subject)
Message-ID: <20020916105728.66744.qmail@mail.com>

Hi all

I just migrated from perl to python and was wondering if python socket can be read line by line too?
How can I get only one line of input from the server without losing info?

Thanxs in advance


-- 
__________________________________________________________
Sign-up for your own FREE Personalized E-mail at Mail.com
http://www.mail.com/?sr=signup




From marklists at mceahern.com  Thu Sep 19 10:25:16 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 19 Sep 2002 09:25:16 -0500
Subject: strong/weak - dynamic/static [Was: Getting started]
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F06233024CC847@mail.hebmex.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOIEOJDPAA.marklists@mceahern.com>

[sismex01 at hebmex.com]
> Python is definitely NOT weakly typed, it's dynamicly-strong
> typed. A variable, being only a "reference to *any* object"
> may refer to any object, but that object (see the difference?)
> may not change it's functionality or type during runtime.

Whatchoo talkin' bout Willis?

I don't have any practical use for the following, but it seems to
demonstrate that what you said is not true...

$ python
Python 2.2.1 (#1, Jun 25 2002, 10:55:46)
[GCC 2.95.3-5 (cygwin special)] on cygwin
Type "help", "copyright", "credits" or "license" for more information.
>>> class foo:
...     def fooize(self):
...             print self.__class__.__name__
...

>>> class bar:
...     def barize(self):
...             print self.__class__.__name__
...
>>> f = foo()
>>> f.fooize()
foo
>>> f.__class__ = bar
>>> f.fooize()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: bar instance has no attribute 'fooize'
>>> f.barize()
bar

// m

-




From grabiller at 3dvf.net  Sun Sep 29 19:49:22 2002
From: grabiller at 3dvf.net (Guy Rabiller)
Date: Mon, 30 Sep 2002 01:49:22 +0200
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com> <874rc8d9c9.fsf@nospam.eml.cc> <mailman.1033319977.7828.python-list@python.org> <3d976fbc$0$23666$91cee783@newsreader02.highway.telekom.at> <8428na.7vj.ln@ix.netcom.com>
Message-ID: <3d979184$0$20332$79c14f64@nan-newsreader-03.noos.net>

[Dennis Lee Bieber wrote:]
> Don't know if MS-DOS supports redirecting stderr.
> On linux you'd use
>
>         python -v script.py 2>output.txt

yes it works on MS-DOS too.

--
guy rabiller
3d animator / td
grabiller at 3dvf.net
http://grabiller.3dvf.net

Dennis Lee Bieber wrote:
> Jurie Horneman fed this fish to the penguins on Sunday 29 September
> 2002 02:26 pm:
>
>> Actually, I just tried this in a DOS shell under Windows 98:
>>
>> python -v script.py >output.txt
>>
>> It lists the verbose output straight to the DOS shell, not to
>> output.txt. The output is useless, I can't scroll back.
>>
>         I suspect -v output uses stderr, not stdout.
>
>         >output.txt redirects stdout
>
>         Don't know if MS-DOS supports redirecting stderr. On linux
> you'd use
>
>         python -v script.py 2>output.txt
>
>
>         However, for this situation -- ie, trying to debug your import
> statement -- don't specify the full script on the command line. Do it
> interactively with the output going to the screen.
>
>         Oh, BTW... you NEED TWO "v"... python -vv
>
>
>>>> import tmp.some     #does not exist
> # trying tmp.so
> # trying tmpmodule.so
> # trying tmp.py
> # trying tmp.pyc
> # trying /usr/lib/python2.2/tmp.so
> # trying /usr/lib/python2.2/tmpmodule.so
> # trying /usr/lib/python2.2/tmp.py
> # trying /usr/lib/python2.2/tmp.pyc
> # trying /usr/lib/python2.2/plat-linux-i386/tmp.so
> # trying /usr/lib/python2.2/plat-linux-i386/tmpmodule.so
> # trying /usr/lib/python2.2/plat-linux-i386/tmp.py
> # trying /usr/lib/python2.2/plat-linux-i386/tmp.pyc
> # trying /usr/lib/python2.2/lib-tk/tmp.so
> # trying /usr/lib/python2.2/lib-tk/tmpmodule.so
> # trying /usr/lib/python2.2/lib-tk/tmp.py
> # trying /usr/lib/python2.2/lib-tk/tmp.pyc
> # trying /usr/lib/python2.2/lib-dynload/tmp.so
> # trying /usr/lib/python2.2/lib-dynload/tmpmodule.so
> # trying /usr/lib/python2.2/lib-dynload/tmp.py
> # trying /usr/lib/python2.2/lib-dynload/tmp.pyc
> # trying /usr/lib/python2.2/site-packages/tmp.so
> # trying /usr/lib/python2.2/site-packages/tmpmodule.so
> # trying /usr/lib/python2.2/site-packages/tmp.py
> # trying /usr/lib/python2.2/site-packages/tmp.pyc
> # trying /usr/lib/python2.2/site-packages/Numeric/tmp.so
> # trying /usr/lib/python2.2/site-packages/Numeric/tmpmodule.so
> # trying /usr/lib/python2.2/site-packages/Numeric/tmp.py
> # trying /usr/lib/python2.2/site-packages/Numeric/tmp.pyc
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> ImportError: No module named tmp.some
> --




From sadams123 at optushome.com.au  Thu Sep 12 12:25:08 2002
From: sadams123 at optushome.com.au (Steven)
Date: Fri, 13 Sep 2002 02:25:08 +1000
Subject: win ce install
References: <newscache$bonb2h$5w2$1@lnews.actcom.co.il>
Message-ID: <3d80bfe4$0$23171$afc38c87@news.optusnet.com.au>

"Ell" <ell at cav.co.il> wrote in message
news:newscache$bonb2h$5w2$1 at lnews.actcom.co.il...
> wince install
> 1.
> how I install python on wince ?

Assuming you have the zip file with all the components

copy

python.exe to \Windows\Start Menu

 -this file might be called 'pocket pc python.exe' or something else, either
way, it'll be the only executable file in there. You can copy python.exe to
other locations in the Start menu e.g. Programs - don't use a shortcut to
it, it won't work.

Create a directory "\Program Files\Python" and copy the 'Lib' directory (And
files) there - this directory is like the Lib directory on a normal python
distribution, it holds all the modules.

Copy the .pyd files win32event, win32api to the "\Program Files\Python"
directory - I guess they could go in the Lib directory, but this is just my
setup.

Lastly, copy the dlls to \WIndows - this is PyWinTypes.dll and PYTHON22.dll

I also have a copy of Python.exe in \Windows - I use that as the target for
the python file associations I created in the registry. Having it in
\Windows as well gives me a non-moving target, just in case I change where
in the menu structure i keep python.

I also went through and removed all the CVS related files to save space,
other than a few modules, I only keep the compiled versions ('.pyc') in Lib,
since it saves on space.

> 2.
> has somebody experience , python on wince ?
> are all objects work as on win32 ?

Which version of PythonCE are you using?

1.5.2 - Brian Loyds (?) port
1.5.2 - Mark Hammonds port
2.2/2.3 Brad Clements port

Different versions had different parts working.Since i have an iPAQ there
was no choice, version 2.2 for me - it does pretty much everything I want -
except 'select'.

Don't expect any easy to use GUI toolkits like wxPython or Tkinter, the only
thing available at the moment is the win32 gui stuff - nowhere near as nice
to use.

> how is the performance ?

depends, its much slower than normal Python on a desktop, but I've found
Python slower on a lot of machines - whether that is the machines fault, or
more work being down on Python on the x86, who knows.

For comparison:

AthlonXP 1800, 256MB, XP Pro - 21375.1 pystones/second
Intel P133, 84 MB, Win98SE - 566.959 pystones/second
iPAQ H3870, PocketPC 2002 - 100 pystones/second

Benchmarks only tell so much, basically, yes, its slower, but its more than
fast enough for using on the PocketPC.

Try what you want to do, thats the only way to really find out if it can do
what you want.

hope tha helps

Steven







From cliechti at gmx.net  Tue Sep 17 17:20:11 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 17 Sep 2002 23:20:11 +0200
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com>
Message-ID: <Xns928CED692541cliechtigmxnet@62.2.16.82>

lpgray at uop.com (Lee Gray) wrote in 
news:96904b50.0209171305.6d65f574 at posting.google.com:
> My boss is pretty open-minded, but also needs to see something working
> to be convinced Python is even a viable platform (he had never heard
> of it).  Otherwise, .Net is a given, whether it's any good or not,
> since corporate is MS all the way.

with python you have to choice of diffrent GUI libs. if you consider 
wxWindows you can play around with the wxPython demo application that comes 
with the download. on the other hand when you support win32 only the 
win32all package has support for all the windows APIs and it has some 
samples that get installed.

chris
-- 
Chris <cliechti at gmx.net>



From tonyl at pillarsoftware.com  Sun Sep  8 00:43:08 2002
From: tonyl at pillarsoftware.com (Tony Leotta)
Date: Sun, 8 Sep 2002 00:43:08 -0400
Subject: C++ and Python
Message-ID: <001401c256f2$3ab5e9f0$f46ede3f@THOTH>

Does any one know of any examples written in C++ that use Python as a
scripting language?
 
I have been looking at VBA.but I don't know all that COM interfacing
gets me down. :(
 
I want to be able to export variables from C++ to the space that all the
scripts will run.and have 
many different python scripts running inside different threads
access/modify that data.
so I will need locking mechanisms as each thread modifies the global
data.
 
I have been written extremely fast C++ servers have expanding pools of
worker threads.and in each thread
I'd like to be able to run scripts.I guess almost how Apache or IIS call
PHP or VB Script.
 
The difference is that I want each Python "script" to be able to
communicate to other threads.something that Apache and IIS do not do.
These will not be short lived threads.some be run for days.
 
Thanks
Tony Leotta
http://www.pillarsoftware.com/
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020908/55d909b3/attachment.html>

From loewis at informatik.hu-berlin.de  Wed Sep 18 05:49:57 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Sep 2002 11:49:57 +0200
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <mailman.1032329366.6813.python-list@python.org>
Message-ID: <j4ptvbaapm.fsf@informatik.hu-berlin.de>

Oren Tirosh <oren-py-l at hishome.net> writes:

> They may be dangerous in some circumstances. IIRC, if a C extension
> calls the Python interpreter recursively with pointers to things on
> its stack it can cause memory corruption.

Depends on your notion of "danger"; I usually associate it with fear
and uncertainty :-)

If you know it can crash when you do certain things, and you document
that limitation, there is no danger. There are already ways to crash
Python, the dlmodule being an example in the core distribution.

Also, readers should notice that your example requires Stackless to be
used: if no stackless module is imported, then accessing variables on
the stack won't cause memory corruption.

Regards,
Martin


From aleax at aleax.it  Tue Sep 17 02:40:09 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 17 Sep 2002 06:40:09 GMT
Subject: threads or queue for this task
References: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com> <mailman.1032202279.8360.python-list@python.org> <Qgqh9.117134$pX1.4203010@news2.tin.it> <mailman.1032241763.6933.python-list@python.org>
Message-ID: <d7Ah9.103593$ub2.2156613@news1.tin.it>

James J. Besemer wrote:
        ...
>>>Generally, Queues never get "full" and you generally don't care if
>>>they're empty.
>>
>>You can define Queues that can hold a maximum number of items; [...]

And I proceed to give a very detailed and rather realistic example.
How come you don't address it at all?


>>Setting size is just an optional embellishment, though
>>sometimes it does pay dividends.
> 
> Setting size has it's place but otherwise introduces unnecessary
> complications.  

If it's not necessary (or at least helpful) it's unnecessary.  Monsieur
de La Palice would be proud of you.

> Among other things, it forces the queue writer to deal
> with the situation of what to do if the queue is full.  Signal error
> back to its client?   Sleep and try again later?  Silently discard the
> request?  Abort?

Almost invariably, just as I said (and you snipped) and in the
example I have, the reason you set a Queue to hold a maximum
number of items is to have threads that post to it implicitly
wait if it's full when they try to post.  The situation is
strongly parallel to dealing with an _empty_ queue -- normally,
you just get items, and implicitly wait if it's empty when
you try to get.


> "Generally," NOT having a limit is by far the more elegant solution.

I dispute the generality you assert, and, again, I gave a
specific example: where the time to produce work-requests
(posted to a queue) is always far less than the time to
consume/process a work-request taken from the queue.  This
is hardly a RARE situation: on the contrary, it's quite a
normal state of affairs that processing a work-request is
much more time-consuming than producing it.

Therefore, I consider your assertion "Generally, Queues never
get *full*" a dangerous over-simplification: they get full
if you specifically define them to have a maximum size (and
post more items than you peel off), and it's not a rare case
to want that ("an optional embellishment", as I said, but one
that does happen quite normally in general cases).


> Even in the general case, the queue may eventually grow to consume all
> available memory if one doesn't have the CPU power to keep up with
> requests.  But, like most Python errors, this eventually will raise a
> specific exception.

If the number of work requests is not unbounded, you may be
thrashing your computer into disgraceful performance rather
than having your program fail with an exception -- wasting
unbounded amounts of memory on an unbounded queue, where a
single, simple call to make it bounded would solve it all
like magic.  It's not a matter how how much CPU power you
have: strictly a ratio between how much time it takes to
produce a work-request, and how much time it takes to process
it.  If that ratio, which may not depend on CPU power at all,
is low enough (it was 1:10 in my example, certainly anything
but an extreme case), using a queue of unbounded size will
degrade performance horrendously (thrashing can slow your
whole machine down almost unlimitedly, if you have virtual
memory larger enough than physical memory) if there are enough
work-requests.


> Anyway, I felt the "general" case was more applicable to this novice's
> line of questioning.  He/she already was confounded by some
> non-essential details.

We disagree on didactics.  I think over-simplification is a
bad way to teach: everything must be kept as simple as possible,
*but no simpler than that*.  The parallels between a queue
becoming empty, and one becoming full (when set to have a max
size), are very strong, and may indeed HELP a bright novice
grasp things better than if one tried to sweep half of the
issue under the carpet.


Alex



From loewis at informatik.hu-berlin.de  Mon Sep 23 05:45:38 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 23 Sep 2002 11:45:38 +0200
Subject: PyXML Building on OS X?
References: <B9B3CED8.3B9%jwold@min.midco.net>
Message-ID: <j4k7ldghtp.fsf@informatik.hu-berlin.de>

Jarrett Wold <jwold at min.midco.net> writes:

>   If you want some sample output of the build I can certainly rerun
> it and post it.

Please do. No need to report the entire output, though - just the
error messages and a little context is needed, such as the compiler
line that caused the error message.

Regards,
Martin



From mak at trisoft.com.pl  Thu Sep 12 07:47:24 2002
From: mak at trisoft.com.pl (Grzegorz Makarewicz)
Date: Thu, 12 Sep 2002 13:47:24 +0200
Subject: win ce install
References: <newscache$bonb2h$5w2$1@lnews.actcom.co.il>
Message-ID: <3D807ECC.9000009@trisoft.com.pl>

Ell wrote:
> wince install
> 1.
> how I install python on wince ?

Just unpack downloaded file with subdirectories and You will see where 
they cam from.

> 2.
> has somebody experience , python on wince ?
> are all objects work as on win32 ?

Python win has some troubles with unicode and not all win32* modules are 
ported to wince esp. gui,api.
Depending on device type You may get problems with SH api - wince is not 
pocketpc, Jornada is not Symbol ,...

Better get sources and compile for Your platform.

> how is the performance ?

:) ~ 20-80 % of C speed

mak





From fafhrd at datacom.kz  Fri Sep  6 00:53:14 2002
From: fafhrd at datacom.kz (Nikolay Kim)
Date: 06 Sep 2002 11:53:14 +0700
Subject: tuple problem
Message-ID: <1031287998.6307.44.camel@fafhrd>

hi.

i have some problem.

simple code

>>> t = '\xba\xd0\xd7\xd0\xe5'
>>> t
'\xba\xd0\xd7\xd0\xe5'
>>> print t
?????
>>> print "%s" % t
?????
>>> 

t is string in utf-8 encoding
but after print it iso-8859-5 
it's ok or error in python?





From joel.quinet at be.unisys.com  Thu Sep 26 04:06:16 2002
From: joel.quinet at be.unisys.com (Quinet, Joel)
Date: Thu, 26 Sep 2002 03:06:16 -0500
Subject: Getting last char in string.
Message-ID: <8D7B1A876CBBD2119E6A00105AC57EF005A38E72@BE-BRU-EXCH-1>

Here under an example:

s = 'abcd'
print s[-1]
d
print s[-2]
c

Joel
-----Original Message-----
From: CheapSkate [mailto:gua81 at XXXyahoo.com]
Sent: jeudi 26 septembre 2002 9:49
To: python-list at python.org
Subject: Getting last char in string.


Hi, im writing a simple method that takes in a string input and check if the
last charcter is an '/'
then it will do A else do B

my Q now is how to get the last char???

Thanks


-- 
http://mail.python.org/mailman/listinfo/python-list



From tdelaney at avaya.com  Tue Sep  3 20:33:34 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 4 Sep 2002 10:33:34 +1000 
Subject: Pyrex 0.4.4
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A567@natasha.auslabs.avaya.com>

> From: pinard at iro.umontreal.ca [mailto:pinard at iro.umontreal.ca]
> 
> I noticed too.  To get going, I got Plex installed 
> separately, and removed the
> 'Plex' line in `setup.py' for Pyrex.

I decided against that, since one of the ficed bugs was in Plex.

Tim Delaney



From sgclift at earthlink.net  Mon Sep  2 17:23:35 2002
From: sgclift at earthlink.net (Steve Clift)
Date: Mon, 02 Sep 2002 14:23:35 -0700
Subject: urllib bug?
Message-ID: <3.0.6.32.20020902142335.007d7bc0@mail.earthlink.net>

I'm running Python 2.2.1 on a venerable W98 box. Attempts to use
urllib.urlopen() to open trivial URLs fail with an exception 'no host given'.

>From tracing through urllib:

urllib.urlopen() instantiates an instance of FancyURLopener, which
subclasses URLopener().

To figure out whether I'm using a proxy, urllib.URLopener.__init__() calls
getproxies(), which is an alias for getproxies_registry() under Windows.

getproxies_registry() ferrets around in the registry, using the _winreg
module to extract values for ProxyEnable and ProxyServer.

On this box, regedit lists the value for ProxyEnable as 00 00 00 00, and
the value for ProxyServer as "". _winreg.QueryValueEx() returns
('\x00\x00\x00\x00', 3) for ProxyEnable and (u'', 1) for ProxyServer.
Clearly, I'm not using a proxy. 

urllib tests the truth of '\x00\x00\x00\x00' and, because it's a
non-zero-length string, incorrectly decides I *am* using a proxy and
constructs bogus http and ftp proxy entries that include no host name.

The absence of a proxy host name causes urllib to grumble.

So,

a) It's a bug.
b) How come no-one else is complaining about it? What's special about this
box?

-Steve



From deblnononospammmmmyyy at theworld.com  Thu Sep 19 00:05:38 2002
From: deblnononospammmmmyyy at theworld.com (David Lees)
Date: Thu, 19 Sep 2002 04:05:38 GMT
Subject: Thanks guys
Message-ID: <3D894D13.6029344C@theworld.com>

OK, I see the PCbuild directory and the readme.txt.

Thanks everyone for the pointers.

David Lees


From loewis at informatik.hu-berlin.de  Fri Sep 20 16:11:19 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 20 Sep 2002 22:11:19 +0200
Subject: httplib 'getaddrinfo failed'
References: <mailman.1032544106.11629.python-list@python.org>
Message-ID: <j4vg50a0bc.fsf@informatik.hu-berlin.de>

"Gordon Williams" <g_will at cyberus.ca> writes:

> Same problem.  Any ideas?  I'm using Sympatico high speed edition which was
> developed by Efficient Networks Inc. - if that is any help.

Not really. More interestingly: What host are you trying to pass to
the httplib? When you are trying to access that host, is your DSL
connection up? Does a regular name lookup for this host (e.g. via
tracert) work? etc.

Regards,
Martin



From ehab_teima at hotmail.com  Thu Sep 19 12:08:37 2002
From: ehab_teima at hotmail.com (Ehab Teima)
Date: 19 Sep 2002 09:08:37 -0700
Subject: Database Modules to Connect to MSSQL
Message-ID: <17aafe08.0209190808.3be7c11@posting.google.com>

Hello,

I am trying to use DB modules in Python to connect to MSSQL server. I
found different modules but I could not find one for MSSQL. Anyone
know which module I can use without having troubles.

Thanks,
Ehab


From dgallion1 at yahoo.com  Fri Sep 13 08:39:21 2002
From: dgallion1 at yahoo.com (Darrell)
Date: 13 Sep 2002 05:39:21 -0700
Subject: Piper Pickling?
References: <dg9g9.2373$Lg2.427869@news2.news.adelphia.net>
Message-ID: <b0c5da3a.0209130439.5963053@posting.google.com>

Here's one approach I've used.
Create an interface class for each class you plan to store.
Also create the schema.

--Darrell


class A:
    """
    Some data class
    """
    _X1=9.2
    _Y1="text"

    def __init__(self, *arg, **kw):
        self._x=1
        self._y="hello"
        self._z=arg[0]
        self._range=zip(range(10),range(10,20))

    def show(self):
        print '*'*10,self.__class__.__name__,'*'*10
        for k, v in self.__class__.__dict__.items():
            if not callable(v):
                print "%s: %s"%(k,v)
        for k, v in self.__dict__.items():
            if not callable(v):
                print "%s: %s"%(k,v)

class DbIforA(DbI):
    def __init__(self, objClass, schema, dbName=None, tableName=None):
        DbI.__init__(self, objClass, schema, dbName, tableName)

    def insert(self, *v, **kw):
        excludes= kw.setdefault("excludeCols",{})
        excludes["_dbLocalTimeStamp"]=0
        excludes['_dbLocalSerialNum']=0
        apply(DbI.insert, (self,)+v,kw)

    def getValues(self):
        res=DbI.getValues(self)
        return res

schema={
    '_x':{'default':55, 'type':'integer'},
    '_y':{'default':65, 'type':'varchar(30)'},
    '_z':{'default':75, 'type':'integer NOT NULL',
'constraints':'Primary key '},
    '_X1':{'default':None, 'type':'double'},
    '_dbLocalTimeStamp':{'default':None, 'type':'TIMESTAMP(10)'},
    '_dbLocalSerialNum':{'default':None, 'type':'integer not null
auto_increment', 'constraints':'key'},
    '_range':{'type':'pickle'}
}


dbI=DbIforA(A, schema, 'testDb', 'testTable')
a1=A(1, x=3, y=4)
a2=A(22, x=32, y=42)
a3=A(33, x=33, y=43)

# Some simple test cases
dbI.dropDataBase(IF_EXISTS="IF EXISTS")
dbI.createDb(IF_NOT_EXISTS="IF NOT EXISTS")
dbI.use()
dbI.createTable(IF_NOT_EXISTS="IF NOT EXISTS")
dbI.insert(objs=[a1, a2, a3])
dbI.use()
dbI.select()
dbI.select(WHERE="where _X1 < 10.0 and _X1 > 9.0")
ret=dbI.getValues()
for i in ret:
    i.show()
    i._y='Dog'
dbI.update(objs=ret)
dbI.select()





"Robert Oschler" <Oschler at earthlink.net> wrote in message news:<dg9g9.2373$Lg2.427869 at news2.news.adelphia.net>...
> Forgive me :)
> 
> If Peter Piper pickles a peck of Python packages, can he pack the pickled
> Python peckers in a proper packing place?
> 
> Translated: Anybody seen a module that can override the standard pickling
> process and store and retrieve Python objects to a MySQL database?
> 
> thx
> 
> P.S. - Whell hell the language is named after the world's best comedy
> troupe, what did you expect, green eggs and spam? :)


From steve at ferg.org  Thu Sep 12 21:45:27 2002
From: steve at ferg.org (Stephen Ferg)
Date: 12 Sep 2002 18:45:27 -0700
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <ac677656.0209111225.4a141bc5@posting.google.com> <1a3a9ffa.0209120856.1de06f57@posting.google.com>
Message-ID: <b16e4ef7.0209121745.44e91aa@posting.google.com>

aicolburn at yahoo.com (Alan C.) wrote in message news:<1a3a9ffa.0209120856.1de06f57 at posting.google.com>...

> I enjoyed this tutorial. 
Thanks!

> perhaps include a brief explanation early on about why you're
> creating the program within a class structure.  

I'm embarrassed to say that I was just following a lot of the other
tutorial materials that I have read.  But to be honest, I can't really
say WHY using a class structure is desirable.  I have a vague
intuition that the program will be more maintainable that way, but I
can't clearly articulate the reasons.

But you're absolutely right -- this is just the kind of issue I wanted
to cover in "Thinking in Tkinter".  Now I'll have to try to come up
with an answer!
 
> I think it
> would be educational to distinguish between, say,
> "self.button1Click()" and "self.button1Click" I don't know for sure,
> but I bet there are folks who get confused distinguishing between
> them.

I think you're right.  A good idea.


From erikprice at mac.com  Thu Sep 12 00:22:23 2002
From: erikprice at mac.com (Erik Price)
Date: Thu, 12 Sep 2002 00:22:23 -0400
Subject: Use glob not re for regexp?
In-Reply-To: <DQQf9.1761$Lg2.233877@news2.news.adelphia.net>
Message-ID: <3CA22EAB-C607-11D6-9A79-00039351FE6A@mac.com>

On Wednesday, September 11, 2002, at 08:02  PM, Robert Oschler wrote:

> I saw a post where someone said to use glob not re for regular 
> expressions?
> Is this true and if so why?  I use Secret Labs regexp package.

Filename globbing syntax is completely different from regular 
expression syntax.  In some cases, generally when dealing with 
filenames, it is more familiar to some people (particularly Unix folk 
or shell programmers).  But in general regular expressions provide far 
more control and precision over your patterns and matches.

So, use whichever one you want.



Erik





--
Erik Price                                                  (zombies 
roam)

email: erikprice at mac.com
jabber: erikprice at jabber.org




From tim.one at comcast.net  Thu Sep  5 13:36:47 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 05 Sep 2002 13:36:47 -0400
Subject: Installing Python 2.2.1 on WinNT
In-Reply-To: <75848aa7.0209050830.4024ebef@posting.google.com>
Message-ID: <BIEJKCLHCIOIHAGOKOLHOEELDKAA.tim.one@comcast.net>

[Phil Rittenhouse]
> Thanks for the feedback Armin, that tells me there must be something
> about my install of NT that's not quite right (or at least, not quite
> the same as everyone else).
>
> I assume that the download file is OK because it installed fine
> on 95, 98, ME, 2K, and XP without problems.
>
> In fact, I have just found that if I install Office 2000 first, the
> python install on my NT works cleanly.   There must be some magic that
> the Office install does, maybe a DLL update, or permission change, that
> allows it to work.

Delighted to hear it finally worked!  I'm afraid we're not much help with
these when they go wrong:  the PLabs installer is 100% vanilla Wise
directives, with no custom code of any kind.  So when something goes wrong,
I'm afraid we're as clueless as the unfortunate user.  Luckily, if you keep
on buying enough new versions of Windows, the problem eventually fixes
itself <wink>.




From sfeil at io.com  Mon Sep 23 21:45:29 2002
From: sfeil at io.com (Steven Feil)
Date: Tue, 24 Sep 2002 01:45:29 GMT
Subject: Three dumb questions (ordered by dumbness descending)
References: <amnvit$7kqdt$1@ID-77524.news.dfncis.de>
Message-ID: <pan.2002.09.23.20.43.59.461504.1216@io.com>

On Mon, 23 Sep 2002 16:03:21 -0500, Thorsten Kampe wrote:

> Okay, here they are:

Ok, I'll take a stab at the first two questions. I do think that a
little more clarity on your part could make it a bit easer.

The first thing to remember is that Python is a dynamically typed
language. It is typed and each object has a type, It would be a little
easer if you would disclose the type for each object you are referring
to.


> 1. Why is 'zip(zip(x)) != x' (should be the same because it's the
> transposed)

Humm, what is is zip(), a function in the standard Python library or
is it something you or someone else made up? I did a quick search in
the Python docs and the only zip I found referred to zip
compression. If this is something from the standard Python libraries
please state the library it comes from.

If zip is some general function, then in general, zip(zip(x)) is not
equal to x. This comes from fundamental abstract algebra. In trig
sin(sin(x)) is not x. There do exist functions who are there own
inverse and if zip is one of them then zip(zip(x)) is equal to x.

You mentioned transpose. If x is a matrix and the function zip only
transposes x, then zip(zip(x)) is equal to x. This is because a
transpose function is it's own inverse. I'm not sure where you got
this zip function, but zip is suppose to transpose x and zip(zip(x))
!= x then there is a bug in zip. Also if zip is a transposition
function it can only act on arrays and it should raise an exception if
x is not of the required type.

I will need to know more about where you got zip from.


> 2. Peter Norvig mentions in "Python for Lisp Programmers" some
> "don'ts": "[x] + y" and "x[1:]". Are there more things to avoid

humm, "[x] + y", again it comes down to type. [x] will always be a
list. It will contain a single element x of whatever type x is. y on
the other hand is whatever type y is (you haven't told me). In Python
when I use "[x] + y" y must be another list, because when used with a
list the + operated must have a second list to concatenate to the
first. It is the way + is defined. If you wish to appended the list
[x] with y you must have a reference to [x] such as z=[x] then append
to it with z.append(y). A quicker way would be to start with [x,y]
which would be equal to the z after you appended to it.

z=[x]
z.append(y)
z == [x,y]

The statement "x[1:]" this is legal if x is a list, or if x is some
type in which index slicing has been defined. basically x[1:] means
something that has every element that x has except for the first
element. So if x = "steve" then x[1"] is "teve".

========================================================================
 Steven Feil               | Gram-pa, back at the turn of the      .~.  
 Programmer/Developer      | century, why did people use an        /V\  
 sfeil at io.com              | operating system, when they were not // \\ 
                           | allowed to see the source code?      (X_X) 
========================================================================


From daniel.dittmar at sap.com  Mon Sep 30 07:29:11 2002
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Mon, 30 Sep 2002 13:29:11 +0200
Subject: is cPickle really this slow?
References: <slrnapfs9u.8re.dag@omally.animagicnet.no>
Message-ID: <an9ci8$s68$1@news1.wdf.sap-ag.de>

Dag wrote:
> I have a script which starts by parsing a large file and turning into
> a data stucture (a dictonary of lists) which I then use in the script.
> This operation however takes about 12-15 seconds.  Since the file
> hardly ever changes , and if it changes it's because I changed it,
> thought I'd save some time by pickeling the datastructure and and
> simply getting that at the start of each script.  So I pickle my
> dictionary, which ends up being a 3.2MB file and try to run my script
> again, this time getting the pickled dictonary instead of building it
> at runtime.  This time however the script takes over 35 seconds to
> run, close to three times as long.  Am I doing something wrong is
> loading and parsing larger pickled files really very slow.

Have you used the binary option to 'dump'?
dump(object, file[, bin]) : If the optional bin argument is true, the binary
pickle format is used; otherwise the (less efficient) text pickle format is
used (for backwards compatibility, this is the default).

Daniel




From kkuroda at crl.ucsd.edu  Mon Sep  2 22:41:00 2002
From: kkuroda at crl.ucsd.edu (Kow Kuroda)
Date: Mon, 2 Sep 2002 19:41:00 -0700
Subject: RedHat 7.3 and Python2.2.1
In-Reply-To: <3d74176d@news.sentex.net>
Message-ID: <95469BCC-BEE6-11D6-B4E4-00050287C0A6@crl.ucsd.edu>

On Monday, September 2, 2002, at 06:59 , Peter Hansen wrote:

> Kow Kuroda wrote:
>> On Monday, September 2, 2002, at 11:43 , Peter Hansen wrote:
>>> Doug Farrell wrote:
>>>> I'm trying to install the latest version of Python
>>>> (python2-2.2.1-1.i386.rpm) from the www.python.org download page, and
>>>> I'm having problems. When I run
>>>>
>>>> rpm -i python2-2.2.1-1.i386.rpm
>>>>
>>>> I get all kinds of conflict messages like this:
>>>>
>>>> file /usr/bin/pydoc2 from install of python2-2.2.1-1 conflicts with
>>>> file from package python2-2.2-16
>>>
>>> Sounds like you already have a previous version of Python2.2 installed.
>>> You should probably be using the "-U" option to update instead of
>>> the "-i" option to install... or uninstall the previous one first.
>> I just wanted to make a comment on Peter's post.
>> I think -U or --upgrade option works, but it's dangerious to uninstall 
>> Python 2.1, which is now part of Red Hat Linux 7.3, unlike in prevous 
>> releases where Python 1.5.2 or something is used for installation scripts.
>
> (Note he had a conflict with Python 2.2, not 2.1.  It should work fine.)

I actuall *did* have a big problem with 2.1 when I tried to install 2.2.1.
  I don't really remember all the error messages I got then.
> Your comments about RH7.3 don't quite fit with my experience.  I just
> installed it (posting from the machine it's on, in fact) and "python"
> executes 1.5.2 still, while "python2" executes 2.2, not 2.1.  What am
> I missing?

Hmmmm, that's odd ... But unfortunatley, I cannot be very precise about 
what I exlained partly because I'm back to RHL 7.2 now. I had some 
embarassing experiences with RHL 7.3 including the problem I described in 
my comments.

All I can say right now is just that I nearly gave up the idea of 
installing 2.2.1 on RHL 7.3 because of the overwhelmingly complicated 
situation with dependencies.

Oh yes, I also tried to update "redhat-config-network" to the most recent 
version, (I needed it to add the wireless interface to "neat"). It requires 
a newer version of Alchemist package, writtin in Python 2.x. And Alchemist 
depends on a newer version of PyGnome, which depends on something, etc, etc.

I admit that it's possible that my situation wasn't typical, but believe me,
  it's true that I got stuck in a very complicated situation caused by 
Python 2.2.1 installation on RHL 7.3.

As a last thought, doesn't Red Hat have different versions of the 7.3 
distribution?


Kow




From sandysj at juno.com  Fri Sep 20 19:10:59 2002
From: sandysj at juno.com (Jeff Sandys)
Date: Fri, 20 Sep 2002 23:10:59 GMT
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com>
Message-ID: <3D8BAB03.D65F3B7E@juno.com>

Benjamin wrote:
> hi everybody
...
> i've been learning python not for becoming a programmer
> ..., but for having fun.
...

There was a Python inititive called CP4E, 
Computer Programming for Everyone.  I believe that 
with the proliferation of computers in society that 
everyone should know how to write a computer program. 

Examples of some of Python's successes are in 
research organizations where computer scientists wrote 
the programs to do data analysis for the research 
scientist, now the research scientist can write thier 
own programs using Python and the computer scientist 
can work on bigger issues.

Python is also being used in some schools like Logo 
was, to introduce all students to programming.

Good Luck on your journey.

Thanks,
Jeff Sandys


From gregory.p.green at boeing.com  Fri Sep 20 12:54:52 2002
From: gregory.p.green at boeing.com (Greg Green)
Date: Fri, 20 Sep 2002 16:54:52 GMT
Subject: Python COM and variable number of arguments
References: <mailman.1032529407.31552.python-list@python.org>
Message-ID: <sn04lhyb.fsf@boeing.com>

In this case I was using the makepy generated wrapper. I have
M. Hammonds book and used the method of ensuring I was using the
early-bound object.

So am I out of luck even if I have the wrapper?
-- 

Greg Green


From uwe.schmitt at procoders.net  Fri Sep  6 08:40:12 2002
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 6 Sep 2002 12:40:12 GMT
Subject: What database should I use
References: <mailman.1031257921.29400.python-list@python.org>
Message-ID: <ala7nc$l74on$4@hades.rz.uni-sb.de>

A <printers at sendme.cz> wrote:
> Hello,
> I am going to program a small application (in Python and wxPython) which consists about 
> 2000 - 3000 records . Can you please recommend which database is the best for that. I 
> use Windows.
> Maybe MS Access  or FoxPro or MySQL or better?
> I want to keep the program as small as possible.
> Thanks
> Ladislav

If your search tasks just rely on one index, the berkley db bindings
can be recommended. It is very fast, leightweight and runs on many platforms. 
Look at anydbm, dbhash or bsdd in the Python module documentation.

Greetings, Uwe



-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           


From salmonia.nospam.please at cardiff.ac.uk  Wed Sep 18 08:07:29 2002
From: salmonia.nospam.please at cardiff.ac.uk (Alan James Salmoni)
Date: Wed, 18 Sep 2002 13:07:29 +0100
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com>
Message-ID: <am9qa6$hsf$1@news.swman.net.uk>

Lee,

I don't know if this is the kind of thing you are looking for, but I've 
written an application for statistical analysis (the scientific kind) 
which is available for download from http://salstat.sunsite.dk 
(screenshots are available too). It's GPL'd, but rather new so there are 
still some rough edges. The GUI was written using wxPython (absolutely 
fantastic IMHO - the learning curve is a bit steep to begin with, but 
shallows extremely quickly). Features:

* A spreadsheet like grid for data entry/display
* A html display window for the output of the analysis and the online 
help pages
* Custom dialog boxes for each set of analyses
* Standard dialogs (file open/save and font choser)
* The GUI took a few weeks to write part-time and that was from scratch 
while learning wxPython. And I'm a psychologist, not a programmer, which 
means you will probably be up to speed in much less time ;).
* A windows installer containing all the required program (ie, non-icon 
or help) files for a Windows box in one file (using the McMillan installer)
* Cross-platform capability - Windows, Linux and OSX (the OSX wasn't 
even planned - it just happened to work!)

Have fun!

Alan.

Lee Gray wrote:
> I'm a fairly new programmer with experience in ASP/VBScript/DHTML.
> Other than small utilities, so far I've mostly written ASP apps to
> talk to an Oracle database on HP/UX, with Windows NT/2000 clients, and
> I've done maintenance work on VB and Oracle Forms apps.
> 
> We're currently looking at .Net, and I'm not very impressed, for a
> variety of reasons. A little research makes me feel that either Java
> or Python would be a good alternative, and given my level of
> experience, I'd much prefer Python over Java.  I *think* I could
> actually get up to speed faster with Python than with VB and ASP.Net,
> in spite of my ASP/VBS experience.
> 
> My boss is pretty open-minded, but also needs to see something working
> to be convinced Python is even a viable platform (he had never heard
> of it).  Otherwise, .Net is a given, whether it's any good or not,
> since corporate is MS all the way.
> 
> Can someone point me to a good Python GUI app to demonstrate (other
> than IDLE)?
> 
> It may very well be that .Net is best for me/us, but from what I've
> seen of Python, I *really* like the language and I'd at least like to
> see a good demonstration of a full-blown app before ruling it out.
> (Something along the lines of what I mentioned above would be ideal,
> but not a necessity.)
> 
> Thanks!
> 
> Lee Gray



From trash at terastat.com  Tue Sep 17 18:47:52 2002
From: trash at terastat.com (bap)
Date: Tue, 17 Sep 2002 22:47:52 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com>
Message-ID: <siOh9.451392$me6.54442@sccrnsc01>

Actually both may be a solution. I've recently had good success at
developing an with  a VB (not .net) front end and doing most of the
work in Python. Use each for what it's best at. Based on my experience with
project, I'd say that Python take about 25% of the code of VB to accomplish
the same objective. Plus Python objects tend to do what is expected so that
you can often guess at the syntax and be right whereas this is seldom true
with
VB objects.

Lee Gray <lpgray at uop.com> wrote in message
news:96904b50.0209171305.6d65f574 at posting.google.com...
> I'm a fairly new programmer with experience in ASP/VBScript/DHTML.
> Other than small utilities, so far I've mostly written ASP apps to
> talk to an Oracle database on HP/UX, with Windows NT/2000 clients, and
> I've done maintenance work on VB and Oracle Forms apps.
>
> We're currently looking at .Net, and I'm not very impressed, for a
> variety of reasons. A little research makes me feel that either Java
> or Python would be a good alternative, and given my level of
> experience, I'd much prefer Python over Java.  I *think* I could
> actually get up to speed faster with Python than with VB and ASP.Net,
> in spite of my ASP/VBS experience.
>
> My boss is pretty open-minded, but also needs to see something working
> to be convinced Python is even a viable platform (he had never heard
> of it).  Otherwise, .Net is a given, whether it's any good or not,
> since corporate is MS all the way.
>
> Can someone point me to a good Python GUI app to demonstrate (other
> than IDLE)?
>
> It may very well be that .Net is best for me/us, but from what I've
> seen of Python, I *really* like the language and I'd at least like to
> see a good demonstration of a full-blown app before ruling it out.
> (Something along the lines of what I mentioned above would be ideal,
> but not a necessity.)
>
> Thanks!
>
> Lee Gray




From news at nospam.eml.cc  Tue Sep 24 18:06:10 2002
From: news at nospam.eml.cc (Mahesh Padmanabhan)
Date: Tue, 24 Sep 2002 22:06:10 GMT
Subject: List/Tuple bug or feature?
Message-ID: <8765wv6o0d.fsf@nospam.eml.cc>

Hi,

I spent a lot of frustrating hours tracking this down and I am
confused as to why I see this behavior.

>>> x = ('a' 'b')
>>> x
'ab'

>>> x = ['a' 'b']
>>> x
['ab']

As you might have guessed, I missed a comma between two elements in
a fairly large list.

If this is a concatenation feature, how can I prevent something like
this happening in the future ?

This is with Python 2.1.3 on Debian Linux.

Python 2.1.3 (#1, Sep  7 2002, 15:29:56) 
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "copyright", "credits" or "license" for more information.

TIA,

-- 
Mahesh Padmanabhan

-------

Replace nospam with prana to email me.


From spoermeg at voldelig.com  Fri Sep 20 16:12:14 2002
From: spoermeg at voldelig.com (Terje Johan Abrahamsen)
Date: Fri, 20 Sep 2002 15:12:14 -0500
Subject: Python Compiling
Message-ID: <amfveg$5g64k$1@ID-53055.news.dfncis.de>

First of all, I am going to warn you. I will ask some stupid questions. I
haven't programmed too much, and don't really know how it all links
together.

But, I guess my question boils down to, why is not Python a compiled
language?

To my best knowledge, the computer understand 0s and 1s. So, any programming
language must be brought down to that. Even assembler must be somewhat
remade into computerlanguage. A compiler for for example C, will take my C
code, and transform it into computerlanguage. The same does the intepreter
for Python. However, the interpreter does that step every time I run the
program, versus the compiler does it once, and gets it over with. So, for
the computer, the computerlanguage that it runs doesn't really differ
between if I wrote my code in Python and run it thru the intepreter, wrote
it in C and compiled it, or wrote assambler and did whatever you do with
assambler. So, when Guido made the first Python, he deceided he wanted an
intepreted language, and wrote an intepreter. Could someone write a compiler
and get Python to be a compiled language like C? I guess Py2Exe does to some
extent that, but to my understanding it is not exactly the same as a
complete compiler. It more just copies the files from the intepreter that
the program needs.

Is it not possible to write a complete compiler like the C compiler for
Python? Or is it just that nobody has done it? Or must the language be
constructed differently?

Thanks for any help to clearify the subject. But please try to keep it
understandable for also one with somewhat more limited understanding of the
subject....




From see_reply_address at something.invalid  Tue Sep 10 02:04:22 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Tue, 10 Sep 2002 18:04:22 +1200
Subject: random.shuffle question
References: <3k4f9.7500$yt3.3831288@newssrv26.news.prodigy.com> <3D7CE713.93BC9FAF@alcyone.com>
Message-ID: <3D7D8B66.6070903@something.invalid>

Erik Max Francis wrote:

> You could simply chop up the list, shuffle the part that you want, and
> then reassemble it:


Or for more fun, you could use the handy class
attached below, and simply say:

   random.shuffle(sublist(mylist, 4, 8))

#-------------------------------------------

class sublist:
   """sublist(x, i, j) is an object which
   behaves like x[i:j] except that changes
   to it are reflected in x. (This version
   only handles getting and setting single
   items.)"""

   def __init__(self, base, begin, end):
     self.base = base
     self.begin = begin
     self.end = end

   def __len__(self):
     return self.end - self.begin

   def __getitem__(self, i):
     return self.base[self.begin + i]

   def __setitem__(self, i, x):
     self.base[self.begin + i] = x

if __name__ == "__main__":
   import random
   l = range(20)
   print "Before:", l
   random.shuffle(sublist(l, 5, 15))
   print "After:", l

#-------------------------------------------
-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From see_reply_address at something.invalid  Wed Sep 11 22:30:27 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Thu, 12 Sep 2002 14:30:27 +1200
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid> <alln71$1s2s$1@nntp6.u.washington.edu> <b16e4ef7.0209110616.48dfe04c@posting.google.com>
Message-ID: <3D7FFC43.10002@something.invalid>

Stephen Ferg wrote:

> So it looks like the
> correct mouse-button binding to replace "command" would be to bind to 
> <ButtonRelease-1>.


That's not quite right either, because the action
should only be triggered if the mouse is released
inside the button.

> Is there any place where definitions or explanations
> or discussions of standard button behavior are written down
> and available? A book? A Web page?


Apple publishes a Human Interface Guidelines document,
which ought to be on their website somewhere. There's
probably something similar from Microsoft.

The generally accepted behaviour for a button on
most platforms nowadays is:

(1) Mouse-down in the button causes it to be highlighted
     in some way (changing colour, becoming "depressed", etc.)

(2) As long as the mouse is held down, the button is
     unhighlighted when the mouse moves out of it, and
     re-highlighted when it moves back in.

(3) If the mouse is released inside the button, the
     button's action is performed. If the mouse is
     releases outside the button, nothing happens.

There's no single event binding in Tk that will give
you all of this. There is code built into the Button
widget that implements it, but to take advantage of
it you have to use the "command" option.

It's a fairly universal principle in Tk that the
"command" of a widget is invoked according to the
standard behaviour of that widget, so I think it
deserves treatment as a fundamental principle
on the same footing as event binding.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From kennethg at pd.jaring.my  Fri Sep 27 11:44:49 2002
From: kennethg at pd.jaring.my (Kenneth Gomez)
Date: Fri, 27 Sep 2002 15:44:49 GMT
Subject: Local Python Server Installation
References: <3d8b47c6.277669@news.jaring.my> <slrnaomn6m.39s.gerhard.haering@haering.opus-gmbh.net> <3d8c993c.2190426@news.jaring.my> <mailman.1032625889.13783.python-list@python.org> <3d8d4009.375344@news.jaring.my> <bQYk9.1630$6g7.1090@fe03>
Message-ID: <3d947c26.2656654@news.jaring.my>

Steve,

I stumbled onto this problem while going through the first exmaple in
the book Python:How to Progra, by Deitel, Deitel, Liperi and
Wiedermann.

My little code sniipet, copied exactly from the book except for the
path to my python exe is :

________________________________________________

#!C:\prog\Python22\python.exe

import time

def printheader(title):
	print """Content-type: text/html

	<?xml version = "1.0" encoding = "UTF-8"?>
	<!DOCTYPE html PUBLIC
		"-//W3C//DTD XHTML 1.0 Strict//EN"
		"DTD/xhtml1-strict.dtd">
	<html xmlns = "http://www.w3.org/1999/xhtml">
	<head><title>%s</title></head>

	<body>""" % title

printheader("Current Date & Time")
print time.ctime(time.time())
print "</body></html>"
___________________________________________________


I get the following error in IE6 :

The XML page cannot be displayed 
Cannot view XML input using style sheet. Please correct the error and
then click the Refresh button, or try again later. 


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

Invalid at the top level of the document. Error processing resource
'http://localhost/ken.py'. Line 1, Position 1 

#!C:\prog\Python22\python.exe
^
 


and when using Mozilla 1.0, all I get is the listing of the script.

Bummer.

Any advise?

Thank you.

Regards,
Kenneth.


On Fri, 27 Sep 2002 09:16:41 -0400, "Steve Holden"
<sholden at holdenweb.com> wrote:

>"Kenneth Gomez" <kennethg at pd.jaring.my> wrote in message
>news:3d8d4009.375344 at news.jaring.my...
>> Chris, you're right. It's an Apache config problem. I'll dig into the
>> doc's.
>
>
>The Wrox book "Professional Apache 2.0" is helpful if you need to understand
>how Apache works and, despite its title, it alos covers older versions of
>Apache.
>
>If you want to write servers in pure Python you could look at "Python Web
>Programming", but you will probably find you can get a lot further than you
>thought with simple CGI, then you'll know enough about other directions
>you'd like to go.
>
>regards
>-----------------------------------------------------------------------
>Steve Holden                                  http://www.holdenweb.com/
>Python Web Programming                 http://pydish.holdenweb.com/pwp/
>Previous .sig file retired to                    www.homeforoldsigs.com
>-----------------------------------------------------------------------
>
>
>

Cheers,
Kenneth Gomez.


From iqbal at colorado.edu  Mon Sep  9 20:41:56 2002
From: iqbal at colorado.edu (Naveed Iqbal)
Date: 9 Sep 2002 17:41:56 -0700
Subject: How do i print to a printer using python?
References: <6afb9d2c.0209041044.43c5f5f1@posting.google.com> <al73gj$lbt$1@wanadoo.fr>
Message-ID: <6afb9d2c.0209091641.1d9b652e@posting.google.com>

Thank you all for your help. I am really a novice when it comes to
python. Will someone tell me how to install the win32ui module in
windows?
thanka
naveed


Eric Brunel <eric.brunel at pragmadev.com> wrote in message news:<al73gj$lbt$1 at wanadoo.fr>...
> Naveed Iqbal wrote:
> > I have to print to a standard printer. I take the data from the
> > keyboard and print it using the printer. Please help me....my job
> > depends on this!!!!
> 
> Things are quite different on Unix and Windows. Just a sample (untested) 
> program to print "Hello world!" to the default printer on both platforms:
> 
> ----Unix----------------------
> import os
> p = os.popen('lpr', 'w')
> p.write("Hello world!\n")
> p.close()
> ------------------------------
> 
> ----Windows-------------------
> import win32ui
> 
> dc = win32ui.CreateDC()
> dc.CreatePrinterDC()
> dc.SetMapMode(4)        # This is UI_MM_LOENGLISH
> ## With this map mode, 12 points is 12*100/72 units = 16
> font = win32ui.CreateFont({'name' : 'Arial', 'height' : 16})
> dc.SelectObject(font)
> 
> dc.StartDoc("HelloWorld")
> dc.StartPage()
> dc.TextOut(40, 40, "Hello world!")
> dc.EndPage()
> dc.EndDoc()
> ------------------------------
> 
> And now you wish you were on Unix ;-)!
> 
> For more information on the Windows version, see the documenattion for 
> win32ui @ 
> http://aspn.activestate.com//ASPN/Python/Reference/Products/ActivePython/PythonWin/modules.html 
> and Micro$oft MFC documenation @ 
> http://msdn.microsoft.com/library/default.asp
> 
> Good luck!
> 
> PS: I'm actually a little biased towards Unix... In fact, if you have the 
> printer directly connected to your PC, opening LPT1: and just writing the 
> text to it should work. But if it's a network printer, the only solution is 
> the code above...
> 
> We also can make things a little more complicated on Unix by using 
> PostScript. But it's still shorter:
> 
> import os
> p = os.popen('lpr', 'w')
> p.write('%!PS\n')
> p.write('/Helvetica findfont 12 scalefont setfont\n')
> p.write('/cm { 72 mul 2.54 div } def\n')
> p.write('1 cm 28 cm moveto\n')
> p.write('(Hello world!) show\n')
> p.write('showpage\n')
> p.close()


From darksyyyde at earthlink.net  Fri Sep  6 23:18:33 2002
From: darksyyyde at earthlink.net (dark)
Date: Sat, 07 Sep 2002 03:18:33 GMT
Subject: Thinking of joining
Message-ID: <3D7970DC.5115BCC8@earthlink.net>

Hello.
Im currently still learning C, but a friend has been trying to turn me
on
to python.
>From what ive seen it looks pretty good.
I was just curious if it will do everything PERL will do, as i
dont care for how PERL reads.
( though a lot of proplr i have talked to try to push PERL on me)
Im going to try to discipline myself to get C down first, but its
tempting
to jump into Python now.
I am a monty Python fan of many years, if that helps lol.




From jhalttun at pp.htv.fi  Thu Sep  5 15:41:23 2002
From: jhalttun at pp.htv.fi (janne halttunen)
Date: Thu, 05 Sep 2002 22:41:23 +0300
Subject: How to get os.popen subprocess PID?
References: <3D775167.6060307@pp.htv.fi> <1031248567.61579.0@doris.uk.clara.net> <1031248831.62067.0@doris.uk.clara.net>
Message-ID: <3D77B363.7000802@pp.htv.fi>

Ricardo wrote:
>>Instead of using os.popen() you could import the popen2 module and use the
>>Popen3 or Popen4 classes.
>>
>>Objects of these classes contain a 'pid' attribute.
>>
> 
> 
> Sorry, forgot to mention these classes are available on Unix only ...

Yeah, I'm on linux, so that's no problem.

Thanks,

janne



From bokr at oz.net  Sat Sep 28 11:57:27 2002
From: bokr at oz.net (Bengt Richter)
Date: 28 Sep 2002 15:57:27 GMT
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <amvcev$cf6$1@s1.read.news.oleane.net> <7nHk9.172208$pX1.6117207@news2.tin.it> <ofLk9.304267$z91.13145351@bin3.nnrp.aus1.giganews.com> <dITk9.175519$pX1.6209886@news2.tin.it> <t_3l9.334924$AR1.14880141@bin2.nnrp.aus1.giganews.com> <mailman.1033165663.5015.python-list@python.org>
Message-ID: <an4jh7$7hn$0@216.39.172.122>

On Fri, 27 Sep 2002 15:24:04 -0700, "Mike Rovner" <mike at bindkey.com> wrote:

>
>"Terry Reedy" <tjreedy at udel.edu> wrote in message
>news:t_3l9.334924$AR1.14880141 at bin2.nnrp.aus1.giganews.com...
>> na=1; np=3; nl=1; ng=0
>> print "Found: %d apple%s, %d plum%s, %d lemon%s, and %d grape%s." %\
>> (na, na!=1 and 's' or '', np, np!=1 and 's' or '', nl, nl!=1 and 's'
>> or '', ng, ng!=1 and 's' or '')
>>
>> Found: 1 apple, 3 plums, 1 lemon, and 0 grapes.
>>
>> Alternatives to this 'risky' construct:
>> 1. forgo the nicety of correct English output
>> 2. nest if/else four levels deep to choose correct variant of 16 print
>> statements (ugh)
or just generate the correct variant:
   fmt = "Found: %%d apple%s, %%d plum%s, %%d lemon%s, and %%d grape%s." % \
         tuple([('','s')[x!=1] for x in [na,np,nl,ng]])
   print fmt % (na, np, nl, ng)

or pre-generate all sixteen:
 >>> fmtList=["Found: %%d apple%s, %%d plum%s, %%d lemon%s, and %%d grape%s." % (a,p,l,g)
 ... for a in ('','s') for p in ('','s') for l in ('','s') for g in ('','s') ]

and select the correct one by index:
 >>> fmt = fmtList[(na!=1)*8+(np!=1)*4+(nl!=1)*2+(ng!=1)]
 >>> print fmt % (na, np, nl, ng)
 Found: 1 apple, 3 plums, 1 lemon, and 0 grapes.

>> 3. precalculate 4 suffixes and store in 4 temp vars with 4 if/else's
>> (8 or 16 lines);
>>    though clearly superiour to 2), this is still tedious with some
>> risk of typo error.
Why so many lines? If you really want the temps:
    tmpa, tmpp, tmpl, tmpg = [('','s')[x!=1] for x in [na,np,nl,ng]]
or
    tmpa, tmpp, tmpl, tmpg = map(lambda x:'s'[x==1:], [na,np,nl,ng])
>>
>4. Try:
>use_s=('','s')
>print "Found: %d apple%s, %d plum%s, %d lemon%s, and %d grape%s." %\
>(na, use_s[na!=1], np, use_s[np!=1], nl, use_s[nl!=1], ng, use_s[ng!=1])
>
5.:
 >>> na=1; np=3; nl=1; ng=0
 >>> def s(n): return 's'[n==1:]
 ...
 >>> print "Found: %d apple%s, %d plum%s, %d lemon%s, and %d grape%s." % (
 ...               na,  s(na), np, s(np), nl,  s(nl),     ng, s(ng)
 ... )
 Found: 1 apple, 3 plums, 1 lemon, and 0 grapes.

6.:
 >>> class NQ:
 ...     def __init__(self, name, quant):
 ...         self.name = name
 ...         self.quant = quant
 ...     def __str__(self):
 ...         return '%s %s%s' % (self.quant, self.name, ('', 's')[self.quant!=1])
 ...
 >>> print "Found: %s, %s, %s, and %s." % (
 ...     NQ('apple',na), NQ('plum',np), NQ('lemon',nl), NQ('grape',ng)
 ... )
 Found: 1 apple, 3 plums, 1 lemon, and 0 grapes.

7.:
 >>> print "Found: %d apple%s, %d plum%s, %d lemon%s, and %d grape%s." % \
 ... tuple([[x, x!=1 and 's' or ''][y] for x in [na,np,nl,ng] for y in [0,1]])
 Found: 1 apple, 3 plums, 1 lemon, and 0 grapes.

... ;-)

Regards,
Bengt Richter


From bokr at oz.net  Thu Sep  5 14:36:09 2002
From: bokr at oz.net (Bengt Richter)
Date: 5 Sep 2002 18:36:09 GMT
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com> <j4it1k4ufy.fsf@informatik.hu-berlin.de>
Message-ID: <al886p$lfj$0@216.39.172.122>

On 05 Sep 2002 12:11:45 +0200, loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=) wrote:

>Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:
>
>> >x = 1
>> >
>> >def foo():
>> >  x = 2
>> >  def bar():
>> >    x = 3
>> >    def baz():
>> >      global x
>> >      x = 4
>> >
>> >If you invoke baz(), which of the variables is changed? How do you
>> >denote that you want to change the others?
>> 
>> I think that crawling up the scope hierarchy and stopping at the first x
>> that one finds (in the above case, the one defined in bar) is a sensible
>> solution. 
>
>This solution is not backwards-compatible. global always refers to
>globals(), programs may rely on that.
>
How about introducing an optional "degree of globalness" for the global declaration? E.g.,

    x = 1
    def foo():
      x = 2
      def bar():
        x = 3
        def baz():
          global(1) x
          x = 4

would go outwards 1 nested scope and thus target the x in bar,
rebinding it from 3 to 4.

plain "global x" would be a backwards-compatible spelling of "global(-1) x",
where the implication of the -1 is to index the last in a list of
scopes starting with global(0) for local and ending with the outermost as
the last in the list, at global(-1). Since all the variants would be syntax
errors now, there would be no old-code breakage.

It might be interesting to let the globals() function take an optional argument
corresponding to the above relative scope parameter. Thus globals(0) would return
the same as locals(), and globals() the same as globals(-1), but in the context of
baz above, globals(1) would return {'x': 3}.

Regards,
Bengt Richter


From larry.tami at gte.net  Sun Sep 29 17:46:58 2002
From: larry.tami at gte.net (Larry Wright)
Date: Sun, 29 Sep 2002 21:46:58 GMT
Subject: Python PAM module
Message-ID: <mxKl9.6230$1F3.3753@nwrddc02.gnilink.net>

Hello, I am wondering if anyone can point me to the location of the Python 
PAM module. My google searches turn up references to 
www/ftp.pangalactic.org, but this site seems down. 

As a side note, I am wanting to use this to do authentication (for Medusa 
FTP) on a system using shadow passwords. Has anyone tried this, and if so, 
do you have any advice?


Thanks in advance.


From bkc at Murkworks.com  Wed Sep 25 18:33:17 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Wed, 25 Sep 2002 18:33:17 -0400
Subject: a python puzzle
References: <mailman.1032989974.25321.python-list@python.org>
Message-ID: <3d923704$1_6@goliath.newsgroups.com>

It's time to test your brainpower! I will give
six months of free web hosting to the first person
to solve this cryptogram. Can you figure it out?

Since you have taken my challenge and decrypted
the secret message, I will even give you the
cryptogram-making source code to amaze your
friends!


import sys, string, random
cipher = list(string.letters[:26])
random.shuffle(cipher)
cipher = "".join(cipher)
cipher += cipher.upper()
map = {}
for plain, crypt in zip(string.letters, cipher):
    map[plain] = crypt
for line in open(sys.argv[1]):
    print "".join([map.get(ch,ch) for ch in line])[:-1]



--
Novell DeveloperNet Sysop #5

_
"Michal Wallace" <sabren at manifestation.com> wrote in message
news:mailman.1032989974.25321.python-list at python.org...
>
>
>
> Here's a python puzzle. Solve it and you win a prize. :)
>
>
>
> """
> Ha'f ahoy ad ayfa wdqr srxhekdpyr! H phmm zhby
> fhv odeajf dc cryy pys jdfahez ad ajy chrfa kyrfde
> ad fdmby ajhf irwkadzrxo. Ixe wdq chzqry ha dqa?
>
> Fheiy wdq jxby axgye ow ijxmmyezy xet tyirwkayt
> ajy fyirya oyffxzy, H phmm ybye zhby wdq ajy
> irwkadzrxo-oxghez fdqriy idty ad xoxuy wdqr
> crhyetf!
> """
> hokdra fwf, farhez, rxetdo
> ihkjyr = mhfa(farhez.myaayrf[:26])
> rxetdo.fjqccmy(ihkjyr)
> ihkjyr = "".ldhe(ihkjyr)
> ihkjyr += ihkjyr.qkkyr()
> oxk = {}
> cdr kmxhe, irwka he uhk(farhez.myaayrf, ihkjyr):
>     oxk[kmxhe] = irwka
> cdr mhey he dkye(fwf.xrzb[1]):
>     krhea "".ldhe([oxk.zya(ij,ij) cdr ij he mhey])[:-1]
>
>
>
> Cheers,
>
> - Michal   http://www.sabren.net/   sabren at manifestation.com
> ------------------------------------------------------------
> Switch to Cornerhost!             http://www.cornerhost.com/
>  Low Priced, Reliable Blog Hosting, With a Human Touch. :)
> ------------------------------------------------------------
>
>




-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From zopestoller at thomas-guettler.de  Fri Sep 27 08:28:44 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Fri, 27 Sep 2002 14:28:44 +0200
Subject: comp.lang.python...
References: <mailman.1033054480.24506.python-list@python.org>
Message-ID: <an1ip7$rgs$00$1@news.t-online.com>

Vincent A. Primavera schrieb:
> Hello,
> 	Stupid question...  What information do I need to set up
> comp.lang.python with my news client(Pan)?
> 

You need to subscribe this newsgroup and
you need a newsserver.

But how could you post this message without it?

  thomas



From jkraska at san.rr.com  Fri Sep 20 00:32:13 2002
From: jkraska at san.rr.com (Courageous)
Date: Fri, 20 Sep 2002 04:32:13 GMT
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <mailman.1031377930.4225.python-list@python.org>
Message-ID: <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com>

>Well, Python code may look awkward --- if not awk-like ---
>if you don't know how to customize the tab width in your
>editor of choice.

Well, there's your problem already. One should use spaces
and not tabs in Python code. Otherwise, any python file where
tabs and spaces are mixed is also a choice of which indentation
level other viewers of the code MUST use in their editors. That's
heinous.

C//



From martin at v.loewis.de  Tue Sep 17 18:23:31 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 18 Sep 2002 00:23:31 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org>
Message-ID: <m3sn08cl24.fsf@mira.informatik.hu-berlin.de>

Michael Chermside <mcherm at destiny.com> writes:

> I am going to pose a hypothetical situation... please don't argue
> about whether these conditions are true now, or whether they will be
> true, but take these as assumptions:
[...]
> The question is this:  Would the advantages of making Stackless part
> of standard Python outweigh the disadvantages of non-portability and
> C-Python specificity?

Answering this question is tricky, because the answer may depend on
the exact feature set that Stackless would provide. 

It is also very difficult not to argue about these conditions, because
it is certain (for me) that some of them are inherently false, in all
possible universes; a false precondition allows to imply anything.

That said, I do think the advantages would outweigh the disadvantages:
users like the features, and, under those conditions, there is no harm
done integrating them.

Regards,
Martin


From Oschler at earthlink.net  Sat Sep 14 20:49:19 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sun, 15 Sep 2002 00:49:19 GMT
Subject: AspectP?
Message-ID: <jOQg9.6057$Lg2.882928@news2.news.adelphia.net>

Anybody seen a robust (fairly ready for primetime) implementation of Aspect
Oriented Programming for Python?

thx





From shalehperry at attbi.com  Wed Sep 25 19:35:10 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 25 Sep 2002 16:35:10 -0700
Subject: transform strings list in a int lists or other type???
In-Reply-To: <872170378.1032996587856.JavaMail.nobody@webmail2.brturbo.com>
References: <872170378.1032996587856.JavaMail.nobody@webmail2.brturbo.com>
Message-ID: <200209251635.10059.shalehperry@attbi.com>

On Wednesday 25 September 2002 16:29, jubafre at brturbo.com wrote:
>                   Can i tranform a list of strings in a list of int, or
> other type in python??? and also write in a file? I think the
> open(path,'wb') in __builtin__ dont?t write in a file other types, just
> strings, rigth???
>
> Juliano Freitas
> www.gebrasil.hpg.com.br

string_ling = ['14, '42', '83']
int_list = map(int, string_list)

or in the newer pythons:

int_list = [int(i) for i in string_list]

there are equivalent conversion functions for the other types as well.



From magnus at thinkware.se  Fri Sep 20 08:46:47 2002
From: magnus at thinkware.se (=?ISO-8859-1?Q?Magnus_Lyck=E5?=)
Date: Fri, 20 Sep 2002 14:46:47 +0200
Subject: monetary applications [was: Python GUI app to impress the bos
 s?]
References: <31575A892FF6D1118F5800600846864DCBD817@intrepid> <mailman.1032362843.10509.python-list@python.org> <23891c90.0209190014.26c5bb9@posting.google.com> <s2fdma.pp3.ln@ix.netcom.com> <7xk7lhb943.fsf@ruckus.brouhaha.com>
Message-ID: <3D8B18B7.2040203@thinkware.se>

Paul Rubin wrote:
> I'm finding it hard to imagine how any real-world financial
> calculation using IEEE doubles could be off by as much as a penny.

Apart from the rounding problems etc that
the Decimal FAQ describes, I've seen people
getting into trouble when they compare floats
for equality. I've seen that several times
over the years.

In one system it worked well for years, until
the application was ported from VMS to HP-UX.
It would probably have worked on HP-UX as well,
but values calculated on one platform and stored
in the database didn't match values calculated
on the other.

There is always a risk for problems with cancellation
errors, but unless the subtracted values are equal
in the decimal representation and different in the
floating point representation, it will still be a
problem in a decimal representaion. But with a flexible
system for decimal representation, you can adjust
your accuracy to the present needs. With floats the
precision is the same regardless of our needs.



From steve at ferg.org  Wed Sep 18 15:20:42 2002
From: steve at ferg.org (Stephen Ferg)
Date: 18 Sep 2002 12:20:42 -0700
Subject: (Newbie) Menu Not Functioning
References: <uoh10cjh4sct0e@corp.supernews.com>
Message-ID: <b16e4ef7.0209181120.2573add9@posting.google.com>

I can't help with your code, but I'd like to suggest something
that may be helpful.

Go to http://home.att.net/~stephen_ferg/easygui/
and download easygui.  

Then the following program would put up a
nice graphic listbox from which you could choose.

For my own testing, I  made the run() procedure 
just print the command, rather than run it.  But if you
change the code in the run() procedure, it should
work just fine for you. 
-- Steve Ferg (steve at ferg.org)

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
HERE'S THE CODE
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
import os
import sys

from easygui import *


menu = [
    "1. Copy Python Files from Office Computer to Floppy",
    "2. Copy Python Files from Home Computer to Floppy",
    "3. Copy Python Files from Floppy to Office Computer",
    "4. Copy Python Files from Floppy to Home Computer",
    ]

def run(command):
    print "Running", command # demo code
    # os.system(command)  # the real code

def officetoflop():
    run("xcopy C:\MyDocu~1\*.py a:")

def hometoflop():
    run("xcopy E:\MyDocu~1\*.py a:")

def floptooffice():
    run("xcopy a:\*.py C:\MyDocu~1")

def floptohome():
    run("xcopy a:\*.py E:\MyDocu~1")


while 1:  # do forever
    # use the easygui choicebox to make the selection
    choice = choicebox("Choose your selection", "Copy Utilities", menu)

    if choice == None:
        print "Goodbye"
        sys.exit(0)

    menu_choice = choice[0] # get the first character of the choice

    if menu_choice == "1":
        officetoflop()
    elif menu_choice == "2":
        hometoflop()
    elif menu_choice == "3":
        floptooffice()
    elif menu_choice == "4":
        floptohome()
    else:
        print "Program Logic error.  Choice:", choice
        sys.exit(0)


From jubafre at brturbo.com  Tue Sep 10 15:07:01 2002
From: jubafre at brturbo.com (jubafre at brturbo.com)
Date: Tue, 10 Sep 2002 16:07:01 -0300 (GMT-03:00)
Subject: read() file??
Message-ID: <-1076673467.1031684821080.JavaMail.nobody@webmail2.brturbo.com>

     I have a file with assembler instruction and i want read the lines of the file and put in a list of string, this is easy to do, but i don?t want to consider the caracteres "/" and the othes after him.

file.asm:
ADD D1  /coment1
LDA D2  /coment2


list of strings:
X=[ 'ADD D1' , 'LDA D2' ]
	
Juliano Freitas
www.gebrasil.hpg.com.br 

From tim.one at comcast.net  Fri Sep  6 10:04:19 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 06 Sep 2002 10:04:19 -0400
Subject: Python "compiler" is  too slow for processing large data files???
In-Reply-To: <Nv1e9.208900$On.8191536@bin3.nnrp.aus1.giganews.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEGOBCAB.tim.one@comcast.net>

[someone]
> list1 = [
>     (323, 870, 46, ),
>     (810, 336, 271, ),
>     (572, 55, 596, ),
>     (337, 256, 629, ),
>     (31, 702, 16, ),
> ]
>
> Anyway, as my data files went from just a few lines, up to about
> 8000 lines (with 10 values in each line for total of about 450KB of text),
> the time to 'exec' the file became too slow (e.g. 15 seconds) and used too
> much memory (e.g. 50MB) (using ms-windows, python 2.2.1).  It is the
> "compile" phase, because if I re-run, and there is *.pyc file available,
> the import goes very fast (no compilation required).

The parse tree created is huge, so the memory use isn't likely to go down in
the near future (addressing this requires a fundamentally different parser,
and nobody is working on that).  Depending on platform, though, there are
two things that can hurt speed a lot:

1. Oodles of tiny memory allocations and deallocations for the leaf
   nodes of the parse tree.

2. Repeated realloc of a high-level parse node to reflect the ever-growing
   number of top-level children.

#1 has been repaired in 2.3 by using pymalloc in the parser instead of the
platform malloc.

#2 has been repaired in 2.2.2 and 2.3 by changing the parse node growth
strategy to sidestep what's currently quadratic-time behavior in some
platform realloc()s.

Regardless, it would be saner to break any file with 450KB of source code
into smaller files and paste them together with import.




From bokr at oz.net  Wed Sep  4 15:08:23 2002
From: bokr at oz.net (Bengt Richter)
Date: 4 Sep 2002 19:08:23 GMT
Subject: ANN: New PEP Format: reStructuredText
References: <mailman.1030681307.29889.python-list@python.org>
Message-ID: <al5ln7$9de$0@216.39.172.122>

On Fri, 30 Aug 2002 00:23:15 -0400, David Goodger <goodger at users.sourceforge.net> wrote:

>With many thanks to Barry Warsaw for his help and patience, I am
>pleased to announce that a new format for PEPs (Python Enhancement
>Proposals) has been deployed.  The new format is reStructuredText, a
>lightweight what-you-see-is-what-you-get plaintext markup syntax and
 ^^^^^^^^^^^ -- well, the rudiments are light weight ;-)

>parser component of the Docutils project.  From the new PEP 12:
>
>    ReStructuredText is offered as an alternative to plaintext PEPs,
>    to allow PEP authors more functionality and expressivity, while
>    maintaining easy readability in the source text.  The processed
>    HTML form makes the functionality accessible to readers: live
>    hyperlinks, styled text, tables, images, and automatic tables of
>    contents, among other advantages.
>
I like it, but I can see symptoms of "we-can-do-anything-with-this" enthusiasm ;-)

I think the "cheat sheet" at

    http://docutils.sourceforge.net/docs/rst/quickref.html

is a better intro than the other intro's, because it's concise and visual
(though some table stuff appears broken in my NS4.5 browser).

[...]
>Docutils and reStructuredText are under active development.  Input is
>very welcome, especially HTML rendering/stylesheet issues with
>different browsers.  We welcome new contributors.  If you'd like to
>get involved, please visit:
>
>    http://docutils.sourceforge.net/
>

Whenever I encounter yet another markup language (BTW, are you guys monitoring
http://yaml.org/spec/ ;-), I look for fodder for my pet b?te, and if I find an
un-nestable and/or otherwise restricted literal-content quoting construct, my
pet grunts happily (which it is doing ;-)

So I'd like to suggest an optional variant of the literal block, to mimic "here document"
functionality. E.g.,

This::end_of_literal::
<This is arbitrarily indented and might be verbatim copied & pasted from anywhere>.
 end_of_literal ?? -- no! It has to be by itself on a final line.
end_of_literal

For rendering in HTML, I might expect to get

        <p>This:
        <pre>
&lt;This is arbitrarily indented and might be verbatim copied &amp; pasted from anywhere&gt;.
 end_of_literal ?? -- no! It has to be by itself on a final line.
</pre> 

(I.e., '&', '<' and '>' converted to entity refs and otherwise no change between the <pre></pre>).

("end_of_literal" is of course an arbitrary string chosen or generated so as not to occur within
the quoted text).

Re YAML, perhaps a .. yaml:: yamlDoc_ directive to include a rendering of YAML content might
do something useful? (I won't mention .. xml:: xmlDoc_ ;-)

BTW, you could also use doc-here quoting to delimit the scope of a directive, e.g.,

    .. doit::xxx::
<arbitrary-except-not-including-xxx-on-line-by-itself>
xxx

A last comment/question: It-would-be-nice-if-there-were/is-there-a grammar file for
reStructuredText like, e.g., Python's (comes with source distro)

    D:\Python-2.1\Grammar\Grammar

file !/?

Sorry if this repeats stuff covered in project docs/threads/faqs/todos. Have only browsed
spec and PEP and cheat sheet so far...

Regards,
Bengt Richter


From sholden at holdenweb.com  Tue Sep 10 18:49:22 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 10 Sep 2002 18:49:22 -0400
Subject: Larry Wall's comment on python...
References: <mailman.1031579787.7378.python-list@python.org> <3D7E1E28.5040709@bogus.thisisabogusdomain.com>
Message-ID: <kHuf9.123170$l_4.13730@atlpnn01.usenetserver.com>

"Jack D. Hammer" <thisisabogusemail at bogus.thisisabogusdomain.com> wrote ...
>
> One good argument for having block delimiters is for embedded situations
like
> ASP.  One can conditionally put code into a block as such (ASP is rusty):
>
I can't help feeling that if ASP is the answer I'm asking the wrong question
:-)

> <%
> if (some_condition) {
> %>
> <tr>
> <td>some data</td>
> <td>some more data</td>
> </tr>
> <%
> } // closing curly

I want to thank you for adding this valuable comment to your code. Without
it, I might have thought you were trying to close the parenthesis you
unopened above...

> %>
>
> I would think that for embedding python, having block delimiters would be
> beneficial.
>
Beneficial in the same way as giving $_ and $< and similar constructs
"meaning"? I use the inverted commas deliberately. They no longer have much
meaning for me. However, I do not deny the pragmatic value of the language,
not the value of the many pieces of software written in it.

If something works for us, we are entitled to use it. I just lost contact
with Perl when I looked at its object-oriented features. Of course, that may
not be Perl's fault. None of us is getting any younger.

[...]

I can't see "{%>" and "<%}" making it into the "Top 20 Delimiters" list,
perhaps I underestimate the pervasiveness of keyboard skills, of which I
have relatively few.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From peter at engcorp.com  Wed Sep  4 22:43:20 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 04 Sep 2002 22:43:20 -0400
Subject: 4Suite / McMillan Installer / py2exe
References: <doc.m3admxkclw.fsf@goodweb.de> <al5nkp$1n0m2t$1@ID-59885.news.dfncis.de>
Message-ID: <3d76c4ac$1@news.sentex.net>

Thomas Heller wrote:
>>Does someone know how to use 4Suite with the McMillan Installer or
>>py2exe on Windows? Any help or hints would be highly appreciated!
> 
> I never used 4Suite, so no. I even heard ;-) that py2exe
> doesn't work so good with _xmlplus.

We appear to have successfully worked around some of these issues
by simply renaming _xmlplus to xml and removing the original
xml directory.  We're about to "finish" this task and verify that
things are still working, but at first glance it seemed to solve
the problems.

For what it's worth.... (and in case anyone wants to say "uh, that
won't work, moron". :-)

-Peter



From skille at ameritech.net  Wed Sep  4 11:16:59 2002
From: skille at ameritech.net (Stevek)
Date: Wed, 04 Sep 2002 15:16:59 GMT
Subject: embeddable ide?
Message-ID: <Ltpd9.6669$yt3.3168986@newssrv26.news.prodigy.com>

We're planning on making extensive use of Python to provide our users with a
powerful scripting mechanism.  I've successfully embedded and extended
Python but now I want to give the user a decent development environment for
writing their scripts.

I know there are several IDEs available for Python development, but I'm
looking for one that can be used from within another application (embedded).

Any suggestions would be appreciated.





From phr-n2002b at NOSPAMnightsong.com  Thu Sep 26 17:40:23 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 26 Sep 2002 14:40:23 -0700
Subject: Is mmap usable for IPC?
References: <ba1e306f.0209260038.37ab940c@posting.google.com>
Message-ID: <7xwup8whtk.fsf@ruckus.brouhaha.com>

janeaustine50 at hotmail.com (Jane Austine) writes:
> Is mmap module usable for inter-process communication?
> 
> Does it provide lock, or semaphore? Is it safe to write
> from a process and read from multiple other processes
> simultaneously?
> 
> Any information appreciated.

Mmap doesn't provide lock.  Use shm instead.

http://gigue.peabody.jhu.edu/~mdboom/omi/source/shm_source/shm.html



From alanmk at hotmail.com  Thu Sep 26 10:16:30 2002
From: alanmk at hotmail.com (Alan Kennedy)
Date: 26 Sep 2002 07:16:30 -0700
Subject: Strange values in Accept-Language header?
References: <f72dac0d.0209231303.12349f2e@posting.google.com> <mailman.1032906153.7507.python-list@python.org>
Message-ID: <f72dac0d.0209260616.523721ce@posting.google.com>

Alan Kennedy wrote:

>> On looking at my server logs for the Accept-Language headers (to
>> answer another post in this group), I came across some strange
values
>> that are definitely not ISO language codes. For example,
>> 
>> Accept-Language: en-us,x-ns1rDGeT4e2FpA,x-ns2f67971hgDw1
>> Accept-Language: x-ns13V8cN8S1Xz9,x-ns2T329gfxKa7d

Jeff Epler wrote:

> Searching on google turns up two hits in quoted mail message headers,
> both with Mozilla 4.7x.
>     http://www.google.com/search?q=accept-language+x-ns1
> 
> I wonder whether really ancient mozilla cvs might have some clues.  I
> groveled around a bit in bonsai but didn't come up with anything.

Jeff,

Thanks for the reply.

I've had some more time to analyse the logs. User agents that
submitted
these strange values are

Mozilla/4.75C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; U; PPC)
Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.0)
Mozilla/5.0 (Macintosh; U; PPC; en-US; rv:0.9.4.1) Gecko/20020508
Netscape6/6.2.3
Mozilla/4.77C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; U; PPC)
Mozilla/4.73 (Macintosh; U; PPC)
Mozilla/5.0 (Macintosh; U; PPC; ja-JP; rv:0.9.4) Gecko/20011130
Netscape6/6.2.1
Mozilla/4.0 (compatible; MSIE 5.5; Windows 98; VZ_IE5; YComp 5.0.0.0)
Mozilla/4.7 [en]C-CCK-MCD   (Win98; I)
Mozilla/4.0 (compatible; MSIE 5.5; Windows 98)
Mozilla/4.79 (Macintosh; U; PPC)
UCmore

Approx 25% of these hits come through a (non-transparent) proxy, with
proxies from

marketscore.com
integrity.com

being the most common. Which is interesting, because marketscore.com
offer "faster surfing at no cost"

http://www.marketscore.com/e/surffaster.asp

Presumably they are modifying requests in some way to achieve that
"faster surfing"?

integrity.com is a filtering ISP that vets all content for obscenity

http://www.integrity.com/

So it is conceivable that they are implementing some form of custom
"signalling" to servers?

Of course, all of this is pure guess-work, but I'd bet that the 75% of
such hits that appear not to have come from a proxy are actually
coming through a transparent proxy of some kind.

That guess aside, I cannot see a single thing in common between all
the requests that contain these values.

I'd still be interested to hear from anyone who knows what these
values signify. I might email integrity.com and merketscore.com to see
if they can shed any light.

Sincerely,

alan kennedy
----------------------------------------------
check http headers here: xhaus.com/headers
email alan:              xhaus.com/mailto/alan


From jonbeebe85 at yahoo.com  Wed Sep 18 09:15:01 2002
From: jonbeebe85 at yahoo.com (Jonathan Beebe)
Date: Wed, 18 Sep 2002 13:15:01 GMT
Subject: running external programs
Message-ID: <p%_h9.507714$UU1.83878@sccrnsc03>

Is there a function or library i can import that i can use to run external
programs from within my program?

example:

runprogram("iexplore.exe")

thanx,
-JONATHAN BEEBE





From helm at dnv.com  Thu Sep 26 06:13:09 2002
From: helm at dnv.com (Jens Bloch Helmers)
Date: Thu, 26 Sep 2002 10:13:09 GMT
Subject: os.makedirs: difference between nt4 and w2k
Message-ID: <u3crxdpmx.fsf@dnv.com>

Suppose unc_path is a UNC path refering to a non existing
directory on another computer.
os.makedirs( unc_path ) throws an exception if this path
is on a windows 2000 machine but creates the directory if
it is on a win-nt machine.

Is this a feature of win2k, or could I change the setup
of that win2k computer to accept the creation of unc_path?
Could this be done from Python?
(I have administrator privileges on both computers)

Jens Helmers


From trash at terastat.com  Tue Sep 17 13:22:58 2002
From: trash at terastat.com (bap)
Date: Tue, 17 Sep 2002 17:22:58 GMT
Subject: PIL 1.1.3 and grabscreen
Message-ID: <SxJh9.344271$_91.442831@rwcrnsc51.ops.asp.att.net>

I'd like to use the grabscreen function from Lundh's site
(http://effbot.org/downloads) however it won't install or run (I use Python
2.2, it may be compiled for 2.1). The function is now included (according to
the documentation) in PIL 1.1.3 -- but the PIL site only has binary versions
of 1.1.2, for v1.1.3 the site refers to Lundh's site -- where again the
installation routine doesn't work (asks for a directory but will accept no
input). So -- does anyone know of any working sources of either grabscreen
or a binary distributable for windows/activestate python 2.2 of PIL 1.1.3?




From sholden at holdenweb.com  Wed Sep 25 08:31:43 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 25 Sep 2002 08:31:43 -0400
Subject: Getting started
References: <uojinmcfapjf45@corp.supernews.com> <mailman.1032442884.18747.python-list@python.org>
Message-ID: <Z2ik9.50939$b%3.38733@fe03>

"holger krekel" <pyth at devel.trillke.net> wrote ...
> Lance wrote:
> > Hi All,
> >
> > Last night I attended a Weak typing, C++ Templates, and Python talk by
Bruce
> > Eckel. It was great, my introduction to Python.
> >
> > I'm sold on Python and weak typing.
>
> I am not. But i much prefer dynamic over static typing :-)

Python is comfortable for me because the data is strongly-typed and no type
is associated with names. This gives maximum flexibility and minimum
requirements for compile-time checking, both good attributes for a
rapid-development environment.

> What is 'weak typing' supposed to mean, anyway?  I have heard that a
couple
> of times from java-people because they tend to think that python
> is not strongly typed.
>
There are languages which use type information to allow a relaxed
polymorphism, and there are those (like Java) where either syntax or
predominant style anally forces the programmer to prove they are in tune
with the framework's requirements. The latter might still be programming,
but it sure isn't fun [note: I do not insist that programming *must* be fun,
but if you don't enjoy it you'll probably not be reading this].

> > Would you be able to suggest an indispensable learning Python book?
>
Alex is, of course, far too modest to promote his forthcoming "Python in a
Nutshell", but that's going to be extraordinarily good for Python newbies. I
think its appearance will also make other O'Reilly books on Python more
valuable.

> David Beazley wrote "Python essential reference" which is a very good
> book for programmers. It doesn't cover the most recent developments
> but it still is one of the best IMO.
>
Beazley is indeed helpful; soon a Python book won't need to labor to
introduce the language but simply list the appropriate tomes where the
knowledge can be acquired. This is a good thing.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From duncan at NOSPAMrcp.co.uk  Tue Sep 24 09:40:59 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 24 Sep 2002 13:40:59 +0000 (UTC)
Subject: list problem
References: <mailman.1032818129.8126.python-list@python.org> <Xns929364D72B214duncanrcpcouk@127.0.0.1> <tCYj9.157570$pX1.5675986@news2.tin.it> <Xns92938CC236BA2duncanrcpcouk@127.0.0.1> <_FZj9.158106$pX1.5692316@news2.tin.it>
Message-ID: <Xns92939499298F5duncanrcpcouk@127.0.0.1>

Alex Martelli <aleax at aleax.it> wrote in 
news:_FZj9.158106$pX1.5692316 at news2.tin.it:

> I _do_ think that reduce(operator.__add__, ...) is the only
> really frequent use of reduce, with such dominance that it might
> even be worth having as a built-in -- clearest and fastest.  It
> might even specialcase _strings_ in particular, turning the op
> into a ''.join in this case... just musing...

The simplest thing could be to modify operator.add to accept any number of 
arguments. Then you could use operator.add(*aList) to add up or concatenate 
as appropriate all the elements of the list.

It still wouldn't be as fast as ''.join though unless you special cased it, 
and it might be hard to special case it effectively: you would have to 
check the types of all the objects in the list before deciding whether or 
not to use join.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From Oschler at earthlink.net  Mon Sep  2 11:34:36 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Mon, 2 Sep 2002 11:34:36 -0400
Subject: Must have Python modules?
References: <mailman.1030971902.24931.python-list@python.org>
Message-ID: <0pLc9.32066$FJ1.5655132@e3500-atl1.usenetserver.com>

"Denis S. Otkidach" <ods at strana.ru> wrote in message
news:mailman.1030971902.24931.python-list at python.org...
> On Mon, 2 Sep 2002, Robert Oschler wrote:
>
> RO> development!)  I want to know what the other favorites are
>
> mx, PIL, DocumentTemplate, TAL, MySQLdb
>

For those of you newbies like me, curious about what TAL is, try this link:

http://dev.zope.org/Wikis/DevSite/Projects/ZPT/TAL

thx






From duncan at NOSPAMrcp.co.uk  Fri Sep  6 08:32:45 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Fri, 6 Sep 2002 12:32:45 +0000 (UTC)
Subject: tree data structure
References: <FM0e9.2860$cP3.6165@news.iol.ie>
Message-ID: <Xns9281891D87291duncanrcpcouk@127.0.0.1>

Padraig Brady <Padraig at Linux.ie> wrote in 
news:FM0e9.2860$cP3.6165 at news.iol.ie:

> Hi,
> 
> I've a graph like stucture represented
> by a list of tuples:
> 
> [(L0,S2), (S2,S0), (S2,S1), (S1,Q2), (S0,Q1), (S2,S3)]
> 
> And would like to convert it to
> a tree structure like:
> 
>            S0 - Q1
>          /
> L0 - S2 - S3
>          \
>            S1 - Q2
> 
> Any tips appreciated.
> 
Does this help?

data = [('L0', 'S2'), ('S2', 'S0'), ('S2', 'S1'), ('S1', 'Q2'), ('S0', 
'Q1'), ('S2', 'S3')]

# Produce a tuple where each node contains its name
# and a list of its child nodes:
#  (node, [child1, ...])
nodes = {}
for parent, child in data:
    nodes.setdefault(parent, (parent, 
[]))[1].append(nodes.setdefault(child, (child, [])))

result = nodes[data[0][0]]

assert result==('L0', [('S2', [('S0', [('Q1', [])]), ('S1', [('Q2', [])]), 
('S3', [])])])


-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From robin at execulink.com  Sat Sep 14 01:59:43 2002
From: robin at execulink.com (robin at execulink.com)
Date: Sat, 14 Sep 2002 01:59:43 -0400
Subject: threads or queue for this task
Message-ID: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com>

I am missing something I think with the difference between the
threading and Queue modules. These could really do with some updated
docs with examples.

I wish to have a server continually cycle (until interrupted by
keyboard, signal, or whatever). It needs to perform several (say 6)
different tasks, each of which may take significantly different
lengths of time to perform.

My first implementation used threading.py. I wrote a dispatcher to
start each of the threads, and then wait until the thread count was
back down to 1. This I put in a while loop, and surrounded it with the
necessary exception code.

Each cycle through the while has to wait for the slowest process to
finish before completing. During this time, the other processes just
sit around doing nothing. What I really want is for each process to
start up again when they are done, so that there are no more than 6
going at once. In other words, I want to be sure I don't start a new
process 3 before the old process 3 is done.

I have looked at Queue but it doesn't seem to have enough control to
make this happen.

I am sure I am overlooking something obvious. Thanks ahead of time for
any help!

-- robin


From cliechti at gmx.net  Wed Sep 25 15:06:40 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 25 Sep 2002 21:06:40 +0200
Subject: Python and Scite
References: <mailman.1032973593.30002.python-list@python.org>
Message-ID: <Xns9294D6CCCC159cliechtigmxnet@62.2.16.82>

"Sperry Russ" <sperry at newcreations.info> wrote in
news:mailman.1032973593.30002.python-list at python.org: 

> Hello,
> 
> I am new to programming.  I have downloaded Python 2.2.1 and Scite
> 1.48 to learn how to program.  I am trying to configure my Scite so
> that it will run the python files by using go.  When I try to do this
> i get a message in the second window at the bottom -- unable to find
> file.  I have tried configuring the global options and have failed can
> anyone help me? I need a step by step example if you can because I
> cannot understand it according to the documentation that came with
> Scite 

works out of the box for me. if you're on windows, make sure that python is 
found in the PATH. i.e. edit PATH that it contains the python directory.
and you could look at PythonWin, it uses the editing component used in 
scite and has some additional features for the python programmer. 
additionaly it comes also with useful, de-facto standard, windows specific 
extensions.

chris

-- 
Chris <cliechti at gmx.net>



From a_salnikov at yahoo.com  Mon Sep  2 14:11:12 2002
From: a_salnikov at yahoo.com (Andy Salnikov)
Date: Mon, 2 Sep 2002 11:11:12 -0700
Subject: Python GUI
References: <3d8d3c0.0208262339.e7bc02c@posting.google.com> <r794nugam0su9metpaas0nobn4jsm84j3j@4ax.com> <akv3b8$7dv$1@usenet.Stanford.EDU> <mailman.1030954205.15207.python-list@python.org>
Message-ID: <al09m3$qsa$1@usenet.Stanford.EDU>

"Bo M. Maryniuck" <b.maryniuk at forbis.lt> wrote in message
news:mailman.1030954205.15207.python-list at python.org...
> On Monday 02 September 2002 09:16, Andy Salnikov wrote:
> > Problem with X11 is that there is no "OS style manager" (whatever
> > this could mean) for X11, or there is no _single_ style manager.
>
> No. Problem with the license. Use wxWindows instead.
>
  Qt for X11 is free and license for X11 is not restrictive. Good example
is whole KDE built on top of Qt. Trolltech guys are indeed nice when
it comes to disscussing license terms. (Qt/Windows is not free though.)
wxWindows may be good too, but I consider Qt more mature, whole
KDE community works with Qt (although it does not mean Qt is free
of bugs.)

  Andy.




From lee100 at sport.rr.com  Thu Sep 19 19:38:12 2002
From: lee100 at sport.rr.com (Lee Gray)
Date: Thu, 19 Sep 2002 23:38:12 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <96904b50.0209181149.429e8533@posting.google.com> <Xns928E8CC3E7680PaCmAnRDLM@161.48.128.20>
Message-ID: <Edti9.281959$eK6.8089354@twister.austin.rr.com>

Fred Pacquier wrote:

> lpgray at uop.com (Lee Gray) said :
> 
>> Thanks for all the suggestions, folks!
>> 
>> I've collected lots of links to a variety of very nice screenshots,
>> and I've been playing with the wxWindows demo and Boa Constructor. Boa
>> was a bit disappointing, not as an app, but because of its current
>> alpha status - but it's got great potential and I'll definitely look
>> forward to future releases. I'll check out PythonCard too.
> 
> This just in : the Python Daily URL (http://www.pythonware.com/daily/)
> mentions this collection of classes and tools for UML modeling and
> programming. I don't understand it all, but the sceenshot of the graphical
> model editor sure is a killer :-)
> 
> http://www.abo.fi/~iporres/Projects/fog0000000023.html

Hehe... yep, that was one of 'em I bookmarked.

Lee


From thomas at xs4all.net  Fri Sep 13 09:27:30 2002
From: thomas at xs4all.net (Thomas Wouters)
Date: Fri, 13 Sep 2002 15:27:30 +0200
Subject: Cursed newlines and readline()
In-Reply-To: <6u5g9.11312$Da3.1133980@kent.svc.tds.net>
References: <6u5g9.11312$Da3.1133980@kent.svc.tds.net>
Message-ID: <20020913132730.GD797@xs4all.nl>

On Thu, Sep 12, 2002 at 06:59:14PM +0000, Edward K. Ream wrote:

> According to one of Leo's Linux users, the Python's readline() routine on
> Linux delivers "\r\n" line end strings verbatim, while the windows versions
> force the string to use the Unix convention of using only "\n". This seems
> unfortunate.  Is there a justification for it?  If not, I would like to
> suggest that readline() work the same on all platforms and that it deliver
> only "\n" as a line terminator.

You can open the file in 'binary' mode (adding 'b' to the mode string) and
the file will contain '\r\n' on both platforms (and any other platforms.)

-- 
Thomas Wouters <thomas at xs4all.net>

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



From skip at pobox.com  Tue Sep  3 23:11:11 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 3 Sep 2002 22:11:11 -0500
Subject: can distutils build simultaneously on two architectures?
Message-ID: <15733.31183.590119.495095@12-248-11-90.client.attbi.com>

Can I safely execute the standard distutils command

    python setup.py build

simultaneously on two different architectures (Solaris and Linux in this
case) without fear of clashing?  I know about the build/{temp,lib}*
directories for compiling and linking C/Fortran code.  What about the
directories containing .py[co] files?  Does any sort of locking take place?

-- 
Skip Montanaro
skip at pobox.com



From remi at cherrypy.org  Thu Sep 12 09:15:50 2002
From: remi at cherrypy.org (Remi Delon)
Date: Thu, 12 Sep 2002 15:15:50 +0200
Subject: Online XML/XSL transformation tool available (based on 4Suite and CherryPy)
Message-ID: <alq428$kim$1@wanadoo.fr>

Hi everyone,

I just made available a very simple online XML/XSL transformation tool.
You just input your XML document and your XSL stylesheet (in textareas), and
it runs the transformation and displays the result.
This can sometimes come handy to debug XSL stylesheets.

This tool is part of the CherryPy online demo and it is used to demonstrate
the integration of the 4Suite module into CherryPy.

The direct URL is http://www.cherrypy.org/demo/xmlXslOnline

Cheers.

Remi.

PS: Let me know if you have any comments or if you would like more features.

PS2: Thanks to the Fourthought guys for their good work on 4Suite.





From mfranklin1 at gatwick.westerngeco.slb.com  Fri Sep 13 09:45:42 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Fri, 13 Sep 2002 13:45:42 +0000
Subject: Pmw.ScrolledText programatically scrolling
In-Reply-To: <3D81D44E.C5FC4A8B@mill.co.uk>
References: <3D81D44E.C5FC4A8B@mill.co.uk>
Message-ID: <200209131345.42294.mfranklin1@gatwick.westerngeco.slb.com>

On Friday 13 Sep 2002 12:04 pm, Joe Connellan wrote:
> I've got a Pmw.ScrolledText() that I'm using as a log and I've written
> an addToLog() method like so:
>
>  def addToLog(self, logString):
>   self.logScr.configure(text_state = 'normal')
>   self.logScr.insert(END, logString)
>   self.logScr.configure(text_state = 'disabled')
>
> I now want it to scroll when I add a line so the user doesn't have to
> constantly scroll down to view the most recent lines.
>
> I can't find any info on doing this with ScrolledText or its text
> component - do I need to reposition the scrollbars for it to scroll?
>
> Thanks
>
> Joe

Joe,

Try this:-


st.insert('end',stuff)
st.yview_pickplace('end')

where 'st' is a Pmw.ScrolledText instance


Regards
Martin

-- 
### Python Powered Signature
I started writting this email on 
Fri Sep 13 13:44:27 2002




From rpw3 at rpw3.org  Wed Sep 18 22:29:35 2002
From: rpw3 at rpw3.org (Rob Warnock)
Date: Thu, 19 Sep 2002 02:29:35 -0000
Subject: DNS problem re: International Lisp Conference 2002
References: <am5m8l$ai8$1@news.franz.com>
Message-ID: <uoidkfh10u1f9@corp.supernews.com>

Raymond de Lacaze <delaray at hotmail.com> wrote:
+---------------
|  Registration deadline:  September 31st, 2002
+---------------

Oh, is a new calendrical system being presented at the Conference?  ;-}

+---------------
| Registration information is available at:
| http://www.international-lisp-conference.org/Overview/LC02-Registration.html
+---------------

Unfortunately, while the Internet root servers claim that
"{ns1,ns2}.franz.com" are authoritative for the domain
"international-lisp-conference.org", the "{ns1,ns2}.franz.com"
servers deny any knowledge of it!!!  (Makes it a little bit
hard to access...)

*Help!*


-Rob

-----
Rob Warnock, PP-ASEL-IA		<rpw3 at rpw3.org>
627 26th Avenue			<URL:http://www.rpw3.org/>
San Mateo, CA 94403		(650)572-2607



From hst at empolis.co.uk  Fri Sep 13 03:41:59 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Fri, 13 Sep 2002 08:41:59 +0100
Subject: Help with regular expression using findall and .*?
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220D18@hendrix.empolisuk.com>

czrpb wrote
> 
> Could anyone help out with rewriting (still using regular expressions)
> the following so that it does not cause an exception:
> 
> import re
> 
> s1=('macro\n'+'a'*200+'\norcam\n')*10
> s2=('macro\n'+'a'*20000+'\norcam\n')*10
> 
> p=re.compile(r'macro.*?orcam',re.DOTALL)
> 
> for x in re.findall(p,s1):
>     print x
> 
> for x in re.findall(p,s2):
>     print x
> 
> thanks!! Quentin Crain
> 

You need to be very careful about using .*? as the engine "only" allows 10,000 backtracks

Try this

p = re.compile('macro(?:[^o]+|o(?!rcam))*orcam')
for x in p.findall(s2):
    print x

HTH

Harvey

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.



From whisper at oz.net  Fri Sep  6 15:25:04 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 6 Sep 2002 12:25:04 -0700
Subject: Larry Wall's comment on python...
In-Reply-To: <Pine.LNX.4.44.0209061302050.1523-100000@ns0.tbye.com>
Message-ID: <GCEDKONBLEFPPADDJCOECEHEENAA.whisper@oz.net>

The man obviously can't see the blocks for the indenting ;)

David LeBlanc
Seattle, WA USA 

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of brueckd at tbye.com
> Sent: Friday, September 06, 2002 13:04
> To: python-list at python.org
> Subject: Re: Larry Wall's comment on python...
> 
> 
> On 6 Sep 2002, Rod Stephenson wrote:
> 
> > "Python is cool to look at small bits of, but I think the "outline"
> > syntax breaks down with larger chunks of code. I'm with Aristotle on
> > the structure of discourse--a story should have a beginning, and
> > middle, and an end. So should blocks"
> > 
> > I'm not quite sure what he's trying to get at here - I guess that for
> > a long heavily indented chunk of code, you could lose track of the
> > overall structure, but I don't write code this way.
> > 
> > Any comments?
> 
> The very idea of Larry Wall lecturing on code readability is hilarious.
> 
> -Dave
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list



From marklists at mceahern.com  Thu Sep 19 13:21:17 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 19 Sep 2002 12:21:17 -0500
Subject: strong/weak - dynamic/static [Was: Getting started]
In-Reply-To: <87ofaut2xt.fsf@pokey.henrik-motakef.de>
Message-ID: <JHEOKEOOLIGLDHCMAHMOIEPIDPAA.marklists@mceahern.com>

[Henrik Motakef]
> There's a difference between a type and a class, I guess...

Alex made the point already, but the following code demonstrates that what
you demonstrated relies on old-style classes, which gives you a misleading
answer imho:

**

class foo(object):

    def changeme(self, newclass):
        self.__class__ = newclass

class bar(object):

    pass

f = foo()
b = bar()

print "type(f) = %s" % type(f)
print "type(b) = %s" % type(b)
print "type(f) == type(b) ==> %s" % (type(f) == type(b))

f.changeme(bar)

try:
    # Try to change back...
    f.changeme(foo)
except AttributeError:
    print "Whaddya know, we can't change f anymore."

print "type(f) = %s" % type(f)
print "type(b) = %s" % type(b)
print "type(f) == type(b) ==> %s" % (type(f) == type(b))

**

Output:

type(f) = <class '__main__.foo'>
type(b) = <class '__main__.bar'>
type(f) == type(b) ==> 0
Whaddya know, we can't change f anymore.
type(f) = <class '__main__.bar'>
type(b) = <class '__main__.bar'>
type(f) == type(b) ==> 1

**

My point in posting this is merely to counter the false claim that you
cannot change an object's type at runtime.  I'm not asserting that you can
change any object's class to anything whatsoever without experiencing
problems or runtime errors, of course.

Cheers,

// m

-




From news at datatailors.com  Mon Sep 23 19:08:24 2002
From: news at datatailors.com (Peter van Kampen)
Date: Tue, 24 Sep 2002 01:08:24 +0200
Subject: Solved (was: Cookbook example, DAO error)
References: <slrnaotvc2.4hm.news@datatailors.com>
Message-ID: <slrnaov7n7.59q.news@datatailors.com>

In article <slrnaotvc2.4hm.news at datatailors.com>, Peter van Kampen wrote:
> Hi,
> 
> I am trying to run the jet2sql script from the Python cookbook but it
> fails. 

It turns out it's the the mdb I wanted to process caused the problems.
I created a new one from scratch and jet2sql worked fine.

PterK


From pyth at devel.trillke.net  Wed Sep 25 19:15:34 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 26 Sep 2002 01:15:34 +0200
Subject: a python puzzle
In-Reply-To: <Pine.LNX.4.44.0209251817040.17059-100000@hydrogen.sabren.com>; from sabren@manifestation.com on Wed, Sep 25, 2002 at 06:21:56PM -0400
References: <20020926000315.K30315@prim.han.de> <Pine.LNX.4.44.0209251817040.17059-100000@hydrogen.sabren.com>
Message-ID: <20020926011534.L30315@prim.han.de>

Michal Wallace wrote:
> On Thu, 26 Sep 2002, holger krekel wrote:
> 
> > OK.  something like 
> > 
> > def decrypt(yourstring):
> >     dedict = {'F': 'S', 'I': 'C', 'H': 'I', 'a': 't', 'c': 'f', 'b': 'v', 'e': 'n',
> >     'd': 'o', 'f': 's', 'i': 'c', 'h': 'i', 'k': 'p', 'j': 'h', 'm': 'l',
> >     'l': 'j', 'o': 'm', 'q': 'u', 'p': 'w', 's': 'b', 'u': 'z', 't': 'd',
> >     'w': 'y', 'v': 'x', 'y': 'e', 'x': 'a', 'z': 'g', 'g': 'k'}
> >     return "".join([dedict.get(i,i) for i in yourstring])
> > returns
> ...
> >     It's time to test your brainpower! I will give
> >     six months of free web hosting to the first person
> >     to solve this cryptogram. Can you figure it out?
> 
> 
> Hi Holger,
> 
> Yep, that's it! I figured it would be easy to solve -
> especially given the python code, but I didn't expect a
> solution THAT quick. Next time I'm using RSA!

No problem :-)

Just letting out the python code would make it a bit harder.
One might even have to start to seriously program ...

> > but i don't know if i need free webhosting :-)
> 
> Hahaha. Well, if you change your mind, just let me know. :)

cheers,

    holger



From aleax at aleax.it  Mon Sep 23 02:58:57 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 23 Sep 2002 06:58:57 GMT
Subject: PyQt setup on Linux Mandrake 8.2
References: <3D8B4898.20808@nospam.attbi.com> <3d8e7141$1@news.nz.asiaonline.net> <3D8E74A4.7000404@nospam.attbi.com>
Message-ID: <RYyj9.128716$ub2.2801139@news1.tin.it>

djw wrote:
        ...
> modules). Unfortunately, when I try to install new py modules, they all
> want to install into /usr instead of /usr/local, since /usr/bin/python
> is ahead of /usr/local/bin/python on $PATH, causing sys.prefix to point

The PATH is not etched in stone: edit /etc/profile or wherever it is
you set your PATH, and just put /ust/local/bin ahead of /usr/bin there.


Alex



From huaiyu at gauss.almadan.ibm.com  Tue Sep  3 20:25:15 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Wed, 4 Sep 2002 00:25:15 +0000 (UTC)
Subject: default tuple unpacking?
Message-ID: <slrnanaknb.1gh.huaiyu@gauss.almadan.ibm.com>

It is often very convenient to have default arguments such as

    def func(a, b, *c): ...

Is there a way to use this in statements too?  Examples:

    a, (b, *c), *d = e
    for a, *b in c: ...
    return a, *b

I often use small utilities such as

    def split1(x): return x[0], tuple(x[1:])
    def split2(x): return x[0], x[1], tuple(x[2:])

to simulate this in easier situations, but I often find the need to do a
more general unpacking by tuple pattern matching.  They would make the code
so much more readable.  Python can already handle all the following

    def func(a, (b,c), *d)
    a, (b, (c, d), e) = f
    for (a,b), (c,d) in e: ...

I would guess it should not be too difficult to handle the combinations.
Any ideas?  Has this been done?


Huaiyu


From fperez528 at yahoo.com  Fri Sep 27 17:38:11 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Fri, 27 Sep 2002 15:38:11 -0600
Subject: watching mutables?
References: <mailman.1033137396.14482.python-list@python.org> <an211j$oi$1@news.hccnet.nl> <Xns9296C58C9ADFCcliechtigmxnet@62.2.16.82> <k7l7qpvf.fsf@ion-tof.com> <Xns9296D3AB53777cliechtigmxnet@62.2.16.82> <mailman.1033154856.20758.python-list@python.org> <an2doh$ajo3p$1@ID-59885.news.dfncis.de> <an2hnt$nh2$1@news.hccnet.nl>
Message-ID: <an2j43$8fg$2@peabody.colorado.edu>

Anton Vredegoor wrote:

> But...
> 
> I specifically asked for something that doesn't touch the original
> object. In eventobj.py's docstring there is the following warning:
> 
> <warning>
> Caution : As event management is performed by changing the class of
> the instance, you use eventobj with critical objects at your own
> risk... !
> </warning>
> 
> So, with a lot of thanks for the results sofar ...
> 
> Is it possible?

Well, somehow you have to hook into the object to trap modifications to it, 
no? The other option is via a debugger interface, which works at the bytecode 
level with the interpreter itself. Look into the python debugger (pdb) for 
this. Incidentally, ddd knows how to interface to pdb and can do what you are 
asking about. It's been a long time since I've used it, but I've done exactly 
that with ddd in the past.

Cheers,

f


From Oschler at earthlink.net  Sun Sep 29 12:52:34 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sun, 29 Sep 2002 16:52:34 GMT
Subject: Linker can't find library (PyProlog setup)
Message-ID: <mdGl9.14197$Lg2.4064162@news2.news.adelphia.net>

I'm having trouble getting gcc/ld to 'see' the existence of a known shared
library.  I quadruple-checked and I do indeed have the files 'libtermcap.so'
and 'libtermcap.a' in my '/usr/lib/termcap' directory.  'gcc' is being
passed the following library directive '-ltermcap'.  I tried the following
'export' commands in my build script (the echoed results do indeed show
/usr/lib/termcap being properly added to the respective environment
variables):

    export LD_LIBRARY_PATH="/usr/lib/termcap:$LD_LIBRARY_PATH"
    echo "LD_LIBRARY_PATH=$LD_LIBRARY_PATH"
    # run gcc command here

and I also tried:

    export PATH="/usr/lib/termcap:$PATH"
    echo "PATH=$PATH"
    # run gcc command here

and lastly, both:

    export LD_LIBRARY_PATH="/usr/lib/termcap:$LD_LIBRARY_PATH"
    export PATH="/usr/lib/termcap:$PATH"
    # run gcc command here

In all three cases I still get a complaint from 'ld', 'cannot
find -ltermcap'.  What am I doing wrong?

thx




From sholden at holdenweb.com  Tue Sep 17 18:01:42 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 17 Sep 2002 18:01:42 -0400
Subject: Will "Python in a Nutshell" be too short?
References: <mailman.1032288261.15607.python-list@python.org>
Message-ID: <LMNh9.15079$T4.1903@fe04>

"Hamish Lawson" <hbl at st-andrews.ac.uk> wrote ...
> O'Reilly doesn't yet list the forthcoming "Python in a Nutshell" by Alex
> Martelli, but according to Amazon.com's page for the book, it will have
400
> pages. Is that indeed the plan? I know that O'Reilly aims to keep fat out
> of its books, and that Alex is striving to be comprehensive yet concise,
> but I worry that the book's proposed length will simply be too short to
> cover the material I'm hoping will be included.
>
I wil be very surprised if Alex' book does not become the canonical
reference for the language, since his treatment is exceptionally thorough
and (even for him) more than usually precise.

> I regularly consult David Beazley's "Python Essential Reference", but the
> usefulness to me of this otherwise excellent book is compromised by the
> number of (platform-independent and not obsolescent) standard modules and
> packages that had to be left out of the book's 380 pages (htmllib,
> telnetlib, urllib2, CongigParser, unittest, tkinter, xml, etc.).
>
They clearly aren't "Essential" :-)

> Many of us were hoping that "Python in a Nutshell" would not only be a
> comprehensive reference of pretty much all of the modules in the standard
> library that aren't obsolescent or platform-specific, but would also cover
> some of the most popular third-party libraries - e.g. DB-API, PIL,
> mxDateTime, Numeric, win32all, wxPython, mod_python, ReportLab.
>
The omission of some material that particular readers consider desirable is
more or less inevitable, given the different requirements we all have of
this single language. Just the same, I'm sure you will be impressed by what
they *have* found space to include.

> My congratulations to Alex and O'Reilly if they reckon they can somehow
> manage to fit this all into just 400 pages; but if this won't be possible,
> as I worry must be the case, then may I make a plea for the book's length
> to be reconsidered. It would be good to have a one-stop Python reference
> book that would be a suitable companion to the task-oriented "Python
> Cookbook" - I'm hoping "Python in a Nutshell" can still be that book.
>
It's far too late into the production process for such pleas to be
effective, no matter what the individuals concerned may think of their
merits. It's probably at the printers already...

> By comparison, O'Reilly's "Perl in a Nutshell" is 800 pages and has the
> following chapters: Introduction to Perl; Installing Perl; The Perl
> Executable; The Perl Language; Function Reference; Debugging; Packages,
> Modules, and Objects; Standard Modules; CGI Overview; The CGI.pm Module;
> Web Server Programming with mod_perl; Databases and Perl; XML and Perl;
> SOAP; Sockets; Email Connectivity; Usenet News; FTP; Lightweight Directory
> Access with Net::LDAP; The LWP Library; Perl/Tk; Win32 Modules and
> Extensions; OLE Automation; ODBC Extension for Win32.
>
Guess you'll just have to wait and see ...

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From TuxTrax at fortress.tuxnet.net  Fri Sep 27 18:05:07 2002
From: TuxTrax at fortress.tuxnet.net (TuxTrax)
Date: Fri, 27 Sep 2002 22:05:07 -0000
Subject: cursor positioning and color
References: <slrnaoqv2q.hn4.TuxTrax@fortress.tuxnet> <1b6dca50.0209231022.6fae7f6d@posting.google.com> <30770aa4.0209232242.1e91d9d5@posting.google.com> <1b6dca50.0209260132.6d1d333a@posting.google.com>
Message-ID: <slrnap9llp.v98.TuxTrax@fortress.tuxnet>

On 26 Sep 2002 02:32:41 -0700, brrrent Wrote in
Steve Ballmers hair grease:

> bogusdrop at myself.com (TuxTrax) wrote in message news:<30770aa4.0209232242.1e91d9d5 at posting.google.com>...
>> bfarwick at juno.com (brrrent) wrote in message news:<1b6dca50.0209231022.6fae7f6d at posting.google.com>...
>> > TuxTrax at fortress.tuxnet.net (TuxTrax) wrote in message news:<slrnaoqv2q.hn4.TuxTrax at fortress.tuxnet>...
>> > > Being new to python, I would like to know how one goes about such 
>> > > tasks as precise cursor positioning and output to screen in color.
>> > > 
>> > 
>> > Hi Mathew,
>> > 
>> > Curses is the way to go when using Linux/Unix.
>> > 
>> > Wconio is the way to go when using Microsoft.
>> > You can find in the Vaults of Parnassus.  If you have any trouble
>> > locating it, let me know, I can email it to you.  Small library, works
>> > well.
>> > 
>> > All the best,
>> > 
>> > brrrent
>> 
>> Thanks Brent. Appreciate it. Am I to understand that there are no
>> python libraries that do this as a standard part of the python
>> distribution?
> 
> Mathew,
> 
> A linux or unix distribution without curses is just about unheard of.
> About the only chance of finding a linux installation without curses
> would be in a standalone firewall box.
> 
> WConio is a "tiny" download and very worth it.  Works like a champ. 
> One heck of a lot easier to grasp than curses.  Many thanks to Mr.
> Gonnerman for such a useful library.  If you write a python WConio
> app, why not just send WConio along with the python code.  As I said,
> it's "tiny".
> 
> brent
> 
>  
>> That would be preferable, as it would eliminate the requirement of
>> making
>> the user have curses on their system before they can use the code.
>> 
>> Cheers,
>> 
>> Mathew

Thanks for the info.

I won't be needing WConio, because I don't use windows.

However, I have been tinkering with curses (I use Linux exclusively) and
have found it frustrating. It looks like what I need, but I can't find much
by way of examples on line, as to how to use it.

for instance, when run from a file, this generates an attribute error:

import curses

stdscr = curses.initscr()

stdscr.addstr(0, 0, "Current mode: Typing mode", curses.A_REVERSE) 

_____________

Now that is exactly from the python docs. It croaks on line 3,

stdscr = curses.initscr()

but according to the docs, that initialization should work. Geeze, what
I want to do is so simple, I want to be able to position the cursor and
maybe use some color, in a text application. It shouldn't be this hard.
I had less trouble using NNTPlib.

Cheers,

Mathew

-- 
TuxTrax   (n.) An invincible, all terrain, Linux driven armored assault
vehicle that can take as much fire as it gives ;-)

Yes, I am a Penguin cult high priest. Flipper readings upon request.

ROT13 this email address to mail me:
bar jbeq abg guerr, uvtu qrfreg zna, ng lnubb qbg pbz



From astrophels at yahoo.com  Tue Sep  3 10:17:10 2002
From: astrophels at yahoo.com (Micah Mayo)
Date: Tue, 03 Sep 2002 14:17:10 GMT
Subject: trailing newline in string.splitfields()
Message-ID: <Gv3d9.287721$me6.36853@sccrnsc01>

I was splitting up some fields in a textfile using string.split, which 
works fine, but that it usee ' ' as the delimiter. I've found that some 
of the things I want to split include a space or two, so I've gone to 
using string.splitfields, which also works fine, but for one thing. 
Unlike string.split() it also grabs the newlne('\n') at the end of each 
line.

The way I've worked around this is as such:

var1,var2,var3 = string.splitfields(stringtosplit,',')
list.append(var1)
list.append(var2)
list.append(var3[:-1])

So, I slice of the character, easy enough. And I'm generally satisfied 
with that - but I was wondering if there may be, for the cause of more 
beautiful code, a way to tell string.splitfields() to do it automatically.

Thanks,
Micah



From thomas.heller at ion-tof.com  Thu Sep 26 09:25:00 2002
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: 26 Sep 2002 15:25:00 +0200
Subject: Getting started
References: <00a801c26511$4e084750$df7e4e18@brianspiv1700> <mailman.1033021239.13549.python-list@python.org> <iaDk9.145545$ub2.3207724@news1.tin.it>
Message-ID: <lm5olw7n.fsf@ion-tof.com>

Alex Martelli <aleax at aleax.it> writes:

> James J. Besemer wrote:
>         ...
> > Types are otherwise pretty secure in C++ except the "cast" mechanism
> > "foils" the type system and thus allows type errors to occur.
> 
> Actually, unions are at least as bad -- they're not often used
> for type-punning because casts are handier for that, but if you
> took cast aways, people would (ab-)use unions (just like they
> abuse EQUIVALENCE in Fortran for similar purposes).
> 
> Technically I believe that in C++ (not sure about the two ISO
> C standards -- I'm more definite that this was the case in
> Fortran-77 and I believe in Fortran-IV too) storing one union's
> member then retrieving another one puts your program in
> undefined-behavior land.  In practice, I know of no existing
> compiler that goes to the bother of checking (e.g. by storing
> a hidden tag when you store into a union and checking whenever
> you retrieve from it -- even harder if you consider _pointers_
> to the various members of the union).

Hidden fields in a union by the discretion of the compiler?
Don't think that would be allowed by any C standard.

Thomas


From syver-en+usenet at online.no  Mon Sep  9 11:48:14 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Mon, 09 Sep 2002 15:48:14 GMT
Subject: win32com.client.gencache and py2exe
Message-ID: <uit1ftba1.fsf@online.no>

When using py2exe to make package a python program that uses gencache
to generate wrapper files from COM type libraries, the resulting
installation fails because the win32com system is unable to find the
generated files. Is there any workaround to this problem?z


-- 

Vennlig hilsen 

Syver Enstad


From fgibbons at hms.harvard.edu  Wed Sep 25 09:09:16 2002
From: fgibbons at hms.harvard.edu (Frank Gibbons)
Date: Wed, 25 Sep 2002 09:09:16 -0400
Subject: Python-list digest, Vol 1 #12538 - 15 msgs
In-Reply-To: <20020925052600.22801.4196.Mailman@mail.python.org>
Message-ID: <5.0.0.25.2.20020925090517.02717c70@hms.harvard.edu>

Thanks for the comments, Donn. I was just about to try something like that 
myself. The toy example I posted does not do what I really want to, which 
is to avoid using a temporary file at all, and echo the output from gzip 
directly to the web browser, with appropriate content-type, of course. Your 
blocked input/output seems the way to go.

Thanks again,

Frank

At 01:26 AM 9/25/2002, you wrote:
>It isn't going to be fixed, either, because there are just some
>axiomatic truths in operation here, like "I/O buffers aren't
>infinitely elastic." Since your example doesn't do anything with
>the output, you don't have to shoot yourself in the foot this way.
>Try something like this -
>
>   zip_input = os.popen('gzip -c > popen.gz', 'w')
>   while 1:
>       data = f.read(16000)
>       if not data:
>           break
>       zip_input.write(data)
>   zip_input.close()
>
>The main point is that this allows gzip to write directly to disk.
>If your actual application needs to process the data further at this
>point, you may read it back from disk and write to a new file if you
>like.   Otherwise, if you really want a second pipe, you have to figure
>out how to read some, write some, etc., to keep gzip's output from
>filling up while you're trying to write all its input.

PhD, Computational Biologist,
Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA.
Tel: 617-432-3555       Fax: 
617-432-3557       http://llama.med.harvard.edu/~fgibbons




From bobx at linuxmail.org  Thu Sep 12 21:14:21 2002
From: bobx at linuxmail.org (Bob X)
Date: Fri, 13 Sep 2002 01:14:21 GMT
Subject: Coding Style: Quotes
References: <bt1g9.17429$Pf7.590136@news1.west.cox.net> <mailman.1031842455.20916.python-list@python.org>
Message-ID: <NZag9.2396$Lg2.449107@news2.news.adelphia.net>

Michael Stenner wrote:
<snip>
> Since there is no "physical" difference between them, I can't imagine
> why it would matter.  The only reason there "needs" to be two is for
> nested quoting.  Sure, it would be annoying to alternate for each
> string, but beyond that...
Don't you have to escape a single quote within a single quote whereas 
you do not while using a double quote?

i.e. 'It\'s raining'  "It's raining"



From b.maryniuk at forbis.lt  Tue Sep  3 08:59:09 2002
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Tue, 3 Sep 2002 14:59:09 +0200
Subject: To all (Re: Servizi Internet)
In-Reply-To: <B0005002240@mailserver2.ede.it>
References: <B0005002240@mailserver2.ede.it>
Message-ID: <200209031459.09717.b.maryniuk@forbis.lt>

Somebody can moderate this annoying and <censored/> SPAM here?!


On Tuesday 03 September 2002 14:38, Servizi Internet wrote:
> Salve.
> Sono il titolare di "Web Ceating.it", una delle web agency piu' in vista
> sul mercato europeo. Volevamo informarLa che la nostra azienda si
> mette a Sua completa disposizione per qualsiasi esigenza o servizio Lei
> desideri. Ci occupiamo principalmente di:
>
> * Servizio Hosting e registrazione domini .com/.net/.org/.biz/.it;
> * Registrazione nei principali motori di ricerca;
> * Posizionamento garantito in testa ai motori di ricerca;
> * Creazione Siti Personali e Aziendali;
> * Consulenza Web Marketing;

[SKIPPED]

-- 
Regards, Bogdan

If the facts don't fit the theory, change the facts.
                -- Albert Einstein




From max at alcyone.com  Wed Sep 25 15:06:56 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 25 Sep 2002 12:06:56 -0700
Subject: Getting started
References: <uojinmcfapjf45@corp.supernews.com> <mailman.1032442884.18747.python-list@python.org> <H3mi9.133635$pX1.4751676@news2.tin.it> <mailman.1032567390.29374.python-list@python.org> <AFik9.51323$b%3.25248@fe03> <mailman.1032979596.18965.python-list@python.org>
Message-ID: <3D920950.323B25B3@alcyone.com>

"James J. Besemer" wrote:

> However "lame" or "weak" you may consider this feature to be, it does
> NOT constitute "weak typing" per the formal definition.  This is
> simple
> polymorphism, as you seem to acknowledge, same in Perl as in Python.

The problem here is that there is no official "formal definition."  Some
use "weak typing" to mean any type system where the typing can be foiled
by casts; some use it to mean any system where operations, not the
objects themselves, determine the types of the objects.  (I favor the
latter definition, since it carries more information about the
language.)  This is simply a case of differing terminology.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nothing is true -- all is permissible.
\__/ Hassan i Sabbah
    Maths reference / http://www.alcyone.com/max/reference/maths/
 A mathematics reference.


From greek_bill at yahoo.com  Wed Sep 18 16:56:37 2002
From: greek_bill at yahoo.com (Morfeas)
Date: Wed, 18 Sep 2002 21:56:37 +0100
Subject: [newbie] Calling Python from C++
Message-ID: <8L5i9.841$Ew4.62552@newsfep2-gui>

my struggle to C++ and python to like each other continues....

I've now managed to make my C++ classes visible from python, so I can write 
python scripts that use my C++ code. great!

Now what I need to do is execute python from within C++. Reading from the 
manual this involves embedding the python interpreter into my app. That in 
itself doesn't sound too bad for simple cases. For example, I think I know 
how to call a python script to instantiate some C++ objects and return them 
to the main app.

What I want to do though, is to execute a python script that operates on an 
existing object (that may or may not have been instantiated from Python). 
So my question is : How do I pass an object into the python script?

I've seen in the manual how a python script can access the main app, but 
it's not really what I want.

I suppose a hack would be to place the object I want to operate on on some 
default position (say within a holder class), and then the python scripts 
gets that object from that holder class. Is there a clear / more elegant 
way?

I'm kind of relying on SWIG to do all the data conversion to and from 
python, so I don't really know much on this topic!

Thanks again,

Morfeas


-- 
www.freecfm.com/m/morfeas


From stephen.boulet at motorola.com  Fri Sep  6 17:51:13 2002
From: stephen.boulet at motorola.com (Stephen Boulet)
Date: Fri, 06 Sep 2002 16:51:13 -0500
Subject: py2exe and a single exe file?
Message-ID: <3D792351.3070008@motorola.com>

I'm trying to get py2exe to work for me, and I must be doing something 
wrong.

I have a script that uses easygui (which uses Tkinter) and Numeric. It 
also imports a small module from the same directory that the script 
resides in.

I try building it like this:

    python setup.py py2exe -d N:\ -i easygui

It builds and runs, but only with the other stuff in the build directory 
with the generaged exe. The other stuff is:

   tcl directory containing tcl8.3 and tk8.3 directories
   _numpy.pyd
   _sre.pyd
   _tkinter.pyd
   multiarray.pyd
   umath.pyd

What I would like to accomplish is just having a single exe file lumping 
everything together. Is this possible?

-- Stephen



From mcfletch at rogers.com  Fri Sep 27 18:20:13 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Fri, 27 Sep 2002 18:20:13 -0400
Subject: Better terminal for windows?
References: <3D94B037.2020202@motorola.com> <mailman.1033158756.25404.python-list@python.org> <3D94D1DD.40407@motorola.com>
Message-ID: <3D94D99D.9050109@rogers.com>

8<_____ setup command completion.reg ______
Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Command Processor]
"CompletionChar"=dword:00000009
"EnableExtensions"=dword:00000001
"PathCompletionChar"=dword:00000009

HTH,
Mike

Stephen Boulet wrote:
> Gerhard H?ring wrote:
...
>> cmd.exe isn't *that* bad once you activate tab completion in the
>> registry. Have you done that?
> 
> 
> Nope. How do you do that? I couldn't figure it out just looking through 
> regedit.
...
_______________________________________
   Mike C. Fletcher
   Designer, VR Plumber, Coder
   http://members.rogers.com/mcfletch/





From duncan at NOSPAMrcp.co.uk  Tue Sep 24 09:00:27 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 24 Sep 2002 13:00:27 +0000 (UTC)
Subject: list problem
References: <mailman.1032818129.8126.python-list@python.org> <mailman.1032828726.20071.python-list@python.org> <amp6i1$7n9bi$1@ID-77524.news.dfncis.de> <Xns929364D72B214duncanrcpcouk@127.0.0.1> <tCYj9.157570$pX1.5675986@news2.tin.it>
Message-ID: <Xns92938CC236BA2duncanrcpcouk@127.0.0.1>

Alex Martelli <aleax at aleax.it> wrote in 
news:tCYj9.157570$pX1.5675986 at news2.tin.it:

>> If you are going to use reduce, then you would be better off writing:
>> 
>>    import operator
>>    reduce(operator.add, x)
> 
> Maybe (marginally).  But timing things is often a good idea:

Apologies for not being clearer. I actually felt that using operator.add 
read more cleanly than using a lambda. I should have made it more obvious 
that was what I meant. I wouldn't particularly worry which is faster unless 
one was actually proven to be a bottleneck, at which point the OP might 
well find the best way to optimise would be to not end up with nested lists 
in the first place.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From peter at engcorp.com  Fri Sep  6 08:55:40 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 06 Sep 2002 08:55:40 -0400
Subject: .py to .pyc
References: <mailman.1031300768.16629.python-list@python.org>
Message-ID: <3d78a5b2$1@news.sentex.net>

andrea.baglioni at redturtle.net wrote:
> Hi everybody,
> 
> I'm a python newbie and i've a problem.
> 
> I want to compile a python file called File.py and i want to produce a file
> called File.pyc, so i do
> 
> python File.py

Actually, that won't do it.  Python only caches the bytecode in a .pyc
file for modules imported by the main routine.  Even if the above
worked properly, it still wouldn't leave you with a .pyc.  Besides,
why do you care about the .pyc anyway?  It's just an optimization
that Python even creates it for you in some cases.

That said, you should be able to get one by just running the interactive
prompt and doing "import File", if File.py has no errors.

> I always find the following error:
> 
> Traceback (most recent call last):
>   File "File.py", line 18, in ?
>     import Globals, StructuredText, string, utils, re
> ImportError: No module named Globals
> 
> Can anybody help me to find the problem? I think it's a problem of
> $PYTHONPATH and $PYTHONHOME but i'm not sure.

Why do you think that's the problem?  It seems more like that the
problem is that you are trying to import modules that are actually
part of the Zope infrastructure (if I recognize the above line),
and you'll likely have far more problems doing that than just what
you've faced so far.  Maybe some more background on what you are
trying to accomplish would help us answer more thoroughly.

-Peter



From loewis at informatik.hu-berlin.de  Mon Sep 16 11:13:24 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 16 Sep 2002 17:13:24 +0200
Subject: python performance
References: <3D84E66F.6020701@linux.ie> <zm7h9.51182$AR1.2015178@bin2.nnrp.aus1.giganews.com> <IThh9.4238$cP3.10401@news.iol.ie> <j4r8fumezm.fsf@informatik.hu-berlin.de> <uk7lmm0sj.fsf@online.no>
Message-ID: <j4n0qiklwr.fsf@informatik.hu-berlin.de>

Syver Enstad <syver-en+usenet at online.no> writes:

> What about Psyco?

Psyco does not optimize access to global variables.

Regards,
Martin


From mark at mceahern.com  Tue Sep  3 20:22:09 2002
From: mark at mceahern.com (Mark McEahern)
Date: Tue, 3 Sep 2002 19:22:09 -0500
Subject: how do i know i built python --with-threads
Message-ID: <JHEOKEOOLIGLDHCMAHMOOECLDMAA.mark@mceahern.com>

Suppose somebody else installed python on my system--e.g., my evil twin--and
I wanted to determine whether the python was built like so:

  ./configure --with-threads --without-pymalloc

How would I do it?

I'm just trying to do some sanity checking of the Python 2.1.3 I installed
for Zope.

Thanks,

// mark

-




From syver-en+usenet at online.no  Tue Sep 17 08:58:11 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Tue, 17 Sep 2002 12:58:11 GMT
Subject: threads or queue for this task
References: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com> <mailman.1032241763.6933.python-list@python.org> <d7Ah9.103593$ub2.2156613@news1.tin.it> <mailman.1032260122.29578.python-list@python.org> <6dFh9.120102$pX1.4321019@news2.tin.it>
Message-ID: <uptvc4vua.fsf@online.no>

Alex Martelli <aleax at aleax.it> writes:

> > Bounded queues are A solution to some problems but not The solution
> for
> 
> > all situations.
> 
> Of course not!  Never did I assert nor imply that you should
> make ALL of your Queue instances bounded -- that would be
> silly indeed.  But the ability to make SOME Queue's bounded,
> and still handle them without fuss, IS important in a vast
> enough number of cases that ruling this possibility out is
> a serious didactic mistake, IMHO.

When should they not be bounded? 

Would it perhaps be easier to program
with Queue.Queue's if they had a default maxsize other than infinity?
That would force the programmer to consider unblocked put's and
setting the maxsize to number reasonable for his/her application.
 

-- 

Vennlig hilsen 

Syver Enstad


From michaels at one.net  Wed Sep 11 22:08:35 2002
From: michaels at one.net (Michael Schneider)
Date: Wed, 11 Sep 2002 22:08:35 -0400
Subject: [FEEDBACK] Is this script efficient...is there a better way?
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <mailman.1031781254.1680.python-list@python.org> <HPRf9.1780$Lg2.247316@news2.news.adelphia.net>
Message-ID: <3D7FF723.3050005@one.net>


Bob X wrote:

>
>> 1) readlines() loads the entire file into a list so if you have a 30+ 
>> mb file you just ate 30+mb of memory.  Try using xreadlines() 
>> instead, it reads the file line by line and is much more memory 
>> friendly.
>
> Very cool...I had missed that! 


It is much slower though.  I am not sure about your config, but I am 
running with a Gig of memory.
I would much rather have the speed then the  30 mb of memory.  Again, 
check your config.




From jacek.generowicz at cern.ch  Thu Sep 19 09:30:30 2002
From: jacek.generowicz at cern.ch (Jacek Generowicz)
Date: 19 Sep 2002 15:30:30 +0200
Subject: strong/weak - dynamic/static [Was: Getting started]
References: <uojinmcfapjf45@corp.supernews.com>
Message-ID: <tyf8z1y5cp5.fsf@pcitapi22.cern.ch>

Sorry to go off at a tangent ...

"Lance" <lbrannma at cablespeed.com> writes:

> Hi All,
> 
> Last night I attended a Weak typing, C++ Templates, and Python talk by Bruce
> Eckel. It was great, my introduction to Python.
> 
> I'm sold on Python and weak typing. I want to write a graphics module that
> will link to a C application, permitting scatter plots, bar charts, log
> scales, etc. I suppose the Python Imaging Library will be used for this
> purpose.

A quick google search suggests that Bruce Eckel really claims that
python is weakly typed (at least more weakly typed than C++).

I've seen these claims made so often now (a quick google search
suggests that the majority of opinions in netspace favour the "Python
is weakly typed" thesis) that I am seriously beginning to doubt my
understanding of the concepts of weak, strong, dynamic and static
typing.

Could someone provide a serious reference (i.e. something which would
have gone through some sort of editorial/peer/technical review, rather
than someone's random opinions on the web) explaining or defining the
concepts of weak, strong, dynamic and static typing.

Thanks,


From erikprice at mac.com  Thu Sep  5 11:51:09 2002
From: erikprice at mac.com (Erik Price)
Date: Thu, 5 Sep 2002 11:51:09 -0400
Subject: newbie problem: use socke lib to retrieve one web page:
In-Reply-To: <czAd9.6803$yt3.3242044@newssrv26.news.prodigy.com>
Message-ID: <4BA03A21-C0E7-11D6-984F-00039351FE6A@mac.com>

On Wednesday, September 4, 2002, at 11:53  PM, koko wrote:

> I write this to retrieve one web page using socket lib.
>
> # retrieve webpage
> import socket
> HOST='www.uic.edu' # remote host
> PORT=80     # remort port
> s=socket.socket(socket.AF_INET, socket.SOCK_STREAM)
> m="""GET /home/events.shtml HTTP/1.0
> From: hh at uic.edu
> User-Agent: test/1.0
> """
> s.connect((HOST,PORT))
> s.send(m)
> data=s.recv(4096)
> print data
> s.close()

Works for me:

 >>> import socket
 >>> host = 'www.uic.edu'
 >>> port = 80
 >>> s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
 >>> s.connect((host, port))
 >>> header = """HEAD /home/events.shtml HTTP/1.0
... From: hh at uic.edu
... User-Agent: test/1.0
...
... """
 >>> s.send(header)
72
 >>> data = s.recv(4096)
 >>> print data
HTTP/1.1 200 OK
Date: Thu, 05 Sep 2002 15:40:00 GMT
Server: Apache/1.3.26 (Unix) PHP/4.1.2 mod_perl/1.27 mod_ssl/2.8.10 
OpenSSL/0.9.6
Connection: close
Content-Type: text/html


 >>> s.close()


I used the HEAD method instead of GET for brevity.  But GET works too.



Erik






--
Erik Price

email: erikprice at mac.com
jabber: erikprice at jabber.org




From jb at cascade-sys.com  Sat Sep 28 04:38:57 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sat, 28 Sep 2002 01:38:57 -0700
Subject: How do I set-up Multiple socket connections
References: <002a01c26604$b515c3c0$0100a8c0@mshome.net>
Message-ID: <3D956AA1.5070806@cascade-sys.com>

Kat & Huw wrote:

> Ok, I want to iterate through a server IP list and establish
> multiple socket connections to the various IP address' in the list.
>  
> I can do the iterations and I can make single connections, but I can't
> work out how I can connect to several address at once.
>  
> I understand that a socket is closed, or closes after its done getting 
> the ACk or
> whatever after sending a request, but how on earth do I connect to like 5
> addresses at once?

This is a complicated problem.

Two general solutions suggest themselves.

A. You can use the select system call to wait on multiple socket I/O 
operations at once.  It waits until one is ready to proceed and it tells 
you which one(s) are ready.  Read the doc.  You would have some 
non-trivial state machine logic to go with it to translate the select 
results into the various i/o functions on individual sockets.  On Unix, 
select will multiplex socket AND regular I/O.  On Windows select only 
works for sockets.

B. You can dispatch multiple threads, one for each host you wish to 
communicate with.  Each thread would do the initial socket() and 
connect() to connect to the remote host and then send() or recv() as 
appropriate.  Then it's up to you to coordinate the results.  If there 
is no coordination to do, e.g., if the results all go to different 
files, then you can skip that part.  If you do need to coordinate 
results, e.g., to funnel all results to a single sink, the Queue class 
likely will be useful.

My preference is alternative B and I have used it successfully on a 
number of projects on Linux.  I have heard rumors that threads and 
socket I/O is problematic on some platforms but I don't know which ones.

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From glingl at aon.at  Thu Sep  5 17:02:39 2002
From: glingl at aon.at (Gregor Lingl)
Date: Thu, 05 Sep 2002 23:02:39 +0200
Subject: German Umlaut Problem with raw_input in IDLE
Message-ID: <3D77C66F.4040305@aon.at>

Hi

This message concerns using IDLE with German
alphabet.

Some month ago I copied sitecustomize.py with
appropriate locale-setting to C:\Python22\Lib
thanks to an advice from some other mailing list.

This worked well until today I made the observation,
that in IDLE raw_input doesn't work correctly:

 >>> a = raw_input("H?? ")
H?? Tsch?s
Traceback (most recent call last):
  File "<pyshell#2>", line 1, in ?
    a = raw_input("H?? ")
TypeError: object.readline() returned non-string
 >>>

Printing H?? works correctly, reading in Tsch?s doesn't!

Consequently I cannot execute programs with
raw_input from IDLE, if the expected answers (inputs)
may contain "Umlaute". Those same programs,
however, can be executed from the commandline.
And also the interactive Python interpreter can execute
raw_input statements even when Umlaute are input.

So I think, this must be a problem of IDLE.
I must confess, I do not understand the error message.

Does somebody know a patch or a workaround?
Thanks

Gregor Lingl



From aleax at aleax.it  Mon Sep 16 10:07:30 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 16 Sep 2002 14:07:30 GMT
Subject: threads or queue for this task
References: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com> <cqp7ou8d0jjcdtrqetu9v9ls9gst2cp4qi@4ax.com> <mailman.1032076038.5654.python-list@python.org> <h2jbouo4j9f8ee646bhfejt3erqku76d61@4ax.com>
Message-ID: <CAlh9.101156$ub2.2068257@news1.tin.it>

robin at execulink.com wrote:
    ...
> For example, I really don't know what "proper design" I should use in
> this case. Any sample code?

IMHO, proper design of a multithreaded app in Python almost invariably
uses instances of Queue to for inter-thread communication.  I do not
understand all the details of your requirements, but, roughly speaking,
I gather: you have some dedicated working threads, each of which must
repeatedly run the same function on all "work-units" sutable for that
thread; and another thread that is able to determine what work units
are suitable for what other threads.  In this case, you can have a
separate Queue for each dedicated thread.  Each such thread peels the
next work-unit destined for it from its Queue, performs the work, then
goes back to peeling.  The supervisor thread just posts work units to
the appropriate queues.

If you can post more precise/accurate/detailed specs, I'll be glad
to try to help in more detail.

> The interface to Queue is deceptively simple. Are there constructs to
> avoid or performance trade-offs to different methods? Again, any
> explicit code (or at least explicit help as to which methods are
> preferable) would be appreciated.

The no-wait methods, or explicit timeouts on the normal methods that
do wait, are often symptoms that you're polling.  IF you can avoid
polling (that's not always possible, alas), your performance will be
better: a sleeping thread consumes close to no resources, one that is
polling consumes resources in rough proportion to the frequency with
which it's polling (and suffers response delays in inverse proportion
to it).


> I am also put off by the description of the Queue Object, particularly
> methods like empty() which says: "Return 1 if the queue is empty, 0
> otherwise. Because of multithreading semantics, this is not reliable."
> 
> If it's not reliable, then what's the point of it existing? Why is it

Because some people like "purely indicative" methods to exist.

> not reliable? (I imagine simply because the thread doing the checking
> could be superceded by the thread that implements the Queue in the
> time it takes for the method to return and be acted upon.) Is it just

Of course.  Basically, when method empty returns 1, all you know is
"there was an instant in the reasonably recent past in which this
queue instance was empty" -- this, of course, has nothing to do with
whether the queue is empty now.  Similarly, when empty returns 0, then
there was an instant in which the queue was non-empty, but, again,
this doesn't necessarily matter much.  Things may be a bit better
when your overall architecture is such that only one thread ever
posts to a given Queue, and/or only one thread ever gets stuff from
it: in such case, you may be able to prove stronger and more useful
statements.  For example, if the thread calling the empy method is
the only thread that ever posts stuff to the Queue, then if the Queue
was empty when empty() was running, and the thread knows it has
posted nothing more, then the Queue is still empty.

> in certain cases that it's not reliable? How can one code if you don't
> reliably know when a Queue is empty or full?

By just trying to post, and/or to get, stuff from the Queue, and
sleeping until the action can be performed; or, polling issues
apart, from performing e.g. a get-nowait operation in a try/except and
treating the exception, if it comes, as a VERY reliable indication
that, yes, the Queue WAS indeed empty at the time we tried to peel
an entry off it.


Alex



From jknapka at earthlink.net  Wed Sep 11 13:45:23 2002
From: jknapka at earthlink.net (Joseph A. Knapka)
Date: Wed, 11 Sep 2002 17:45:23 GMT
Subject: Variables vs. names
Message-ID: <3D7F8132.2D5CEC24@earthlink.net>

Hi folks,

I've been using Python for some time now, and I am
completely cognizant of and comfortable with its
semantics. However, I'm curious about something:

I understand that a Python "variable" is really
just a dictionary key; fine. In other languages,
instead of "naming" a dictionary entry, a
variable might directly "name" a chunk of
storage. So other than the layer of indirection
introduced by the dictionary mapping, what is
the practical significance of "names bound to values"
vs "variables with values"? Are there languages that
achieve Python's semantics using "real"
variables? Java's semantics seem pretty much
identical to Python's in this respect, yet
nobody in the Java community will object if
you call a thing-with-a-name-and-a-value a
"variable".

(And does an instance of a class with __slots__
have "names bound to values" for its members, or does
it have actual "variables"?)

Just curious,

-- Joe
  "I'd rather chew my leg off than maintain Java code, which
   sucks, 'cause I have a lot of Java code to maintain and
   the leg surgery is starting to get expensive." - Me


From SBrunning at trisystems.co.uk  Wed Sep  4 10:53:34 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 4 Sep 2002 15:53:34 +0100 
Subject: How to force built-in commands over imported.
Message-ID: <31575A892FF6D1118F5800600846864DCBD721@intrepid>

> From:	Maurice van de Rijzen [SMTP:maurice at akst.tn.tudelft.nl]
> There is a built-in commans called open(fileName). There is also a 
> open() command in the os-module open which needs more than the fileName. 
>   Originally I didn't import the os-module and everything worked fine. 
> However, when I imported the os-module I received an error that open 
> needed more arguments. How can I force to take the built-in open() 
> instead of the open() in the os-module.
 
You are, I assume, importing 'os' as follows:

from os import *

*Don't* *do* *this*. Use:

import os

instead. There *are* reasons for using the former method, but they are few
and far between. The latter form should always be the one you use unless you
have a good reason.

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




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



From wilk-spamout at flibuste.net  Mon Sep 30 15:35:12 2002
From: wilk-spamout at flibuste.net (William)
Date: 30 Sep 2002 21:35:12 +0200
Subject: redhat8
References: <873crr4fy6.fsf@flibuste.net> <mailman.1033404219.15052.python-list@python.org>
Message-ID: <87ofafxocv.fsf@flibuste.net>

Anthony Tuininga <anthony at computronix.com> writes:

> According to the release documentation:
> 
> Python 2.2.1 is on the CD.

great, but i would like to know if it's the default version, linked with
/usr/bin/python...

-- 
William Dod? - flibuste.net
http://wikipython.tuxfamily.org


From gerhard.haering at gmx.de  Fri Sep 27 18:20:29 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sat, 28 Sep 2002 00:20:29 +0200
Subject: Is Boost Python version 2 ready?
In-Reply-To: <6D4l9.12184$Lg2.3654433@news2.news.adelphia.net>
References: <MvZk9.12124$Lg2.3584877@news2.news.adelphia.net> <CG2l9.263192$216.10040127@bin4.nnrp.aus1.giganews.com> <6D4l9.12184$Lg2.3654433@news2.news.adelphia.net>
Message-ID: <20020927222029.GB7781@lilith.ghaering.test>

* Bob X <bobx at linuxmail.org> [2002-09-27 22:05 +0000]:
> Is there a page up saying what 2.2.2 will give us?

Bugfixes only. There's a wiki page, probably mostly interesting for
Python committers: 

http://manatee.mojam.com/py222wiki/

-- Gerhard



From dubal at khimjiramdas.com  Thu Sep  5 02:55:53 2002
From: dubal at khimjiramdas.com (J Dubal)
Date: 4 Sep 2002 23:55:53 -0700
Subject: How to install informixdb-1.3 for python2.2 on RH7.3?
Message-ID: <c345ff6f.0209042255.65dd5824@posting.google.com>

Hello good people!

As mentioned in its readme, informixdb-1.3 compiles, installs and
works with the stock python 1.5.2 of RH7.3. However, when we try to
compile it for python 2.2.1, informixdb's ./configure fails because it
is looking for Makefile.pre.in which is not there in python 2.2.1. It
appears that someone has to repackage informixdb with distutils. How
to do this? In the meanwhile can we install this manually? I do not
understand Makefile.pre.in. Please help. Thanks in advance.
Dubal.


From jb at cascade-sys.com  Tue Sep 17 02:54:15 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 16 Sep 2002 23:54:15 -0700
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <pUdh9.9377$yt3.4604908@newssrv26.news.prodigy.com> <mailman.1032184821.5180.python-list@python.org> <c%qh9.102794$ub2.2120685@news1.tin.it> <mailman.1032212180.11973.python-list@python.org> <3d868ccc$1@news.sentex.net> <mailman.1032240440.21101.python-list@python.org> <7xfzw9qgo7.fsf@ruckus.brouhaha.com>
Message-ID: <3D86D197.9050806@cascade-sys.com>

Paul Rubin wrote:

>"James J. Besemer" <jb at cascade-sys.com> writes:
>  
>
>>Furthermore, even if you formally assign the copyright to somebody
>>else, I understand ownership eventually reverts to the original author
>>after a lengthy period (45 years or something).  Evidently it
>>preserves the copyright for the heirs of the author.
>>    
>>
>
>I heard the opposite.  In fact part of the debate before the 1976
>

That was over 25 years ago and a considerable amount of copyright 
legislation has been passed since.

My above blanket statement is not 100% correct.  It IS possible to 
permanently transfer rights to a copyrighted work.  However, the 
contract has to be "just so" to meet all of the provisions of the law. 
If the contract does not meet all of the statutory criteria, then the 
original author retains the right to terminate transfer of the copyright 
after 35 years (though it's not automatic, as I implied).

I dunno when the various laws came into effect (mid- to late-80's I'm 
guessing) but since before I started my own company in 1994, lawyers 
were very particular about the wording of contracts where copyrightable 
works were being created.  

Anyway, here's a document outlining the legal hoops relevant to 
retaining or transferring copyrights per current US legislation:

    http://tenonline.org/art/ip1/0001.html

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From jb at cascade-sys.com  Wed Sep 25 14:45:32 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Wed, 25 Sep 2002 11:45:32 -0700
Subject: Getting started
References: <uojinmcfapjf45@corp.supernews.com> <mailman.1032442884.18747.python-list@python.org> <H3mi9.133635$pX1.4751676@news2.tin.it> <mailman.1032567390.29374.python-list@python.org> <AFik9.51323$b%3.25248@fe03>
Message-ID: <3D92044C.9050801@cascade-sys.com>

   
Steve Holden wrote:

>"James J. Besemer" <jb at cascade-sys.com> wrote in message
>news:mailman.1032567390.29374.python-list at python.org...
>  
>
>>Alex Martelli wrote:
>>
>>>$a = "2.3";
>>>$b = 45;
>>>print $a+$b;
>>>
>>>emitting 47.3 is pretty close to what I consider "weak typing".
>>>
>>I don't see why you call this "weak typing".  In Perl, the "+" operator
>>is defined to work with strings or numbers.
>>    
>>
>
>Well, it seems pretty weak to me that the *operator* has to examine the
>types of its arguments in order to know which operation to perform. Can we
>say polymorphism?
>

However "lame" or "weak" you may consider this feature to be, it does 
NOT constitute "weak typing" per the formal definition.  This is simple 
polymorphism, as you seem to acknowledge, same in Perl as in Python.

The proper definition of "weak typing" is that in an otherwise strong 
typing system there are ways to cheat the system and apply illegal 
operations to certain types.  The quintessential example is in C++, 
where, despite a fairly robust and secure type system, you always can 
cast objects to the wrong type and force illegal operations on them.   
This is in contrast to Strong typing, where such "type errors" are 
strictly prohibited.  These type errors are not possible in Python nor 
in Perl.  Thus neither is "weakly typed" in the formal sense.  They're 
both strongly, dynamically typed.  

You say "the operator has to examine the types..."  as if this is 
somehow different from Python.  But, e.g., Python's "*" operator has to 
examine it's arguments in order to know which operation to perform, does 
it not?   You don't like to think of it that way but if the notion 
applies to Perl then it applies equally well to Python..  

In FACT, if you examine the source code, this is EXACTLY how both 
languages are implemented.  There is but one each binary "+" and "*" 
operators in Python's byte codes.  At runtime, the interpreter takes the 
operator and then LITERALLY "examines the types of it's arguments (on 
the top of stack) in order to know which operation to perform".  E.g., 
binary_multiply "looks" at it's arguments to decide if  it's to perform 
a string operation, a list operation or a numeric computation. 
 Similarly, the binary subscript has a special test inline if the left 
argument is a List and the right one is an Integer, in which case it 
directly executes PyList_GET_ITEM(); otherwise, it calls the more 
generic PyObject_GetItem(), which in turn continues the decision tree 
down for each valid combination of operand types, ultimately throwing an 
exception if the combination is invalid.

Now that's how it WORKS.  Underneath the hood, Python is implemented in 
flat C code that emulates objects.  The code "switches" on operators 
that in turn inspect object types to decide what specific operation to 
perform.  Now, in principle, you also can "say" that operations are 
defined on a per object basis, since that is the abstraction ultimately 
provided and because we all prefer to view the universe from that 
perspective.  But in any case, the implementation and the fundamental 
abstraction are the same for both Python and Perl.  You can call it OO 
or you can call it operator-inspection.  The difference is merely your 
point of view.

The only real difference between the two languages is the particular 
choice of operators and valid operands  -- and arguably the relative 
wisdom of those choices.

>>This all is no different in principle from Python's:
>>
>>    a = 2.3
>>    b = 45L
>>    print a + b
>>
>>and getting 47.3.  The "+" operator is defined for floats and also for
>>longs.  Since they're all "numbers" it would seem really strange if they
>>did not interact naturally like this.  But fundamentally they're
>>distinct types as "type( 2.3 )" and "type( 45L )" illustrate.
>>
>>    
>>
>It seems very different to me. 
>

I agree "+" is too subtle an example.  "*" is better.

>For example, how does the Perl "+" operator
>find out about a new addable type? 
>

I don't know.  Look it up and you tell me.   In Perl, the greater number 
of builtins are keyword based and Perl classes definitely CAN 
override/overload builtin keywords.

>Is Perl even able to integrate its
>intrinsic data objects with user-defined types? 
>

No.  I don't think so.  Perl's OOP features have lagged behind Python's. 
 And in all fairness, it's a new feature to Python.

>Can we say inheritance hierarchy? 
>

Yes, Perl's OO mechanism includes inheritance.

>In Python it's easy to define methods to integrate your own
>classes into the language's architecture. I could never discern enough
>architecture in Perl to see how to do that, though I willingly accept that
>this deficiency might be in me rather than in Perl.
>

Certainly, Perl's object model is not as good as Python's.  At best, 
it's more like Python's was before type/object unification (and it's not 
even that good).  Nevertheless, the Quality of the object model (and the 
questions you raise above) is IMMATERIAL to my original claim, regarding 
weak vs. strong typing (in the formal sense).  In fact, weak vs. strong 
typing is a distinction that applies equally well to languages without 
any OOP support at all.  It has to do with "type safety", not the 
OOP-ed-ness of the language.  It's not a sliding scale, 0 to 100%; it's 
binary, black vs. white.  Can the user violate the type rules of the 
language (whatever they may be), yes or no?  If yes then Weak; if not 
then Strong.  Perl, like Python, protects its objects with dynamic 
runtime checks, thus it is strong not weak.

I think this is the KEY misunderstanding when you and others attempt to 
argue with me.  Read Cardelli's paper.  Either you concede this point 
and thus my original premise, or else you're simply wrong and I'll have 
done all I can to set you straight.


>>Perhaps an even better example is the "*" operator:
>>
>Not really. Python has strongly-typed data with late binding of methods and
>operators. This is an organised way to obtain the results you have so far
>quoted.
>

Perl and Python work exactly the same way in this regard.  

The confusion is that Perl defines many more primitive operations and 
defines meaning for many operator/type combinations where Python does 
not.  Now you can argue that Perl is flawed in this way but in principle 
there is no difference in how the two work.  Both are absolutely 
deterministic in what specific operation is applied to what object in 
every legal context.  Only Perl defines extra ones, such as arithmetic 
on strings, that Python does not.

One way to think about Perl's "+" and "." operators formally is that 
there actually are several of them, all producing similar (but subtly 
different) results.  

    string + string -> number
    string + number -> number
    number + string -> number
    number + number -> number
    string . string -> string
    string . number -> string
    number . string -> string
    number . number -> string
    list + list -> list
    string * number -> string
    list * number -> list

    (many others)

In principle, it's no different from Python (except in choice of operators):

    string + string -> string
    number + number -> number
    list + list -> list
    number * number -> number
    string * number -> string
    list * number -> list

    (many others)

Fundamentally it's just polymorphism.  This way, you don't have to think 
about the "operators making a decision," if you don't like.   Whether 
the operator or the object "decides" really is an implementation detail. 
 The abstraction is that there exists a thorough enumeration of  valid 
combinations of operations and object types that maps onto a set of 
operations.  In this fundamental sense, the languages are the same.

Of course, each may have some back-door C module that implements "peek" 
and "poke" or some mechanism that does in fact allow type safety to be 
violated.  But I'm intentionally disregarding limit cases such as that; 
certainly they do not constitute the overall intent of either language.


>>Anyway, why should "2" + 5 == 7 be all that much stranger than "2" * 5
>>== "22222"?  If the operation is unambiguous and useful what rule would
>>it break?  Personally, I think Python would be improved slightly by
>>adding this conversion.  It's convenient and there's ample precedent in
>>other languages (Awk, Snobol, Icon, Perl, etc.).  But I am sure that
>>notion will provoke howls of protest from the True Believers.
>>
>Not at all. If I thought such a suggestion would be taken seriously I might
>go so far as to advise against that, but I don't see this being a religious
>debate.
>

Many other languages have worked this way.  And it certainly has some 
utility.  If you don't like it, fine, but there's ample precedent by 
smarter inventors than you and me that indicate it's not a crazy idea.

Anyway, I'm not arguing for the feature's inclusion into Python.  My 
only point is it does not make Perl "weakly typed" in the formal sense.


>For some meaning of strong, I suppose, but please don't bother to define
>what this is.
>

EEEK!  Sarcasm.  ;o)

These terms has been bandied about so much on this list that I thought 
it was generally understood.  Geeze, they're part of the standard 
mantra.  And I didn't start the discussion; I only took issue with 
Alex's misclassification of Perl.  Curiously, this thread reveals that a 
lot of people don't have the first idea what these terms mean..

Anyway, I did give a clear definition at the start of this post, with 
references to back it up.  Also, if you do a google search, you'll find 
a bunch of hits that are consistent with my definition (along with some 
that equate "weak typing" with "dynamic typing", which I'm sure 
everybody here will agree is practically the Opposite of what we all 
mean (perhaps this is why we're confused ourselves)).  

Equates weak and dynamic [Bogus]

   http://www.lispme.de/lispme/doc/lm_sema.htm
   http://www.perlmonth.com/perlmonth/issue5/modules.html
   
http://www.usenix.org/publications/library/proceedings/%20tcl98/full_papers/howlett/howlett.pdf 

   http://www.ruby-lang.org/en/testimony.html

Strong is Safe vs. Weak is Unsafe (specifically preventing vs. allowing 
"type errors"):

   
http://www-lp.doc.ic.ac.uk/UserPages/staff/ft/alp/net/typing/strong.html 
(weak is unsafe)
   http://www.cise.ufl.edu/~jnw/OOCourse-95/Lectures/Jan.31.html (strong 
is safe)
   http://www.cs.auc.dk/~normark/prog3-01/html/notes/fu-intr-1-book.html 
(lead to errors)
   
http://216.239.39.100/search?q=cache:oBq2r3PwR1kC:www.cs.cornell.edu/courses/cs412/2002sp/lectures/lec12.ps+definition+weak+typing&hl=en&ie=UTF- 



Weak == otherwise Strong + Means to override e.g., C:

   http://www.wkonline.com/d/weak_typing.html
   http://www.wkonline.com/d/strong_typing.html ("strict enforcement")

"Strong types for weak minds" [a cute quote I thought y'all'd enjoy] -- 
strong typing is either overly restrictive (Pascal) or allows overrides 
(C) [which is definition of] "weak typing".

   http://people.cs.uct.ac.za/~gaz/teach/cpl/lect2.ppt


Read and understand.

>Well, considering that the %r and %s conversion types are currently
>documented as
>
>"""
>r String (converts any python object using repr()).
>s String (converts any python object using str()).
>"""
>
>it should hardly be surprising that conversions take place. 
>

Perl documents the (surprising to you) behavior of "+".  If it's working 
as advertised it's standard behavior cannot be considered a "type error" 
and thus it's not a sign of weak typing.

>Such
>conversions, however, are not intrinsic to the *language*, but to the *data*
>(where the type is significant) - Python goes so far as to expose explicitly
>how the conversions are performed, and they will clearly work for any object
>with a suitable (returns a string) __repr__() and/or __str__() method. 
>

True but irrelevant to strong vs. weak typing.

>Just
>to clarify the nature of this, try
>
>    "%6.2f" % 2.6
>
>and then try
>
>    "%6.2f" % "2.6"
>
>and we'll see if you still feel that string formatting 'treats numbers as
>strings in a string "context"'.
>

I said it treats numbers as strings, not the reverse.  I'll let this go, 
though, as I am NOT advocating such a change to Python.

>By contrast Perl seems a little muddled, and I personally think it's no
>accident that Wall's own writing tends to stress the significance of
>*context*. The rules are quite explicit in Python, clouded in Perl. Of
>course you must take this as a dim memory frmo one who never claimed to
>understand Perl that well in the first place (I used to program Perl, but
>I'm all right now ;-).
>

I agree Perl is muddled.  It's a pile of crap.  It's a hack on a hack, 
raised to the 1e5th power.  Perl's notion of scalar vs. array "context" 
is the worst idea ever.  Blah blah blah.  I agree it's a crock.  But 
that does not make the language "weakly typed".  It's strongly typed, 
like Python, as it does not allow "type errors."

>It's also possible to enjoy some of the benefits of prison by simply
>refusing to go outside. Your very confession of C's and C++'s weaknesses
>highlights the fact that even "strongly typed" languages have problems
>enforcing the rules (whose value I anyway challenge except for the specific
>promotion of execution efficiency). Python is just a bit less anal :-)
>
These deficiencies of C and C++ make them "weakly typed" by definition. 
 They're good examples while Perl is not one.  Yes, otherwise "strongly 
typed" languages are instead weak if they permit cheating.  Pascal would 
be strong if a few tiny cracks were plugged.

I had an even longer reply to this thread that was rejected for being 
too long.  I think this reply sums things up better in less space.

Anyway, I'm just picking a nit.  Perl is inferior to Python in just 
about every conceivable manner -- except this one, highly technical sense.

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From chris.gonnerman at newcenturycomputers.net  Wed Sep  4 23:46:08 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Wed, 4 Sep 2002 22:46:08 -0500
Subject: heeelp
References: <001f01c25357$53cb7ca0$2500a8c0@maxitec.co.za>
Message-ID: <00b301c2548e$c9863420$0101010a@local>

First of all, Morne, you should NEVER EVER post to public
mailing lists in HTML format; or to newsgroups either.

Bad poster, no biscuit.

Second, don't hammer us.  You posted here at least twice,
and I read another poster stating you have posted this in
other newsgroups as well.

ALSO, you didn't indicate your Python version or your OS.
You MUST do this, for any but the most general questions.

Now that I've busted your chops, here's the answer to your
question (see attachment).  I'm not sure where your ppp.log
came from (i.e. what OS) so I guessed that a connection
where an IP was not assigned was an error.  If the criteria
for an error is different you might explain it.  Your code
is looking for a string of "ERROR" which did not appear in
your sample.

Here's what my code produced on my system:

----------------------------------------------------------------------------
--
Jul  5 18:45  Connected to Internet (IP 155.239.150.146 )
Jul  5 18:48  Closed connection, connect time = 197 secs
              (5091 octets in, 1864 octets out)
----------------------------------------------------------------------------
--
Successful Connections:        1
Failed Connections:            0
Total Time Online:           197 seconds
Total Data In:              5091 octets
Total Data Out:             1864 octets

Hope this helps.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net

> ----- Original Message -----
> From: Morne
>
> Could someone please help me with this?
> Im really new to any kind of programming
>
> If posible would you be
> able to answer in detail.
> Thanx
> morne
>
> is it wrong of me to send the ppp.log file to just anyone?
>
> HERE IS WHAT I HAVE:
>

--- code sample rudely hacked away here ---

>
> THIS IS WHAT I NEED THE PROGRAM TO DO:
>
> I want the program to read the attached ppp.log file and give me the
> following type of report:
>
>
> PPP Log file summary for 5 Jul 2002
> -----------------------------------
> Jul  5 18:15:01 Connected to internet (IP 155.239.150.146)
> Jul  5 18:18:08 Closed connection. ( Connect time: 197 secs: 5091 octets
in,
> 1864 octets out )
> -----------------------------------
> Successful connections: 5
> Failed connections: 2
> Total time online: 456 seconds
> Total data in: 66576 octets
> Total Data out: 66349 octets


-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: logproc.py
URL: <http://mail.python.org/pipermail/python-list/attachments/20020904/90b16e37/attachment.ksh>

From imbosol at vt.edu  Tue Sep  3 14:13:32 2002
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 3 Sep 2002 14:13:32 -0400
Subject: Would you like a Snobol pattern matcher? (Re: Making regex suck less)
References: <3d725881.345921@news.t-online.de> <un7n06qrl1gm93@news.supernews.com> <3D740DFC.1040009@something.invalid> <un9bkdrqeh305c@news.supernews.com>
Message-ID: <al30tq$aud$1@solaris.cc.vt.edu>

John Roth wrote:
> I don't know much about Snobol, unfortunately.
> I think my biggest issue here is that we shouldn't
> reinvent the wheel unless there is a good reason.
> In other words, Larry is taking Perl in a specific
> direction. Assuming we want to make major
> changes to regex,

I don't think anyone wants to make changes to regular expressions, per
se.  I think this discussion is about adding a higher level syntax to
regular expressions, that makes them more readable and versatile, for
cases where that's important.


> is there any _good_ reason
> for doing something conceptually different and
> consequently adding to the cacophony?

I think there is a good reason to have a higher level syntax.  It
serves a purpose heretofore unserved: that complex regular expressions
no longer have to be unbearably unreadable.

I could live without it, though.


-- 
CARL BANKS
http://www.aerojockey.com


From mhammond at skippinet.com.au  Fri Sep  6 23:10:57 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 07 Sep 2002 03:10:57 GMT
Subject: Win32 question: Problem with NetLocalGroupDelMembers
References: <3D78A350.4010006@btk.utu.fi>
Message-ID: <57ee9.1357$_l1.3640@news-server.bigpond.net.au>

Marten Hedman wrote:
> Hi,
> 
> Im writing a script for administrating local user groups on a remote 
> WinNT computer, using Python 2.2.1 and the Win32 extensions, build 148. 
> Getting information about users in a group and adding users to a group 
> works fine, but when I try to delete users from a group I get an 
> exception, 'TypeError: The object can not be converted to a Unicode 
> object'. An interactive example session looks like this:

This is a bug in the documentation.

>  >>> win32net.NetLocalGroupDelMembers(server, group, mem)
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> TypeError: The object can not be converted to a Unicode object

This function expects a list of string usernames, *not* 
PyLOCALGROUP_INFO structures.  Thus, the following will work:

names = [ m['domainandname'] for m in mem ]
win32net.NetLocalGroupDelMembers(server, group, names)

I added the following function to win32netdemo.py.

Mark.


def LocalGroup(uname=None):
     "Creates a local group, adds some members, deletes them, then 
removes the group"
     level = 3
     if uname is None: uname=win32api.GetUserName()
     if uname.find("\\")<0:
         uname = win32api.GetDomainName() + "\\" + uname
     group = 'python_test_group'
     # delete the group if it already exists
     try:
         win32net.NetLocalGroupDel(server, group)
         print "WARNING: existing local group '%s' has been deleted."
     except win32net.error:
         pass
     group_data = {'name': group}
     win32net.NetLocalGroupAdd(server, 1, group_data)
     try:
         u={'domainandname': uname}
         win32net.NetLocalGroupAddMembers(server, group, level, [u])
         mem, tot, res = win32net.NetLocalGroupGetMembers(server, group, 
level)
         print "members are", mem
         if mem[0]['domainandname'] != uname:
             print "ERROR: LocalGroup just added %s, but members are %r" 
% (uname, mem)
         # Convert the list of dicts to a list of strings.
         win32net.NetLocalGroupDelMembers(server, group, 
[m['domainandname'] for m in mem])
     finally:
         win32net.NetLocalGroupDel(server, group)
     print "Created a local group, added and removed members, then 
deleted the group"



From jwold at min.midco.net  Sun Sep 22 20:47:36 2002
From: jwold at min.midco.net (Jarrett Wold)
Date: Sun, 22 Sep 2002 19:47:36 -0500
Subject: PyXML Building on OS X?
Message-ID: <B9B3CED8.3B9%jwold@min.midco.net>

  I downloaded PyXML source distribution 0.8.1 and attempted to build it
using Python 2.2 distribution from Fink (fink.sourceforge.net).  Needless to
say it failed.  Any ideas?  I dug around Google Groups attempting to find an
answer and found a message from September 2000 that briefly touched on the
desire to support OS X.

  If you want some sample output of the build I can certainly rerun it and
post it.

Jarrett L Wold



From fetus at AUTO  Thu Sep 26 16:00:24 2002
From: fetus at AUTO (A. Dimsdale Fetus A. Moser)
Date: Thu, 26 Sep 2002 20:00:24 GMT
Subject: Pythonwin (win32all build 148) on Python 2.2.1 on Win2K, bogus syntax e
References: <u47k9.46401$V7.11091779@twister.socal.rr.com> <CJ8k9.16479$6g7.52569@news-server.bigpond.net.au>
Message-ID: <sHJk9.66556$U7.23747920@twister.socal.rr.com>

"Neil Hodgson" <nhodgson at bigpond.net.au> writes:

>Aaron Dimsdale and Alex Moser:

>> Pythonwin barfed on the quote marks marked by carets, returning: "Failed
>to run script -
>> syntax error - invalid syntax".

>   Try turning on View | End-of-line markers to see if you have non-standard
>line endings such as pure carriage returns, which cause just this sort of
>error. On Windows, you want line endings to be carriage return + line feed.

>   Neil

Well, I did that, and that seems to be the problem. Is there a Pythonwin
setting I can set to make it do standard line endings? Every line except for
whitespace lines seems to end in just a carriage return. How do I fix this?

Thanks.

--Aaron L. Dimsdale



From andreas at andreas-jung.com  Fri Sep 20 11:45:19 2002
From: andreas at andreas-jung.com (Andreas Jung)
Date: Fri, 20 Sep 2002 16:45:19 +0100
Subject: bit operations ?
In-Reply-To: <amfbuq$56f28$1@ID-146704.news.dfncis.de>
References: <amfbuq$56f28$1@ID-146704.news.dfncis.de>
Message-ID: <727165.1032540319@[10.10.1.2]>

You should check this with the Python documentation
See '|', "&" operators.

-aj

--On Freitag, 20. September 2002 16:40 +0200 Shagshag13 
<shagshag13 at yahoo.fr> wrote:

> hello,
>
> i'm looking for any libs which could handle bit operations on integer.
> (i must be able to get 0 to n bits from any int., only bit j, etc.)
>
> thanks in advance,
> s13.
>
>
>
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From kalle at lysator.liu.se  Mon Sep  2 03:50:40 2002
From: kalle at lysator.liu.se (Kalle Svensson)
Date: Mon, 2 Sep 2002 09:50:40 +0200
Subject: ascii value
In-Reply-To: <akv4fn$eo2$03$1@news.t-online.com>
References: <akv4fn$eo2$03$1@news.t-online.com>
Message-ID: <20020902075039.GA2776@i92.ryd.student.liu.se>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

[Axel Grune]
> How can I get the ascii value of char which is in a string,
> e.g. getting the 97 if the char was 'a'.

>>> ord('a')
97

Peace,
  Kalle
- -- 
Kalle Svensson, http://www.juckapan.org/~kalle/
Student, root and saint in the Church of Emacs.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)
Comment: Processed by Mailcrypt 3.5.6 <http://mailcrypt.sourceforge.net/>

iD8DBQE9cxg7dNeA1787sd0RAsgUAJ9nPP6vQWq1N5tCn+BuZBuV2eMVuACfbl3l
P5Zo5MqtM2jCf+AU/aCWBJo=
=cIb4
-----END PGP SIGNATURE-----



From tismer at tismer.com  Fri Sep 27 22:29:50 2002
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 27 Sep 2002 19:29:50 -0700
Subject: Python threading?
References: <f9Pi9.10275$Lg2.2191294@news2.news.adelphia.net> <amogmi$r1l$1@panix1.panix.com> <T5Rj9.13058$6g7.39977@news-server.bigpond.net.au> <7xfzw0avrg.fsf@ruckus.brouhaha.com>
Message-ID: <3D95141E.6080302@tismer.com>

Paul Rubin wrote:
> Mark Hammond <mhammond at skippinet.com.au> writes:
> 
>>A "one thread per connection" model is doomed from the start -
>>therefore, in most applications, I would consider 5 threads a large
>>number - particularly when all are actively doing something.  OTOH, if
>>you are willing to accept that your first version will accept a max of
>>~20 connections, the brain-power saved by using threads beats the
>>additional OS overhead by a mile ;)
> 
> 
> The new Linux native thread implementation was recently benchmarked
> running 100,000 simultaneous threads on a 1 GB machine.  Starting all
> the threads and closing them took around 2 seconds.  

Not too bad.
How about 1,000,000 in 100 MB?
Started and shut down in 15 seconds, from Python?

This was just a rough test of new features, will
give exact timings when I write the article...

ciao - chris
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From gerhard.haering at gmx.de  Sun Sep  8 13:29:43 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sun, 8 Sep 2002 19:29:43 +0200
Subject: [ANN] pyPgSQL 2.2 released
Message-ID: <20020908172943.GA624@lilith.ghaering.test>

Announce: pyPgSQL - Version 2.2 is released.
===========================================================================

pyPgSQL v2.2 has been released. It is a bug fix release to version 2.1.

It is available at http://pypgsql.sourceforge.net.

pyPgSQL is a package of two (2) modules that provide a Python DB-API 2.0
compliant interface to PostgreSQL databases. The first module, libpq,
exports the PostgreSQL C API to Python. This module is written in C and
can be compiled into Python or can be dynamically loaded on demand. The
second module, PgSQL, provides the DB-API 2.0 compliant interface and
support for various PostgreSQL data types, such as INT8, NUMERIC, MONEY,
BOOL, ARRAYS, etc. This module is written in Python and works with
PostgreSQL 7.0 or later and Python 2.0 or later.

Note: It is highly recommended that you use PostgreSQL 7.1 or later and
Python 2.1 or later.

PostgreSQL is a sophisticated Object-Relational DBMS, supporting almost all
SQL constructs, including sub-selects, transactions, and user-defined types
and functions. It is the most advanced open-source database available any-
where More information about PostgreSQL can be found at the PostgreSQL home
page at http://www.postgresql.org.

Python is an interpreted, interactive, object-oriented programming lang-
uage. It combines remarkable power with very clear syntax. It has mod-
ules, classes, exceptions, very high level dynamic data types, and dynamic
typing. There are interfaces to many system calls and libraries, as well
as to various windowing systems (X11, Motif, Tk, Mac, MFC). New builtin
modules are easily written in C or C++. Python is also usable as an exten-
sion language for applications that need a programmable interface. Python
is copyrighted but freely usable and distributable, even for commercial
use. More information about Python can be found on the Python home page at
http://www.python.org.

---------------------------------------------------------------------------
ChangeLog:
===========================================================================

Changes since pyPgSQL Version 2.1
=================================

The following source code files were added to Version 2.2 of pyPgSQL:

    pyPgSQL.spec	- RPM spec file, contributed by Sean Reifschneider.

Changes to README
-----------------
* Added note about case-insensitiveness of column access in PgResultSet.

Changes to PgSQL.py
-------------------
* Fixed various problems with the PgResultSet: Column (attribute and
dictionary) access is now case-insensitive.  A __contains__ method was added
and the __setattr__ method was fixed. The get method got an optional default
value parameter.

* Fixed various problems with the PgNumeric type:
  - Added code to allow a float as an argument to the PgNumeric constructor.
  - You can now change the precision/scale of a PgNumeric by:
      a = PgNumeric(pgnumeric, new prec, new scale).
    This can be used to 'cast' a PgNumeric to the proper precision and scale
    before storing it in a field.
  - The arithmatic routines (__add__, __radd__, etc) now ensure that the
    arguments are properly coerced to the correct types.
  - Added support for the augmented arithmetic operations (__iadd__, etc).
  - The math routines would lose precision because the precision/ scale were
    set to be the same as the first operand.  This is no longer the case all
    precision is retained for the +, -, and * operations.

* Fixed problem that occurs when a query on an OID field doesn't return any
rows. [Bug #589370].
	
* Applied patch #569203 and also added __pos__ and __abs__ special methods to
PgNumeric.

* Ensure proper SQL-quoting of long ints.

Changes to PgSQLTestcases.py
----------------------------
* 14 new tests, mostly for PgNumeric and PgResultSet.



From fgranger at alussinan.org  Fri Sep 20 13:29:10 2002
From: fgranger at alussinan.org (=?ISO-8859-1?Q?Fran=E7ois_Granger?=)
Date: Fri, 20 Sep 2002 19:29:10 +0200
Subject: [Help]: mailbox classes
References: <jylimd-DAD12F.17305418092002@skates.gsfc.nasa.gov> <Ndfi9.130668$pX1.4682176@news2.tin.it>
Message-ID: <1fitpvg.gi8n0y1iem93tN%fgranger@alussinan.org>

Alex Martelli <aleax at aleax.it> wrote:

> J. Li wrote:
> 
> > Hi,
> > 
> > I am trying to extract a message from a unix mailbox file.
...
> 
> > what is it relationship with email.message classes.
> 
> Not much, intrinsically -- the items in a mailbox object by
> default are rfc822.Message instances.  But you can pass as
> the second argument to the constructor a factory with a
> compatible interface, which should let you get email.Message
> instances instead, if that's what you want.

email does not handle the unix mailbox properly. It sees only the first
message.

I was going the way you suggest by first using mailbox.UnixMailbox, then
i'd like to pass each message given by the next() method to email for
further analysis. But this is way over my current level of understanding
of Python objects and classes. ....

-- 
Une faq de frj: http://faq.jardin.free.fr/, pr?s de l'encyclop?die
http://nature.jardin.free.fr/
MacPython beginners http://francois.granger.free.fr/MacPython/faq/


From loewis at informatik.hu-berlin.de  Wed Sep 18 06:33:14 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Sep 2002 12:33:14 +0200
Subject: python does not start
References: <mailman.1032337162.20546.python-list@python.org>
Message-ID: <j4d6rba8ph.fsf@informatik.hu-berlin.de>

Brian Lee <senux at senux.com> writes:

> On SunOS 5.7, python which is compiled on other sun machine does not start.
> How can I fix this?

You need to copy libdb-3.2.so from that machine as well.

Regards,
Martin


From bobx at linuxmail.org  Wed Sep 18 13:52:00 2002
From: bobx at linuxmail.org (Bob)
Date: 18 Sep 2002 10:52:00 -0700
Subject: Python equivalent to perldoc?
References: <Olvh9.7898$Lg2.1389903@news2.news.adelphia.net> <mailman.1032226040.2391.python-list@python.org> <9ywh9.7955$Lg2.1406050@news2.news.adelphia.net> <am7ngs$rk9$1@peabody.colorado.edu>
Message-ID: <1001ff04.0209180952.57e7b865@posting.google.com>

Fernando P?rez <fperez528 at yahoo.com> wrote in message news:<am7ngs$rk9$1 at peabody.colorado.edu>...
> Bob X wrote:
> 
> > Skip Montanaro wrote:
> >>     Bob> In Perl you can run "perldoc warnings" and get the definition of
> >>     Bob> what "warnings" does. Is there a equivalent in Python?
> >> 
> >> Try "pydoc warnings".
> >> 
> > What a funny guy!  :-)
> 
> $ pydoc warnings
> NAME
>     warnings - Python part of the warnings subsystem.
> 
> FILE
>     /usr/lib/python2.2/warnings.py
> 
> CLASSES
>     exceptions.Exception
>         _OptionError
> 
>     class _OptionError(exceptions.Exception)
>      |  Exception used by option processing helpers.
>      |
> ... snipped rest.
> 
> What's funny about it? Looks like exactly what you asked for.
> 
> Cheers,
> 
> f.
Actually it works on *nix as was the above...but not on Windows XP as was mentioned.

Cherrios,

Bob


From davidis at diebots.davidis.diebots.plus.com  Sun Sep 29 19:48:27 2002
From: davidis at diebots.davidis.diebots.plus.com (David Harrison)
Date: Mon, 30 Sep 2002 00:48:27 +0100
Subject: windows xp command prompt
References: <576c1752.0209271411.6d7bb4c4@posting.google.com>
Message-ID: <AjMl9.11444$J47.995513@stones>

mike henley wrote:
> i have installed the activestate perl, python, tcl, and pragmatic
> programmer's ruby. when i go to the command prompt on windows xp and
> type perl for example... (or python, ruby... etc) i get the following
> message.
> 
> 'perl' is not recognized as an internal or external command,
> operable program or batch file.
> 
> in the control panel| system | advanced | environment variables | Path
> i have
> 
> C:\ruby\bin;C:\Python22;C:\Perl\bin;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem
> 
> does anyone know what i should do to fix this problem? (apart from the
> obvious of restarting the machine)


Is it like Win2K, where you have environment variables for 'this user' 
and 'all users' (easiest found by right-clicking on My Computer, 
selecting 'Manage', then right-clicking again on 'Computer Management 
(Local)', and selecting the 'Advanced' Tab and clickign on the 
'Environment Variables' button. You want to edit the 'System' path 
variable).

Just a thought.

Dave H

-- 


"When the tough get going, I'll steal their sandwiches."

If you need to contact me direct, please remove the obvious from the 
reply address



From martin at v.loewis.de  Mon Sep  9 18:56:49 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 10 Sep 2002 00:56:49 +0200
Subject: how to build wxPython w/ Sun's C++ compiler?
References: <mailman.1031607011.28525.python-list@python.org>
Message-ID: <m3k7luojpq.fsf@mira.informatik.hu-berlin.de>

Skip Montanaro <skip at pobox.com> writes:

> How do I convince it to use CC for files with .cpp extensions.

Currently, you cannot: distutils does not support usage of separate
tools for C and C++; feel free to contribute patches.

If you can compile all source files as C++, try setting CC to CC in
the installed Makefile.

Regards,
Martin


From marklists at mceahern.com  Mon Sep 16 21:20:39 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 16 Sep 2002 20:20:39 -0500
Subject: Source code
In-Reply-To: <3d867f4d$0$23173$afc38c87@news.optusnet.com.au>
Message-ID: <JHEOKEOOLIGLDHCMAHMOEEBGDPAA.marklists@mceahern.com>

bah, to heck with usability; the OP should consult this fine oracle for 
pearls of wisdom on securing code against all interested parties:

  http://www.mindprod.com/unmain.html

// m
-




From Oschler at earthlink.net  Fri Sep 20 20:35:23 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sat, 21 Sep 2002 00:35:23 GMT
Subject: Python threading?
Message-ID: <f9Pi9.10275$Lg2.2191294@news2.news.adelphia.net>

How can I tell if the version of Python I am running is thread-enabled?
Also, what is a reasonable number of threads to expect to be able to run
before context switching overhead becomes a problem (I'm using a PIII 500
Mhz with 512MB ram if that helps).

thx





From astrophels at yahoo.com  Fri Sep  6 11:05:20 2002
From: astrophels at yahoo.com (Micah Mayo)
Date: Fri, 06 Sep 2002 15:05:20 GMT
Subject: Getting a proccesses ID
Message-ID: <Qu3e9.188987$_91.214523@rwcrnsc51.ops.asp.att.net>

Hi,

I'm writing a little tool for SysAdmins on FreeBSD - I'm trying to make
it as portable as possible(as far as the unices go), and have come upon
a problem - one of the things this tool does is restarts various deamons
on the system, in order to do that I have to send a killsignal to the
pid. The problem is getting the process id - right now I've got a hack
using ps, grep, and awk, which I doubt will translate well to other
flavors of *nix because they all seem to have their own method of
delivering info via ps. Can anyone tell me a cross platform way of
grabbing a particular processes ID

Here's what I'm doing now, by the way:

pid = os.popen("ps -x | grep sshd | grep -v grep | awk '{ print $1 }'")
pid = pid.atoi()



From Padraig at Linux.ie  Wed Sep 11 09:13:31 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Wed, 11 Sep 2002 13:13:31 GMT
Subject: regular expression for space seperated quoted string
References: <PqFf9.3627$cP3.7118@news.iol.ie> <alnbpb$ce5$1@wanadoo.fr>
Message-ID: <%jHf9.3635$cP3.7161@news.iol.ie>

Eric Brunel wrote:
> Padraig Brady wrote:
> 
> 
>>Hi, I'm trying to split a string that is seperated
>>by spaces and also contains double quoted words which
>>can contain spaces:

[snip]

> What about:
> 
>>>>p = r'[^ \t\n\v\f"]+|"[^"]*"'
>>>>re.findall(p, '1 2 3')

FYI, I'm using the following:

import fileinput, re
for line in fileinput.input():
     #split fields
     fields = re.findall('[^ "]+|"[^"]+"', line[:-1])
     #remove quotes
     fields = map(lambda field: field.replace('"', ''), listLine)

thanks again,
P?draig.



From quiteblack at yahoo.com  Mon Sep  2 23:27:24 2002
From: quiteblack at yahoo.com (black)
Date: 2 Sep 2002 20:27:24 -0700
Subject: what about things like __*** ?
Message-ID: <7153651c.0209021927.16a3efaf@posting.google.com>

Hi, all~

when reading the tutorial I met some trouble.(Maybe trifle to you all
but really trouble to me)

That is I couldnt figure out what __*** is. Here is the description of
that article but even confused me:

There is limited support for class-private identifiers. Any identifier
of the form __spam (at least two leading underscores, at most one
trailing underscore) is now textually replaced with _classname__spam,
where classname is the current class name with leading underscore(s)
stripped.

What's it trying to say please ???


From tjreedy at udel.edu  Mon Sep 23 09:17:12 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 23 Sep 2002 13:17:12 GMT
Subject: convention for Pickler file extension
References: <918bc22f.0209230208.4eaca702@posting.google.com>
Message-ID: <svEj9.226013$z91.9600123@bin3.nnrp.aus1.giganews.com>

"Donnal Walter" <donnal at donnal.net> wrote in message
news:918bc22f.0209230208.4eaca702 at posting.google.com...
> When one uses Pickler.dump(object) to serialize an object to a file
> associated with the Pickler, is there a Python convention for the
file
> extension name?

If there is not currently, I think .pyp is available.




From chris.gonnerman at newcenturycomputers.net  Fri Sep 13 00:38:01 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Thu, 12 Sep 2002 23:38:01 -0500
Subject: problems with getpass.getpass() under Linux
References: <008601c25ada$d1a5f580$3101a8c0@borg.fielden.com.au>
Message-ID: <00b501c25adf$5d7a53e0$0101010a@local>

----- Original Message -----
From: "DG" <david at fielden.com.au>


> Greetings,
>
> I have Slackware Linux 8.0 with the (as distributed) Python 2.0.1
installed.
> When I import getpass and try to getpass.getpass() I get the following
error
> and the password is echoed:
>
> "Warning: Problem with getpass. Passwords may be echoed."
>
> Looking into getpass.py it does this:
>
> try:
>         import termios, TERMIOS
>
> If that fails, it goes on to try to import Windows and then Mac-specific
> modules, with a fallback to just calling raw_input (effectively).
>
> I have TERMIOS.py in /usr/lib/python2.0/plat-linux2 but I do not have
> termios.py anywhere.

That's because it is termios.so; it should be in
/usr/lib/python2.0/lib-dynload.

> I have the Python 2.2.1 tarball (not (yet) installed), it contains
> TERMIOS.py and but not termios.py (altho it does have a termios.c).  I
> notice that the 2.2.1 getpass.py only tries to import termios.
>
> What do I need to do to have getpass.getpass() work correctly?

If you build the 2.2.1 system it should make the termios.so module for you;
I'd think that would work fine.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From xscottgjunk at yahoo.com  Mon Sep 23 18:33:03 2002
From: xscottgjunk at yahoo.com (Scott Gilbert)
Date: 23 Sep 2002 15:33:03 -0700
Subject: numpy array assignment
References: <mailman.1032793169.23743.python-list@python.org>
Message-ID: <79b2b0bb.0209231433.11f26388@posting.google.com>

John Hunter <jdhunter at ace.bsd.uchicago.edu> wrote:
>
> I have a 2 dimensional array X which is NxM and a 1 dimensional array
> v which is Nx1.  I want to assign v to the k-th column of X
> 
> import Numeric
> import RandomArray
> 
> Z = Numeric.zeros( (3,10), Numeric.Float  )
> v = Numeric.ones( (3,1), Numeric.Float )
> Z[:,4] = v
> 
> But I get the ValueError: Object too deep for desired array.  I have
> read over the numpy manual on slices and array methods and the closest
> thing I saw was the 'put' method, but this seems to be designed for 1d
> indexing.
> 
> What's the best way to assign an arbitrary Nx1 array to a column of X?
> 

Is this what you want?

     Z[:,4] = v[:,0]

v is actually a 2D array from Numeric's perspective, it's just the
second dimension isn't very deep.

Cheers,
    -Scott


From tjreedy at udel.edu  Sat Sep  7 18:23:14 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 07 Sep 2002 22:23:14 GMT
Subject: Converting python object reference to int and back
References: <3D79AC95.1000000@NOSPAMbitbuilder.com> <m38z2e2qef.fsf@mira.informatik.hu-berlin.de>
Message-ID: <m%ue9.240806$On.9517391@bin3.nnrp.aus1.giganews.com>

"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m38z2e2qef.fsf at mira.informatik.hu-berlin.de...
> Ahmad Baitalmal <ahmad at NOSPAMbitbuilder.com> writes:
>
> > How can I get an id (pointer/reference) from a python object and
then
> > get that object back using that integer?
>
> To get an integer, use the id function. To get the object back,
create
> a dictionary mapping the ids to the real objects.

While this answers the OP's question as presented, it seems pretty
useless (and unlikely to solve the real question not asked).  Since
the 'real objects' are represented internally by the same id used as
the key, such a dict would amount to an identity mapping.

Terry




From jslowery at hotmail.com  Fri Sep 13 18:21:46 2002
From: jslowery at hotmail.com (Jeremy Lowery)
Date: Fri, 13 Sep 2002 17:21:46 -0500
Subject: CPython detecting read-only attributes.
Message-ID: <qktg9.1085$AY5.284165@e3500-atl1.usenetserver.com>

I'm implementing an extension type in CPython, using the new getsetlist
tp_getset.

I have defined my own tp_setattro function, and my question is: What would
be the easiest way to detect the attempt to write to a readonly attribute.

Scouring PyObject_GenericSetAttr, I see that it always returns -1 on fail,
but that is if it doesn't find a corresponding match or not, and the same
exception is raised PyExc_AttributeError.

I am trying to use the builtin functionality and then, if the attribute is
not found and only if it is not found, I would like to store it in a PyDict
myself.

Is there a better function around to do the Attribute Lookup/setting for the
builtin? I haven't been able to find documentation for these new API
functions (added 2.2) in the reference.

Thanks,
J Lowery






From grabiller at 3dvf.net  Sun Sep 29 19:47:28 2002
From: grabiller at 3dvf.net (Guy Rabiller)
Date: Mon, 30 Sep 2002 01:47:28 +0200
Subject: Python's import: why doesn't it work?
References: <mailman.1033336717.3399.python-list@python.org>
Message-ID: <3d979112$0$18477$79c14f64@nan-newsreader-02.noos.net>

[ Jurie Horneman]
> python -v script.py >output.txt
>
> It lists the verbose output straight to the DOS shell, not to
> output.txt. The output is useless, I can't scroll back.

try:

python -v script.py 2> output.txt


--
guy rabiller
3d animator / td
grabiller at 3dvf.net
http://grabiller.3dvf.net

Tim Peters wrote:
> [ Jurie Horneman]
>> Actually, I just tried this in a DOS shell under Windows 98:
>>
>> python -v script.py >output.txt
>>
>> It lists the verbose output straight to the DOS shell, not to
>> output.txt. The output is useless, I can't scroll back.
>>
>> Oddly,
>>
>> python -h >output.txt
>>
>> correctly sends the output to output.txt.
>>
>> Does anybody know if this is a bug,
>
> Arguably so, but it's a command.com (the Win9x shell) limitation.
> Your first example printed to stdout, your second to stderr, and
> command.com simply doesn't allow redirecting stderr, or setting the
> DOS box history to a usable value.  cmd.exe (the WinNT/2K/XP shell)
> does both.
>
>> and how I can work around it?
>
> Programming life under command.com is impossible without tcap:
>
>     http://www.simtel.net/pub/pd/11141.html
>
> It's free, tiny, and works great.  It's the only Win9x "console
> capturing" utility I've tried that actually works, and doesn't cause
> system instability.




From skip at pobox.com  Fri Sep 27 13:03:03 2002
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 27 Sep 2002 12:03:03 -0500
Subject: watching mutables?
In-Reply-To: <an211j$oi$1@news.hccnet.nl>
References: <mailman.1033137396.14482.python-list@python.org>
        <an211j$oi$1@news.hccnet.nl>
Message-ID: <15764.36679.516534.436308@12-248-11-90.client.attbi.com>

    Anton> Variables should be 'marked' to install this function for
    Anton> them. For example 'watch(m,f)' indicates that function f is
    Anton> called if mutable variable m is changed. 

You could implement a proxy object which delegates to the real object and
catches all __setattr__ and __setitem__ calls.  Something like:

    class MutableProxy:
        def __init__(self, obj):
            self.__dict__['obj'] = obj

        def __setattr__(self, attr, val):
            print "Hey! My %s attribute got diddled!" % attr
            setattr(self.__dict__['obj'], attr, val)

        def __getattr__(self, attr):
            return getattr(self.__dict__['obj'], attr)

    if __name__ == "__main__":
        class Dummy:
            pass

        d = MutableProxy(Dummy())
        d.x = 1
        print d.x

__setitem__ and __getitem__ are similar, but you have to worry about ranges.
You'd probably have to wrap __call__ as well.

There's probably a way to do it with metaclasses, but ...  ooh! I just got a
headache, better back off before my brain expl ... <* blam! *>

-- 
Skip Montanaro - skip at pobox.com
"Airplanes don't fly until the paperwork equals the weight of the
aircraft. Same with i18N." - from the "Perl, Unicode and i18N FAQ"



From pearu at cens.ioc.ee  Mon Sep  9 05:24:36 2002
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Mon, 9 Sep 2002 12:24:36 +0300 (EEST)
Subject: Grumble about too strict an attitude about backward compatibility...
In-Reply-To: <15733.30396.670049.67330@12-248-11-90.client.attbi.com>
Message-ID: <Pine.LNX.4.21.0209091157060.2238-100000@cens.kybi>

On 4 Sep 2002, Skip Montanaro wrote:

> 
> I don't want to wake any sleeping dogs, however, as I'm sitting here
> watching a bunch of Fortran function names fly by as SciPy builds, I'm
> reminded of the fairly recent threads about backward compatibility,
> Python-in-a-tie, etc.   Here I am compiling with a Fortran 90/95 compiler
> (Sun's Forte thing-a-ma-bob) and see function names like
> 
>         lpni:
>         klvna:
>         chgubi:
>         cyzo:
>         klvnb:
>         rmn2so:
>         csphik:
> 
> spew forth.  So, while it's great that this same large library compiles and
> runs on compilers back to at least Fortran 77 (and probably earlier),
> programmers are still stuck with the same cryptic function and data names
> they had to deal with 30+ years ago, all in the name of backward
> compatibility.

I don't think that its is due to the backward compatibility.
There is huge amount of Fortran 77 code available and if one wants to fix
the cryptic names of functions to something more meaningful, then one has
to switch to newer Fortran standard (that allows longer names) and edit
huge amount of F77 code. I think nobody is willing to take this task due
to enormous amount of work (and it cannot be fully automated, some
brain has to deside what are meaningful names), the backward compatibility
is a secondary issue (if an issue at all).

> What's the Python connection?  Other than a reminder not to get to slavish
> about backward compatibility, I note that these same function names will
> then go on to pollute the Python namespace because all this Fortran code is
> automatically wrapped using f2py.

The Fortran function names, that you refer to, are visible only in the
extension modules that are wrapping these functions (this is due to how
Python imports shared modules). All functions that f2py generates,
have the f2py_ prefix to avoid name collision with names from Python
or any other library.
So, I don't understand your argument about name pollution.

And btw, f2py supports mapping function names so that wrapping a Fortran
function with a name `tiarcf' can be accessed in Python side as
`this_is_a_really_cool_function', for example.

Pearu




From mhammond at skippinet.com.au  Tue Sep 24 00:33:42 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 24 Sep 2002 04:33:42 GMT
Subject: Python threading?
References: <f9Pi9.10275$Lg2.2191294@news2.news.adelphia.net> <amogmi$r1l$1@panix1.panix.com> <T5Rj9.13058$6g7.39977@news-server.bigpond.net.au> <7xfzw0avrg.fsf@ruckus.brouhaha.com>
Message-ID: <GWRj9.13330$6g7.40516@news-server.bigpond.net.au>

Paul Rubin wrote:
> Mark Hammond <mhammond at skippinet.com.au> writes:
> 
>>A "one thread per connection" model is doomed from the start -
>>therefore, in most applications, I would consider 5 threads a large
>>number - particularly when all are actively doing something.  OTOH, if
>>you are willing to accept that your first version will accept a max of
>>~20 connections, the brain-power saved by using threads beats the
>>additional OS overhead by a mile ;)
> 
> 
> The new Linux native thread implementation was recently benchmarked
> running 100,000 simultaneous threads on a 1 GB machine.  Starting all
> the threads and closing them took around 2 seconds.  

Just because you *can* doesn't mean you *should* <wink>.  Are you 
suggesting you would use a threading model to support 100,000 
simultaneous connections to your program on such a system?

Mark.



From daniel.dittmar at sap.com  Wed Sep  4 08:43:30 2002
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Wed, 4 Sep 2002 14:43:30 +0200
Subject: newsgroup for zope
References: <al4tru$o3j$03$1@news.t-online.com>
Message-ID: <al4v5j$jtp$1@news1.wdf.sap-ag.de>

stephan wrote:
> but does there exist a newsgrup for zope centric items?
news://news.gmane.org/gmane.comp.web.zope.general. This
is a gateway to the mailing list, so don't expect your ISP
to provide it.

Daniel





From Oschler at earthlink.net  Sat Sep 28 22:08:49 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sun, 29 Sep 2002 02:08:49 GMT
Subject: Best way to create a class dynamically on the fly?
References: <tFml9.13043$Lg2.3865279@news2.news.adelphia.net> <3D96511D.2040407@4-am.com>
Message-ID: <Rgtl9.13598$Lg2.3950931@news2.news.adelphia.net>

"Evan" <evan at 4-am.com> wrote in message news:3D96511D.2040407 at 4-am.com...
> Robert Oschler wrote:
> > What is the best way to create a Python class dynamically at runtime?
>
> ALL Python classes are created dynamically at runtime.  The following is
> not a declaration in the sense that you're accustomed to from C/C++, it
> is normal code executed when the containing file is run (or imported):
>
> class Foo:
>    def meth1(self):
>      return "I am a method"
>
<snip>

Evan,

I understand but what I'm after is building the class itself dynamically,
akin to the way you can manufacture a prolog predicate on the fly.  For
example,  if wanted to create a class whose name was provided in a string
variable called ClassName, at runtime.  Or did I misunderstand your reply?

thx





From SBrunning at trisystems.co.uk  Wed Sep  4 08:33:16 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 4 Sep 2002 13:33:16 +0100 
Subject: Please help in  calling python fucntion from 'c'
Message-ID: <31575A892FF6D1118F5800600846864DCBD71E@intrepid>

> From:	Praveen Patil [SMTP:praveen.patil at silver-software.com]
 
(snip)
 
> Please anybody help me solving the problem.
 
Is there anywhere you *didn't* post this?

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




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



From tdelaney at avaya.com  Thu Sep 26 01:12:27 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 26 Sep 2002 15:12:27 +1000
Subject: revenge of the python puzzle
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5E4@natasha.auslabs.avaya.com>

May I note of course that I shamelessly stole Holger's code for the basic
framework ;)

Tim Delaney



From printers at sendme.cz  Thu Sep  5 09:20:31 2002
From: printers at sendme.cz (A)
Date: Thu, 05 Sep 2002 15:20:31 +0200
Subject: What database should I use
Message-ID: <3D77763F.23080.BFAFA3@localhost>

Hello,
I am going to program a small application (in Python and wxPython) which consists about 
2000 - 3000 records . Can you please recommend which database is the best for that. I 
use Windows.
Maybe MS Access  or FoxPro or MySQL or better?
I want to keep the program as small as possible.
Thanks
Ladislav




From tismer at tismer.com  Sun Sep 15 14:24:54 2002
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 15 Sep 2002 20:24:54 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <j43csbruq7.fsf@informatik.hu-berlin.de>
Message-ID: <3D84D076.1000006@tismer.com>

Martin v. L?wis wrote:
> "David Abrahams" <david.abrahams at rcn.com> writes:
> 
> 
>>I think it still matters. You don't want people who need stackless'
>>capabilities to have to use a non-standard Python forever, do you?
> 
> 
> Unfortunately, the current implementation is processor-dependent; it
> supports x86, SPARC, and PPC only (not really sure about PPC).

Since today it supports X86, Sparc, PPC and Cygwin.

> It also
> limits extensions: if they call back to Python, they must not keep
> "global" variables on the stack (global in the sense that their
> address is known outside their function).
> 
> These limitations require Stackless users to be more aware of the
> inner workings of Python, or they risk crashes - something that would
> be considered a bug in standard Python.

Yes, but :-)
Stackless is built in a way that by default there is no difference
to the standard. No stack switching occours, all is identical.
If you want stackless behavior, you must switch it on.
And if you want to use task switching, you need to import the
stackless extension (which is now apart).

Therefore I'd claim that putting Stackless into the core trunk
would not do any harm. Users who want to use it can use it,
others may ignore it (and it is very tiny, about 50k).

This could be something like the thread package, which also
does not work on all platforms.

ciao - chris
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From diltonm at pacbell.net  Tue Sep 24 23:29:47 2002
From: diltonm at pacbell.net (Dilton McGowan II)
Date: Wed, 25 Sep 2002 03:29:47 GMT
Subject: PythonWin CVS problem
References: <mailman.1032921753.8359.python-list@python.org>
Message-ID: <L4ak9.478$vJ5.43427085@newssvr21.news.prodigy.com>

David LeBlanc wrote:

> Second day it's inaccessable with "permission denied"
> 
> David LeBlanc
> Seattle, WA USA

Are you using cvs.pythonpros.com? It worked for me yesterday and again just 
now. They seem to have the latest source. Just a friendly suggestion.


From fbarbuto at telusplanet.net  Fri Sep 20 13:27:08 2002
From: fbarbuto at telusplanet.net (Fausto Arinos de A. Barbuto)
Date: Fri, 20 Sep 2002 17:27:08 GMT
Subject: What the heck? open() requires an integer?
References: <FCBi9.13802$Zk4.708378@news2.telusplanet.net> <ZLBi9.136908$pX1.4869246@news2.tin.it>
Message-ID: <MTIi9.14708$Zk4.751630@news2.telusplanet.net>

      
Hi Alex and Henrik,

    Thanks so much for your help. I thought it might be
    that open() had been superseded by some other
    function having the same name (but different syntax),
    but didn't know which one.

    Also, it is a good idea to learn more Python before
    venturing to develop code. Or to adapt someone
    else's source code, which was exactly the case.

---Fausto


Alex Martelli <aleax at aleax.it> wrote:

> Fausto Arinos de A. Barbuto wrote:
> 
>>     The open() function is returning me a strange error.
>>     An attempt to create an output file with
>> 
>>             f = open("output.txt","w")
>> 
>>     yields the following error:
>> 
>>     Traceback (most recent call last):
>>     File "<interactive input>", line 1, in ?
>>     TypeError: an integer is required
>> 
>>     That happens when I am using either Pythonwin or
>>     IDLE 0.8, but not in interactive mode. In this latter
>>     case I can open, write to and close a file OK.
>> 
>>     What's the matter? Thanks in advance!
> 
> I'd be willing to bet that you did what you SHOULDN'T do,
> specifically a statement
>         from somewhere import *
> before this point in your interactive session.  I'd lay
> good odds that the 'somewhere' is os.
> 
> Don't do it.  "from ... import *" tramples all over your
> namespace: after you use it, how can you know WHAT
> function object is referred to by any given name?
> 
> Well, if you "print open" you could be able to tell, at
> least in Python 2.2...:
> 
>>>> print open
> <type 'file'>
>>>> from os import *
>>>> print open
> <built-in function open>
>>>> 
> 
> So, when you're calling open after "from os import *",
> you're really calling os.open -- and THAT one does indeed
> require an integer (as its second argument).
> 
> 
> A suggestion: forget all about the existence of statement
> "from" until you feel you have totally mastered Python.  I
> think you'll have an easier time learning, using, and
> mastering Python, if you ALWAYS use the import statement
> when you need to do imports, rather than using from.  (The
> "from ... import *" is the main issue, but there are others:
> e.g., built-in reload won't do what you expect if you've
> used statement from -- only if you've used statement import).
> 
> 
> Alex



From dpbmisc at pobox.com  Tue Sep 17 15:31:12 2002
From: dpbmisc at pobox.com (Don Bennett)
Date: Tue, 17 Sep 2002 19:31:12 GMT
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de>
Message-ID: <4qLh9.1185$%U1.52435185@newssvr21.news.prodigy.com>

Terje Johan Abrahamsen wrote:
 > If I would like to write some code that would not be possible for 
others to
 > read, after I distribute the program, would Python be suitable?

Yes.

 > I know that
 > there are a few programs like Py2Exe, that creates .exe files. But, I 
also
 > saw this:
 >
 > --------------------
 >  A zip-compatible archive is built, containing all files from this
 >   directory as well as your script, and appended to a custom python
 >   interpreter supplied with py2exe.
 > --------------------
 >
 > Doesn't sound like it is protected very well. Is there some 
alternative to
 > Py2Exe that can do the job, or should I look for a compiled language
 > instead?
 >
Take a look at Tools/freeze.py.

With this tool, you can compile your python files in to C files, and
then link the C files with the python interpreter to create a
stand-alone executable that represents your program.







From teaandbikkie at aol.com  Fri Sep  6 10:48:48 2002
From: teaandbikkie at aol.com (TeaAndBikkie)
Date: 06 Sep 2002 14:48:48 GMT
Subject: example for using calldll with python2.2
References: <20020906104210.15840.00000001@mb-cv.aol.com>
Message-ID: <20020906104848.15840.00000002@mb-cv.aol.com>

a better link is:

http://www.nightmare.com/~rushing/dynwin/

:)


From tibirna at kde.org  Wed Sep 18 19:24:58 2002
From: tibirna at kde.org (Cristian Tibirna)
Date: Wed, 18 Sep 2002 19:24:58 -0400
Subject: RegEx editors
References: <amaq0d$4aton$1@ID-144898.news.dfncis.de>
Message-ID: <sW7i9.5931$w72.1246680@news20.bellglobal.com>

On Wednesday, 18 September 2002 17:05, Eric Arnold wrote to 
comp.lang.python:

> Does anybody know of any good graphical regex editors out there?

There is one in KDE. It's available in KDE-3.1 from most "Find" tools of KDE 
applications. The editor itself is called "KRegExpEditor" :-)

-- 
Cristian Tibirna .. tibirna at sympatico.ca
PhD student .. ctibirna at giref.ulaval.ca .. www.giref.ulaval.ca/~ctibirna
KDE developer .. tibirna at kde.org .. www.kde.org


From ezcoder at mailcity.com  Sat Sep 14 12:20:26 2002
From: ezcoder at mailcity.com (Glenn Scott)
Date: 14 Sep 2002 09:20:26 -0700
Subject: Question regarding mysqldb module and variable table names
References: <45ac92e5.0209131920.23b1c22a@posting.google.com> <slrnao5bik.8ni.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <45ac92e5.0209140820.7d5089f3@posting.google.com>

Ahh...cool trick.  Works great.  Thanks!

Gerhard H?ring <gerhard.haering at gmx.de> wrote in message news:<slrnao5bik.8ni.gerhard.haering at lilith.my-fqdn.de>...
> Glenn Scott wrote in comp.lang.python:
> >  c.execute("""INSERT INTO %s \
> >         SET playerkey=%s, link=%s, snippet=%s, time_created=%s, \
> >         last_seen=%s, relevancy=%s, doc_title=%s """, \
> >          (infoTable, playerKey, urlLink, snippetText,
> > self.getCurrentTime(), self.getCurrentTime(),\
> >          noOfMentions, re.escape(str(self.doc_title))) )
> 
> sql = """INSERT INTO %s (playerkey, link, snippet, time_created,
>          last_seen, relevancy, doc_title)
>          VALUES (%%s, %%s, %%s, %%s, %%s, %%s, %%s)""" % infoTable
> 
> c.execute(sql, (...))
> 
> The reason why your original approach didn't work is that DB-API
> quoting is different from Python-quoting. So MySQLdb quotes your table
> name for SQL insertion, and adds apostrophes around it. You can use
> Python quoting as above to work around this.
> 
> Also, forget about re.escape. The DB-API module will quote the values
> appropriately for insertion into SQL.
> 
> -- Gerhard


From pyth at devel.trillke.net  Thu Sep 26 05:24:11 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 26 Sep 2002 11:24:11 +0200
Subject: revenge of the python puzzle
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8BF0A5E4@natasha.auslabs.avaya.com>; from tdelaney@avaya.com on Thu, Sep 26, 2002 at 03:12:27PM +1000
References: <B43D149A9AB2D411971300B0D03D7E8BF0A5E4@natasha.auslabs.avaya.com>
Message-ID: <20020926112411.M30315@prim.han.de>

Delaney, Timothy wrote:
> May I note of course that I shamelessly stole Holger's code for the basic
> framework ;)

that was released into the public domain anyway :-)

    holger



From kkto at csis.hku.hk  Sun Sep  8 10:21:50 2002
From: kkto at csis.hku.hk (Isaac To)
Date: 08 Sep 2002 22:21:50 +0800
Subject: python program to watch a file - example.
References: <mailman.1031486051.16547.python-list@python.org> <3D7B5650.4050304@zute.net>
Message-ID: <7iadms8su9.fsf@enark.csis.hku.hk>

>>>>> "Tim" == Tim Hoffman <timh at zute.net> writes:

    Tim> If it is always the same file and seeing you are looking at *nix
    Tim> varient, one of the unix like ways would be to tail -f the file,
    Tim> and hook it up to your process which just does blocking reads on
    Tim> stdin.

    Tim> ie tail -f somfile | python my_watch.py

tail -f is doing a polling loop, polling each second by default.  In python
you are probably better off coding even the polling loop in Python as well,
rather than calling tail.

Regards,
Isaac.


From syver-en+usenet at online.no  Thu Sep 12 16:39:15 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Thu, 12 Sep 2002 20:39:15 GMT
Subject: [win32all] Out parameter in outgoing interface
Message-ID: <u8z27x7s3.fsf@online.no>

It does not seem possible use out parameters in an outgoing interface.

Scenario:

outgoing dispinterface, atl implementation of server, makepy support
generated, python client which implements the outgoing interface.

Tried with both win32com.client.getevents and
win32com.client.DispatchWithEvents

Debugged through it, the event gets fired from atl code the python
code gets called, I thought maybe returning a tuple (the return value
+ the outgoing value) would work but the out parameter, a BSTR* (ie a
OLECHAR**) is untouched back in ATL land.

I guess this is unsupported, or??

-- 

Vennlig hilsen 

Syver Enstad


From max at alcyone.com  Thu Sep 19 21:08:15 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 19 Sep 2002 18:08:15 -0700
Subject: lists in dictonaries ?
References: <mailman.1032426629.1933.python-list@python.org> <slrnaojc79.c05.Andreas.Leitgeb@pc7499.gud.siemens.at> <d111eafa.0209191646.6429dcb9@posting.google.com>
Message-ID: <3D8A74FF.A27BF330@alcyone.com>

xtian wrote:

> Andreas.Leitgeb at siemens.at (Andreas Leitgeb) wrote in message
> news:<slrnaojc79.c05.Andreas.Leitgeb at pc7499.gud.siemens.at>...
>
> > Oh, can I create a dictionary in Python that contains all those
> > dictionaries that don't contain themselves ... ?-)
> >
> > *scnr*'ly yours.
> 
> Would it contain itself?

[universe collapses in on itself]

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Everybody's playing the game / But nobody's rules are the same
\__/ Florence, _Chess_
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From lu_gio at hotmail.com  Wed Sep  4 17:43:22 2002
From: lu_gio at hotmail.com (Max)
Date: Wed, 04 Sep 2002 21:43:22 GMT
Subject: Database BSD on Mac?
References: <3144be3c280c062e5e71a98ab191f664.98737@mygate.mailgate.org> <mailman.1031095985.2997.python-list@python.org>
Message-ID: <_7vd9.57894$ub2.928923@news1.tin.it>

> > I was looking on BSD but I'm not sure it is ported to all platforms.
> > Wich db should I utilize?
>
> Do you really need one? Or do either pickle or shelve suffice?

I read shelve can be used for small amount of data... will it work with 5mb?




From ataylor at ucsd.edu  Fri Sep 27 15:59:42 2002
From: ataylor at ucsd.edu (Adam Taylor)
Date: 27 Sep 2002 12:59:42 -0700
Subject: Getting started
References: <mailman.1032917851.23864.python-list@python.org> <Sbbk9.262614$z91.11379176@bin3.nnrp.aus1.giganews.com>
Message-ID: <e15b6fa3.0209271159.7b83f3d8@posting.google.com>

> Python types objects and their values rather than names and the fixed
> chunks of memory that names usually refer to in name-typed languages.
> Given a definition of type that sees typeness as inhering only in
> names rather than objects, then yes, Python is 'untyped'.  But how
> useful is such a viewpoint?  It obscures a real difference between
> data models.

Yes, I agree.  I think a better way to classify things (as pointed out
by Alex Martelli, referencing the Scheme spec) is as statically typed
vs. dynamically typed vs. untyped.  Basically, I think the best way to
talk about this stuff is summarized in this diagram:

                      Statically typed   Dynamically typed   Untyped
Strongly checked      ML                 Lisp,Python,Perl
Weakly checked        C                                      Assembler

With the term "weakly typed" being a morass best avoided (IMHO).

Adam

-----
"Terry Reedy" <tjreedy at udel.edu> wrote in message news:<Sbbk9.262614$z91.11379176 at bin3.nnrp.aus1.giganews.com>...
> "Adam Taylor" <ataylor at ucsd.edu> wrote in message
> news:mailman.1032917851.23864.python-list at python.org...
> > Based on this definition, it seems to me that Python _is_ untyped,
> > since the type of a variable can change arbitrarily in mid-program.
> 
> Python types objects and their values rather than names and the fixed
> chunks of memory that names usually refer to in name-typed languages.
> Given a definition of type that sees typeness as inhering only in
> names rather than objects, then yes, Python is 'untyped'.  But how
> useful is such a viewpoint?  It obscures a real difference between
> data models.
> 
> If a language has nested scopes and allows the type of a variable to
> be different in different scopes,  then the type associated with a
> particular name can change arbitrarily in mid-program (at scope
> changes) even if the name is strongly and statically typed within each
> scope.  One can say that the name represents different variables in
> the different contexts (scopes), but the net result is similar to what
> Python does.  Ie, one can view the rebinding of a name to an object of
> different type as implicitly defining a new scope with respect to that
> name.
> 
> Terry J. Reedy


From answer at tnoo.net  Mon Sep  9 13:23:38 2002
From: answer at tnoo.net (Martin =?iso-8859-1?q?L=FCthi?=)
Date: 09 Sep 2002 19:23:38 +0200
Subject: Binary input into NumPy.
References: <alil2o$mmk$1@news.tudelft.nl>
Message-ID: <m3y9abgjqd.fsf@yak.ethz.ch>

>From the manual:

fromstring(string, typecode) 

Will return the array formed by the binary data given in string of the
specified typecode. This is mainly used for reading binary data to and from
files, it can also be used to exchange binary data with other modules that use
python strings as storage (e.g. PIL). Note that this representation is
dependent on the byte order. To find out the byte ordering used, use the
byteswapped() method described on page47.

Hope this helps

Martin

Maurice van de Rijzen <maurice at akst.tn.tudelft.nl> writes:

> Dear all,
> I just downloaded Numerical Python and I think it will be very usefull.
> However I have one major broblem:
> I don't know how to import binary data from a file on disk into an array!
> Can onyone help me with this problem?

-- 
Martin L?thi                 answer at tnoo.net


From sholden at holdenweb.com  Thu Sep  5 21:43:55 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 5 Sep 2002 21:43:55 -0400
Subject: Gratitude due to Red Hat? More work for the PBF?
References: <Iyn89.11545$aC5.4453@atlpnn01.usenetserver.com> <mailman.1030142061.12228.python-list@python.org>  <xuyptw3knfx.fsf@halden.devel.redhat.com> <mailman.1031250971.7837.python-list@python.org> <al8dbo$408$0@216.39.172.122>
Message-ID: <QMTd9.11262$%M3.3512@atlpnn01.usenetserver.com>

"Bengt Richter" <bokr at oz.net> wrote in message
news:al8dbo$408$0 at 216.39.172.122...
> On 05 Sep 2002 11:28:14 -0700, Cliff Wells
<LogiplexSoftware at earthlink.net> wrote:
> [...]
> >
> >The aggravating thing is that this issue would be a snap for RedHat to
> >fix.  Instead of preempting /usr/bin/python for their own use, they
> >could maintain a /usr/bin/python-rh and leave python for user-installed
> >versions (instead of requiring the user version to be called "python2").
> >
> >Problem solved, everyone happy.
> >
> So why don't they do this? Do they need help?
>

Nobody seems to know. It's such an obvious solution to a really annoying
problem that I can only assume they either refuse to uinderstand there *is*
a problem or they don't appreciate how annopying it can be. Pity they didn't
write Anaconda (their maintenance utility) in Perl, then we wouldn't have
had to suufer the results of their success :-)

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                        pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From Patrick.Carabin at SciencesNaturelles.be  Thu Sep 19 08:42:56 2002
From: Patrick.Carabin at SciencesNaturelles.be (Patrick Carabin)
Date: Thu, 19 Sep 2002 14:42:56 +0200
Subject: How do I reply to an item in the list ?
Message-ID: <02091914425601.01191@pc20_118>

I 'd like to answer a message, e.g. on the topic ?Installation of 
multiple versions of Python on Unix?, but on the page
http://mail.python.org/pipermail/python-list/2002-September/120875.html
which contains the message i 'd like to ?reply? to, there is NO 
?replybutton?. How can i do ?
TIA


Patrick Carabin.

Institut Royal des Sciences Naturelles de Belgique 
http://www.SciencesNaturelles.be/
Koninklijk Belgisch Instituut voor Natuurwetenschappen 
http://www.NatuurWetenschappen.be/ 

?Het geluk is niet op het einde van de weg, het geluk is de weg.?
?Le bonheur n'est pas au bout du chemin, le bonheur est le chemin.?
						Dala? Lama.



From paul.pichaureau at nospam.laposte.net  Sun Sep 29 19:49:43 2002
From: paul.pichaureau at nospam.laposte.net (Paul Pichaureau)
Date: Mon, 30 Sep 2002 01:49:43 +0200
Subject: Python's import: why doesn't it work?
References: <mailman.1033322378.3697.python-list@python.org>
Message-ID: <an83an$2qun$1@norfair.nerim.net>

"Tim Peters" <tim.one at comcast.net> a ?crit dans le message de news:
mailman.1033322378.3697.python-list at python.org...

> The docs are exactly as good as volunteers have made them (well, actually
> better, thanks to Fred <wink>).  When you find an explanation inadequate
to
> your needs, suggesting specific improvements in the right place is the
only
> way it will improve.

  Then, these are my suggestions to improve documentation on installing and
managing module :

** I have found no explanation of any kind about what I shall put in the
__init__.py file.
** I have seen many module where this file contain just one "import" line,
e.g.

  import the_module_in_this_directory

This seems silly, and I suggest somebody documents the exact way how one
specific function is eventually loaded by the interpreter.

I have read all the documentation about docstrip, and I don't know at all
what I have to do to import a given fonction.

** Is the __name__ list indispensable ? Nobody seems to use it...
** How the sys.path is created in a run, and how can it be cleaned (for
example, when I uninstall a old package, or an obsolete version of a
package).

Suggestion for the distutil package :

** First a complete documentation. I don't like the pseudo-pedagocical style
of the actual documentation. I prefer a complete and clear explanation of
all the command avalaible. The two type of documentation should be
avalaible.

** A clean and safe way to uninstall package. I can't understand why this
hasn't been implemented yet.

** The actual beahviour is quite strange : distutil install all the
subdirectory of the current working directory.
For example, if you consider this tree :

---setup.py
(main dir)
---titi.py
------(lib dir)
---------__init__.py
---------toto.py
------(stuff dir)
---------__init__.py
---------tata.py

  If you declare the package main.lib, the package main.stuff will also be
installed. Seems silly to me

                    Paul




From aleax at aleax.it  Wed Sep 25 05:26:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 09:26:24 GMT
Subject: help with print function in extension
References: <tA8k9.270440$AR1.11448440@bin2.nnrp.aus1.giganews.com>
Message-ID: <4jfk9.163306$pX1.5842487@news2.tin.it>

Eric Hagemann wrote:

> I'm handcrafting an extension in C
> 
> If have defined both a str() and a repr() function and when I execute
> either repr(obj), or str(obj) I get what I expect
> 
> however if I execute 'print obj' I get the following error
> 
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> SystemError: error return without exception set
> 
> although 'print str(obj)' works like I think it should
> 
> Am I mistaken in thinking that  'print str(obj)' should be the same as
> 'print obj'

*ALMOST!*  In the PyTypeObject struct, there's a tp_print slot
that is used by preference, if non-null, when you print something
and stdout is a real file.  What do you have in tp_print for
your type?  It's smack in-between tp_dealloc and tp_getattr.


Alex



From gerhard.haering at opus-gmbh.net  Wed Sep  4 13:55:57 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 4 Sep 2002 17:55:57 GMT
Subject: Does BlackAdder really suck?
Message-ID: <slrnanci6i.1s8.gerhard.haering@haering.opus-gmbh.net>

I just checked out the BlackAdder demo and was quite disappointed.

I expected that this is a tight integration of Python and PyQt. But it
looks like it's just Qt Designer embedded in the IDE, while still having to
run pyuic manually. I don't know if at least the debugger is any good,
because the demo was quite limited.

Is there any usable commercial IDE for GUI design? BlackAdder and
PythonWare Pro seem to be the only candidates (yes, I did check out Boa).

-- Gerhard


From Stefan.Schwarzer at tu-clausthal.de  Wed Sep  4 15:10:13 2002
From: Stefan.Schwarzer at tu-clausthal.de (Stefan Schwarzer)
Date: Wed, 04 Sep 2002 21:10:13 +0200
Subject: Strange threads/stack size interaction on FreeBSD
Message-ID: <3D765A95.6090308@tu-clausthal.de>

OS: FreeBSD 4.6.2
Python: 2.2.1 (built from the ports collection)
Hardware: Dual Processor Athlon, 1 GB RAM

I compiled Python while having a default stack size (as shown by ulimit -a)
of 262144 kbytes.

Because of a problem with a Python program (that dumped core, but that's
another problem) I recompiled the kernel with a stack size of 524288 kbytes.
After rebooting the program mentioned succeeded without errors.

A while later I noticed that a unit test case dealing with threads no longer
worked; it hung. Checking with /usr/local/lib/python2.2/test/test_threading.py
I observed that it hung, too, after only printing "waiting for all tasks to
complete".

After rebooting the system with the old kernel with the old stack size, I
saw both my own unit test and test_threading.py succeed! I also tried
compiling Python for the larger stack size but that didn't help either.

It seems, that there is some interaction between the set stack size and
Python's threading code.

I've searched the Python FAQ, Google Groups and the bug database of the
Python project on this, but without success, so I have some questions.

- Has anybody here experienced something similar, perhaps on other
   platforms?

- Is Python's threading code dependent on a stack limit set at compiled
   time (seems so)?

- What are your recommendations to deal with the problem?

- Should I enter a bug report on this topic?

Stefan



From maxm at mxm.dk  Tue Sep  3 04:58:58 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 03 Sep 2002 10:58:58 +0200
Subject: string reading bug in ActiveState Python 2.2 ?
References: <3D74757D.1010405@molgen.mpg.de>
Message-ID: <3D7479D2.60100@mxm.dk>

Axel Kowald wrote:


> I'm using ActiveStates Python 2.2.1 under win2000 to read in the 
> attached text file and it seems that python is only reading part of the 
> string. After
> 
> fp = open("bla.txt")
> z = fp.read()
> print len(z), z
> 
> I see that only 11 of the 15 characters have been read. I know they are 
> non-printable characters, but that shouldn't matter, should it ?  Using 
> another python on one of our unix machines everything works as expected 
> and all 15 chars are read in.  Is that there something special about 
> windows that I should know, or is this a bug in ActiveStates Python ?



Probably you should open it as a binary file:


fp = open("bla.txt","b")


regards max M



From jstavnitzky at atsautomation.com  Fri Sep 20 12:09:08 2002
From: jstavnitzky at atsautomation.com (Jay)
Date: 20 Sep 2002 09:09:08 -0700
Subject: Porting woes
Message-ID: <3a4b2539.0209200809.407d50d5@posting.google.com>

Hello, I am in the process of trying to port Python to a TI
TMS320C6xxx series DSP.  The code now compiles (under the TI Code
Composer IDE), but I have assertions failing at runtime.  The
assertion reads as follows:

Assertion failed, (((cls)->ob_type == &PyClass_Type)), file
typeobject.c, line 649

This failure is occuring during the function classic_mro().  The stack
at that point is classic_mro(),mro_implementation(),mro_internal(),PyType_Ready(),
PyTypeReady(),Py_ReadyTypes(),Py_Initialize(),Py_Main().

Help!! What is happening at this point, and what could be causing this
assertion to fail?

Thanks,
Jay


From pinard at iro.umontreal.ca  Thu Sep 26 10:49:27 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Thu, 26 Sep 2002 10:49:27 -0400
Subject: HTML page for Pymacs 0.18
Message-ID: <oq1y7gajrc.fsf@carouge.sram.qc.ca>

Hi!

There is now a simple HTML page for Pymacs, which repeats the documentation
found in the `README' file of the version 0.18 of the distribution:

   http://www.iro.umontreal.ca/~pinard/pymacs/

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From aleax at aleax.it  Tue Sep 24 08:39:32 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 24 Sep 2002 12:39:32 GMT
Subject: Three dumb questions (ordered by dumbness descending)
References: <amnvit$7kqdt$1@ID-77524.news.dfncis.de>
Message-ID: <82Zj9.136094$ub2.2966822@news1.tin.it>

Thorsten Kampe wrote:

> Okay, here they are:
> 
> 1. Why is 'zip(zip(x)) != x' (should be the same because it's the
> transposed)

I think this has been answered at length (you need the * special form
for the arguments -- and x to be of the right kind -- and then the
equality will hold).


> 2. Peter Norvig mentions in "Python for Lisp Programmers" some
> "don'ts": "[x] + y" and "x[1:]". Are there more things to avoid
> (especially in a loop)?

Strictly a performance issue: these operations are O(N) in len(y)
and len(x) respectively, and Peter is writing for an audience used
to similar operations (cons and cdr respectively) being O(1).  There
is no special reason to avoid such constructs if they best express
what you really want to do.

The only Python performance pitfall that's REALLY easy to fall into:

    hugestring = ''
    for piece in alotoftinypieces:
        hugestring += piece

Now THIS is guaranteed to kill your program's performance.  Use:

    hugestring = ''.join(alotoftinypieces)

and enjoy roughly O(N) performance rather than roughly O(N squared).

More generally, I think that putting strings together with + or +=
IS something "to avoid (especially in a loop)" -- if you don't
already have a sequence of all the tiny pieces, build one up (e.g.
by appending the pieces to an originally empty list) and finally
''.join it.  Strings are immutable, so "hugestring += piece"
forces Python to free the space it originally held for hugestring
and allocate a new one.  Lists are mutable, and in an amortized
sense there's "always" a free slot at the end, so templist.append
is O(1) [amortized].


> 3. random.shuffle: the documentation says: "Note that for even rather
> small len(x), the total number of permutations of x is larger than the
> period of most random number generators; this implies that most
> permutations of a long sequence can never be generated." I translate
> this to "don't shuffle lists with more than 15 elements if you want a
> randomly ordered list".
> 
> Alex Martelli says the same: "For example, Python users who are
> calling random.shuffle on a sequence of substantial length have a need
> for an underlying random generator with a *HUGE* period, as a
> necessary although not sufficient condition towards ensuring that all
> permutations of the shuffled sequence get generated with equal
> likelihood.
> 
> Wichman-Hill has a period of less than 7,000 billions (7e12),
> insufficient to account even for the permutations of a sequence
> whose length is just 16 (16! > 2e13)."
> 
> Otherwise in the "Python Cookbook" (Selecting Random Elements from a
> List Without Repetition) the shuffle version is called "the winner".
> 
> So: can I use random.shuffle or do I have to write my own - slower -
> version to get a random list?

For most practical purposes, you can use random.shuffle.  If you
must guarantee that *all* permutations are equally likely, then
your first issue is ensuring enough randomness -- a generator with
a period of over 10**68 or so to permute 52 objects, for example --
and that's not at all a sufficient condition, either.  If you really
labor under such strict conditions, maybe it's better to use a
"truly random" (hardware) generator rather than a pseudo-random one.

Once you do have a random or pseudo-random generator that fully
satisfies you, Ian Bicking's snippet reimplementing shuffle as a
Decorate-Sort-Undecorate idiom is truly excellent -- a little jewel,
IMHO, despite the little typo in it.  It does, alas, end up a
little slower than Tim Peters' original -- but losing a speed
competition to Tim is no shame, and Ian's version only loses very
very little indeed.  To wit (having corrected Ian's typo):


import time, sys, operator
from random import random

def shuffle_org(x):
    for i in xrange(len(x)-1, 0, -1):
        # pick an element in x[:i+1] with which to exchange x[i]
        j = int(random() * (i+1))
        x[i], x[j] = x[j], x[i]

def shuffle_ian(x):
    newList = [(random(), i) for i in x]
    newList.sort()
    x[:] = [i[1] for i in newList]


biggie = range(200)

def timit(func, N=1000):
    repeat = [None] * N
    start = time.clock()
    for x in repeat: func(biggie)
    stend = time.clock()
    return "%2.2f %s" % (stend-start, func.__name__)

print sys.version
for i in range(2):
    for func in shuffle_org, shuffle_ian:
        print timit(func)


[alex at lancelot linud]$ python -O timo.py
2.2.1 (#2, Jul 15 2002, 17:32:26)
[GCC 2.96 20000731 (Mandrake Linux 8.1 2.96-0.62mdk)]
1.59 shuffle_org
1.87 shuffle_ian
1.60 shuffle_org
1.86 shuffle_ian
[alex at lancelot linud]$



Alex



From duncan at NOSPAMrcp.co.uk  Fri Sep  6 09:12:31 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Fri, 6 Sep 2002 13:12:31 +0000 (UTC)
Subject: tree data structure
References: <FM0e9.2860$cP3.6165@news.iol.ie> <Xns9281891D87291duncanrcpcouk@127.0.0.1> <Us1e9.2865$cP3.6131@news.iol.ie>
Message-ID: <Xns92818FDA1D5F4duncanrcpcouk@127.0.0.1>

Padraig Brady <Padraig at Linux.ie> wrote in
news:Us1e9.2865$cP3.6131 at news.iol.ie: 

>> assert result==('L0', [('S2', [('S0', [('Q1', [])]), ('S1', [('Q2',
>> [])]), ('S3', [])])])
>> 
>> 
> 
> It certainly does!
> Just one thing...Why use lists instead of tuples?

Specifically because you can't append to tuples, but you can append to a 
list.

However, a general rule of thumb is to use tuples where you want a record 
(but don't want the overhead of a full-blown class). So each node here is a 
record with the fields name,children. Use a list where you want to store a 
collection of unknown length: so a list is appropriate for the children.

Note that neither your original list, nor the code I used actually force 
this to be a tree, you could equally well have a graph with or without 
cycles. You will find it impossible to represent a graph containing cycles 
if you use tuples throughout.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From martin at v.loewis.de  Mon Sep 30 17:17:43 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 30 Sep 2002 23:17:43 +0200
Subject: a question about mailman
References: <mailman.1033414779.26934.python-list@python.org>
Message-ID: <m3it0n2n48.fsf@mira.informatik.hu-berlin.de>

"Mark McEahern" <mark at mceahern.com> writes:

> In that context, I wonder, "Wouldn't it be easier if you could configure
> Mailman to insert the archive URL into the footer of outgoing messages?"

I'm not sure I understand this question: Do you want to add the URL of
the list archive to the footer, or the URL for the specific message?

Adding the URL for the entire archive is easy (it only varies with the
list, not with the message), so I assume you want to add the URL for
the message itself.

> 1.  Would this generally be useful--having the ability to put message
> specific information in the footers?
> 2.  Would it specifically be useful to put the archive url into the footers?
> (I personally think so.)
> 3.  Is the vague approach I sketched above the way you'd do it?  That is,
> does it seem likely that it would work?

Yes, yes, and no. I believe the trick is that, at the point of sending
the message, you cannot know the URL that the message will get in the
archive. Finding out that URL is particularly tricky in presence of
external archivers.

Regards,
Martin



From writeson at earthlink.net  Tue Sep  3 10:40:14 2002
From: writeson at earthlink.net (Doug Farrell)
Date: 3 Sep 2002 07:40:14 -0700
Subject: RedHat 7.3 and Python2.2.1
References: <mailman.1031020923.22683.python-list@python.org>
Message-ID: <88bc63c6.0209030640.71e074ff@posting.google.com>

To everyone who responded,

Thanks you for all your help, much appreciated. I ended up solving the
problem (at least I hope so, and didn't create more for myself! <g>)
by doing the following:

rpm -ivh --replacefiles python-2.2*

The --replacefiles option seemed to eliminate the file conflicts and
the RedHat tools (like up2date) still are working.

Again, thanks!
Doug


From aahz at pythoncraft.com  Tue Sep 17 10:36:49 2002
From: aahz at pythoncraft.com (Aahz)
Date: 17 Sep 2002 10:36:49 -0400
Subject: threads or queue for this task
References: <mailman.1032244699.24395.python-list@python.org>
Message-ID: <am7em1$fjp$1@panix1.panix.com>

In article <mailman.1032244699.24395.python-list at python.org>,
Bjorn Pettersen <BPettersen at NAREX.com> wrote:
>
>If you have any amount of data to process (why else would you be
>multithreading?) I've never seen producers and consumers being close
>enough in performance to not have the system end up swapping to death --
>unless, that is, if the queue is bounded. The Python Queue module
>supports most of this out-of-the-box, and adding timeouts is fairly
>trivial...

Well, one instance would be where producer and consumer are the same
threads.  For example, a web spider.  You *want* an unbounded queue in
that case to avoid deadlock.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From tismer at tismer.com  Mon Sep 23 17:53:59 2002
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 23 Sep 2002 14:53:59 -0700
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <LNBBLJKPBEHFEDALKOLCAEOJBFAB.tim.one@comcast.net>
Message-ID: <3D8F8D77.1090500@tismer.com>

Tim Peters wrote:
> [Christian Tismer]

[you cannot discuss it since they want it]

> Bless you, Christian!  That's all the truth.  It reminds me of my first job
> in this business, where after a few months my boss took me aside and
> reminded me that I was hired to be a prostitute, not a missionary <wink>.

Hee hee, yes we are prostitutes. Well, Open Source is a way
out. But when you want people to pay you for it, it slips in, again.

> otoh-the-bankers-need-stackless-but-just-don't-know-it-yet-ly y'rs  - tim

I'll let them know, teling them that Stackless is pure decimal.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From johanfo at essay.org  Mon Sep 23 17:01:54 2002
From: johanfo at essay.org (Johan Fredrik Øhman)
Date: Mon, 23 Sep 2002 23:01:54 +0200
Subject: Dynamic loading of modules
References: <PrJj9.12293$Uu.3503@news1.bredband.com> <87sn00zeke.fsf@pokey.henrik-motakef.de>
Message-ID: <amnvgg$71f$1@maud.ifi.uio.no>

I tried this below.  But as you see, even if the Password.py has a "def init()"
it doesn't work.

Python 2.2 (#1, Mar 26 2002, 15:46:04)
[GCC 2.95.3 20010315 (SuSE)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> z = __import__("plugins.Password")
>>> z.init()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: 'module' object has no attribute 'init'

--
Johan Fredrik ?hman


"Henrik Motakef" <henrik.motakef at web.de> wrote in message
news:87sn00zeke.fsf at pokey.henrik-motakef.de...
> "Andreas Lydersen" <al at metroic.no> writes:
>
> > Hi,
> > I'd like to load a set of modules from a directory.  I have tried to use the
> > __import__(...) statement to do so, but with no success.
> > ?
> > The static approch works (of course), but the dynamic does not!
> >
> > static:
> > import plugins.Password
> >
> > dynamic atttempt:
> > __import__("plugins.Password")
>
> __import__ doesn't really import the module, it returns a module
> object. So do something like
>
> >>> pwd = __import__("plugins.Password")
> >>> pwd.spam()
> Spam!
>
> hth
> Henrik




From torkil at fast.no  Tue Sep 17 05:03:30 2002
From: torkil at fast.no (Torkil Grindstein)
Date: Tue, 17 Sep 2002 11:03:30 +0200
Subject: Regular expression help needed
Message-ID: <3D86EFE2.CFF57171@fast.no>

Hi.

It's been a while since I played around with regexps, and I
realize that I really need some help here.

Mission: I have a string containing a whole WML document. I want
to extract data from following occurences:

<meta name="michael" content="owen"/>

That is, I want to extract the content of any occurences of
the meta "michael" key. There may be several occurences in one
document.
<meta name="michael" content="owen"/>
<meta name="michael" content="jackson"/>
will give me the results "owen" and "jackson".

Of course, it is possible that the document has written the
tags in uppercase (META Name, META NAME, etc), but I could
lowercase the whole document string prior to searching. (The
content is case insensitive for my purposes.)

I would really be glad if someone out there took this challenge..:)

Cheers,
Torkil


From wilk-spamout at flibuste.net  Mon Sep 30 05:24:43 2002
From: wilk-spamout at flibuste.net (William)
Date: 30 Sep 2002 11:24:43 +0200
Subject: Why does QT Windows need Visual Studio?
References: <3ljepusnje8iuca90im2r9vceja5tfc5l7@4ax.com>
Message-ID: <87fzvr7rtw.fsf@flibuste.net>

John Hall <wweexxsseessssaa at telusplanet.net> writes:

> I was about to download the Windows non-commercial version of QT, but
> noticed that it requires MS Visual Studio, so didn't.
> This requirement surprises me - is there a way around that?

It will work even without visual studio... Just install it, and pyqt.

But this strange licence make me prefer wxpython for now.

-- 
William Dod? - flibuste.net
http://wikipython.tuxfamily.org


From steve at ferg.org  Thu Sep 12 09:04:54 2002
From: steve at ferg.org (Stephen Ferg)
Date: 12 Sep 2002 06:04:54 -0700
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid> <alln71$1s2s$1@nntp6.u.washington.edu> <b16e4ef7.0209110616.48dfe04c@posting.google.com> <3D7FFC43.10002@something.invalid>
Message-ID: <b16e4ef7.0209120504.62d3c5bd@posting.google.com>

Thanks for the tips on where to find Human Interface Guidelines.

> It's a fairly universal principle in Tk that the
> "command" of a widget is invoked according to the
> standard behaviour of that widget, so I think it
> deserves treatment as a fundamental principle
> on the same footing as event binding.

Absolutely!  This observation is GOLD! This is just the kind of thing
I'm trying to get at in "Thinking in Tkinter".

Let me make sure I've got it right.  We have two fundamental, and
different, approaches to specifying the behavior of a widget.

(1) event binding, in which you get to "roll your" own bindings by
binding callbacks to particular events that you specify.

(2) command binding, in which a standard behavior for a widget is
already defined for you by Tkinter (and Tk), and you bind a callback
to that standard behavior.

This explains why "command" binding doesn't automatically pass an
event object the way that event binding does.  The standard behavior
of a widget  might be defined across multiple kinds of events  (e.g.
Button's standard behavior is defined across both a button-press and a
button-release), so there is no *single* event that could be passed.

-- Steve


From Oschler at earthlink.net  Sat Sep 28 14:37:13 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sat, 28 Sep 2002 18:37:13 GMT
Subject: Best way to create a class dynamically on the fly?
Message-ID: <tFml9.13043$Lg2.3865279@news2.news.adelphia.net>

What is the best way to create a Python class dynamically at runtime?  A
sample or doc would be great.  I'm a veteran C++ programmer so this concept
is very alien to me as you can imagine (dynamically created classes).  An
example that would show the creation of a descendant (sub-class), and a
another example showing the creation of an ancestor class would be perfect.

thx





From loewis at informatik.hu-berlin.de  Wed Sep 18 02:48:52 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Sep 2002 08:48:52 +0200
Subject: Detecting IPv6 support
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032245506.8580.python-list@python.org> <j4n0qhjbkf.fsf@informatik.hu-berlin.de> <pvsn08lw9k.fsf_-_@agni.research.nokia.com> <j43cs8ft6x.fsf@informatik.hu-berlin.de> <pvhegnj2lc.fsf@agni.research.nokia.com>
Message-ID: <j44rcndc8b.fsf@informatik.hu-berlin.de>

Pekka Pessi <Pekka.Pessi at nokia.com> writes:

> 	I have Redhat 7.3. The prepackaged RPMs have disabled IPv6, but
> 	the underlying OS supports it. So, this is what happens with the
> 	bind() system call, for instance:

I see. Testing for that isn't really supported. A reliable way is to
inspect pyconfig.h, whose location is available through
distutils.sysconfig.get_config_h_filename(); check whether it contains
a line

#define ENABLE_IPV6 1

(more precisely, check whether the line containing ENABLE_IPV6 does
not contain undef, and starts with whitespace*#.

> 	So AF_INET6 is there, and I can create a socket, but everything
> 	dealing with addresses fails. 

That, perhaps, gives a reliable test, too: See whether
socket.getaddrinfo("::",80) succeeds.

Determining whether the system "supports" IPv6 is more tricky: If you
can create an INET6 socket, the system supports the protocol. I don't
know how to find out whether there are any configured non-local
interfaces. I don't think there is a way to determine whether the
system has v6 connectivity to "the world".

Regards,
Martin


From sholden at holdenweb.com  Fri Sep 13 07:56:24 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 13 Sep 2002 07:56:24 -0400
Subject: [FEEDBACK] Is this script efficient...is there a better way?
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <L9Rf9.168314$l_4.127784@atlpnn01.usenetserver.com> <4ZRf9.1782$Lg2.248733@news2.news.adelphia.net> <3hSf9.168947$l_4.98525@atlpnn01.usenetserver.com> <8PSf9.1796$Lg2.257378@news2.news.adelphia.net> <_oUf9.170870$l_4.126599@atlpnn01.usenetserver.com> <Am7g9.2344$Lg2.402676@news2.news.adelphia.net>
Message-ID: <3pkg9.3313$WV.2782@fe03>

"Bob X" <bobx at linuxmail.org> wrote ...
> Steve Holden wrote:
> > "Bob X" <bobx at linuxmail.org> wrote ...
> > [ ... ]
> >
> >>Thank you very much...your shorter script did work. The keywords are all
> >>"normal" english words that would not have any special characters. I
> >>only added a re.IGNORECASE (since case doesn't matter to me) after a
> >>quick search of the python docs.
> >>
> >
> > Was there any significant difference in the timings for the two
different
> > methods, and how does the number of different keywords affect the
timings?
> >
> I haven't tested...is there an easy way to time it...besides my going
> "that one ran faster"?
>
> However, I had a chance to run your RE version today and it did very
> well. BTW..the line that looked like a debug statement was just a visual
> for me to see the script was actually doing something. I put in a
> counter that prints "lines written:", a where "a" is the counter. I do
> have one question on that. It writes each line sequentially like:
>
> lines written: 1
> lines written: 2
> lines written: 3
>
> Can I just get it to do it in place? Where the numbers are just replaced
> instead of scrolling? Just a thought...
>

That shouldn't be too difficult. Somebody recently posted code for Python
"spinners", that could be the bases of your efforts. See [... googles ...]

Aargh. Apparently I didn't save any attribution for the attached -- it's not
my code, somebody else wrote this and published it in c.l.py, but I don't
have a record of who it was. As far as timing goes, just subtract
time.time() at the beginnig from time.time() at the end for a naiive
approach. If this doesn't show any differences  it's probnably not worth
measuring more closely...

import threading, time, sys


class Spinner( threading.Thread ):

    DELAY = 0.1
    DISPLAY = [ '|', '/', '-', '\\' ]

    def __init__( self, before='', after='' ):
        threading.Thread.__init__( self )
        self.before = before
        self.after = after

    def run( self ):
        write, flush = sys.stdout.write, sys.stdout.flush
        self.running = 1
        pos = -1
        while self.running:
            pos = (pos + 1) % len(self.DISPLAY)
            msg = self.before + self.DISPLAY[pos] + self.after
            write( msg )
            flush()
            write( '\x08' * len(msg) )
            time.sleep( self.DELAY )
        write( ' ' * len(msg) + '\x08' * len(msg) )
        flush()

    def stop( self ):
        self.running = 0
        self.join()


class BarberPole( Spinner ):
    DISPLAY = [ '|/__/__/__|', '|_/__/__/_|', '|__/__/__/|' ]

class HAL( Spinner ):
    DELAY = 0.2
    DISPLAY = [ ' ', '.', '+', '*', '+', '.', ' ', ' ' ]

class MySpin(Spinner):
 DELAY = 0.08
 DISPLAY = [' ', '.', ':', '!', "'"]

def test():

    spinner = MySpin("My own spinner: [", ']')
    spinner.start()
    time.sleep(5)
    spinner.stop()

    spinner = Spinner( "[syncing... ", "]" )
    spinner.start()
    time.sleep( 5.0 )
    spinner.stop()
    print "Disks synced."

    pole = BarberPole( "Shutting down: " )
    pole.start()
    time.sleep( 5.0 )
    pole.stop()
    print "You may now switch off your computer."

    hal = HAL( "[", "]  I'm sorry Dave" )
    hal.start()
    time.sleep( 5.0 )
    hal.stop()
    print "I'm afraid I can't do that."


if __name__ == '__main__':
    test()

Works well, though, and will show you how to update your message in place.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From bdesth at free.fr  Thu Sep  5 08:18:41 2002
From: bdesth at free.fr (Laotseu)
Date: 5 Sep 2002 05:18:41 -0700
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com>
Message-ID: <4f49d232.0209050418.7668c21@posting.google.com>

Benjamin <phncontact.libero.it@> wrote in message news:<uafbnugdli4c3f8ns6qr33mna5urdno78g at 4ax.com>...
> Hi guys (and girls)
> I just begun to learn how to program. And I choosed Python for
> doing that. surfing around the web i found an article about this
> language, they said it's the ideal language for beginners to the
> world of programming. and i admit it, the syntax is very clear making
> the code very readable, i learn easily the concepts of programming
> and i soon wrote some nice programs. but since i finished my
> newbie-tutorial, i am wondering about one question: is it worth to go
> deeper into python? i talked to a friend of mine, he told me that i
> should try php..
> what should i do?? ^_^ in other words: 

> what is python good for?
Python is a 'general', hi level programming language. You can use it
from simple scripting tasks to full-blown applications.

Now what it may not be good for :
- really intensive computing, needing to take advandage of
multithreading on a multi-processor system
- writing low level code, like OS Kernel, Drivers, etc.

For those kind of task, C (or Ada or... ) could be a better choice.

> why should i continue with python?
Because it's easy, powerful, and really funny, and it's ok for most of
what you could have to do as a beginner.

Now you should *also* learn other languages. But you'd better go with
a lo level languages and a functionnal one - C and Lisp could be good
choices.

Once you'll know Python *and* C *and* Lisp, you should not have big
troubles learning many other (Php included) !-)

laotseu

> thank you!
> 
> _____________________________________________________
> I'm going to prove the impossible really exists


From dbasch at yahoo.com  Mon Sep 30 11:59:40 2002
From: dbasch at yahoo.com (Derek Basch)
Date: Mon, 30 Sep 2002 08:59:40 -0700 (PDT)
Subject: Catching user defined exceptions from foreign modules
Message-ID: <20020930155940.23964.qmail@web20805.mail.yahoo.com>

Everyone,
You were correct that the test1 module shouldn't have
worked at all. I needed to change the first line to:

from test2 import TestError, TryItOut

I had test2 open in the PythonWin IDE so it was
already in the namespace.
Which brings me back to my original issue of why it
works here and not in my larger module. Is it normal
to have to import user defined exceptions? I dont
rememeber having to import any of the built in
exceptions. So, here is the output of the fixed test1:

test2.TestError
fudge

Am I doing this all wrong?
Thanks,
Derek Basch




 
Derek Basch fed this fish to the penguins on Sunday 29
September 2002 
07:22 pm:

> I know that TestError has been imported because I
can
> put a "print TestError" before the except clause in
> test1 and I get "test2.TestError". Can anyone tell
me
> why I am having to put the test2.TestError instead
of
> just TestError? It confuses me even further that my
> small test modules work like I would expect and it
> doesn't work on the larger modules.
>

I don't even see why "test1" is working. You
explicitly import the 
exception by name, but you didn't bring in
"TryItOut"... That alone 
should have raised an exception.


[wulfraed at b... wulfraed]$ python test1.py
Traceback (most recent call last):
File "test1.py", line 14, in ?
letitrip.run()
File "test1.py", line 7, in run
tryitout = TryItOut()
NameError: global name 'TryItOut' is not defined


__________________________________________________
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
http://sbc.yahoo.com



From noone at nowhere.com  Fri Sep 13 02:43:34 2002
From: noone at nowhere.com (Marcus)
Date: Fri, 13 Sep 2002 06:43:34 GMT
Subject: Tkinter experts out there?
Message-ID: <mOfg9.461067$UU1.75649@sccrnsc03>

I am trying to get some answers on using Tkinter, does anyone know if there
are channels or forums being actively used?

Thanks
Matthew Bonig




From gkrause at psipenta.com  Fri Sep 13 05:39:02 2002
From: gkrause at psipenta.com (Gernot Krause)
Date: Fri, 13 Sep 2002 11:39:02 +0200
Subject: dynamic includes from classes...
References: <Vz3g9.246505$kp.845705@rwcrnsc52.ops.asp.att.net>
Message-ID: <alsasa$1s7j$1@innferno.news.tiscali.de>

Hi
> First, is there a better way than exec to dynamically include a module?
> i.e.
> importvar = "foo"
> exec("from %s import %s" % (importvar, importvar))
(Sorry for my horrible english)
In my mind,its better - you implement a interface class.
If you know the criteria - on script parse time - which class would be used
,
then try this.

DependClass = importvar

class Interface (DependClass):
    def __init__(self):
       DependClass.__init__(self)
    def __del__(self):
       DependClass.__del__(self)
...
Gernot







From dgallion1 at yahoo.com  Sat Sep 14 09:28:56 2002
From: dgallion1 at yahoo.com (darrell)
Date: Sat, 14 Sep 2002 13:28:56 GMT
Subject: Help with regular expression using findall and .*?
References: <mailman.1031934617.4693.python-list@python.org>
Message-ID: <sQGg9.50185$xm.10721692@twister.nyroc.rr.com>

Here's an example with backtracking turned off
>>> s="""a\nb\n1""" 
>>> re.findall("[^\n]+?\d", s) 
['a\nb\n1'] 

Which is not correct.
What ever pattern proceeds the '+' must be reevaluated as the pattern moves 
forward. sre handles this though recursion.

import re
s2=('macro\n'+'a'*20000+'\norcam\n')*10
s2split=re.split("macro\n|\norcam\n",s2)
for r in s2split:
    print r


This should be fast also.
--Darrell


czrpb wrote:
> Harvey:
> 
> Great thanks!! And thanks for sticking to my question's requirements.
> <wink!>
> 
> Ok, this is what we thought around here. But what I do not understand is
> why any backtracking data is being kept? The '?' in '.*?' means it is
> non-greedy right? When would backtracking ever occur using '.*?'? What am
> I missing?
> 



From cliechti at gmx.net  Sun Sep  1 07:19:22 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 1 Sep 2002 13:19:22 +0200
Subject: Import error depending on where importing from
References: <20020831.231529.1350573793.13153@cybermesa.com>
Message-ID: <Xns927C877E79D25cliechtigmxnet@62.2.16.82>

"Jay O'Connor" <joconnor at cybermesa.com> wrote in 
news:20020831.231529.1350573793.13153 at cybermesa.com:
> For certain reasons, I had to put the MySQLdb files in my own directory
> (called 'lib').   Now, when I'm in the lib directory I can import MySQLdb
> and use it fine.  However, when I'm one directory up, and I try to do
> "from lib import MySQLdb" I get 
> ImportError: _mysql: init failed

you need to make sure that the module is in a directory that is in 
"sys.path". you can add you "lib" with 
"sys.path.append(os.path.abspath('lib'))" or whatever is appropriate for 
you.
 
chris

-- 
Chris <cliechti at gmx.net>



From Nicola.Musatti at ObjectWay.it  Mon Sep 16 07:56:30 2002
From: Nicola.Musatti at ObjectWay.it (Nicola Musatti)
Date: Mon, 16 Sep 2002 13:56:30 +0200
Subject: Good book choice?
References: <e6c4872c.0209150635.4816ef98@posting.google.com> <05k9ou4o2m9e41puce9ni3953b5hbhn8c0@4ax.com>
Message-ID: <3D85C6EE.CB7A1F52@ObjectWay.it>


robin at execulink.com wrote:
[...]
> Can't help you with those, but have had luck with Learning Python
> (Mark Lutz & David Ascher, pub. O'Reilly), a great introduction to the
> language that doesn't pad out its length with stuff you can read in
> the reference documentation. I think it's the single-best Python book
> out there. So do the other developers I work with.

While this is definitely a good book, I feel it is a bit slow paced for
someone with programming experience. It is also based on a rather old
version of the language. I'm not sure whether a new edition is in the
works or not.

Cheers,
Nicola Musatti


From skip at pobox.com  Tue Sep 10 14:52:37 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 10 Sep 2002 13:52:37 -0500
Subject: how to build wxPython w/ Sun's C++ compiler?
In-Reply-To: <m3k7ltitat.fsf@mira.informatik.hu-berlin.de>
References: <mailman.1031607011.28525.python-list@python.org>
        <m3k7luojpq.fsf@mira.informatik.hu-berlin.de>
        <mailman.1031621231.16841.python-list@python.org>
        <23891c90.0209092359.11acb607@posting.google.com>
        <mailman.1031662034.25815.python-list@python.org>
        <m3k7ltitat.fsf@mira.informatik.hu-berlin.de>
Message-ID: <15742.16245.999727.145361@12-248-11-90.client.attbi.com>

    >> Agreed.  However, disutils commands have this structure:
    >> 
    >> python setup.py [ options ] command [ more options ] ...
    >> 
    >> which I think can get kind of confusing.

    Martin> You forgot 

    Martin>   python [ options ] setup.py [ options ] command [ more options ] ...

    Martin> :-)

I finally concluded that you either need to use command line switches or ini
files plus a setup.py command line switch which enables the .ini file.  If
you use just .ini files which are always enabled, you might well wind up
screwing up a Python install.

Skip



From alienoid at is.lg.ua  Tue Sep  3 06:03:05 2002
From: alienoid at is.lg.ua (Ruslan Spivak)
Date: Tue, 3 Sep 2002 13:03:05 +0300
Subject: code doesn't work: question
Message-ID: <1551859543.20020903130305@is.lg.ua>

Hello python-list user,

I want somebody to explain me why next code doesn't work:

var = 5

if (tmp = var) == 5:         # this is the error line
   print "ok"
else:
   print "error"

Best regards,
 Ruslan                          mailto:alienoid at is.lg.ua




From whisper at oz.net  Thu Sep  5 20:07:21 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 5 Sep 2002 17:07:21 -0700
Subject: try blocks in Python vs. C++
In-Reply-To: <roy-13BC67.19534805092002@reader1.panix.com>
Message-ID: <GCEDKONBLEFPPADDJCOEKEDEENAA.whisper@oz.net>

I think the biggest reason why try blocks aren't suggested for use in C++ is
that they increase code size substantially (as do some other nice features).

Try...except is good practice in Python IMO.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Roy Smith
> Sent: Thursday, September 05, 2002 16:54
> To: python-list at python.org
> Subject: try blocks in Python vs. C++
>
>
> I'm reading Stroustroup's C++ book, and came upon a piece of advice to
> avoid using try blocks.  I use try blocks a lot in Python and gather
> it's common practice among pythonistas.
>
> Is there something fundamentally different about try blocks in Python
> and C++ which makes them more attractive to use in one vs. the other?
> On the surface, they appear to be virtually identical in just about
> every respect.  Is it simply a cultural difference?  Or just the
> author's own personal style?
> --
> http://mail.python.org/mailman/listinfo/python-list




From lu_gio at hotmail.com  Mon Sep 23 05:20:47 2002
From: lu_gio at hotmail.com (Max)
Date: Mon, 23 Sep 2002 11:20:47 +0200
Subject: PyGame + PyOpenGl + ZBuffer?
Message-ID: <3D8EDCEF.70F827F6@hotmail.com>

Hi!
I found my code is not working well in all computers: with a SiS 300/200
graphic card, the Z buffer seems disabled.

The OpenGL initialization code is:

pygame.init()
pygame.display.set_mode((640,480), OPENGL|DOUBLEBUF|HWSURFACE,32 )
glEnable(GL_DEPTH_TEST)
glDepthFunc(GL_LESS)
glDepthRange(0,1)

How can I force Opengl to utilize ZBuffer?




From andrew.henshaw at mail.com  Sun Sep 15 00:54:47 2002
From: andrew.henshaw at mail.com (Andrew Henshaw)
Date: Sun, 15 Sep 2002 00:54:47 -0400
Subject: Skipping Parts of a For Loop
References: <pan.2002.09.14.12.23.44.103401.723@thomas-guettler.de>
Message-ID: <uo84kscns3l067@corp.supernews.com>

Thomas Guettler wrote:

> Hi!
> 
> Is there a way to skip some parts in a for loop?
> 
> 
> for i in range(3):
> print "i before", i
> i=2
> print "i after", i
>
...snip... 
> 
> Is there a way to skip some parts?
> 
> "continue" skipps one element. But how can
> I skip several?
> 

While not really answering your question... In my opinion, it would be 
better to use a while loop for that purpose.  The code would be much 
clearer.

-- 
Andy Henshaw


From aleax at aleax.it  Wed Sep 18 09:16:28 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 18 Sep 2002 13:16:28 GMT
Subject: Copy with __slots__
References: <mailman.1032349042.19649.python-list@python.org>
Message-ID: <M0%h9.109669$ub2.2318985@news1.tin.it>

Griebel, Peer wrote [I'm only leaving the relevant parts of the script]:

> class C1(object):
>     __slots__ = "s1";
> 
> class C2(C1):
>     __slots__ = "s2";

Funny use of semicolons here.  I'd suggest omitting them!

> o2 = C2()
> 
> import copy
> p2 = copy.copy(o2)
> print p2.s2           # attribute s2 does not exist (has not been copied)
> 
> The last print statement fails with the an AttributeError. Since

Traceback (most recent call last):
  File "sloco.py", line 25, in ?
    p2 = copy.copy(o2)
  File "/usr/local/lib/python2.2/copy.py", line 84, in copy
    y = _reconstruct(x, reductor(), 0)
  File "/usr/local/lib/python2.2/copy_reg.py", line 68, in _reduce
    dict = getstate()
TypeError: a class that defines __slots__ without defining __getstate__ 
cannot be pickled

This is with Python 2.2.1 -- as you see, it doesn't get anywhere
as far as the print statement.  The error message is somewhat
peculiar (who's trying to _pickle_ anything...?) -- and if you
do try to define __getstate__ (and possibly __setstate__), you'll
see they're no use for copy.copy (nor copy.deepcopy).  _reconstruct
gets in the way.

If you only care about copy.copy, the simplest alternative is:

class C2(C1):
    __slots__ = "s2"
    def __copy__(self):
        x = self.__class__()
        x.s1 = self.s1
        x.s2 = self.s2
        return x

You can of course also add try/except AttributeError around each of the 
assignments "x.sN = self.sN", if you want to let an instance of C2 be 
copied even when one or both of its attributes are not yet defined.
For deepcopy, you could use a similar special method __deepcopy__ (needs
to deepcopy the attributes, of course).

The way to cover copy.copy, copy.deepcopy AND pickle is to have the
class supply a special method __reduce__.  __reduce__ can return a
pair where the first item is callable, the second item is the tuple
to pass as arguments to the first.  Thus, for example:

class C2(C1):
    __slots__ = "s2"
    def _build(s1, s2):
        x = C2()
        x.s1 = s1
        x.s2 = s2
        return x
    _build = staticmethod(_build)
    def __reduce__(self):
        return C2._build, (self.s1, self.s2)

Here, _build need not worry about deep-copying -- if needed, it's
been performed previously, on the args _build gets passed.


The nicest thing about it is that a class with slots will often
already supply some nice function to build instances of the class
given values to put in the slots -- you don't necessarily have
to write factory functions such as _build just for the purpose
of supporting copy, deepcopy and pickle... you often have them
lying around!-).  For example, a more typical case might be:

class C2(C1):
    __slots__ = 's2'
    def __init__(self, s1=None, s2=None):
        C1.__init__(self, s1)
        self.s2 = s2
    def __reduce__(self):
        return self.__class__, (self.s1, self.s2)

Here I'm assuming C1 has a similar __init__, but that's a detail.
The point is that __reduce__ can use class C2 itself as the
first item of its result -- the class IS callable, after all.

I'm specifically using self.__class__ rather than a literal C2
here as a tiny enhancement -- if a class C3 subclasses C2 and
does not override __init__ in a way that requires arguments
different from the slots' values, then C3 need not override
__reduce__ either -- it can just inherit it from C2.  Not the
end of the world, but sometimes handy.


Alex



From max at alcyone.com  Wed Sep 25 18:43:06 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 25 Sep 2002 15:43:06 -0700
Subject: a python puzzle
References: <mailman.1032989974.25321.python-list@python.org>
Message-ID: <3D923BFA.B42928F6@alcyone.com>

Michal Wallace wrote:

> Here's a python puzzle. Solve it and you win a prize. :)

tr hokdrafwfezxtculyjqmbipsvHIFg importsysngadfzjehulvcwbxICSk

I started by decoding the (obvious) Python source, which made the job
considerably easier.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Seriousness is the only refuge of the shallow.
\__/ Oscar Wilde
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From use-net at schabi.de  Wed Sep 18 16:24:21 2002
From: use-net at schabi.de (Markus Schaber)
Date: Wed, 18 Sep 2002 22:24:21 +0200
Subject: Remarks to Python 2.3
Message-ID: <20020918222421.328088eb.use-net@schabi.de>

Hello,

I studied http://python.org/dev/doc/devel/whatsnew/contents.html and was
very impressed - you manage to get python even more fascinating with
every release.

I especially like enumerate() and Bool.

But I have some remarks:

- Mutable sets support some in-place operations - should those, just as
the in place operations on lists, like sort(), return None instead of
the new list to make clear that they are in place operations and for the
sake of consistency?

- When replacing list elements using the extended slice syntax - e. G.
in a[::2] = range(0, -2, -1) - the left side must have exactly the same
numer of elements as the right side. So, to cite the example, a[::2] =
range(3) is illegal for a=[0, 1, -1, 3]. 

Now I could image that it makes sense to allow one single additional
element to be added to the list in case only the step is specified and
the operation doesn't produce a gap. So the given example above would be
valid, giving a sequence of five elements, whereas with a=[0, 1, 2] it
would stay illegal (as it would produce a gap). a[0:3:2]=range(3) would
stay illegal as it explicitly specifies the length of the range,
a[0:3:5]=range(3) could be allowed.

Thanks,
Markus

-- 
Schabi's Flohmarkt: http://www.uni-ulm.de/~s_mschab/flohmarkt/

"XML combines all the inefficiency of text-based formats with most of
the unreadability of binary formats :-)" - Oren Tirosh


From gerhard.haering at opus-gmbh.net  Wed Sep 18 14:13:01 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 18 Sep 2002 18:13:01 GMT
Subject: isFloat: Without Exception-Handling
References: <ama6nt$n6i$06$1@news.t-online.com> <pan.2002.09.18.11.40.50.766802.29524@presidency.com> <%82i9.1366$bh1.943@fe03> <pan.2002.09.18.13.21.24.660912.29524@presidency.com>
Message-ID: <slrnaohg7l.2jg.gerhard.haering@haering.opus-gmbh.net>

Rajarshi Guha wrote:
>>> def isFloat(string):
>>>   if type(string) == type(1.0):
>>>     return 1
>>>   else
>>>     return 0
>>>
>>> This should do what you want without exceptions
>
> If you call it as isFloat(1.2) it works

It answers no interesting question, however. Cos if I want to test if something
is a float or not i use type() or isinstance() anyway.

The original question was how to test if a /string/ can be converted to a
float.

-- Gerhard


From bokr at oz.net  Wed Sep  4 20:51:03 2002
From: bokr at oz.net (Bengt Richter)
Date: 5 Sep 2002 00:51:03 GMT
Subject: Memory fault - core dumped   while compiling python 2.2.1 on  SCO
References: <3D757EFA.9050401@nospam.yahoo.com> <yu99sn0qqw1r.fsf@europa.research.att.com> <m3bs7etem9.fsf@mira.informatik.hu-berlin.de> <3D765A05.6030004@nospam.yahoo.com> <j48z2h61j9.fsf@informatik.hu-berlin.de> <3D768753.8030302@nospam.yahoo.com> <m3n0qx8j3w.fsf@mira.informatik.hu-berlin.de> <3D76A139.1030304@nospam.yahoo.com>
Message-ID: <al69pn$sg$0@216.39.172.122>

On Wed, 04 Sep 2002 17:11:37 -0700, Anton Graph <aglyportREMOvethispart at nospam.yahoo.com> wrote:

>Martin v. Loewis wrote:
>> Anton Graph <aglyportREMOvethispart at nospam.yahoo.com> writes:
>> 
>> - What is the precise version of the operating system you are using?
>> 
>
>SCO_SV bongo 3.2 5.0.5 i386
>
>> - I'm troubled by the fact that in both configurations (with an
>>   without gcc), the compiler is always referred-to as "cc". Are you
>>   sure you are not using gcc?
>> 
>I'm positive it's SCO cc:
>
>Wed Sep  4 16:17:04 PDT 2002 bongo:/u/src/agl/txb/txt $ gcc -v
>Reading specs from /usr/local/lib/gcc-lib/i386-pc-sco3.2v5.0.5/2.95.2/specs
>gcc version 2.95.2 19991024 (release)
>Wed Sep  4 16:17:07 PDT 2002 bongo:/u/src/agl/txb/txt $ /usr/local/bin/cc
>bash: /usr/local/bin/cc: No such file or directory
>Wed Sep  4 16:17:14 PDT 2002 bongo:/u/src/agl/txb/txt $ /usr/bin/cc
>Usage: cc [ options ] files ...
>Wed Sep  4 16:17:18 PDT 2002 bongo:/u/src/agl/txb/txt $ /bin/cc
>Usage: cc [ options ] files ...
>
How about
    cc -v
?

what does

    which cc

give you? And

    which cc | xargs -ld
?


Regards,
Bengt Richter


From tundra at tundraware.com  Wed Sep  4 18:15:07 2002
From: tundra at tundraware.com (Tim Daneliuk)
Date: 04 Sep 2002 22:15:07 GMT
Subject: [ANN] 'abck'  2.2 Intrusion Reporting System Released
Message-ID: <3D768506.50200@tundraware.com>

'abck' 2.2 is released at:

     http://www.tundraware.com/Software/abck/

Several new and nifty features/options since the last release (1.99).

'abck' is a Python program designed to manage attack attempts on your Unix
machine.  It examines the contents of /var/log/messages for evidence
of an intrusion attempt.  When it finds such a record, it helps you
determine the source and allows you to send email to the system
administrator in the responsible domain.

'abck' was written for and tested on FreeBSD 4.6R, but it should be
easily adapted to Linux or one of the other *nix platforms.

Upgrade to the FreeBSD port has also been submitted.
-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com



From gerhard.haering at gmx.de  Sat Sep  7 09:35:26 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 7 Sep 2002 13:35:26 GMT
Subject: Python, CGI, MySQL -> 'Error 2004: Can't create TCP/IP socket (10106)'
References: <mailman.1031403490.14051.python-list@python.org>
Message-ID: <slrnank04u.12l.gerhard.haering@lilith.my-fqdn.de>

Doru-Catalin Togea wrote in comp.lang.python:
> I am using Python 2.2 on a win 2000 Pro computer to interact with
> mysql, by means of the MySQLdb module. My non-CGI scripts work as
> expected. 
> 
> I have now installed a webserver (Savant 3.1) am I try to connect to
> mysql by means of CGI and still MySQLdb. My scripts fail now with
> the 'Error 2004: Can't create TCP/IP socket (10106)' error message.

> Anyone who can recommand a free webserver which interacts well with
> MySQL by means of pythonic CGI scripts on Win 2000 Pro?

There's nothing special about MySQL, but I can recommend Apache, even
on Windows.

> [...] These functions work when the script connects to the
> mysql-server directly, but fail when a webserver is in between.

Just guessing: Maybe you're using IIS and the CGI scripts don't have
the privilege to create sockets?

-- Gerhard


From marklists at mceahern.com  Mon Sep 23 17:38:11 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 23 Sep 2002 16:38:11 -0500
Subject: Elementary confusion
In-Reply-To: <878z1s1kqf.fsf@smtp.blueyonder.co.uk>
Message-ID: <JHEOKEOOLIGLDHCMAHMOOENJEAAA.marklists@mceahern.com>

[Keith]
> I am working through O'Reilly's "Learning Python". I am up to page 33,
> and already I have a problem. The text outlines a cli entry about math
> operators, all really straight forward;
>
> .>>>a=3
> .>>>b=4
> .>>>b/2+a
>  5
> .>>>b/(2.0+a)
>  0.8
>
> This is so straight forward that I wasn't going to type it in, but I
> did anyway, and got instead of '0.8' something I cannot account for. I
> got '0.80000000000000004'.
>
> So after 32 pages, I am already flumoxed. What is going on

This is a FAQ.  Please read the FAQ.  This has to do with floating point
math:

  http://www.python.org/doc/FAQ.html#4.98

// mark




From sismex01 at hebmex.com  Fri Sep 27 10:41:33 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Fri, 27 Sep 2002 09:41:33 -0500
Subject: Python Process that can survive
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F0623302639DA6@mail.hebmex.com>

> From: peeryc at hotmail.com [mailto:peeryc at hotmail.com]
> 
> Alright... I love python but I think I may have found a definite
> problem. I can't seem to run a python process, kick it to the
> back-ground, and then kill the terminal that started it. With most
> other things, the python process would keep running in the
> back-ground regardless. For me, the process always dies with the
> terminal.
> 
> So my question: is there a way to separate a python script from the
> underlying terminal so that this will not happen.
> 
> I'm hoping for some scheme that's supported by the language. I've
> found a few hacks on-line where the python process is daemonized
> but this requires shutting down all the standard io streams.
> There's got to be a better way especially since it can be done in
> Perl very easily with no special tricks.
> 
> By the way, I'm doing this on a linux box.
> 
> Chris
> 

This same question I asked a few weeks ago, and got a few
answers back from lotsa nice peepl, so I pass this function
on to you for it's safekeeping (tag, your it ;-)

def Daemonize():
   "Detach the current process from the console."
   import os, sys
   if os.fork(): os._exit(0)

   # Close connection to current console.
   os.close(sys.__stdin__.fileno())
   os.close(sys.__stdout__.fileno())
   os.close(sys.__stderr__.fileno())

   # Reopen std handles to /dev/null.
   dev_null = os.open("/dev/null", 0)
   os.dup2(dev_null, 0)
   os.dup2(dev_null, 1)
   os.dup2(dev_null, 2)

   # Create new OS session.
   os.setsid()

   # Reset UMASK.
   os.umask(0)

   # Refork to have init as parent.
   if os.fork(): os._exit(0)


This is a synthesis of several replies I received, and it
seems to work, but others might have some observations
to make to this little function.

HTH!

-gustavo










Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From pinard at iro.umontreal.ca  Tue Sep  3 15:41:10 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Tue, 03 Sep 2002 15:41:10 -0400
Subject: Servizi Internet
In-Reply-To: <slrnan9qi1.25g.gerhard.haering@haering.opus-gmbh.net> (Gerhard
 =?iso-8859-1?q?H=E4ring's?= message of "3 Sep 2002 16:59:40 GMT")
References: <mailman.1031056925.7383.python-list@python.org>
	<3bl9nu8924gihfj2nlpog874vfm73qqbh9@4ax.com>
	<slrnan9qi1.25g.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <oqptvu6euh.fsf@titan.progiciels-bpi.ca>

[Gerhard H?ring]

> The only means you have is complain to their ISP (abuse@{name_of_ISP}).

This is rather discouraging, you know.

Most ISP around here, and especially the big ones, could not care any less.
Thinking about it, some international providers are not fully pink either.
They are far more interested into protecting the anonymity of those from which
they receive money.  Aren't they in business, after all?

They will almost never take action unless forced to do so by police
intelligence services, and such services are so horribly heavy or slow to get
in motion that it is almost a lost cause as far as practice is concerned.

Police people affected to computer crimes told me that they are severely
understaffed, and fully overwhelmed by requests emanating from big banks or
military agencies, which are to be served in priority.  Moreover, in far too
many cases, they have youth protection law in their way.

We pay all the way.  Spammers use our network connectivity, our CPUs, our
disks and our own time for bothering us; they misredirect rejects so companies
pay postmasters or sysadmin services to handle their crap; and everybody
invest effort and energies trying to escape the constant annoyance.  While
politicians, from their ivory towers, discourse on the virtues of Free Speech!

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



From branded at sendme.cz  Thu Sep 19 08:08:32 2002
From: branded at sendme.cz (A)
Date: Thu, 19 Sep 2002 14:08:32 +0200
Subject: Is there a widget or program  for a database browsing
Message-ID: <3D89DA60.8878.8D30C74@localhost>

Hi,
In my program users will search  a database and I need to display choosen
record(s ) in a nice graphical way. Is there a module( program), that I could use together 
with Python, that  could easily  show the database record(s) and take care of all thinks 
during browsing? Thanks for help.
 Ladislav





From cbbrowne at acm.org  Thu Sep  5 11:33:24 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: 5 Sep 2002 15:33:24 GMT
Subject: Graham's spam filter
References: <mailman.1029992314.27908.python-list@python.org> <3D647B4B.28D9BF04@alcyone.com> <mailman.1030050253.11330.python-list@python.org> <3D655AE6.A097AA25@alcyone.com> <ak44j1$1f2ft4$2@ID-125932.news.dfncis.de> <c2ef5940.0209041912.16e102fb@posting.google.com> <3D770156.2E55864C@alcyone.com>
Message-ID: <al7tg4$1o5a9k$5@ID-125932.news.dfncis.de>

Centuries ago, Nostradamus foresaw when Erik Max Francis <max at alcyone.com> would write:
> Aaron Swartz wrote:
>
>> I've been using bogofilter[1], Eric Raymond's Graham-derived spam
>> filter which threw away base64-encoded data and 90% of all spam that
>> got past the filter was base64-encoded. Therefore, I think that base64
>> content really needs to be decoded. I wrote a base64-decoding filter
>> in Python for it and the problem has gone away.
>
> Indeed.  I've been finding very much the same thing with my rule-based
> filter; about 90% of the spam that's getting through is base64 encoded. 
> I haven't yet taken the next step of automatically decoding the base64
> text parts (and then just processing that), but as you have discovered
> it is an obvious solution to the obvious problem.

Have you considered simply replacing strings that appear to be
base64-encoded with a token like "base64-text"?

That allows the database to at least be aware that the spam commonly
contains base64 data.

-> Supposing there is interesting text encoded (such as source code
   for a virus) inside the base64 stuff, it _would_ be useful to
   decode it; 

-> Supposing the base64 stuff is basically just a GIF/JPEG/PNG, or
   something else that doesn't contain "interesting text," you'll
   have not much of value from the decoding process.

Making the "tokenizing" step a tad smarter (e.g. - recognizing "this
is likely base 64" and collecting stats on numbers of lines of base64
material) requires minimal added effort, and I expect it would buy you
_most_ of the benefits of decoding.
-- 
(concatenate 'string "chris" "@cbbrowne.com")
http://cbbrowne.com/info/linux.html
Let me control a planet's oxygen supply and I don't care who makes the
laws.


From d_blade8 at hotmail.com  Fri Sep 27 15:07:20 2002
From: d_blade8 at hotmail.com (Lemniscate)
Date: 27 Sep 2002 12:07:20 -0700
Subject: Silly question about ungzipping a file
References: <7396d2b2.0209261311.1fdd2d3a@posting.google.com> <kSTk9.149109$ub2.3289741@news1.tin.it>
Message-ID: <7396d2b2.0209271107.2e3bfd21@posting.google.com>

Wow, thanks a bunch.  Let me just say that this did the trick with
flying colors.  Of course, now I have to go buy FL's book (my wife was
giving me grief but now I have an excellent argument in my cause). 
Thanks a bunch,
Lem


Alex Martelli <aleax at aleax.it> wrote in message news:<kSTk9.149109$ub2.3289741 at news1.tin.it>...
> Lemniscate wrote:
>         ...
> >>>> import gzip
> >>>> myFile = gzip.open('LL_tmpl.gz')
> >>>> file('output.txt', 'w').write(myFile.read())
> >>>> 
> > 
> > Now, is there any way to do this so that less memory is used?  I mean,
> 
> import shutil
> shutil.copyfileobj(myFile, open('output.txt', 'w'))
> 
> > 129MB.  I'm sure I'm just missing something simple, but why is that?
> 
> I think the reason you're missing this simple something is that you
> are not entirely familiar with Python's rich standard library -- a
> defect common to us all (I don't recall _every_ useful module either:-).
> 
> A partial help is /F's "Python Standard Library" (O'Reilly) -- it
> won't magically infuse you with knowledge of every such module, but
> by giving at least one example script for each module in the
> library (all in a remarkably concise 280 pages!) it makes it quite
> a bit more likely that more modules will stay in your subconscious
> memory.  Python's online reference, Beazley's New Riders book, my
> Nutshell when it comes (the delay ain't my fault THIS time - the
> complete, technically-revised draft is currently in the editor's
> hands!-), each give a different perspective on Python's standard
> library, and enhance the likelihood that you'll know where to find
> excellent code ripe for reuse.  Probably the best single piece of
> advice is the one in the Python online docs -- "sleep with the
> Library Reference under your pillow" (and use a thin pillow, so that
> the knowledge has an easier time filtering up into your brain:-).
> 
> 
> Alex


From gmuller at worldonline.nl  Sat Sep  7 05:20:03 2002
From: gmuller at worldonline.nl (GerritM)
Date: Sat, 7 Sep 2002 11:20:03 +0200
Subject: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com>
Message-ID: <alcgg0$8p4$1@reader1.tiscali.nl>

"Stephen Ferg" <steve at ferg.org> schreef in bericht
news:b16e4ef7.0209062009.26e218e2 at posting.google.com...
> A small page devoted to "Thinking in Tkinter"
> at http://www.ferg.org/thinking_in_tkinter/index.html
> for whoever might find it useful.
>
<...snip...>
I enjoyed reading it! Good work.

regards Gerrit
--
www.extra.research.philips.com/natlab/sysarch/





From anthony_barker at hotmail.com  Wed Sep  4 10:17:02 2002
From: anthony_barker at hotmail.com (Anthony_Barker)
Date: 4 Sep 2002 07:17:02 -0700
Subject: Python critical mass /turning point
Message-ID: <899f842.0209040617.711664fe@posting.google.com>

I have been trying to determine what is holding python back from 
reaching the critical mass/success that perl/c++/Java have experienced.
Caveate - I am not a python guru - so some of the technical stuff may be wrong.

I believe most of the issues are not technical - If technical features were 
key to critical mass then LISP would be the main programming 
language.

1. Interpreter Available on every desktop/server
Perl has this on the server
Java had this in the browser space

Solution(s)
Parrot/Perl6 integration - the if python is included
with the perl interpreter it would be everywhere
Lobby Sun & IBM to include Python on their Unix CDs
Lobby Microsoft to get added to their resource kit
Other options are already available such as freeze....

2. Suitability for commercial products
A couple of the problems I believe are technical and
reputation.
1) Speed issue compared when compared with c/c++
2) The ability to enforce interfaces
Allow enforced type checking and encapsulation - people are often
surprised when they find out data is hidden in python.
3) Access c libraries/apis seamlessly

Solutions(s)
Speed issue I believe may be solved by psycho, and time. As python
matures it seems to be getting faster, while computers are getting
so fast speed matters less.

Having a compiler would allow python to be know not only as a 
scripting language. Unfortunately "scripting language" for some
is equivalent to - not a real language. Enforced interfaces
and the ability to enforce data hiding

SWIG is doing a good job with the last issue.


3) Big corporate sponsor(s)
IBM sponsors Java/Perl/C/C++ & Cobalt 
by having IDEs and corporate support available for those languages.

IBM supports PERL?
Surprise - Tivoli products are glued together with Perl
Lotus Unix server products use perl under the hood
Perl is shipped on every AIX box
Perl courses are available as is support
etc

Microsoft Supports
VB/C/C++/c#/Java(kind of)/Perl etc

Perl is included in the NT/Win2K resource kits
Also is part of the tools for Unix

Sun
Java/Perl/etc

Perl is bundled with Solaris
Sun Solutions has Perl courses available


Cheers,

Anthony
http://xminc.com/anthony/


From franke at ableton.com  Tue Sep  3 08:41:13 2002
From: franke at ableton.com (Stefan Franke)
Date: Tue, 03 Sep 2002 14:41:13 +0200
Subject: Better tracebacks for exec and execfile
Message-ID: <3i99nusjbgpiepva9pkuv3mtl7equ7h88h@4ax.com>

1. I'm toying araound with a text templating engine containing
embedded Python scripts. These scripts are run with exec.
If they throw an exception, the traceback looks like

  File "<string>", line 5
     e = a+c+b+
               ^
 SyntaxError: invalid syntax

Since the scripts are extracted from a bigger file, the filename
"<string>" and line number are almost useless. Is there a way
to pass both to the interpreter and make them appear in
tracebacks?

2. If a command file passed to execfile throws an exception,
again the traceback looks like this:

  File "D:\Programme\Python22\del.py", line 7, in ?
    execfile(filename)
  File "<string>", line 5
     e = a+c+b+
               ^
 SyntaxError: invalid syntax

Why isn't the filename shown instead of "<string>"?




From fperez528 at yahoo.com  Sun Sep 29 19:41:29 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Sun, 29 Sep 2002 17:41:29 -0600
Subject: [Mac] Editor on Macintosh?
References: <1fj7za2.1erbqps1uvzpj8N%news@agapow.net> <an7u1s$33de$1@nntp6.u.washington.edu>
Message-ID: <an8339$ld6$1@peabody.colorado.edu>

In article <1fj7za2.1erbqps1uvzpj8N%news at agapow.net>, news at agapow.net 
(Paul-Michael Agapow) wrote:

>Sadly, it seems like the excellent editor Pepper is going away. (See
><http://www.hekkelman.com>.) I found it the best editor for Python on
>the Macintosh, having never been really happy with the alternatives. 
>
>So what are other MacPython using? There are some Java-based editors,
>but I'm not persuaded that that they'll perform adequately. There's are
>numerous X11 based options, but who wants to swap into X everytime they
>need to edit a file? The once excellent editor Alpha seems to have
>disappeared permanently into testing for its next release.
>
>Ideas? Anyone? Bueller?

Well, now that the mac has a real operating system, you can use:
http://fink.sourceforge.net/pdb/search.php?s=emacs
http://fink.sourceforge.net/pdb/search.php?s=vim

or you can probabbly compile jed (http://space.mit.edu/~davis/jed/) in a 
minute if you want something a bit lighter than emacs but with the same 
overall feel.

All three fantastic tools, powerful beyond anyone's needs and different enough 
that one of them hopefully will meet your tastes.

Since all three will run happily in a terminal, no need for X11.

Cheers,

f.


From yaipa at aol.com  Fri Sep 13 04:19:44 2002
From: yaipa at aol.com (yaipa)
Date: 13 Sep 2002 01:19:44 -0700
Subject: [Tutor] Paradox  table
References: <mailman.1031848942.4346.python-list@python.org>
Message-ID: <8d148763.0209130019.57fcef08@posting.google.com>

montana <sarmxiii at knology.net> wrote in message news:<mailman.1031848942.4346.python-list at python.org>...
> Is there a good example on how to use a CSV parser in Python?
> So you have this comma seperated file, does the CSV parser break data 
> up into groups based upon where they are positioned in the file or does 
> it just generate a list of data or both?
> 
> Is there any good sites out there that deal with CSV parsing and python 
> that have some good examples or can someone on this list give some?
> 
> Thanks.
> 
> SA
> 
> "I can do everything on my Mac I used to do on my PC, plus alot more 
> ..."
> --Me
> 
> On Thursday, September 12, 2002, at 09:35 AM, Derrick 'dman' Hudson 
> wrote:
> 
> > On Wed, Sep 11, 2002 at 03:14:30PM +0200, A wrote:
> > | Hi,
> > |  Is there a module for reading Paradox database (.DB) files with 
>  Python?
> > | I need to run the script under Linux, so I can not use ODBC.
> > | Any suggestion would be appreciated
> >
> > One way to access the data is to export it from paradox in CSV format.
> > CSV is an open standard and several (python) parsers exist for it.
> > However, using this method won't allow you to easily access the data
> > from both paradox and from your script, instead you would need to
> > re-export it every time you modified it in paradox and you would not
> > be able to modify it using your script.  If you can, it would probably
> > be easiest to migrate to a different db such as postgresql.  Postgres
> > databases are much more accessible than paradox is (at least with the
> > paradox version I've seen, 3.5).
> >
> > -D
> >
> > -- 
> > Dishonest money dwindles away,
> > but he who gathers money little by little makes it grow.
> >         Proverbs 13:11
> >
> > http://dman.ddts.net/~dman/
> > <mime-attachment>

Not sure if this is what you are looking for, but maybe this...

--- <snip> --- <snip> --- <snip> --- <snip> --- <snip> ---
#! /usr/local/bin/python

def cfgLstToLst(_cfgTags = [], tokenSeparator = None):
# Note: If tokens are separated by whitespace ONLY 
# (whitespace includes newlines)
# then leave tokenSeparator == None.


   _cfgTagList      = []

   for _i in _cfgTags:
      # split the string at the newline char(s)
      # while consuming the newline char
      #~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      _cfgTagList = _cfgTags.split("\n")

      # Rejoin the list into one string w/o
      # newline char(s).
      #~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      _cfgTags =  " ".join(_cfgTagList)

      # split the string once more and remove
      # the commas. Presto! We now have
      # a list of tags. We will strip the
      # whitespace from each element next.
      #~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      if None:
         # do Nothing
         pass
      else:
         _cfgTagList = _cfgTags.split(tokenSeparator)
     
      
      # This is a List Comprehension.
      # The function of THIS list comprehension is to 
      # strip each list element of whitespace.
      #~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      _cfgTagList = [item.strip() for item in _cfgTagList] 

      if _cfgTagList:
         # remove any empty list items, [''], we may have
         # picked up.
         _cfgTagList = [_x for _x in _cfgTagList if _x ]
         del _x

   return _cfgTagList

if __name__ == "__main__":
    
   # turn a string of comma seperated tokens
   # into a python list.

   data = "this, that, the, other,"

   print data

   print cfgLstToLst(data, ",")

--- <snip> --- <snip> --- <snip> --- <snip> --- <snip> ---

Viking1% cfgLstToLst.py

>> this, that, the, other,
>> ['this', 'that', 'the', 'other']

Viking1%


From just at xs4all.nl  Mon Sep  9 11:42:56 2002
From: just at xs4all.nl (Just)
Date: Mon, 09 Sep 2002 17:42:56 +0200
Subject: Windows/DOS: double clicking a .py file
References: <just-A936EC.17062909092002@news1.xs4all.nl> <alie4t$1q4jol$1@ID-59885.news.dfncis.de>
Message-ID: <just-F18BA3.17425609092002@news1.xs4all.nl>

In article <alie4t$1q4jol$1 at ID-59885.news.dfncis.de>,
 "Thomas Heller" <theller at python.net> wrote:

> "Just" <just at xs4all.nl> wrote in message 
> news:just-A936EC.17062909092002 at news1.xs4all.nl...
> > Sorry if this is a Windows FAQ, I'm a newbie to Windows, but not to
> > Python.
> >
> > If I double-click a Python script (or use the start command in the dos
> > prompt) a new dos prompt window is opened and my program is run. The
> > window goes away when the program is done. However, the window _also_
> > goes away when the program raises an exception, making it virtually
> > impossible to actually _see_ the exception... Is there a way to keep the
> > dos prompt window in this case?
> One solution is outlined here:
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/65287
> This is useful for the development machine. It keeps the DOS window
> open, and leaves you in the Python debugger, where you can play around
> even more to find out about the state of your program.
> >
> > A py2exe-fied version of the program shows the same behavior, which
> > leads my users to report problems like "the output file is mysteriously
> > truncated" instead of "I'm getting a traceback. Here, have a look."...
> >
> > Any suggestions?
> 
> For py2exe-packages apps you should redirect sys.stdout and sys.stderr
> into a log-file.

Thanks for both suggestions. Another workaround just occured to me: do a 
raw_input("type enter to exit") in an exception hook. That should work 
just fine in a py2exe-fied app as well I suppose.

Just


From zopestoller at thomas-guettler.de  Wed Sep 18 11:41:47 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Wed, 18 Sep 2002 17:41:47 +0200
Subject: isFloat: Without Exception-Handling
Message-ID: <ama6nt$n6i$06$1@news.t-online.com>

Hi!

Is there a way to write the following
method without using exceptions?

def isFloat(string):
     is_float=1
     try:
         float(string)
     except:
         is_float=0
     return is_float

print isFloat("asdf") # --> 0
print isFloat("0.1")  # --> 1

Somehow I don't like to catch exceptions. Maybe
that's because my first language was C.

  thomas



From bokr at oz.net  Thu Sep 26 19:55:52 2002
From: bokr at oz.net (Bengt Richter)
Date: 26 Sep 2002 23:55:52 GMT
Subject: e vs exp()?
References: <3a838998.0209191933.1f22e977@posting.google.com> <mailman.1032497547.8453.python-list@python.org>
Message-ID: <an06q8$67j$0@216.39.172.122>

On Thu, 19 Sep 2002 21:51:25 -0700, "James J. Besemer" <jb at cascade-sys.com> wrote:

>
>Buckshot wrote:
>
>>Can anyone here explain why I'm getting different results from
>>e**2 and exp(2)?  Which is more accurate?  
>>
>>Python 2.2 (#1, Mar 27 2002, 14:56:58)
>>[GCC 2.95.3 20010315 (release)] on sunos5
>>Type "help", "copyright", "credits" or "license" for more information.
>>  
>>
>>>>>from math import *
>>>>>e**2
>>>>>        
>>>>>
>>7.3890560989306522
>>  
>>
>>>>>exp(2)
>>>>>        
>>>>>
>>7.3890560989306504
>>  
>>
>Mathematica says e**2 to 30 digits is:
>
>    7.3890560989306502 272304274606
>
We can also brute force the series for e with longs and maybe ignore the last few digits:

 >>> def calce(nd):
 ...     numi = n = 10L**nd
 ...     deni = 1L
 ...     i = 0; s = ''
 ...     while 1:
 ...         i += 1
 ...         numi = numi*i + n
 ...         deni *= i
 ...         sl, s = s, str(numi/deni)
 ...         if sl[:nd+1] == s[:nd+1]: return numi, deni
 ...
 >>> n,d = calce(60)
 >>> n/d
 2718281828459045235360287471352662497757247093699959574966967L
 >>> (n/d)**2
 7389056098930650227230427460575007813180315570551847324087124...
Comparing side by side:
 7.3890560989306502 27230427460575007813180315570551847324087124...
vs
 7.3890560989306502 272304274606 from matlab
 7.3890560989306522              for e**2, and
 7.3890560989306504              for exp(2)

The latter seems more accurate.

>in which case the second form would happen to be more accurate.
>
>Since Python math uses your machine's native floating point formats, 
>precision is necessarily limited and one should not be surprised to 
>encounter minor errors in the least significant bits or digits.
>
>Although Mathematica recognizes that the two forms are identical, Python 
>uses different formula for computing the two results.  The fact that the 
>base is "e" is of no consequence to the computation.  exp() is computed 
>from C's exp() runtime library routine, whereas e**2 is evaluated using 
>C's "pow()" function.  
>
>In general, I would expect pow() to be less accurate, as pow(x,y) 
>probably is computed by exp(log(x)*y).
>
Looks plausible.

Regards,
Bengt Richter


From freddie at madcowdisease.org  Fri Sep 13 20:00:49 2002
From: freddie at madcowdisease.org (Freddie)
Date: Sat, 14 Sep 2002 00:00:49 GMT
Subject: Argh! Python 2.2 + threading + sockets = trouble
Message-ID: <R_ug9.753$Ee4.1406@news-server.bigpond.net.au>

Hi,

I'm a fairly new Python programmer, but I've written a decent sized 
project (threaded IRC offer bot), and I wrote a threaded IPv6 capable 
IRC bouncer yesterday. Of course, it doesn't work quite as I expected.

MadCowOffer (offerbot) has rarely works properly with Python 2.2. It 
will sometimes manage to start all the threads, then it's fine, but it 
will usually just stop working after the IRC connection thread is started.

cowbounce (bouncer!) is almost the same. It will load fine, start the 
daemon sockets, wait for connections, etc etc. After a while (1 
connection, 3, it varies), the main thread also just stops doing 
anything. Some of the already connected BounceClient() threads will keep 
working, and some won't.

I have absolutely no idea what's causing this, or how to find out what 
exactly is making it unhappy :\


Using Python 2.2.1, from the Debian 'testing' packs python-2.2 2.2.1-6. 
Using 2.1.x isn't an option, as it doesn't seem to have working IPv6 
support.


Links to code:
   cowbounce   - http://www.madcowdisease.org/Files/cowbounce-broken.tar.gz

   MadCowOffer - http://www.madcowdisease.org/Files/MadCowOffer-0.1.2.tar.gz


Any help would be GREATLY appreciated :)

Freddie



From ruediger.maehl at web.de  Mon Sep  2 01:59:56 2002
From: ruediger.maehl at web.de (Rüdiger Mähl)
Date: 2 Sep 2002 05:59:56 GMT
Subject: Document Conversion and Python??
References: <mailman.1030594135.17322.python-list@python.org>
Message-ID: <Xns927D515F03A9Bruedigermaehlwebde@193.101.67.2>

Derek Basch <dbasch at yahoo.com> wrote in 
news:mailman.1030594135.17322.python-list at python.org:

> Hello,
> Does anyone have any suggestions for tools for
> converting documents between different formats
> (html,pdf,tab delimeted, etc...)?

The docutils (docutils.sf.net/docs/rst/quickstart.html) 
might also be interesting for you.

R?diger


From printers at sendme.cz  Mon Sep 23 03:47:08 2002
From: printers at sendme.cz (A)
Date: Mon, 23 Sep 2002 09:47:08 +0200
Subject: How to call a procedure in exe file
Message-ID: <3D8EE31C.32426.219719@localhost>

Hi,
I know that it is possible to call a function in DLL library( even though I have never tried it) 
from Python but is it possible to call a procedure in an exe program?
I have a program that downloads some files( that was written in Delphi, we bought that 
program compiled already in exe).
I must start the program and then push a button that runs a procedure for downloading. 
And I must do it al the time when I want to download a file. So I am thinking about 
downloading automatically from Python. I can start the program
with os.startfile but I do know know how to start the procedure that is started when I push 
the button for downloading.(I may possiblly find address of the procedure
Can you please help?
Ladislav






From martin at v.loewis.de  Mon Sep 16 01:48:11 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Sep 2002 07:48:11 +0200
Subject: Are generator functions thread-safe?
References: <T%dh9.6427$Lg2.1180859@news2.news.adelphia.net>
Message-ID: <m31y7uzdr8.fsf@mira.informatik.hu-berlin.de>

"Robert Oschler" <Oschler at earthlink.net> writes:

> I know that a call to a generator function actually returns an independent
> iterator object (mea culpa if I'm wrong on this), but are there any hidden
> gotchas that would cause the use of the same generator function from
> multiple threads, to result in a sharing (especially local storage) or
> memory collision issue?

Terminology is tricky here: If you call a function that is a
generator, you get a generator object. So it is not clear to me
whether you have the same function in multiple threads, or the same
generator object.

If you access the same generator object from multiple threads, they
clearly share state.

If you call the same generator function from multiple threads, you get
a new generator object each time. You also get a new set of local
variables for each invocation. There is no "hidden" shared state. So
yes, generator functions are thread-safe.

Regards,
Martin



From marklists at mceahern.com  Wed Sep  4 11:01:27 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 4 Sep 2002 10:01:27 -0500
Subject: How to force built-in commands over imported.
In-Reply-To: <3D761C39.4020802@akst.tn.tudelft.nl>
Message-ID: <JHEOKEOOLIGLDHCMAHMOOEEODMAA.marklists@mceahern.com>

[Maurice van de Rijzen]
> There is a built-in commans called open(fileName). There is also a
> open() command in the os-module open which needs more than the fileName.
>   Originally I didn't import the os-module and everything worked fine.
> However, when I imported the os-module I received an error that open
> needed more arguments. How can I force to take the built-in open()
> instead of the open() in the os-module.

How are you importing os.open?  If you do it like this:

  import os
  f1 = os.open(...)
  f2 = open(...)

you can keep the two open methods separate.  However, if you do it like
this:

  from os import open
  ...

you can't.  Unless you do this:

  from os import open as open_from_os
  ...

But then why would you do that?  ;-)

The idea here is namespaces.  Using 'from module import function' is not
recommended unless you know what you're doing.

Also, for what it's worth, you may want to consider using:

  f = file(...)

instead of:

  f = open(...)

since the method name matches the type:

  >>> type(f)
  <type 'file'>

// m

-




From mwh at python.net  Fri Sep 27 06:18:14 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 27 Sep 2002 10:18:14 GMT
Subject: SET_LINENO
References: <3D92A46E.2030706@cygnus-software.com>
Message-ID: <lk8z1nwxjp.fsf@pc150.maths.bris.ac.uk>

I'm the one responsible for SET_LINENO going away in 2.3, so here's
my take...

Bruce Dawson <bruce_deletethis_dawson at cygnus-software.com> writes:

> Executive summary: Has the SET_LINENO removal been tested with the HAP
> debugger, 

No.  I'd never heard of it for one thing!

> and wouldn't a breakpoint instruction give much larger
> benefits?

I fail to see how.

> As a partial author of the HAP Python Debugger (available on source
> forge) I read about the imminent removal of SET_LINENO with some
> worry. It *sounds* like it has been done in such way to not cause
> debugger problems, but I'm curious if anyone has tested HAP with 2.3.

I tried really hard to make it transparent to debuggers.  There are a
few very small differences, but unless you're actually scanning for
SET_LINENO instructions they shouldn't bite.  I'd invite you to test,
of course.

> Also, it sounds like the reason for removing SET_LINENO was to improve
> the performance of Python scripts. However one of the posts on this
> topic (can't find it now) suggested that the hoped for improvements
> had not actually materialized.

It speeds programs run without -O.  On Windows, programs run with -O
got somewhat slower after this patch (on Linux it made essentially no
difference, maybe a very slight speedup).  See Tim's post.

> It seems to me (from the sidelines/peanut gallery/wherever) that the
> biggest performance problems in the SET_LINENO area are when
> debugging, not when running free, and that therefore this optimization
> is targeting the lesser problem.

You've got my motivation entirely backwards there.

> The need to check for breakpoints at each line has always made Python
> debugging performance poor at best. Technically the performance under
> the debugger doesn't matter, and yet sometimes it does. Having just
> helped complete a Python based sports game for PC, where real-time
> performance was critical, I can say that the slowdown in HAP was
> definitely annoying.

You're faced with a hard problem.  Good luck :)

> So, wouldn't a break-point opcode be a more substantial improvement
> than removing SET_LINENO? Just bringing up that old topic again to see
> if it gets any more traction :-)

How would a breakpoint opcode differ from what SET_LINENO was?

> P.S. For reference I include the canonical? link to the SET_LINENO
> discussion.
> 
> http://mail.python.org/pipermail/python-dev/2000-July/007652.html

Note that what I did bears little resemblance to the scheme sketched
in that post.  I hack stuff out of co_lnotab.

> P.P.S. A breakpoint instruction would also be a boon for asserts - I'd
> love to be able to stop on the assert when running under a debugger,
> to see what is wrong before the stack frame unwinds.

Well, you break on the exception, don't you?

Cheers,
M.

-- 
  The "of course, while I have no problem with this at all, it's
  surely too much for a lesser being" flavor of argument always
  rings hollow to me.                       -- Tim Peters, 29 Apr 1998


From joconnor at cybermesa.com  Tue Sep  3 09:58:13 2002
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 03 Sep 2002 13:58:13 GMT
Subject: code doesn't work: question
References: <1551859543.20020903130305@is.lg.ua> <mailman.1031048764.951.python-list@python.org>
Message-ID: <3d74bdc4.281693@news.cybermesa.com>

On 03 Sep 2002 12:28:59 +0200, Geert-Jan Van den Bogaerde
<gvdbogae at vub.ac.be> wrote:

>Python doesn't allow assignment inside expressions, the rationale (from
>the Python tutorial at http://www.python.org/doc/current/tut/node7.html
>- section 5.5 last sentence): 
>
>"""
>Note that in Python, unlike C, assignment cannot occur inside
>expressions. C programmers may grumble about this, but it avoids a
>common class of problems encountered in C programs: typing = in an
>expression when == was intended. 
>"""
>
>Given the number of times I've personally made this mistake, I'd agree
>with Python's rationale for leaving this out.


This is really only a problem in a language that doesn't really have a
boolean data type and then allows boolean evaluation of non-boolean
objects (such as C with numbers and pointers and Python with..almost
everything :)

Smalltalk allows assignment within an expression, but does not suffer
from the "I meant == and accidentally typed =" problem because
Smalltalk has a first class Boolean class and does not allow boolean
comparisons for non-boolean objects

(x :=y) ifTrue: [
	"blah"
]

Only works if y is already a Boolean; if it's an Integer, the ifTrue:
message raises an error, so 

((x := y) == 5) ifTrue: [

]

is perfectly ok, and much safer than the equivalent C


Take care,
Jay
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor


"God himself plays on the bass strings first, when he tunes the soul"


From spam at ob_scure.dk  Sat Sep  7 10:55:21 2002
From: spam at ob_scure.dk (Thomas Jensen)
Date: Sat, 07 Sep 2002 16:55:21 +0200
Subject: Using Python to call a DCOM server - "'Class not registered"
References: <3D787D95.9000907@ob_scure.dk> <5%ce9.1272$_l1.2960@news-server.bigpond.net.au>
Message-ID: <3D7A1359.4010607@ob_scure.dk>

Mark Hammond wrote:
> Thomas Jensen wrote:

[snip - example of not working DCOM]

> Are you sure that the COM object you are using is correctly registered 
> for use as a DCOM server.  The code above insists that the COM object be 
> registered as LocalServer - ie, other than just InprocServer32.

I fought with this 'till 5 AM this morning (0300 CET) and finally got it 
working.
You are absolutely right! The components are made in VB6 (NOT my 
prefered language, but hey - I gotta make a living :-). I don't really 
know very much about the technicalities of COM (or DCOM for that 
matter), but when I saw how simple the testDCOM.py script was, I thought 
I'd give it a try :-)
I don't really know what the relation between an InprocServer32 and a 
LocalServer is, but I somehow figured out, that by changing the project 
type from "ActiveX DLL" to "ActiveX EXE", the component showed up in 
dcomcnfg.exe allowing me to finally get it working!

Now I just need to figure out how to unregister this new .EXE beast 
("regsvr32 /u" doesn't seem to work).

Is it safe to say that "ActiveX DLL" = InprocServer32 and "ActiveX EXE" 
= LocalServer and that I need the later to use DCOM? (I have some 
existing components that I'd rather keep as DLL - don't know why really? :-)

Anyhow, thanks alot for the reply and not least	for the great win32 package!


-- 
Best Regards
Thomas Jensen
(remove underscore in email address to mail me)



From jb at cascade-sys.com  Fri Sep 13 01:42:33 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Thu, 12 Sep 2002 22:42:33 -0700
Subject: function pointer
References: <ffcedbc9.0209110215.644c1788@posting.google.com> <aln5ce$5kt$1@news.netpower.no> <slrnanu7i1.22k.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <3D817AC9.4020302@cascade-sys.com>

Gerhard H?ring wrote:

>Broken quoting - yuck! Your newsreader is broken. Please repair it (using
>third party software) or make Usenet a better place by throwing away that
>crapware.
>

I believe this error occured when the line was wrapped by the 
python-list mailer, not necessarily the submitter's newsreader or email 
program.

>If we want to keep the terminology right, Python doesn't have variables. It
>has names and values. 
>

The fact that The Tutorial and other Python doc uses the term "variable" 
makes it confusing for newcomers.

It's doubly confusing, since the general definitions of a variable seems 
to apply equally well to the corresponding Python thingies.

    http://www.webopedia.com/TERM/V/variable.html

    http://www.techweb.com/encyclopedia/defineterm?term=variable

So, what is the difference between a "name" and a "variable"?  

Or is this merely a distinction without a difference?

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From tdelaney at avaya.com  Wed Sep 25 19:36:18 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 26 Sep 2002 09:36:18 +1000
Subject: comparing all values of a list to regex
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5D8@natasha.auslabs.avaya.com>

> From: Alex Martelli [mailto:aleax at aleax.it]
> 
> num_matches = [ there.match(item) is not None for item in 
> alist ].count(1)
> 
> if num_matches == 0:
>     againanothernewlist.append(alist)
> elif num_matches == 1:
>     anothetnewlist.append(alist)
> elif num_matches == len(alist):
>     anewlist.append(alist)

I would actually prefer to do this another way.

match_all = []
match_one = []
match_some = []
match_none = []

# Expanded out for clarity - could be one line
matches = map(regexp.match, alist)
matches = filter(None, matches)
matches = len(matches)

# Find out which list to append to

match_list = match_all

if not matches:
    match_list = match_none
elif matches == 1:
    match_list = match_one
elif matches != len(alist):
    match_list = match_some

match_list.append(alist)

Note that I added an additional case - where more than one matched, but not
all.

Tim Delaney



From gua81 at XXXyahoo.com  Thu Sep 26 04:29:54 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Thu, 26 Sep 2002 18:29:54 +1000
Subject: IGNORE ABOVE POST!
References: <amug6m$ivm$1@lust.ihug.co.nz>
Message-ID: <amugh7$j8o$1@lust.ihug.co.nz>

Sorry,
my mistake...
my web server took a while o refresh itself.

"CheapSkate" <gua81 at XXXyahoo.com> wrote in message
news:amug6m$ivm$1 at lust.ihug.co.nz...
> This thing worked before.... I dont know why it didn't work now.
>
> this code is in the other page with the link:
> '<a href="parser.cgi?link=%s/%s">%s</a>'% (dirname, filename, filename)
>
> This code is the one that crash:
> all_values=cgi.FieldStorage()
> kkk=all_values["link"].value
> start_parse(kkk)
>
> when I run it, it gives this error:
> Traceback (most recent call last):
>   File "parser.cgi", line 53, in ?
>     direc=all_values["link"].value
>   File "/usr/local/lib/python2.1/cgi.py", line 550, in __getitem__
>     raise KeyError, key
> KeyError: link
>
>




From maurice at akst.tn.tudelft.nl  Wed Sep  4 10:44:09 2002
From: maurice at akst.tn.tudelft.nl (Maurice van de Rijzen)
Date: Wed, 04 Sep 2002 16:44:09 +0200
Subject: How to force built-in commands over imported.
Message-ID: <3D761C39.4020802@akst.tn.tudelft.nl>

Dear all.
I'm a newbie on the field of python, therefore the answer to my question 
may be simple.
There is a built-in commans called open(fileName). There is also a 
open() command in the os-module open which needs more than the fileName. 
  Originally I didn't import the os-module and everything worked fine. 
However, when I imported the os-module I received an error that open 
needed more arguments. How can I force to take the built-in open() 
instead of the open() in the os-module.

Gr,
Maurice



From codic at stasis.org  Wed Sep 11 16:11:48 2002
From: codic at stasis.org (Ian Crawford)
Date: 11 Sep 2002 13:11:48 -0700
Subject: Global Interpreter Lock
Message-ID: <3822aca1.0209111211.8a1538f@posting.google.com>

Hi everyone,
    I'm working on a project with an embedded Python
interpreter which I'd like to be capable of parallel
execution on a multi-processor machine.  I'm still
trying to figure out one thing in particular.  Is
there a global interpreter lock per interpreter (i.e.
if I call Py_NewInterpreter, does that interpreter
have it's own lock), or will I have to embed seperate
interpreters in their own processes to have parallel
execution?

Thanks,

Andrew


From itsmikeytime at hotmail.com  Mon Sep  9 11:22:01 2002
From: itsmikeytime at hotmail.com (Mikey at Work)
Date: Mon, 9 Sep 2002 11:22:01 -0400
Subject: Possible bug in WxPython
References: <9ff6b66.0209090557.3c0e3c94@posting.google.com>
Message-ID: <alieap$6b7$1@news-int.gatech.edu>

"vtail" <vtail at yandex.ru> wrote in message
news:9ff6b66.0209090557.3c0e3c94 at posting.google.com...
> Greetings.
>
> When trying to run an example program from wxWiKi
> http://wiki.wxpython.org/index.cgi/Getting_20Started, section 7.1.2, I
> encounter following error. Could someone please explain it to me?
> I've attached the error message and the program that causes it.
>
> O:\_PUBLIC\Newdbase\Python\wxPython>python test4.pyw > aaa.txt
> 17:52:59: Debug: c:\Projects\wx\src\msw\app.cpp(542):
> 'UnregisterClass(no redraw
>  canvas)' failed with error 0x00000584 (class still has open
> windows.).

I had the exact same problem this morning.  I got around it by adding:

app.MainLoop()

at the end of the program.  It solved my problems.





From cbbrowne at acm.org  Mon Sep 23 10:20:37 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: 23 Sep 2002 14:20:37 GMT
Subject: WSDL for XML-RPC ??
References: <_YAj9.21316$0p1.377838@news2.ulv.nextra.no>
Message-ID: <amn7vk$79if3$2@ID-125932.news.dfncis.de>

"Thomas Weholt" <2002 at weholt.org> wrote:
> SOAP has WSDL to describe a webservice. Doeas XML-RPC have something
> similar?
> Is WDDX the right way to go or have I totally missed the point?
>
> Ref :
> http://www.wddx.org
> http://www.openwddx.org/

The express intent of XML-RPC is to provide a fairly /minimal/ set of
useful functionality.

Contrast with SOAP, where it's kind of "expressly indicated" that
there will be some nebulous set of "extended standards," where it is
supremely unlikely that anyone would actually be able to implement
them all at once.

The result, with SOAP, is that there's something of a "dog's
breakfast" of implementations, where they may assortedly trumpet:
  "WSDL is supported!"
  "UDDI is supported!"
  "We support 8 alternative transport methods!" (SOAP::Lite for Perl)
  "We support these bits of WSSecurity!"

And every implementation is free (and likely) to choose different bits
to support.  They'll probably /try/ to support WSDL (though support
for that in scripting languages like Perl/Python/Ruby is spotty, at
best); they /might/ support UDDI; other stuff scatters all over the
map, likely only being implemented by Big Vendors...

XML-RPC takes a vastly less ambitious road, not claiming any of that
sort of extensibility.  And so the standard fits on about two pages,
and most implementations provide pretty much complete coverage of it.

There are three fairly common XML-RPC extensions:
"system.listMethods", "system.methodSignature", "system.methodHelp"

These are, in theory, able to provide some WSDL-like information,
albeit being vastly simpler, as XML-RPC eschews the use of XML Schemas
and has no support at all for the "complex schema" of SOAP.

It's not totally evident where WDDX fits in, except as a 'possibly
slightly more convenient way of transferring results of DBMS queries'
than SOAP/XML-RPC.  If you really need the differences between
SOAP/XML-RPC and WDDX< you're probably as well off setting up an XML
'schema' specifically for your application, and transferring around
data as XML documents whose generation and degeneration /you/ control.
-- 
(concatenate 'string "cbbrowne" "@acm.org")
http://cbbrowne.com/info/soap.html
"Starting a project in C/C++ is a premature optimization."
-- Peter Jensen


From gerhard.haering at gmx.de  Fri Sep 20 20:59:25 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sat, 21 Sep 2002 02:59:25 +0200
Subject: Python question on cgi
In-Reply-To: <amgf6b$5ob93$1@ID-49758.news.dfncis.de>
References: <amgf6b$5ob93$1@ID-49758.news.dfncis.de>
Message-ID: <20020921005924.GA3318@lilith.ghaering.test>

* Ken <ken at hotmail.com> [2002-09-21 10:41 +1000]:
> Hi all, how do I pass parameters to a backend C++ program and how do I know
> when does the C++ program ends so that I can read the output file produced
> by the C++ program and print to the HTML interface?

Check out the os.popen and/or the functions in the popen2 module. The
most simplistic example would be:

output_of_other_program = os.popen("/bin/ls /").read()

waiting-to-be-bashed-for-not-closing-the-stream-ly yours,

-- Gerhard



From ch_knievel at gmx.de  Sat Sep 21 12:13:29 2002
From: ch_knievel at gmx.de (christopher knievel)
Date: Sat, 21 Sep 2002 18:13:29 +0200
Subject: [newbie] mod_python apache w32
Message-ID: <ami5rr$des$04$1@news.t-online.com>

Hi all,
i got some problems running mod_python with apache under windows.
I am using python 2.2 and apache 1.3.26.
As written in the manual i configured httpd.conf like this:

LoadModule python_module modules/mod_python.dll

AddModule mod_python.c

<Directory "C:/Programme/Apache Group/Apache/python">
  AddHandler python-program .py
  PythonHandler mptest
  PythonDebug on
</Directory>

I copied the mod_python.dll to C:/Python22/lib/mod_python.

The Apache server is starting with "Apache/1.3.26 (Win32) mod_python/2.7.6
Python/2.2.1 running..."
So i guess it should be working....

The test script looks like this:

#! C:/Python22/python.exe

from mod_python import apache

def handler(req):
    req.send_http_header()
    req.write("<html><body>Hello World!</body></html>")
    return apache.OK

my error.log (last entry):
[Sat Sep 21 18:00:13 2002] [error] [client 192.168.1.99] python_handler:
make_obcallback returned no obCallBack!

i've read several newsgroups and googled a lot but i didn't found a working
solution
Thanks a lot in advance...
christopher knievel




From mangwirop at kwekwepoly.co.zw  Mon Sep  9 10:06:45 2002
From: mangwirop at kwekwepoly.co.zw (Phillip Mangwiro)
Date: Mon, 9 Sep 2002 15:06:45 +0100
Subject: Accessing Tk-inter Popup menu entry
In-Reply-To: <alhm24$j17$1@wanadoo.fr>
Message-ID: <001101c2580a$21c87de0$0700a8c4@ict4>

Can anybody help me to learn Python?




From andy.elvey at paradise.net.nz  Thu Sep 12 04:22:07 2002
From: andy.elvey at paradise.net.nz (Andy Elvey)
Date: Thu, 12 Sep 2002 20:22:07 +1200
Subject: Overloading the ">>" operator in Python 2.2.1
Message-ID: <fhYf9.1418$b5.128061@news02.tsnz.net>

Hi all .
   I'm using Python 2.2.1 under Win98.  I'm doing some experimenting with
parsing, and as part of that, I'm trying to overload the right-shift
operator (>>) so that I can use it as follows (testing against a sequence
that I read in) -
  a >> b   would mean  "a is followed by b" .
   So,  you could have a function like seqtest("string_to_test" ,
"foo">>"bar" ) , and that function would return true or false, depending on
whether foo was followed by bar.
 (  I know it's possible to use regexes for this,  but being able to create
a sequence-testing operator like this would be useful .)
    Many thanks in advance ......







From spam07 at thack.org  Mon Sep 23 14:31:14 2002
From: spam07 at thack.org (Tapio Kelloniemi)
Date: Mon, 23 Sep 2002 18:31:14 GMT
Subject: Non-interpreted Python?
Message-ID: <S5Jj9.360$UC2.34013@read2.inet.fi>

Hi
I'm new to Python and would like to know, wheter it is possible to
compile Python code to binary executables (and on which platforms)? Or
better, if it is possible to compile code into object format and then
link with C/C++/whatever code.

Thanks for all answers!

--
Tapio



From marklists at mceahern.com  Mon Sep 23 23:05:35 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 23 Sep 2002 22:05:35 -0500
Subject: confusion about opening files
In-Reply-To: <3D8FD2F3.81F3831C@alcyone.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOCEPAEAAA.marklists@mceahern.com>

[Erik Max Francis]
> Note:  open() is an alias for file().

I read somewhere that 'file' was preferred to the use of 'open' because of
the symmetry to the type:

  >>> f = file("junk.txt")
  >>> type(f)
  <type 'file'>

Anyway, that's how I've trained myself to spell 'open'.

Cheers,

// m

-




From oren-py-l at hishome.net  Thu Sep 12 03:26:46 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Thu, 12 Sep 2002 03:26:46 -0400
Subject: iterators and generators, is this the Python Way???
In-Reply-To: <3D7FF921.2040305@one.net>
References: <3D7FF921.2040305@one.net>
Message-ID: <20020912072646.GA19246@hishome.net>

On Wed, Sep 11, 2002 at 10:17:05PM -0400, Michael Schneider wrote:
> Hello All.
> 
> I just updated my python to 2.2 and noticed that generators, and 
> iterators were in there.
> 
> I often parse through many files (one at a time), and wanted the speed 
> of read,  and
> the convience of   for x in myGen:

"for x in fileobject" uses xreadlines for readahead buffering. It's faster
than reading a file with readline and should be faster than anything you 
can implement in Python.  If you must have backward compatibility to 
Python 2.1 use "for x in fileobject.xreadlines()".

Python 2.3 makes files into their own iterators. It's even faster than
xreadlines. 

> The code below works just fine,  could you review this code against the
> intent of iterators and generators.

IMHO, much of the intent of iterators and generators is to *avoid* reading 
everying into memory in advance. This way the memory requirements can often
remain constant regardless of input size.  Another good reason not to read
everything into memory is if the source is a real-time stream instead of a
file and you don't even know when it's going to end.

	Oren



From rsrchstr at msn.com  Mon Sep 30 15:33:50 2002
From: rsrchstr at msn.com (mike henley)
Date: Mon, 30 Sep 2002 20:33:50 +0100
Subject: idea for a much needed application
References: <576c1752.0209291326.56dcdf31@posting.google.com> <pasepuc0r73mh5vlcsu1l7erl4e2b17qvi@4ax.com> <an86f7$c2veb$1@ID-157863.news.dfncis.de> <fn0gpu8kh44oa1t7ih2qjmu413gvab7dou@4ax.com>
Message-ID: <ana8v7$c2agg$1@ID-157863.news.dfncis.de>

Dear Sir

With all due and deserved respect, I really don't know why you insist on
arguing over this. I said my point was not about tim burners-lee, not about
economics or politics, and yet you insist on discussing them.
As for IRC; i have used IRC for several years, and AIM is not the only thing
i ever used. Yes, I believe IRC compared to modern instant messaging is a
messy medium, and perhaps a dying one even, and i'm not the only one who
believes so. Several years ago, IRC was lively and interesting, nowadays,
not many people go there. There are too many networks, many servers on the
same network don't communicate with each other, net splits are common
occurence, lags happen from time to time, servers come and go like the
passing wind, here today gone tomorrow, some may let you access, some may
not... etc. There are good reasons why modern instant messaging is easier
and more popular.
However, that's not my point, and anyway, it really don't matter now, i
found a webhost, but thanks a lot for your comments and replies, much
appreciated, honestly.

"TGOS" <tgos at invalid.invalid> wrote in message
news:fn0gpu8kh44oa1t7ih2qjmu413gvab7dou at 4ax.com...
>
> On Mon, 30 Sep 2002 01:38:51 +0100 "mike henley" <rsrchstr at msn.com>
> wrote in comp.lang.java.programmer:
>
>  <snip>
>
> > that's the whole idea of giving a personal webserver "presence", so that
> > others don't have to "go there and wait in my channel till i log on".
>
> And how does it work when you use AIM? They connect to the AIM server
> and wait till you log on. So where's the difference? Where is the
> difference whether they are running a tiny AIM client, that constantly
> monitors the AIM network, to detect when you connect to AIM, compared to
> a small IRC client, the permanently monitors a channel (in an IRC
> network) to detect when you connect to IRC? There is no difference at
> all.
>
> Except that the IRC client supports bots and they can even make a bot
> that will automatically open the browser and go to your page when you
> log on, while this is not possible in AIM.
>
> > Also, i want to be able to give a simple AOL or MSN username to
> > anyone without assuming they can write specific scripts.
>
> They don't have to write a script, YOU must write a script. They can
> install pre-written scripts for mIRC, for example, that will notify them
> once a certain user joins the channel (these already exist hundreds of
> times) - or maybe meanwhile mIRC already has this ability hard-coded,
> who knows.
>
> However, they can write a script, which gives much more power any stupid
> AIM client will ever give them. E.g. they can write a script that
> monitors for your presence and at the very second you log in, it will
> immediately resume a download, that was stopped because you logged off
> last time.
>
> If you are unhappy with the abilities of existing IRC software, write
> your own client for users (if you think you can convince them to use
> it). It doesn't even have to "present" IRC to them, IOW, they may not
> even know that they are connected to an IRC network (not chat window or
> anything). The program just sits in the background and when it detects
> that you log into the IRC channel, it will open a window:
>
> User "mike" just went online, _click here_ to visit his/her page.
>
> And if they click there, they are taken to your page.
> Your "address" would then include the information:
>
> IRC server, channel, nickname
>
> And you could choose the format yourself, if you wrote the client
> yourself, the user would input something like:
>
> irc.service.org:dyndns:mike
>
> The channel dyndns could then be also used by your friends. E.g. if the
> user enters:
>
> irc.service.org:dyndns:pete
>
> into the client, and pete goes online, the client window will say:
>
> User "pete" just went online, _click here_ to visit his/her page.
>
> Every user can add as many "watch addresses" into his client as he
> likes. Maybe this will be a completely new task for IRC, nobody has
> considered so far. You just have to write such a client, if you are
> unhappy with the existing clients.
>
> > IRC is a messy medium.
>
> IRC is no messy medium. Methinks it's above your horizon, as you never
> dealt with anything but AIM so far.
>
> --
> TGOS




From anthony at interlink.com.au  Tue Sep 24 00:37:23 2002
From: anthony at interlink.com.au (Anthony Baxter)
Date: Tue, 24 Sep 2002 14:37:23 +1000
Subject: Outlook 2002 and Python 
In-Reply-To: <03F53405C6ED434A986221919785A6A78C9339@nca-postal.ncaustin.com> 
Message-ID: <200209240437.g8O4bNK23193@localhost.localdomain>


> What:  I saw an idea for a way to create a good SPAM filter for email. 
> 
> How: I would like to try to implement the algorithm in Python.
> 
> Why:  I am trying to get familiar with Python, as I will be using it at
> work quite a bit, and would like to keep my skills honed.
> 
> Where:  Windows XP and Outlook 2002

There's work going on in this area at http://spambayes.sf.net/, 
and if you check the mailing list archives, you'll see that the
issue of integrating it with Outlook just came up in the last day
or so (subject was something like "Practical Applications").

Anthony
-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.




From marklists at mceahern.com  Fri Sep 27 15:31:10 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 27 Sep 2002 14:31:10 -0500
Subject: [XML-SIG] well-formed xml
In-Reply-To: <an2b4m$7g9$1@main.gmane.org>
Message-ID: <JHEOKEOOLIGLDHCMAHMOAECFECAA.marklists@mceahern.com>

[Mike Rovner]
> IIRC, double quotes are the must:
> s = '<a href="http://google.com/search?hl=en&amp;q=foobar">search</a>'

Thanks for the reply.  It seems to be the ampersand:

from xml.dom import minidom
s = "<a href='http://google.com/search?hl=en&q=foobar'>search</a>"
s2 = s.replace("&", "&amp;")
for x in (s, s2):
    try:
        doc = minidom.parseString(x)
    except Exception, e:
        print "Failed: %s" % x
    else:
        print "Succeeded: %s" % x

Output:

Failed: <a href='http://google.com/search?hl=en&q=foobar'>search</a>
Succeeded: <a href='http://google.com/search?hl=en&amp;q=foobar'>search</a>

// m



From padraig.brady at corvil.com  Wed Sep 11 08:02:47 2002
From: padraig.brady at corvil.com (Padraig Brady)
Date: Wed, 11 Sep 2002 13:02:47 +0100
Subject: deleting the first and the last character of a string
References: <mailman.1031741971.20624.python-list@python.org> <3D7F2960.20507@Linux.ie> <20020911115118.GE2054@partagas.as.de.cw.net>
Message-ID: <3D7F30E7.6070406@corvil.com>

Manuel Hendel wrote:
> On Wed, Sep 11, 2002 at 12:30:40PM +0100, Padraig Brady wrote:
> 
>>How about: line = line[1:-1]
> 
> 
> This worked quite good, but I found another problem. There seams to be
> trailing whitespaces. I tried to strip(s) to get rid of them, but this
> doesn't work, I get the following error message:
> 
> Traceback (most recent call last):
>   File "./pop3create.py", line 9, in ?
>       print strip(line)
>       NameError: name 'strip' is not defined
> 
> What's wrong here?
> 
> manuel
> 

You need to do: line.strip()
Note the alternative is to 'import string'
before you 'string.strip(line)' however
it's much better IMHO to use the new
string methods on the string object.

P?draig.




From ulbi at ivs.tu-berlin.de  Mon Sep 23 10:24:50 2002
From: ulbi at ivs.tu-berlin.de (Andreas Ulbrich)
Date: Mon, 23 Sep 2002 16:24:50 +0200
Subject: Jython file manipulation?!
References: <2yej9.32454$gA4.4053@sccrnsc02>
Message-ID: <amn87j$73a$1@mamenchi.zrz.TU-Berlin.DE>

Matthew Bonig wrote:
> I am trying to rename a file in Jython...
> the command I am using is just
> os.rename(filename, newFilename)

This works for me (Jython 2.1, JDK 1.4.1)

You should give us some details, i.e. exceptions etc., to get some 
better answers here.



From sholden at holdenweb.com  Mon Sep  9 16:11:02 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 9 Sep 2002 16:11:02 -0400
Subject: Windows/DOS: double clicking a .py file
References: <just-A936EC.17062909092002@news1.xs4all.nl> <alie4t$1q4jol$1@ID-59885.news.dfncis.de> <just-F18BA3.17425609092002@news1.xs4all.nl>
Message-ID: <3h7f9.77472$l_4.28399@atlpnn01.usenetserver.com>

"Just" <just at xs4all.nl> wrote in message
news:just-F18BA3.17425609092002 at news1.xs4all.nl...
[ how top see messages ]
>
> Thanks for both suggestions. Another workaround just occured to me: do a
> raw_input("type enter to exit") in an exception hook. That should work
> just fine in a py2exe-fied app as well I suppose.
>

python -i

will allow you to enter interactive statements after your execution
finishes, which can be useful when you're debugging.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From syver-en+usenet at online.no  Thu Sep 12 16:12:59 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Thu, 12 Sep 2002 20:12:59 GMT
Subject: embedded Python sample
References: <2a0f4cc4.0209100520.1d3a723b@posting.google.com> <ur8g16aiw.fsf@online.no> <2a0f4cc4.0209110609.1bca6901@posting.google.com>
Message-ID: <ulm67x8zv.fsf@online.no>

monnetb at yahoo.com (bruno Monnet) writes:

> This sample looks great, but I'm coding for windows and the library
> used by python for fopen is not compliant with the multithreaded lib
> used by Visual 6.0 projects ...

Oh yes, I use Visual 6.0 myself. I posted here recently on what to do
with the compiler. But the quick fix is:

Don't use the debug version, use release (check my previous post on
how to debug)
Don't use anything but the multithreaded release dll version of the
cruntime libraries.

Compile and run.



-- 

Vennlig hilsen 

Syver Enstad


From LogiplexSoftware at earthlink.net  Wed Sep 11 16:50:07 2002
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 11 Sep 2002 13:50:07 -0700
Subject: deleting the first and the last character of a string
In-Reply-To: <200209111246.03080.shalehperry@attbi.com>
References: <20020911105818.GC2054@partagas.as.de.cw.net>
	<mailman.1031756234.22942.python-list@python.org>
	<alo5df$hoa$03$1@news.t-online.com> 
	<200209111246.03080.shalehperry@attbi.com>
Message-ID: <1031777408.17116.65.camel@software1.logiplex.internal>

On Wed, 2002-09-11 at 12:46, Sean 'Shaleh' Perry wrote:
> On Wednesday 11 September 2002 12:25, Markus wrote:
> > Sean 'Shaleh' Perry wrote:
> > >> Is there anything which says delete the first or last
> > >> character of a string? I checked the string module on the web, but I
> > >> couldn't find anything.
> > >>
> > >>>> s = '|f|o|o|'
> > >>>> s.strip('|')
> > >>>
> > >>> s[1:-1]
> >
> > 'f|o|o'
> >
> > -Markus
> 
> True, but I like that the strip() call handles the case where one of the 
> delimiters is missing.  One of the bugs that always bit me in perl was using 
> chop() to eat the newline character when there was not a newline there to 
> eat.  I try to avoid putting myself in that situation now.

Let me apologize in advance for this:

>>> def chopd(s, d):
...     s = s.strip()
...     return s[s[0] == d : {0:len(s), 1:-1}[s[-1] == d]]
... 
>>> chopd("|f|o|o|", '|')
'f|o|o'
>>> chopd("|f|o|o", '|')
'f|o|o'
>>> chopd("f|o|o|", '|')
'f|o|o'
>>> chopd("f|o|o", '|')
'f|o|o'
>>> 

Of course, since you are worried about missing delimiters, you also have
to consider whether the data being delimited contains whitespace
(strip() might then remove some of the data).

Have you considered using one of the many delimited-file handling
modules that already exist?

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From lee100 at sport.rr.com  Tue Sep 17 20:49:28 2002
From: lee100 at sport.rr.com (Lee Gray)
Date: Wed, 18 Sep 2002 00:49:28 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <Xns928CED692541cliechtigmxnet@62.2.16.82> <C05DA40F31B09329.0165201CAF094801.13AF9332EB64347A@lp.airnews.net>
Message-ID: <s4Qh9.479726$q53.16320265@twister.austin.rr.com>

Cameron Laird wrote:

>>lpgray at uop.com (Lee Gray) wrote in
>>news:96904b50.0209171305.6d65f574 at posting.google.com:
>>> My boss is pretty open-minded, but also needs to see something working
>>> to be convinced Python is even a viable platform (he had never heard
>>> of it).  Otherwise, .Net is a given, whether it's any good or not,
>>> since corporate is MS all the way.
> .
> While we're pointing out individual applications that
> might be of interest, please keep in mind that Python
> is in serious production use at tens of THOUSANDS of
> sites.  Don't judge it solely by what you see of the
> first two demonstrations someone happens to push your
> way.

I read that all the time (and believe it), but that's not going to impress 
my boss, much. He's still going to want something demonstrated up close and 
personal.

> their efforts only "off the record".  I'm willing to
> bet, though, that Python will be around for longer
> than VBScript.

I can certainly agree with that. There's already conflicting information 
about MS' continuing support of VBS, the Windows Script Host and IE's HTAs 
(HTML Applications) in the scripting newsgroups.  That's one reason I want 
to learn Python personally, but if we're going to use it at work, I need to 
show something quickly to convince anyone (including myself!) that that's a 
good direction.

Thanks,
Lee



From no at mail.com  Mon Sep 16 16:17:12 2002
From: no at mail.com (BT)
Date: Mon, 16 Sep 2002 20:17:12 GMT
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <pUdh9.9377$yt3.4604908@newssrv26.news.prodigy.com> <mailman.1032184821.5180.python-list@python.org>
Message-ID: <c%qh9.102794$ub2.2120685@news1.tin.it>

> |If your code is that valuable, run it on your server remote from end
> |users as a trade secret, or force them to agree to a license and protect
> |it w/ copyright.
>
> I would reformulate the issue slightly:  If you BELIEVE your code is
> that valuable, seek psychiatric help.
>

I really don't understand posts like these.
You think I'm fool to protect my work?
Ok, let's say I make a software in two years and I market it on internet.
The software is good and sell some copy.
You find the software, like it, change the name, add some feature and resell
it.
My time: 2 years, yours: 1 month.

I think you need a psychiatric help.
Only Id shared some source code, after years...
Noooooooo!!! IDDDDD!!! Foolish people!!! You want to sell it???!!! You want
to live with that???!!!
Go working in a mine, and write games for fun!!!
I can give you the phone of my clinic, but stop it NOW, join the ideal
world.




From erikprice at mac.com  Wed Sep  4 10:35:40 2002
From: erikprice at mac.com (Erik Price)
Date: Wed, 4 Sep 2002 10:35:40 -0400
Subject: help on log
In-Reply-To: <009201c253e6$b4f47ba0$2500a8c0@maxitec.co.za>
Message-ID: <959E9362-C013-11D6-BB50-00039351FE6A@mac.com>

This message has popped up quite a few times on c.l.p. and on the Tutor 
list, under two different email addresses even.

If you are really new to any kind of programming, I think this script 
is a little bit advanced.  For instance, I see that you're using the 
regular expression module, but I'm not sure that you understand how to 
use it.  It looks like you're trying to compile a regular expression 
for the text that you want generated by the script, but regular 
expressions help you search for text in an already-existing file.  
Also, you're using the string module to find a "LayerUp" string, but I 
don't see that string anywhere in your log -- so is this really even 
necessary?  (Even if it's necessary to your code, it could be omitted 
from your post.)

Also, few people have the time to sit down and troubleshoot an entire 
script, even a small one like this.  If you knew exactly what your 
problem was, you could isolate the problem from the rest of the script 
and then just ask that question.  Once it's solved, you can move on to 
the next problem.  This mailing list gets so much traffic that I end up 
deleting a lot of the messages without reading them, even the short 
ones.  Long ones which vaguely ask for help and then post a lot of code 
go right to my Trash folder.

My advice to you is that you -not- tackle an entire script like this 
right off the bat.  Fire up your interactive Python interpreter.  
Import the regular expression and string modules.  Open up a file 
handle to your ppp.log file.  Then start playing with the various 
features of the regular expression and string modules, and see what you 
can do interactively.  When you can get something to look good, write 
down what you did.  When you have a straightforward question, go ahead 
and ask it here or on the Tutor list.

I'm not trying to sound mean, but you will find this a lot easier if 
you break it down into smaller steps.  The interactive Python 
interpreter is the ultimate tool for doing this.



Erik

PS: don't use HTML-formatted or Rich Text-formatted email, on some mail 
clients (such as mine) it makes your text render too small to read.  
That only exacerbates your problem.




On Wednesday, September 4, 2002, at 03:42  AM, Morne wrote:

> Could someone please help me with this? Im really new to any kind of 
> programming .If posible would you be
> ?able to answer in detail.
> ?
> Thanx
> ?
> ?HERE IS WHAT I HAVE:
>
> ?import re
> ?import string
>
> ?secondsMatcher=re.compile("Connect time: (\d+) secs")
>
> ?def getStuff(file):
> ???? closeline = 0
> ???? errorcount = 0
> ???? secondscount = 0
> ???? connectcount = 0
> ???? data_outcount = 0
> ???? data_incount = 0
> ???? ppplist = file.readlines()
>
> ???? for line in ppplist:
>
> ?????? if string.find(line,"LayerUp") >= 0:
> ?????????? print "found connect"
> ?????????? connectline = string.split(line)
> ?????????? connectcount = connectcount + 1
>
> ?????? elif string.find(line,"ERROR") >= 0:
> ?????????? print "found error"
> ?????????? errorcount = errorcount + 1
>
> ??????? elif string.find(line,"Connect time") >= 0:
> ?????????? print "found close"
> connectSecs=secondsMatcher.search(line).group(1)
> ?????????? print "Connected for",connectSecs
> ????????? closeline = string.split(line)
> ?????????? secondscount = secondscount + string.atoi(connectSecs)
>
> ?def main():
> ??? f=open("C:/my documents/ppp.log","r")
> ???? getStuff(f)
> ???? return None
>
> ?if __name__=="__main__":
> ???? main()
>
> ?THIS IS WHAT I NEED THE PROGRAM TO DO:
>
> ?I want the program to read the attached ppp.log file and give me the
> ?following type of report:
>
>
> ?PPP Log file summary for 5 Jul 2002
> ?-----------------------------------
> ?Jul? 5 18:15:01 Connected to internet (IP 155.239.150.146)
> ?Jul? 5 18:18:08 Closed connection. ( Connect time: 197 secs: 5091 
> octets
> in,
> ?1864 octets out )
> ?-----------------------------------
> ?Successful connections: 5
> ?Failed connections: 2
> ?Total time online: 456 seconds Total data in: 66576 octets
> ?Total Data out: 66349 octets
>
> HERE IS part of the olg file seems noone wants to touch the 
> Attachment(lol)
>
> > Jul? 5 18:00:00 maxigate newsyslog[87300]: logfile turned over
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Phase: Connected to local 
> client.
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Command: /var/run/internet: 
> quit
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Phase: /var/run/internet: 
> Client
> connection dropped.
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Phase: Connected to local 
> client.
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Command: /var/run/internet: 
> set
> timeout 180
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Command: /var/run/internet: 
> dial
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Phase: bundle: Establish
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Phase: deflink: closed -> 
> opening
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Phase: deflink: Connected!
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Phase: deflink: opening -> 
> dial
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Chat: Phone: 0860007249
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Chat: deflink: Dial attempt 
> 1 of
> 1
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Phase: /var/run/internet: 
> Client
> connection closed.
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Chat: Send: AT^M
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Chat: Expect(5): OK
> Jul? 5 18:15:00 maxigate ppp[34607]: tun0: Chat: Received: AT^M^M
> Jul? 5 18:15:01 maxigate ppp[34607]: tun0: Chat: Received: OK^M
> Jul? 5 18:15:01 maxigate ppp[34607]: tun0: Chat: Send: ATE1Q0^M
> Jul? 5 18:15:01 maxigate ppp[34607]: tun0: Chat: Expect(5): OK
> Jul? 5 18:15:01 maxigate ppp[34607]: tun0: Chat: Received: ATE1Q0^M^M
> Jul? 5 18:15:01 maxigate ppp[34607]: tun0: Chat: Received: OK^M
> Jul? 5 18:15:01 maxigate ppp[34607]: tun0: Chat: Send: 
> ATx1DT0860007249^M
> Jul? 5 18:15:03 maxigate ppp[34607]: tun0: Chat: Expect(40): CONNECT
> Jul? 5 18:15:22 maxigate ppp[34607]: tun0: Chat: Received:
> ATx1DT0860007249^M^M
> Jul? 5 18:15:22 maxigate ppp[34607]: tun0: Chat: Received: CONNECT
> 33600/ARQ/V34/LAPM/V42BIS^M
> Jul? 5 18:15:22 maxigate ppp[34607]: tun0: Phase: deflink: dial -> 
> carrier
> Jul? 5 18:15:23 maxigate ppp[34607]: tun0: Phase: deflink: /dev/cuaa0: 
> CD
> detected
> Jul? 5 18:15:23 maxigate ppp[34607]: tun0: Phase: deflink: carrier -> 
> login
> Jul? 5 18:15:23 maxigate ppp[34607]: tun0: Phase: deflink: login -> lcp
> Jul? 5 18:15:23 maxigate ppp[34607]: tun0: LCP: FSM: Using "deflink" 
> as a
> transport
> Jul? 5 18:15:23 maxigate ppp[34607]: tun0: LCP: deflink: State change
> Initial --> Closed
> Jul? 5 18:15:23 maxigate ppp[34607]: tun0: LCP: deflink: State change
> Closed --> Stopped
> Jul? 5 18:15:24 maxigate ppp[34607]: tun0: LCP: deflink: LayerStart
> Jul? 5 18:15:24 maxigate ppp[34607]: tun0: LCP: deflink: 
> SendConfigReq(179)
> state = Stopped
> Jul? 5 18:15:24 maxigate ppp[34607]: tun0: LCP:? ACFCOMP[2]
> Jul? 5 18:15:24 maxigate ppp[34607]: tun0: LCP:? PROTOCOMP[2]
> Jul? 5 18:15:24 maxigate ppp[34607]: tun0: LCP:? ACCMAP[6] 0x00000000
> Jul? 5 18:15:24 maxigate ppp[34607]: tun0: LCP:? MRU[4] 1500
> Jul? 5 18:15:2
>
>
>





--
Erik Price

email: erikprice at mac.com
jabber: erikprice at jabber.org




From aleax at aleax.it  Fri Sep 20 02:45:27 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 06:45:27 GMT
Subject: Destructor never called ???
References: <76710511.0209111450.40999dd8@posting.google.com> <alp8n3$bsd$1@panix1.panix.com> <4e8lou018f9uosl5p9mb0tt6691mfivamj@4ax.com> <ame8es$ppc$1@panix2.panix.com>
Message-ID: <buzi9.136158$pX1.4848554@news2.tin.it>

Aahz wrote:
        ...
> There are several different issues.  First of all, __del__ does not get
> called if an object is destroyed by GC (GC gets used to clean up cyclic
> garbage).  Secondly, Jython does not use refcounting; it's pure GC, just

Right, but, I'd put it differently: GC is inhibited from destroying
objects whose types define a __del__ -- rather, it collects them into
gc.garbage, because it can't know that finalizing them is safe.  I
find this has a strong inhibitory effect against defining __del__ --
if and when I do, then reference loops become a potential problem
again, i.e., may lead to resource leaks once more.


Alex



From mcfletch at rogers.com  Sat Sep 21 16:35:55 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Sat, 21 Sep 2002 16:35:55 -0400
Subject: WxString and python string module?
References: <36de4780.0209211158.2fc3815@posting.google.com>
Message-ID: <3D8CD82B.6070201@rogers.com>

wxString is automatically converted to/from regular Python string (or 
unicode under wxPython 2.3.3, depending on the wxPython build type).

HTH,
Mike

Decent wrote:
> Dear all:
> 
> I'm trying to rewrite my apps using wxpython,
> but how do I convert WxString object to python string?
> I have to use python's default string module, not the wxWindow's WxString.
> 
> Any Idea?

-- 
_______________________________________
   Mike C. Fletcher
   Designer, VR Plumber, Coder
   http://members.rogers.com/mcfletch/





From Gareth.McCaughan at pobox.com  Mon Sep 16 20:13:34 2002
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: Tue, 17 Sep 2002 01:13:34 +0100
Subject: CFP: UK Python Conferece 2003
References: <Xns928BAD1086D00duncanrcpcouk@127.0.0.1>
Message-ID: <slrnaocste.qcp.Gareth.McCaughan@g.local>

Duncan Booth wrote:
>                     *** Call for Participation ***
>                       UK Python Conference 2003
> 
>                             April 2-3 2003
>                          Holiday Inn, Oxford
...
> To submit a proposal, send an email message to duncan at rcp.co.uk by 2nd
> October 2003. In your message, please include the following:
          ^^^^

I'd like to propose a paper on "Novel applications of
Guido's time machine". :-)

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
.sig under construc


From jon+usenet at unequivocal.co.uk  Mon Sep 23 18:35:48 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Mon, 23 Sep 2002 22:35:48 -0000
Subject: Python and cgi and <a> problem!
References: <ammn6e$lj7$1@lust.ihug.co.nz> <Xns9292EA042C259cliechtigmxnet@62.2.16.82>
Message-ID: <slrnaov5q3.nrp.jon+usenet@snowy.squish.net>

In article <Xns9292EA042C259cliechtigmxnet at 62.2.16.82>, Chris Liechti wrote:
> print '<A href="%s">%s</A>' % (filename, filename)

I hope none of your filenames have '"', ">" or "&" in.


From aleax at aleax.it  Mon Sep 16 03:45:32 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 16 Sep 2002 07:45:32 GMT
Subject: raw post data
References: <Wt8h9.9583$Ee4.22928@news-server.bigpond.net.au>
Message-ID: <w_fh9.113093$pX1.4054147@news2.tin.it>

:B nerdy wrote:

> how would i retrieve the raw post data in cgi python scripts ?

I think something like:

import sys
all_posted_lines = sys.stdin.readlines()

should do it.  Raw POST'ed data should be no more and no less
than the contents of the CGI script's standard-input.


Alex



From pyth at devel.trillke.net  Tue Sep  3 06:03:19 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 3 Sep 2002 12:03:19 +0200
Subject: Why it is so dramatical?
In-Reply-To: <t1wptvv3bpu.fsf@kermit.wreck.org>; from list-python@ccraig.org on Tue, Sep 03, 2002 at 01:04:45AM -0400
References: <200209021107.25761.b.maryniuk@forbis.lt> <mailman.1030961043.19190.python-list@python.org> <3D73E035.2040300@nospam.attbi.com> <t1wptvv3bpu.fsf@kermit.wreck.org>
Message-ID: <20020903120319.L3893@prim.han.de>

Christopher A. Craig wrote:
> djw <dwelch91 at nospam.attbi.com> writes:
> 
> > Stupid question:
> > 
> > If ""join() is 100-1000 X faster, why doesn't the Python interperter
> > translate s = s1 + s2 + ... into s = "".join( [ s1, s2, ... ] ) 
> > automatically?
> 
> You won't see that kind of dramatic results if you did that.  In fact,
> my testing shows it would be slower.  
> 
> s = "".join(s1, s2) is slower than s=s1+s2.

That's not really the question.  The OP wanted to build up a huge string
(for XML/XSTL-processing) in single steps.  Using

    something += otherstring

in *loops* is usually slow.  Doing

    stringlist.append(otherstring)

and at the end

    "".join(stringlist)

is orders of magnitude faster (depending on the size of the strings).  

Another speedy option is to use cStringIO.  See my example in another 
part of this discussion.

    holger



From prittenhouse at hotmail.com  Thu Sep  5 12:30:35 2002
From: prittenhouse at hotmail.com (Phil Rittenhouse)
Date: 5 Sep 2002 09:30:35 -0700
Subject: Installing Python 2.2.1 on WinNT
References: <mOSb9.38460$C8.132490@nnrp1.uunet.ca> <ddc19db7.0209050139.34c12ed1@posting.google.com>
Message-ID: <75848aa7.0209050830.4024ebef@posting.google.com>

Thanks for the feedback Armin, that tells me there must be something
about my install of NT that's not quite right (or at least, not quite
the same as everyone else).

I assume that the download file is OK because it installed fine
on 95, 98, ME, 2K, and XP without problems.

In fact, I have just found that if I install Office 2000 first, the
python install on my NT works cleanly.   There must be some magic that 
the Office install does, maybe a DLL update, or permission change, that 
allows it to work.

Thanks again!
Phil

 
> I have recently installed Python 2.2.1 on a NT 4.0 machine with SP6.
> 
> The install process worked flawless. 
> Is your dowloaded install file OK ??
> 
> - Armin


From thorsten at thorstenkampe.de  Wed Sep 11 10:09:02 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Wed, 11 Sep 2002 16:09:02 +0200
Subject: Python primer - comments appreciated!
References: <aljdhe$1q63hj$1@ID-77524.news.dfncis.de> <w3cf9.535714$m91.21695020@bin5.nnrp.aus1.giganews.com> <allpms$1r005a$1@ID-77524.news.dfncis.de> <LIwf9.447074$Aw4.18721391@bin2.nnrp.aus1.giganews.com>
Message-ID: <alnipo$1pqmu8$1@ID-77524.news.dfncis.de>

*  Terry Reedy
> "Thorsten Kampe" <thorsten at thorstenkampe.de> wrote in message
>> * Terry Reedy
>>>      result.setdefault(func(item),[]).append(item)
>>
>> To make every item hashable, I made this:
>> result.setdefault(repr(func(item)),[]).append(item)

Is repr(seq) a "bulletproof", "canonical" way to make a mutable 
sequence hashable and most important for this algorithm:
repr(x) = repr(y) <=> x = y ?

I think so...

>>>      # {}.setdefault(key,default) was added for just this sort of
>>> usage
>>
>> Sorry, I just don't get it: your mixing (chaining?) a dict method
>> and
>> a list method as if it was a f(g(x)) and /it works/!?! Where is that
>> documented? How is is that evaluated (left to right)?
>
> Yes, left to right.  Dict.setdefault returns the value corresponding
> to the key after setting d[key]=default if key not already in d.  In
> this case, *all* values are initialized to [] by such a call.

You mean the value corresponding to the prior non existent key is set 
to '[]':

>>> D = {'0': [3, 6], '1': [4]}
>>> D.setdefault('2', [])
[]
>>> D
{'1': [4], '0': [3, 6], '2': []}

> Then
> append to the list.  This working depends on there being *two*
> references to the list.  The one in the dict (possibly new, possible
> not) and the one returned by the method and used by append().

Okay, let's make this clear:
D = {'0': [3, 6], '1': [4]}
D.setdefault('1', [])
  ...returns: [4]
[4].append(7)
 ...puts [4, 7] on the "stack"

But where is the instruction saying "store [4, 7] back to D['1']"?
Shouldn't this be: D['1'] = D.setdefault('1', []).append(7) ?

>> (Mathematically spoken, the quotient set is the set of all
>> equivalence
>> classes induced by func, which is what we return)
>
> Since the equivalence lists may have duplicates, they are not,
> strictly speaking, sets.  If the set of return values were made to be
> dicts (or, in 2.3, sets) instead of lists, then they would be.

This is correct. Fortunately the /quotient set/ *is* a set and does 
*not* contain duplicates (duplicate equivalence classes). It's a 
partition and partitions are always disjoint and non-empty.

Thorsten


From thorsten at thorstenkampe.de  Sat Sep  7 16:05:38 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Sat, 7 Sep 2002 22:05:38 +0200
Subject: Simulating "getopt" in functions (newbie)
Message-ID: <aldm7h$1orqou$1@ID-77524.news.dfncis.de>

Hi,

I want to simulate a "getopt"-like behaviour within a _function_. This 
is to avoid having to define two functions: "function" and 
"slightly_different_function".

If the function was a standalone program, it would work like this 
("v" for "verbose"):

$ function.py 5
1

$ function.py -v 5
The result is: 1

I've thought of this to simulate it in a function:
def function(a, b=None):
    if b == "v":
        print "The result is: 1"
    else:
        print "1"

>>> function(5)
1

>>> function(5, "v")
The result is: 1

So: what is the "pythonic", "natural" way to do this in Python?


Thanks in advance,
Thorsten


From a.clarke11 at pop.ntlworld.com  Mon Sep 23 19:10:22 2002
From: a.clarke11 at pop.ntlworld.com (a.clarke11)
Date: Tue, 24 Sep 2002 00:10:22 +0100
Subject: list problem
References: <mailman.1032818129.8126.python-list@python.org>
Message-ID: <3D8F9F5E.DC48B13F@pop.ntlworld.com>

How about:

for item in x:
 while item:
  y.append(item.pop(0))
print'y=',y
?
hope this helps
Tony Clarke
Glasgow

jubafre at brturbo.com wrote:

> i have:
> x=[['0020', '0000x9'], ['0030', '0000xa'], ['00B4', '0000x8'], ['0030', '0000xb'], ['00F0'], ['0001'], ['0003'], ['0005']]
>
> and i want:
> y=['0020', '0000x9', '0030', '0000xa', '00B4', '0000x8', '0030', '0000xb','00F0', '0001',
> '0003', '0005']
>
> how to do this?
>
> Juliano Freitas
> www.gebrasil.hpg.com.br



From loewis at informatik.hu-berlin.de  Mon Sep 30 04:34:47 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 30 Sep 2002 10:34:47 +0200
Subject: sizeof PyTypeObject difference
References: <j5Rl9.29807$IO5.4027049@e3500-atl1.usenetserver.com> <m34rc8qamo.fsf@mira.informatik.hu-berlin.de> <mrTl9.30085$IO5.4052423@e3500-atl1.usenetserver.com>
Message-ID: <j4adlz3mfs.fsf@informatik.hu-berlin.de>

"Tom Widrick" <lordmacro at hotmail.com> writes:

> Well, aren't instances of a metatype types?  I know this statement was
> probably made from invalid assumptions, but just as a point.
> Since types are variable sized, as you say, how does this tie in with
> tp_basicsize and metatypes?

The instances of the type type are types, yes. If you have your own
metatype, it is your decision to determine what instances of it are.

The snippet you posted looks consistent, so the error must lie
elsewhere.

As for variable-sized objects: the instances of the type type are
variable-sized; if you create your own metatype, it is your decision
what the layout of them is. Of course, if you use functions of the
type type to implement your meta-type, you have to follow the
conventions of the type type, which are undocumented, and may be
unaccessible to external code.

> > That is hard to believe. Declaring the typedef alone can't cause a
> > core dump. Defining an object of that type, and putting the value of
> > 196 into it can't cause a core dump, either. You must be doing other
> > things, like using the resulting object in some way, to cause a core
> > dump.
> 
> Trying to use the metatype as a metaclass for a python class.

That's still not precise enough: Did you merely use the type as the
__metaclass__ attribute, or did you also try to create instances of
the resulting type?

> As an aside, are there any other quarks about making metatypes in C? the
> xxsubtype.c example in the distrib is helpful, along with the PEPs, but I'm
> yet to find a good example of a metatype implemented in C (besides the 2
> builtin metatypes, but I don't think those serve as  great minimalistic
> examples).

xxsubtype demonstrate how to subtype a type in C. Creating metatypes
in C is not really supported; you need to study the source code, and
may find that the things you want to do cannot be done without
changing typeobject.c.

In any case, I recommend to contact Christian Tismer.

Regards,
Martin


From emile at fenx.com  Thu Sep 26 11:51:31 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 26 Sep 2002 08:51:31 -0700
Subject: revenge of the python puzzle
References: <Pine.LNX.4.44.0209261018590.22263-100000@hydrogen.sabren.com> <JHEOKEOOLIGLDHCMAHMOAEMFEBAA.marklists@mceahern.com>
Message-ID: <amvabt$4gf$1@main.gmane.org>

Mark McEahern
> One man's obvious is another man's enigma.  I thought it was a fine
puzzle.
>
> Another way to make it harder would be to scramble all characters.
>

This is the result of the self test I use in my scramble function:


Wkjbj Uqypfu

Fdmnjn Wpxms
(A ykhwneb pwxezmh nu l jiznjmze zinb rsrwfodpxe l zpxstmby.)
Ykhwneb: Hkfy rmymbufkz.
Wpxstmby: Kmymbufkz.
Ykhwneb: Kp, mkqmmq poj nez'b nb?
Wpxstmby: Kqp, 'ynd yprb.
Ykhwneb: Tkk tmbm fu tkmnpoj?
Wpxstmby: Skrm, Y mnqm 'xwp.
Ykhwneb: Op, akfy gku mkk. Lp... ypfep KJV zpxms rwxu'b ypxq?
Wpxstmby: Tmrp.
Ykhwneb: Xqq, ypfiapb ypxq embm. Oumq, eprb rwp ypxq pknus hs nu ypp
ywxmd?
Wpxstmby: A gonw nixeynfu ruh fup yprb nu bmdmjb emxzd 're jmxu iidp fu
iq injy. Zb'd iq dkjenyxwxy fsnunkz yprb ypxq'bm jmzbnu'.
Ykhwneb: Smzbnus?
Wpxstmby: Kqp.
Ykhwneb: Onpm jnbyd?
Wpxstmby: Vfrkydj. Zb'd iq jmmnxm yprb ypxep zpxms rwp mojkbnz' hupmu
ypp inzorsbmtmjenkz yprb ypxq'bm jnbyd. Ohzmbdp ypxnu jm'rdnku
. Yopm gku l zbrwb ypp zpxmre' ymjyxudq yk 'fs rhfib ypp gnxdh fu ypxnu
'nuh mmae. Skg enyuxed ypxnu rbymisye yk gdj gwfq ywxm yk yw
xm. Skyndm yprb ypxq pk jkb zk iidp gdj re... rdhqimb.
(Coro joro... gdrs gdrs gdrs... epfkzp... yphy.)
Ykhwneb: Tmd, jib epj pk ypxq ypnup ypxq'bm jnbyd?
Wpxstmby: Kufbtmu gonw nixeynfu. Oup ypnus ne gku zibm, ypp zpxms ne jkb
l dwxoyibm fm ypp rnu. Ypxq toqm xufwikhe pngmnkhdyq nu ypp
 dkisrwrbndxdj znismm rkb fm rmbktnz'. (Coro joro... gdrs gdrs gdrs...
epfkzp... yphy.) Ywfijdp ne, zpxms rwp qmbq pnq. Oudm ypxq am
b ru nyxo nu ypxnu 'xope, ypxwp'd jk zpnmynz' nb.
Ykhwneb: Cib epxwp pnh ypxq amb ypp nyxo?
Wpxstmby: Fwfq Xobkmy. Xm'd yprb ikzb pojnxwfid fm dwxoyibmd, l ddxdxw
zpxms. 'V'd bmrdnbxy yprb l zpxms'd mngm dkjeneye fm zbrupnz'
 rwfijy gku l gmg ikjbte ruh ypxu jmnu' xoymz. Kuh yprb'd l pmrwxeznjn
rwfermdb gku ru rqjnynfid zpxms.
Ykhwneb: Vmmd epj pkz'b mkk kizb bmikqm Xobkmy?
Wpxstmby: Cmdohep fm ypp xufwikhe dkiqxwdnrd rkzenhndnbnmd nm 'p
zidkxmpe.
Iknkp Odxw: Kuh eprb xfrkydj rwp ypp dkiqxwdnrd rkzenhndnbnmd fm fdnup
rdnoynfu?

Hmb mkhw nifbxe gwfq WqrwyGhkym - tbys://efg.iczkgb.dk.hz

--

Emile van Sebille
emile at fenx.com

---------







From sismex01 at hebmex.com  Wed Sep 25 19:18:28 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Wed, 25 Sep 2002 18:18:28 -0500
Subject: List/Tuple bug or feature?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062330260434F@mail.hebmex.com>

> 
> sismex01 at hebmex.com fed this fish to the penguins on Tuesday 24 
> September 2002 03:12 pm:
> 
> > It's one of python's features :-)
> > 
> > For when you wanna write a long string, SQL for example:
> > 
> > Query = "SELECT what, when, where, why "
> >         "FROM t1, t2, t2 "
> >         "WHERE t1.id = t2.id_t1 and t3.id_t2 = t2.id and t1.id_t2 =
> >         t3.id " "AND t1.answer = 'OK'"
> >
>         What's wrong with...
> 
> query = """select what, when, where, why
>         from t1, t2, t3
>         where t1.id = t2.id_t1
>                 and t3_id2 = t2.id ...etc """
> 

Hey, don't kill the messenger, I'm just explaining what
happened regarding the auto-concatenation of string
constants. If you don't like, don't use it.

But then, you did ask.

>>> query = """select what, when, where, why
               from t1, t2, t3
               where t1.id = t2.id_t1
                 and t3_id2 = t2.id ...etc """
>>> query
'select what, when, where, why\n               from t1, t2, t3\n
where t1.id = t2.id_t1\n                 and t3_id2 = t2.id ...etc '
>>> 

...an unsightly amount of useless whitespace.

-gus






Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From phr-n2002b at NOSPAMnightsong.com  Sun Sep 15 04:13:09 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 15 Sep 2002 01:13:09 -0700
Subject: How to represent sets
References: <alvd8p$1sq$1@news.hccnet.nl> <mailman.1032075137.18192.python-list@python.org>
Message-ID: <7xn0qjacx6.fsf@ruckus.brouhaha.com>

"James J. Besemer" <jb at cascade-sys.com> writes:
> For dense sets isomorphic with integers, the long integer
> implementation probably will be superior.

Using immutable long integers to represent mutable sets doesn't make
much sense.  Better to use arrays like the ones implemented by the
array module.


From peter at engcorp.com  Tue Sep 10 10:27:57 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 10 Sep 2002 10:27:57 -0400
Subject: run-mailcap in python?
References: <alkcfb$9ql$07$1@news.t-online.com> <alkco4$j0k$06$1@news.t-online.com> <3d7de60b$1@news.sentex.net> <alksap$ekd$02$1@news.t-online.com>
Message-ID: <3D7E016D.2646AD9F@engcorp.com>

Thomas Guettler wrote:
> 
> Peter Hansen schrieb:
> > import  os
> > if os.name == 'posix':
> >     def startfile(name):
> >         os.system('run-mailcap "%s"' % name)
> >     os.startfile = startfile
> 
> Yes, ugly but it works. Thank you!

In Pythonia, "ugly" is sometimes spelled "pragmatic".  <grin>

-Peter


From ataylor at ucsd.edu  Thu Sep 26 15:28:29 2002
From: ataylor at ucsd.edu (Adam Taylor)
Date: 26 Sep 2002 12:28:29 -0700
Subject: Getting started
References: <3D911306.7080609@ucsd.edu> <mailman.1032982114.4627.python-list@python.org>
Message-ID: <e15b6fa3.0209261128.4edcdfe8@posting.google.com>

> My read of page 5 is he equates his internal use of "weakly checked" 
> with "weakly typed" and further suggest that the latter term is commonly 
> used "in the literature," which it does seem to be.

This sounds like the same thing as I said, or damn close.  Am I
missing something?

> I think your read of Cardelli overall is correct -- that Python, like 
> Lisp, would be classified as "untyped" because it has no static type 
> system.  

Yes.  It seems to me that Table 1 in Cardelli:

          Typed   Untyped
Safe      ML      LISP
Unsafe    C       Assembler

Might be better like this:

          Statically typed   Dynamically typed    Untyped
Safe      ML                 LISP
Unsafe    C                                       Assembler

(Python would also be in the (Safe,Dynamically typed) position.)

Although safety and strongly-checked are not the same thing, this
table is also correct:

                      Statically typed   Dynamically typed   Untyped
Strongly checked      ML                 LISP
Weakly checked        C                                      Assembler

> A literal reading furthermore defines "weakly typed" only in 
> terms of static checking.

I don't understand what you mean by this.  Cardelli never uses the
term "weakly typed" himself, he only uses it that one time, to say
that it's the term most people use for the concept that he prefers to
call "weakly checked".

> Interestingly, a google search on the term revealed that about 25% of 
> the meaningful hits equated "weakly typed" and "dynamically typed" (the 
> latter of which is what Cardelli calls "untyped").  The other 75% 
> generally characterized strong vs. weak type systems in terms of "type 
> safety", i.e., a mechanism for preventing  "type errors," regardless of 
> whether the checking is done statically or dynamically.  [The list of 
> references appears in a prior post to this thread.]

So 75% of people, when they talk about strongly-typed vs. weakly-typed
languages, are refering to the strongly-checked vs. weakly-checked
axis in the table above, whereas 25% of people are refering to the
statically-typed vs. dynamically-typed vs. untyped axis.  These
numbers (75% and 25%) seem far enough from 100% and 0% that everyone
would be better off not using the terms "strongly typed" and "weakly
typed", and using "strongly checked", "weakly checked", "statically
typed", "dynamically typed", or "untyped", as appropriate.  Or at
least when people start talking about "strongly typed" vs "weakly
typed" languages, that first thing to get straight is which axis
they're talking about.

Adam

-----
"James J. Besemer" <jb at cascade-sys.com> wrote in message news:<mailman.1032982114.4627.python-list at python.org>...
> Adam Taylor wrote:
> 
> > This last is available at:
> > http://research.microsoft.com/Users/luca/Papers/TypeSystems.pdf 
> 
> I believe your interpretation is accurate for the most part.
> 
> > Cardelli doesn't use the term "weakly typed" himself, but on page 5 he
> > seems to be say that "weakly typed" is the commonly used term for what
> > he calls "weakly checked".  C is weakly checked, in his terminology, and
> > that's why it's unsafe. 
> 
> 
> My read of page 5 is he equates his internal use of "weakly checked" 
> with "weakly typed" and further suggest that the latter term is commonly 
> used "in the literature," which it does seem to be.
> 
> > But people on this thread seem to be using "weakly typed" to mean
> > roughly the same thing as Cardelli's "untyped".  Which is fine, and
> > pretty close to Cardelli's terminology.  But clearly the terms "untyped"
> > (=="weakly typed" on this thread) and "weakly checked" (=="weakly typed"
> > in common usage) are very different things. 
> 
> 
> Well, some people here use the term casually and incorrectly, as a 
> measure of quality of a language's object model, which I am sure you'll 
> agree is completely inconsistent with most uses of the term, including 
> Cardelli's.
> 
> Interestingly, a google search on the term revealed that about 25% of 
> the meaningful hits equated "weakly typed" and "dynamically typed" (the 
> latter of which is what Cardelli calls "untyped").  The other 75% 
> generally characterized strong vs. weak type systems in terms of "type 
> safety", i.e., a mechanism for preventing  "type errors," regardless of 
> whether the checking is done statically or dynamically.  [The list of 
> references appears in a prior post to this thread.]
> 
> I think your read of Cardelli overall is correct -- that Python, like 
> Lisp, would be classified as "untyped" because it has no static type 
> system.  A literal reading furthermore defines "weakly typed" only in 
> terms of static checking.
> 
> However, I submit that his use is inconsistent with the more common use 
> of the term.  Too, I think this is actually a minor oversight in 
> Cardelli's paper.  His focus overall was static systems and I think he 
> simply didn't give enough consideration to dynamically typed languages. 
>  E.g., he says "untyped" languages such as Lisp can be "safe" but he 
> defines "weakly typed" only in terms of languages that do static type 
> checking.  I would argue that in the first place he misclasifies Lisp as 
> "untyped".  I suspect he took the naive view of Lisp, in which 
> everything is a list, when in fact a rather wide variety of "objects" 
> actually exist in a typical, mature Lisp system.  Of course, Python and 
> Smalltalk are better examples, in that they both have robust object type 
> systems and are in principle very secure from a type checking sense. 
>  (Of course, there are probably things you can do with the compiler and 
> debug class objects that violate type safety, making them both unsafe in 
> general.)
> 
> The primary distinction between safe and unsafe, strong and weak type 
> systems revolves around the notion of whether or not "type errors" can 
> be prevented by the type system.  The distinction between typed and 
> untyped is less interesting, except to use "untyped" as a way to 
> eliminate a class of languages from further discussion.  Within typed, 
> there should be two sub-classes, based on whether type checking is 
> primarily static or dynamic.  This is more interesting and useful, and 
> is consistent with "the literature" as far as I can determine.  
> 
> I furthermore have a conjecture that much of Cardelli's work on 
> "judgments" would apply equally well to static and to dynamic systems. 
>  The important criterion is that the type system be secure and that the 
> total collection of function x arguments definitions be unambiguous (at 
> least during the sequence of computation of interest).  Whether the 
> checks take place at compile time or at runtime I don't think materially 
> affects his theories.  But I could be overlooking something in this regard.
> 
> In any case, my own use of the term "weakly typed" and associated terms 
> (strongly typed, type safety, etc.) in this thread have been consistent 
> with "the literature" (though slightly inconsistent with Cardelli), in 
> that I class Python to be "strongly typed" using "dynamic checking".   
> That seems to be the consensus here.  Only confusion has been about 
> "weakly typed," and hopefully we're close to exhausting that topic..
> 
> --jb


From martin at v.loewis.de  Mon Sep 16 01:43:54 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Sep 2002 07:43:54 +0200
Subject: couldnt find __dic__
References: <7153651c.0209151958.4099ac3e@posting.google.com>
Message-ID: <m365x6zdyd.fsf@mira.informatik.hu-berlin.de>

quiteblack at yahoo.com (black) writes:

> my book told me all dictionaries have __dic__ method but I searched
> each dictionaries and find nothing, where is it ???

Your book is wrong; it is "__dict__" not "__dic__" (however, your book
might be right and you just did not read it carefully).

Regards,
Martin


From martin at v.loewis.de  Sat Sep  7 03:46:32 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 07 Sep 2002 09:46:32 +0200
Subject: Converting python object reference to int and back
References: <3D79AC95.1000000@NOSPAMbitbuilder.com>
Message-ID: <m38z2e2qef.fsf@mira.informatik.hu-berlin.de>

Ahmad Baitalmal <ahmad at NOSPAMbitbuilder.com> writes:

> How can I get an id (pointer/reference) from a python object and then
> get that object back using that integer?

To get an integer, use the id function. To get the object back, create
a dictionary mapping the ids to the real objects.

Regards,
Martin


From dwelch91 at nospam.attbi.com  Sun Sep 22 21:55:48 2002
From: dwelch91 at nospam.attbi.com (djw)
Date: Mon, 23 Sep 2002 01:55:48 GMT
Subject: PyQt setup on Linux Mandrake 8.2
References: <3D8B4898.20808@nospam.attbi.com> <3d8e7141$1@news.nz.asiaonline.net>
Message-ID: <3D8E74A4.7000404@nospam.attbi.com>

Thanks to you and the others, I finally got this to work. My remaining 
question/problem/puzzlement is what to do about the fact I now have two 
Python 2.2 installs on my system. Mandrake installed everything during 
its setup into /usr and everything I built/installed went into 
/usr/local. I was going to rip out the version that Mandrake layed down 
in /usr and just use the /usr/local copy, but when I ran the package 
manager and told it to uninstall all the various py* pieces, the list of 
dependencies it wanted to also delete was scary (stuff like filesystem 
modules). Unfortunately, when I try to install new py modules, they all 
want to install into /usr instead of /usr/local, since /usr/bin/python 
is ahead of /usr/local/bin/python on $PATH, causing sys.prefix to point 
to /usr, I suppose.

So, this leaves me wondering... its not possible to install Mandrake 
without Python (as evidenced by the dependencies), but the location it 
is installed at is "wrong" by default - so, what is the "proper" way to 
handle this situation? (It would seem nice to install Linux without 
Python and compile and install it afterwards). I can't tell from your 
list of commands where python is located, but I assume its in /usr? 
Also, I noticed that the copy of python that was installed by the system 
lacked Python.h, whereas the copy I compiled and installed had this 
file. Are you supposed to "overlay" the system copy in /usr with the 
extra files that things like sip/pyqt need (i.e., Python.h)?


Don


Colin Brown wrote:
> "dwelch" <dwelch91 at nospam.attbi.com> wrote in message
> news:3D8B4898.20808 at nospam.attbi.com...
> 
>>I would like to setup PyQt 3.4 on my Mandrake 8.2 box.
>>----snip----
>>
> 
> 
> Hi Don
> 
> I few months ago I installed PyQt under Mandrake Linux 8.2 for i586
> and ended up doing the following:
> ----------------------------------------------------------------------------
> 
> Put tmake_1.8.tar.gz into /usr/local
> Put sip-3.3.2.tar.gz into /usr/local
> Put PyQt_3.3.2.tar.gz into /usr/local
> 
> gunzip tmake_1.8.tar.gz
> tar -xvf tmake_1.8.tar
> gunzip sip-3.3.2.tar.gz
> tar -xvf sip-3.3.2.tar.gz
> gunzip PyQt_3.3.2.tar.gz
> tar -xvf PyQt_3.3.2.tar.gz
> 
> TMAKEPATH=/usr/local/tmake_1.8/lib/linux-g++
> export TMAKEPATH
> 
> cd /usr/local/sip-3.3.2
> python build.py -q /usr/lib/qt2 -i /usr/lib/qt2/include -g
> /usr/local/tmake_1.8/bin/tmake -l qt-mt
> make
> make install
> 
> cd /usr/local/PyQt_3.3.2
> python build.py -q /usr/lib/qt2 -g /usr/local/tmake_1.8/bin/tmake -l qt-mt
> make
> make install
> 
> ----------------------------------------------------------------------------
> 
> I hope you find this useful.
> 
> There was a compile error that have yet to follow up on to do with glu but
> apart from that I ended up with a pyuic that works okay with output from
> Qt_Designer except for the table and custom widgets. When I get time I
> will follow up on these.
> 
> Cheers
> Colin Brown
> PyNZ
> 
> 
> 
> 




From gumuz at looze.net  Tue Sep  3 08:56:21 2002
From: gumuz at looze.net (Gumuz)
Date: Tue, 3 Sep 2002 14:56:21 +0200
Subject: Design question.... parent/child class
Message-ID: <3d74b021$0$234$4d4ebb8e@news.nl.uu.net>

hi all,

i have a SessionManager class, which contains a list of Session
objects(instantiated from a Session class). I have a design problem.

Each session object has a small 'queue' of messages for the specific
session. How can a session send a message to another session without having
some kind of parent-reference to the SessionManager object?

i am a bit puzzled, or is this parent-reference not such a bad idea after
all?


thanx,

Guyon




From t_therkelsen at hotmail.com  Sun Sep  8 15:03:25 2002
From: t_therkelsen at hotmail.com (Troels Therkelsen)
Date: 08 Sep 2002 19:03:25 GMT
Subject: transform a list of lists in a lit of strtings???
References: <mailman.1031507832.24527.python-list@python.org>
Message-ID: <slrnann7o2.ok6.t_therkelsen@abattoir.2-10.org>

jubafre at brturbo.com wrote:
> I have a list of lists like this:
> 
> labels=[['D1'], ['D2'], ['D3']]
> 
> and i want to transform ia a list of strings
> 
> labels2==[ 'D1' ,  'D2' ,  'D3' ] <<<<<<<----------- how i can get it????
> 
> 	
> Juliano Freitas
> www.gebrasil.hpg.com.br 

labels2 = map(lambda x: x[0], labels)

or

labels2 = [ x[0] for x in labels ]


Whichever suits your fancy and/or programming style.

Regards,

Troels Therkelsen


From noone at nowhere.com  Thu Sep  5 22:46:45 2002
From: noone at nowhere.com (c)
Date: Thu, 05 Sep 2002 22:46:45 -0400
Subject: Windows NT login script
Message-ID: <3D781715.5070505@nowhere.com>

I am a newbie to Python and I'm wondering how difficult it would be to 
write a Windows NT/2000 login script using Python?

Can anyone think of a good way to do it?



From aahz at pythoncraft.com  Thu Sep 12 01:29:07 2002
From: aahz at pythoncraft.com (Aahz)
Date: 12 Sep 2002 01:29:07 -0400
Subject: Destructor never called ???
References: <76710511.0209111450.40999dd8@posting.google.com> <tp70ou0r0j11as6enhpsaqul4oqa3mlu6t@4ax.com>
Message-ID: <alp8n3$bsd$1@panix1.panix.com>

In article <tp70ou0r0j11as6enhpsaqul4oqa3mlu6t at 4ax.com>,
Tim Roberts  <timr at probo.com> wrote:
>
>However, Python is a garbage-collected system.  Unlike in COM, an object is
>NOT destroyed as soon as its reference count goes to 0.  Your object will
>only be destroyed if the garbage collector happens to run, and decides that
>it might be a convenient time to actually free up the space your object
>occupies.

Wrong.  CPython uses reference counting backed by GC.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From mmaddox at hcsmail.com  Fri Sep  6 18:58:57 2002
From: mmaddox at hcsmail.com (mmaddox at hcsmail.com)
Date: Fri, 6 Sep 2002 18:58:57 -0400 (EDT)
Subject: Larry Wall's comment on python...
Message-ID: <200209062258.g86MwvVm096127@mail.hcsmail.com>

> So does the fact that C++ forces you to use { } as block delimiters
irritate
> the Libertarian in you?

Of course it does. I need a parser that can decipher the pure randomness
of my own code. Basically, I'd like to run an SHA-512 algorithm against a
poorly-typed text file describing my program. Then, I'd just compile the
message digest.

Of course, most of my current code has one big block, main(), and I just
cut-and-paste the rest. 

--
Michael F. Maddox
Senior Software Engineer
Hayes
Tallahassee, FL, USA




From sebastian at wangnick.de  Sun Sep 22 08:08:55 2002
From: sebastian at wangnick.de (Sebastian Wangnick)
Date: Sun, 22 Sep 2002 14:08:55 +0200
Subject: Trying the impossible?
References: <30292041.0209210032.6e07d9f3@posting.google.com>
Message-ID: <amkbvp$31p$07$1@news.t-online.com>

Dear all,

yesterday I recompiled Tcl/Tk (8.4.0) and pythoncore/python/_tkinter (2.2.1)
in debug mode with VC 6.0. Now, when trying my script, I'm getting the
following error:
    PyThreadState_Get: no current thread
This is happening when NTDLL tries to call my_wndproc via calldll.

It seems to me that the following is happening: Within Tkapp_MainLoop,
before calling Tcl_DoOneEvent, the code saves the current thread state
within tcl_tstate and then releases the interpreter_lock, setting the
_PyThreadState_current to NULL. Whenever Tcl calls back to Python, the
ENTER_PYTHON macro re-acquires the interpreter_lock and restores the thread
state. However, in my script this is not performed, as it is Windows calling
my script, and calldll doesn't know that it has to reacquire the interpreter
lock, nor where to find the thread state!

I undef'd WITH_THREAD in pyConfig.h and recompiled, and then my script
works.

I'm uncertain how to best make calldll thread-safe under these
circumstances. Maybe the author can help?

Regards,
Sebastian

"Sebastian Wangnick" <sebastian.wangnick at eurocontrol.int> wrote in message
news:30292041.0209210032.6e07d9f3 at posting.google.com...
> Dear all,
>
> I'm trying to overload, using dynwin/calldll under Windows, the
> Tkinter drawing routines to perform some additional drawing on top of
> Tk.
>
> My first attempt was to bind to "<Expose>", but this binding seemingly
> is triggered before Tk performs its drawing, so my stuff was erased by
> Tk :(
>
> Now I'm trying to overload the WindowProc itself. To test the
> mechanism, I started by simply passing everything on to
> CallWindowProc:
>
> from Tkinter import *
> import dynwin
> import dynwin.gencb
> import dynwin.windll
> import dynwin.winwin
> user32 = dynwin.windll.module ('user32')
>
> def my_winproc (hwnd, message, wparam, lparam):
>     return user32.CallWindowProc (tk_winproc,
>         hwnd, message, wparam, lparam)
> my_winproc_cb = dynwin.gencb.generated_callback ('llll', my_winproc)
>
> def map (event):
>     global tk_winproc
>     hwnd = frame.winfo_id()
>     tk_winproc = user32.GetWindowLong (hwnd,
>         dynwin.winwin.GWL_WNDPROC)
>     user32.SetWindowLong (hwnd, dynwin.winwin.GWL_WNDPROC,
>         my_winproc_cb.address)
>
> root = Tk()
> frame = Frame(root,width=100,height=100)
> frame.pack(fill=BOTH,expand=1)
> frame.bind("<Map>",map,"+")
> root.mainloop()
>
> However, when starting this it crashes with 'The instruction at
> "0x1e0381fa" referenced memory at "0x00000008". The memory could not
> be "read".'
>
> Any idea what I'm doing wrong? I'm using XP, btw.
>
> Regards,
> Sebastian




From aleax at aleax.it  Wed Sep 18 11:37:22 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 18 Sep 2002 15:37:22 GMT
Subject: Python comment stripper
References: <16Nh9.248$Wc6.17087796@newssvr13.news.prodigy.com> <3d8807b1$1@news.sentex.net> <Eg0i9.561$Y33.34967690@newssvr13.news.prodigy.com>
Message-ID: <S41i9.127467$pX1.4558233@news2.tin.it>

Dennis Reinhardt wrote:

>> This doesn't seem like a good reason to me.  Why not just use .pyc files
>> all the time and be done with it?
> 
> I am looking to package Python programs as EXE files for upload/download
> over the internet.  The "good reason" is faster upload/download over
> dialup lines, not hard disk space.

Makes sense.  Let's see, what about a one-(logical)-liner (warning, 
untested!):

open(outfile, 'w').writelines( [ 
    tok[1] for tok in tokenize.generate_tokens(open(infile).readline) 
           if tok[0] != tokenize.COMMENT ] )

Probably best rewritten as a loop of a few lines, but, I think the
key idea holds -- tokenize.generate_tokens seems to be the best way.


Module gzip is probably a good idea to help you reduce size, too.
I.e., besides stripping comments, you might also use gzip for
compression -- perhaps not quite as good as the compression you
might get from bzip2, but the latter is not part of the standard
Python distribution.  What are you using now for compression?


Alex



From dsavitsk at e-coli.net  Tue Sep  3 20:42:30 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Wed, 04 Sep 2002 00:42:30 GMT
Subject: winsound.Beep() frequencies
References: <3ab719f7.0209031229.33667461@posting.google.com>
Message-ID: <WFcd9.6556$yt3.3086163@newssrv26.news.prodigy.com>

i believe a supplied value of 4000 plays a frequency of 4000Hz.  you should
consult a table of frequency to pitch to convert.  typically A in the 4th
piano octave is 440Hz, and you should be able to derive the rest from that
(or search on google).

-d

"Stuart" <megabytemonster at hotmail.com> wrote in message
news:3ab719f7.0209031229.33667461 at posting.google.com...
> While looking at the python documentation, I noticed that there are no
> frequency/pitch tables for using with the winsound module.  Does
> anyone know whether this documentation exists, and/or what the
> corresponding frequencies to pitches are?
>
> Thanks,
> Stuart




From richie at entrian.com  Tue Sep 24 08:32:26 2002
From: richie at entrian.com (Richie Hindle)
Date: Tue, 24 Sep 2002 13:32:26 +0100
Subject: convention for Pickler file extension
References: <918bc22f.0209230208.4eaca702@posting.google.com> <svEj9.226013$z91.9600123@bin3.nnrp.aus1.giganews.com> <ampfbe$rq0$07$2@news.t-online.com>
Message-ID: <smm0pus5t41l783ffo8ms6uur2umob0ftf@4ax.com>

> I call files *.pickle

<aol>Me too.</aol>

Are there any non-medieval platforms that don't let you have arbitrary
length file extensions?

-- 
Richie Hindle
richie at entrian.com



From gerhard.haering at opus-gmbh.net  Mon Sep  2 08:53:19 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 2 Sep 2002 12:53:19 GMT
Subject: Must have Python modules?
References: <_THc9.42153$vY2.1039887@e3500-atl2.usenetserver.com> <slrnan6k4s.eg.gerhard.haering@haering.opus-gmbh.net> <FCIc9.30407$FJ1.5585669@e3500-atl1.usenetserver.com> <3D735D9C.9060607@mxm.dk>
Message-ID: <slrnan6not.1ko.gerhard.haering@haering.opus-gmbh.net>

In article <3D735D9C.9060607 at mxm.dk>, Max M wrote:
> Robert Oschler wrote:
> 
>> I want to know what the other favorites are "out there".
> 
> win32all (if os in windows), mx*, PIL, reportlab, Nummeric

The same set for me, minus ReportLab (not yet) Plus wxPython and pyPgSQL.

-- Gerhard


From black at flamingo.ru  Fri Sep 20 07:24:24 2002
From: black at flamingo.ru (Vasiliy A. Chernoivan)
Date: Fri, 20 Sep 2002 17:24:24 +0600
Subject: M$SQL connections from linux
Message-ID: <1032528860.599940@news.etel.ru>

Hello all
Is there some module for python which performs the task?
Is it possible at allto access data under M$SQL from linux machine?

Every comment will be highly appreciated
Thanks
                Vasiliy.




From radfaraf at NOSPAMoptonline.net  Mon Sep 30 21:31:49 2002
From: radfaraf at NOSPAMoptonline.net (kingfool)
Date: Tue, 01 Oct 2002 01:31:49 GMT
Subject: Keeping track of variables.
Message-ID: <Xns9299DAFBBB6Bradfaraf@167.206.3.3>

Is there an editor I can use that will help me find the first instance a 
variable was used across an entire program of multiple files? Reason I ask 
is because I'm going through and adding features to a program as a learning 
experience. Coming from C++ as my only programming language I find it a bit 
confusing to not know when something was declared/used first, which can 
reveal a bit more about what the variable is for.


From jpmieville at bluewin.ch  Thu Sep 19 15:08:52 2002
From: jpmieville at bluewin.ch (Jean-Paul Mi=?ISO-8859-1?B?6Q==?=ville)
Date: Thu, 19 Sep 2002 21:08:52 +0200
Subject: Problem with urllib
References: <B9AEA35F.7FF7%jpmieville@bluewin.ch> <dmci9.43613$U7.15655300@twister.socal.rr.com>
Message-ID: <B9AFED64.80F1%jpmieville@bluewin.ch>

Yes,

I use :
Python 2.2.1 (#134, Apr  9 2002, 21:16:52)  [CW CARBON GUSI2 THREADS GC]
Type "copyright", "credits" or "license" for more information.
MacPython IDE 1.0.1

I don't know if this problem is related to MacPython.



> From: Jeff Davis <jdavis at empires.org>
> Organization: RoadRunner - West
> Newsgroups: comp.lang.python
> Date: Thu, 19 Sep 2002 04:26:49 GMT
> Subject: Re: Problem with urllib
> 
> 
> I can't reproduce the error either way (interactive or otherwise). Are you
> use that you're using the same python interpreter version for both?
> 
> Regards,
> Jeff
> 
> Jean-Paul Mi?ville wrote:
> 
>> I don't understand why with some URL I have got [Errno 57] Socket is not
>> connected.
>> 
>> The code is :
>> 
>> import urllib2
>> 
>> page = urllib2.urlopen(' http://www.python.org/doc/NonEnglish.html')
>> print page.read()
>> 
>> If I do the same thing with python interactive, I can't reproduce the
>> error.
>> 
>> Thanks in advance,
>> Jean-Paul
> 



From aahz at pythoncraft.com  Sun Sep 29 11:10:39 2002
From: aahz at pythoncraft.com (Aahz)
Date: 29 Sep 2002 11:10:39 -0400
Subject: ANN: Tim Peters' FixedPoint on SourceForge
References: <mailman.1033085920.21837.python-list@python.org>
Message-ID: <an755f$fm8$1@panix1.panix.com>

In article <mailman.1033085920.21837.python-list at python.org>,
dougfort <dougfort at dougfort.net> wrote:
>
>This project makes Tim Peters' FixedPoint math class available to the community
>as a SourceForge project. 
>
>Check out http://fixedpoint.sourceforge.net

Should I change the link on my web site to point at this?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From mcfletch at rogers.com  Tue Sep 10 10:30:14 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 10 Sep 2002 10:30:14 -0400
Subject: dynamism
References: <mailman.1031660952.2330.python-list@python.org> <Xns928597A0CEC6Dduncanrcpcouk@127.0.0.1>
Message-ID: <3D7E01F6.6030407@rogers.com>

I'm coming in kinda late in this thread... Here's my understanding 
(which might be erroneous):

	It is possible for new-style classes to declare that they have only a 
limited number of instance attributes (via the __slots__ declaration or 
C equivalent) in order to enforce policy or allow for optimisation. 
Derivatives of these classes, however, which do not themselves declare 
__slots__ and which don't use any getattr* setattr* or descriptor 
(property) hooks to prevent the operations, have standard Python 
semantics (i.e. they have a dict object in which instance attributes are 
stored, and you may add/remove keys from this dict with attribute's 
semantics).

So, while it's probably true that "all" of those derivative classes do 
not allow you to add attributes, most will allow it, and AFAIK, all of 
the standard built-in types (list, tuple, object, string, etc.) do 
nothing to prevent their _sub-classes_ from having new attributes 
(though they will likely reject adding attributes to instances of the 
base class).

Don't forget your smiley :) ,
Mike

Duncan Booth wrote:
...
>>[Duncan Booth]
>>
>>>New style classes (which are those with 'object' or other builtin
>>>classes as a base class) do not all allow you to add arbitrary
>>
>                                    ^^^
...
> Read what I wrote? The word 'all' was supposed to be a clue.

_______________________________________
   Mike C. Fletcher
   Designer, VR Plumber, Coder
   http://members.rogers.com/mcfletch/





From jubafre at brturbo.com  Mon Sep 23 00:49:01 2002
From: jubafre at brturbo.com (jubafre at brturbo.com)
Date: Mon, 23 Sep 2002 01:49:01 -0300 (GMT-03:00)
Subject: match or sub problem?
Message-ID: <-2119489446.1032756541985.JavaMail.nobody@webmail1>

I have this list of strings:
final=['20 0x9', '30 0xa', 'B4 0x8', '30 0xb', 'F0', ' 1', ' 3', ' 5']

and i want to erase the nodes with whitespaces in the begining,
i use  "match" but doesn?t work and "sub", i think the problem is in the "sub", why?

final2=[]
for i in range(len(final)):
    p=re.compile(" ")
    t1=p.match(final[i])
    t=sub(p, "", final[i])
    print t
    final2.append(t)
print final2

None
None
None
None
None
<_sre.SRE_Match object at 0x01420DF0>
<_sre.SRE_Match object at 0x01421B00>
<_sre.SRE_Match object at 0x014211D0>	

p.match find just last 3  nodes with whitespaces in the begining, but the sub replace
al the nodes, like that: ['200x9', '300xa', 'B40x8', '300xb', 'F0', '1', '3', '5']

i want just the last 3 nodes replaced
 ['20 0x9', '30 0xa', 'B4 0x8', '30 0xb', 'F0', '1', '3', '5']

Juliano Freitas
www.gebrasil.hpg.com.br 

From manuel at hendel.net  Tue Sep 17 04:23:36 2002
From: manuel at hendel.net (Manuel Hendel)
Date: Tue, 17 Sep 2002 10:23:36 +0200
Subject: don't understand error message
Message-ID: <20020917082336.GD23629@partagas.as.de.cw.net>

I do the following,

domains = []

for line in lines:
    line = string.strip(line)[1:-1]
    fields = string.split(line, "|")
    if fields[3] not in domains:
        domains.append(fields[3])

error:

Traceback (most recent call last):
  File "./pop3create.py", line 16, in ?
    if fields[3] not in domains:
IndexError: list index out of range

Thanks,
    Manuel

-- 
When a man stops trying to prove he is a man, he is a man. 
-D. Talmadge Gosnell 



From peter at engcorp.com  Tue Sep 10 08:31:28 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 10 Sep 2002 08:31:28 -0400
Subject: run-mailcap in python?
References: <alkcfb$9ql$07$1@news.t-online.com> <alkco4$j0k$06$1@news.t-online.com>
Message-ID: <3d7de60b$1@news.sentex.net>

Thomas Guettler wrote:
> Thomas Guettler schrieb:
> 
>> Hi!
>>
>> On Unix there is a script called run-mailcap, which
>> opens the corresponding programm with the given file:
>>
>> run-mailcap foo.html
>> --> Default browser starts
> 
> 
> [cut]
> 
>> run("foo.html")
>>  --> Browser starts
> 
> 
> I just discoverd os.startfile(), but it is unfortunately
> only available on win32.
> 
> A Unix version would be nice.

import  os
if os.name == 'posix':
     def startfile(name):
         os.system('run-mailcap "%s"' % name)
     os.startfile = startfile

?

-Peter



From Oschler at earthlink.net  Wed Sep  4 08:59:31 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Wed, 4 Sep 2002 08:59:31 -0400
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com> <unbp765rbsn809@corp.supernews.com> <3D75EA07.D5EB1F5A@info.unicaen.fr>
Message-ID: <4vnd9.2$2L.307@e3500-atl2.usenetserver.com>

"Jerzy Karczmarczuk" <karczma at info.unicaen.fr> wrote in message
news:3D75EA07.D5EB1F5A at info.unicaen.fr...
> RPM1 wrote:
>
> You know, I am a peaceful, animal-friendly guy who thinks that corridas
> are a pure barbarism. But I refused myself the temptation to go and see
> one, because I am afraid that, all philosophy apart, I may discover
> that - after all - perhaps I adore the public slaughter of bulls?
>

Don't bother messing with corridas, do you adore the current stock market?
:)  Sorry, that one was just too tempting to pass over.

thx






From pyth at devel.trillke.net  Wed Sep 25 18:03:15 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 26 Sep 2002 00:03:15 +0200
Subject: a python puzzle
In-Reply-To: <Pine.LNX.4.44.0209251736470.16806-100000@hydrogen.sabren.com>; from sabren@manifestation.com on Wed, Sep 25, 2002 at 05:42:40PM -0400
References: <Pine.LNX.4.44.0209251736470.16806-100000@hydrogen.sabren.com>
Message-ID: <20020926000315.K30315@prim.han.de>

Michal Wallace wrote:
> Here's a python puzzle. Solve it and you win a prize. :)
> 
> """
> Ha'f ahoy ad ayfa wdqr srxhekdpyr! H phmm zhby
> fhv odeajf dc cryy pys jdfahez ad ajy chrfa kyrfde
> ad fdmby ajhf irwkadzrxo. Ixe wdq chzqry ha dqa?
> 
> Fheiy wdq jxby axgye ow ijxmmyezy xet tyirwkayt
> ajy fyirya oyffxzy, H phmm ybye zhby wdq ajy
> irwkadzrxo-oxghez fdqriy idty ad xoxuy wdqr
> crhyetf!
> """
> hokdra fwf, farhez, rxetdo
> ihkjyr = mhfa(farhez.myaayrf[:26])
> rxetdo.fjqccmy(ihkjyr)
> ihkjyr = "".ldhe(ihkjyr)
> ihkjyr += ihkjyr.qkkyr()
> oxk = {}
> cdr kmxhe, irwka he uhk(farhez.myaayrf, ihkjyr):
>     oxk[kmxhe] = irwka
> cdr mhey he dkye(fwf.xrzb[1]):
>     krhea "".ldhe([oxk.zya(ij,ij) cdr ij he mhey])[:-1]


OK.  something like 

def decrypt(yourstring):
    dedict = {'F': 'S', 'I': 'C', 'H': 'I', 'a': 't', 'c': 'f', 'b': 'v', 'e': 'n',
    'd': 'o', 'f': 's', 'i': 'c', 'h': 'i', 'k': 'p', 'j': 'h', 'm': 'l',
    'l': 'j', 'o': 'm', 'q': 'u', 'p': 'w', 's': 'b', 'u': 'z', 't': 'd',
    'w': 'y', 'v': 'x', 'y': 'e', 'x': 'a', 'z': 'g', 'g': 'k'}
    return "".join([dedict.get(i,i) for i in yourstring])

returns 

    """
    It's time to test your brainpower! I will give
    six months of free web hosting to the first person
    to solve this cryptogram. Can you figure it out?

    Since you have taken my challenge and decrypted
    the secret message, I will even give you the
    cryptogram-making source code to amaze your
    friends!
    """
    import sys, string, random
    cipher = list(string.letters[:26])
    random.shuffle(cipher)
    cipher = "".join(cipher)
    cipher += cipher.upper()
    map = {}
    for plain, crypt in zip(string.letters, cipher):
        map[plain] = crypt
    for line in open(sys.argv[1]):
        print "".join([map.get(ch,ch) for ch in line])[:-1]


but i don't know if i need free webhosting :-)

regards,

    holger



From loewis at informatik.hu-berlin.de  Fri Sep  6 10:20:45 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 06 Sep 2002 16:20:45 +0200
Subject: Difference between static method and a function as a class attribute?
References: <pqchnu40hs1dgqc4bct2shi4lbpl1oe80i@4ax.com>
Message-ID: <j4znuvb3nm.fsf@informatik.hu-berlin.de>

Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:

> What are the differences, if any, between 1 and 2?

In 1, test.funtest will be an unbound method; in 2, it will be a
function.

> And if there are any differences do they really matter, that is, does
> anybody has any test cases where 1 and 2 are not interchangeable?

Sure:

def m():
    print "called ::m"
class t1(object):
    method = m

class t2(object):
    def m():
        print "called t2::m"

    method = staticmethod(m)

t2.method()
t1.method()

gives

called t2::m
Traceback (most recent call last):
  File "a.py", line 13, in ?
    t1.method()
TypeError: unbound method m() must be called with t1 instance as first argument (got nothing instead)

Regards,
Martin


From joconnor at cybermesa.com  Mon Sep  9 12:29:08 2002
From: joconnor at cybermesa.com (Fearless Freep)
Date: 9 Sep 2002 09:29:08 -0700
Subject: None or 0
References: <k7ip4-0p3.ln1@junker.stroeder.com> <jwbaxter-E9C054.09080807092002@corp.supernews.com> <3d7a2b6b.92959728@news.cybermesa.com> <jwbaxter-3756A1.15241707092002@corp.supernews.com>
Message-ID: <7d3dc526.0209090829.2caf4b6b@posting.google.com>

John Baxter <jwbaxter at spamcop.net> wrote in message news:<jwbaxter-3756A1.15241707092002 at corp.supernews.com>...
> In article <3d7a2b6b.92959728 at news.cybermesa.com>,
>  joconnor at cybermesa.com (Jay O'Connor) wrote:
> 
> > >Go into (English-speaking) classrooms, and you're going to hear chants 
> > >of "5 and 1 is 6" not "5 and 1 is 1".
> > 
> > Because in that contexts "and" means "plus" and in this context "and"
> > means "logical and"
> 
> Agreed and understood.  But not helpful to the programming for everyone 
> cause.

Not really.  People also understand "and" meaning a logical and as
well.  "I want the shirt that is read and size 4" or whatever.  "And"
just means that you want to combine two things, whether numbers or
conditions.  It's an ambigious concept  that people use all the time
and let the context indicate what is meant.

Of course, when we try to build that ambiguity into computer languages
and allow context to determine meaning, we usually get into trouble :)

Then again, we keep a lot of lawyers employed for the same reason when
such ambiguity causes problems in human language.

I don't think "and" as a "logical and" and not as "plus" is really a
problem in Python for non-programmers to deal with

Take care,
Jay


From loewis at informatik.hu-berlin.de  Mon Sep  9 05:22:46 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 09 Sep 2002 11:22:46 +0200
Subject: Howto debug self?
References: <alhivr$fom$04$1@news.t-online.com>
Message-ID: <j47khvbjq1.fsf@informatik.hu-berlin.de>

Oliver Eichler <oliver.eichler at dspsolutions.de> writes:

> Is it possible to debug self with pdb? I mean it in the sense of getting a 
> list of vars and their values. 

If you print dir(self), you will get the names instance attributes. If
you print self.__dict__, you will get the instance dictionary.

HTH,
Martin


From kroc at blueyonder.co.uk  Fri Sep 27 05:42:38 2002
From: kroc at blueyonder.co.uk (Keith O'Connell)
Date: Fri, 27 Sep 2002 09:42:38 GMT
Subject: How do you pronounce "__" (double underscore)?
References: <525eaba.0209260956.72bf6786@posting.google.com>
Message-ID: <87wup793aq.fsf@smtp.blueyonder.co.uk>

>>>>> "Pat" == Pat Notz <pknotz at sandia.gov> writes:

         @ = "at"
         ! = "bang"
         . = "dot"
         * = "star"

    Pat> Is there a short (one or two syllable) pronounciation for
    Pat> "__" (double underscore).  "Underscore underscore" and
    Pat> "double underscore" are kind of long.  Somthing like "splat"?
    Pat> Anyone?

        "Skid"
                        It's a skid mark!

-- 
Keith.
 [Debian Woody & Python 2.1.3]


From member at dbforums.com  Fri Sep 13 05:37:36 2002
From: member at dbforums.com (hpyhpy)
Date: Fri, 13 Sep 2002 09:37:36 +0000
Subject: Dynamic loading of c modules,question.
Message-ID: <1810446.1031909856@dbforums.com>

hi,everybody:

I'm puzzled about SWIG.

On sco openserver5,I try swig's example:

/****************************/
example.c

/* File : example.c */

/* A global variable */
double Foo = 3.0;

/* Compute the greatest common divisor of positive integers */
int gcd(int x, int y) {
int g;
g = y;
while (x > 0) {
g = x;
x = y % x;
y = g;
}
return g;
}
/****************************/
example.i

/* File : example.i */
%module example

extern int    gcd(int x, int y);
extern double Foo;
/****************************/
And I do :
#gcc -c example.c
#swig -python example.i
#gcc -c example_wrap.c -I/usr/local/include/python2.2/
#gcc -G -Wl example.o example_wrap.o -o _examplecmodule.so

Then it has these files:
example.c
example.i
example.o
example_wrap.c
example_wrap.doc
example_wrap.o
_examplecmodule.so

So I wonder where is the "example.py"???
So I can't "import example"

Traceback (most recent call last):
File "", line 1, in ?
ImportError: No module named example

*_*   *_*   *_*

Who can tell me why,please help me,I'm a tyro^_^
Thank!

--
Posted via http://dbforums.com



From sholden at holdenweb.com  Wed Sep 18 12:32:59 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 18 Sep 2002 12:32:59 -0400
Subject: timeoutsocket module help
References: <am7r98.h0.1@kserver.org>
Message-ID: <eX1i9.1228$bh1.667@fe03>

"Sheila King" <usenet at thinkspot.net> wrote ...
> OK, I'm doing DNS lookups in a script I've written, to determine whether
> certain IP addresses are listed in DNS BL lists (for blocking spammers).
>
> Have been using the code I've written for a long time now (months) without
> significant problems. However, someone has made the point to me, that my
> code should technically have a default timeout on the socket that tries to
> do the DNS lookup. And they are right, so I'm trying to modify my code.
>
> So, I'm using Tim O'Malley's timeoutsocket module, or trying to. But I am
> not having any success with this. In the following C&P from an interactive
> session, it took 6 or 7 seconds to return the socket error, which is the
> result I would expect in the case that the host name is not in the DNS
list
> (and in this case it is not, so this is a correct result). The problem is,
> that I set the default time out (I thought???) to 0.02 seconds, but the
> lookup went waaaayyyy past that and basically seemed to ignore the
> instruction.
>
> >>> import sys
> >>> import socket
> >>>
sys.path.append(r'j:\web\thinkspotwebsite\xthinkspot\emailfilter_test')
> >>> import timeoutsocket
> >>> timeoutsocket.setDefaultSocketTimeout(0.02)
> >>> socket.gethostbyname('226.200.3.4.bl.spamcop.net')
> Traceback (most recent call last):
>   File "<pyshell#5>", line 1, in ?
>     socket.gethostbyname('226.200.3.4.bl.spamcop.net')
> gaierror: (7, 'getaddrinfo failed')
> >>>
>
> Now, I will note, that the comments in the timeoutsocket module indicate
> that it works for TCP connections, and I'm not sure that TCP is being used
> in the gethostbyname function, as it is my understanding that DNS should
> usually be done by UDP instead. So, maybe in this case, using O'Malley's
> timeoutsocket module is totally inappropriate and won't work.
>
> In any case, I'm hoping to get some clues here.
>
You are correct in noting that the timeoutsocket module won't help if you
use UDP, which by default is used for DNS. I found it was actually quite
easy to effect my own timeouts and do my own DNS lookups -- in my case,
looking for MX records, but there's no reason why you shouldn't extend it to
others. The dnslib module is an immense help here.

Here's some sample code that uses it, and uses select.select() to apply a
fifteen-second timeout to the DNS transaction. I'm afraid the code is
missing a lot of context:

def MXlist(qname, server):
    """Return a list of MX exchangers for a given domain.

    This code is based on the dnslib library, and  is complicated
    by the necessity to handle UDP errors where no response is
    generated by (or at least received from) the remote host."""
    global mxdict       # Cacheing dictionary
    if mxdict.has_key(qname):
        lf.log("[[[ Domain %s MX hosts were cached" % (qname, ))
        return mxdict[qname]
    protocol = 'udp'    # This is bogus: not used anywhere!
    port = 53
    opcode = dnsopcode.QUERY
    rd = 1              # server should recurse
    qtype = dnstype.MX
    m = dnslib.Mpacker()
    m.addHeader(0, 0, opcode, 0, 0, rd, 0, 0, 0, 1, 0, 0, 0)
    m.addQuestion(qname, qtype, dnsclass.IN)
    request = m.getbuf()
    s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
    s.setblocking(0) # We need a non-blocking socket
    s.connect((server, port))
    r, w, e = select.select([s], [s], [s], 15)
    if [r, w, e] == [[], [], []]:
        lf.debug("!!! Connect timed out in MXlist")
        return []
    s.send(request)
    r, w, e = select.select([s], [s], [s], 15)
    if [r, w, e] == [[], [], []]:
        lf.debug("!!! Send timed out in MXlist")
        return []
    r,w,e = select.select([s], [], [], 15.0)
    if r:
        reply = s.recv(2048)
    else:
        s.close()
        raise Timeout("Recv timed out in MXlist")
    u = dnslib.Munpacker(reply)
    (id, qr, opcode, aa, tc, rd, ra, z, rcode,
        qdcount, ancount, nscount, arcount) = u.getHeader()
    for i in range(qdcount):
        qname, qtype, qclass = u.getQuestion()
    MX = []
    for i in range(ancount):
        name, type, klass, ttl, rdlength = u.getRRheader()
        MX.append([u.get16bit(), u.getname()])
    MX.sort()
    servers = []
    for mx in MX:
        servers.append(string.lower(mx[1]))
    s.close()
    mxdict[qname] = servers
    lf.log("[[[ %d candidate mail exchangers for %s: %s" %
                (len(servers), qname, string.join(servers, ", ")))
    return servers

Hope this is enough to get you going.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From nospam at bigfoot.com  Tue Sep 24 18:38:30 2002
From: nospam at bigfoot.com (Gillou)
Date: Wed, 25 Sep 2002 00:38:30 +0200
Subject: List/Tuple bug or feature?
References: <8765wv6o0d.fsf@nospam.eml.cc>
Message-ID: <amqook$23v3$1@norfair.nerim.net>

"Mahesh Padmanabhan" <news at nospam.eml.cc> a ?crit dans le message de news:
8765wv6o0d.fsf at nospam.eml.cc...
>
> Hi,
>
> I spent a lot of frustrating hours tracking this down and I am
> confused as to why I see this behavior.
>
> >>> x = ('a' 'b')
> >>> x
> 'ab'
>
> >>> x = ['a' 'b']
> >>> x
> ['ab']
>

It's a feature that may be helpful for making long strings without '\n' in
it (i.e. long regexps with comments).

 longstring = (
  'this is long ' # the first
...
  'this is long '
  'this is long ' # the hundredth
  'this is long ')

Have a look at the tutorial (don't remember where exactly)

--Gilles





From aglyportREMOvethispart at nospam.yahoo.com  Tue Sep  3 23:33:14 2002
From: aglyportREMOvethispart at nospam.yahoo.com (Anton Graph)
Date: Tue, 03 Sep 2002 20:33:14 -0700
Subject: Memory fault - core dumped   while compiling 2.2.1 on SCO
Message-ID: <3D757EFA.9050401@nospam.yahoo.com>

gcc: file path prefix `dynamic' never used
cc -G -KPIC -Ki486 -belf -Wl,-Bexport 
build/temp.sco_sv-3.2-i386-2.2/timemodule.o -L/usr/local/lib -lm -o 
build/lib.sco_sv-3.2-i386-2.2/time.so
/bin/sh: 19686 Memory fault - core dumped
gmake: *** [sharedmods] Error 139
Tue Sep  3 19:56:22 PDT 2002 bongo:/u5/src/Python-2.2.1 $

I removed build/lib.sco_sv-3.2-i386-2.2/time.so
and tried again: got the same result.

Any ideas?



From cotabas at earthlink.net  Tue Sep  3 21:47:55 2002
From: cotabas at earthlink.net (CotaBas)
Date: Wed, 04 Sep 2002 01:47:55 GMT
Subject: Newby Question: os.rename
References: <AcGc9.68214$Xa.3523883@twister.southeast.rr.com> <tLGc9.68215$Xa.3526150@twister.southeast.rr.com>
Message-ID: <fDdd9.6312$6i4.491988@newsread1.prod.itd.earthlink.net>

>
> "Kevin Bass" <KBASS1 at nc.rr.com> wrote in message
> news:AcGc9.68214$Xa.3523883 at twister.southeast.rr.com...
> > I am attempting to create an archive function within my program. This
> > function should create a directory with a timestamp name (such as
> > /home/oracle/data/20020831) on the fly, list the contents of a current
> > directory then place the contain from the previous directory into the
> newly
> > created directory.
> >
> > The newly created directory with the timestamp name is being created and
> the
> > list of files from the current directory has been retrieved but my
problem
> > comes when moving the files to the newly created directory. I am trying
to
> > achieve this tasks with os.rename but find it difficult to get the
command
> > to recognize the new directory and place the files into it.
> >
> > I have attempted to get os.rename to recognize the 'archloc' variable
that
> > contains the name of the newly created directory but that did not work.
I
> > also tried to join the newly created directory with the file names then
> send
> > it through os.rename and that did not work either.
> > What am I doing wrong? Thanks!! (My code is listed below)
> >
> >
> > def archive_data(archive):
> >
> >    the_date = strftime("%Y%m%d")
> >    archloc = ('/home/oracle/data/' + the_date)
> >    os.mkdir(archloc)
> >
> >    if archive == 'yes':
> >       for file in os.listdir('/home/oracle/data'):
> >          file2 = os.path.join(archloc, file)
> >          #print file2
> >    os.rename(r'/home/oracle/data/' + file, file2)
> >
> > Kevin
> >
> >
> >
> >
>
> I have figured it out. Thanks!
>
>
>

LOL if you wait long enough people solve there own problems....








From gerhard.haering at opus-gmbh.net  Tue Sep 17 14:28:34 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 17 Sep 2002 18:28:34 GMT
Subject: CGI question
References: <slrnaoerk7.mgo.jboy@otaku.freeshell.org>
Message-ID: <slrnaoespa.2i8.gerhard.haering@haering.opus-gmbh.net>

In article <slrnaoerk7.mgo.jboy at otaku.freeshell.org>, John D. Boy wrote:
> Hello there!
> 
> I am working on a simple CGI script to use as a framework for a website
> of mine. Basically, I want it to work as follows: When somebody accesses
> the site, a CGI (called e.g. index.py) is run. It takes an html-template
> and fills it with content from another file, called e.g. index.dat, and
> returns the completed html-document. When I want to present another
> page, I want the url to look like this: http://my.domain.name/?page. In
> that case, index.py would fill the same template with the content of a
> file called e.g. page.dat. 
> 
> Now, my problem is to find out what follows behind the question mark in
> my CGI script. Unless the key has a value
> (http://my.domain.name/?key=value), it is not in the cgi.FieldStorage
> dictionary. Is it simply impossible, or am I not using the right
> function calls? 

You need to configure your webserver to rewrite URLs from your desired form to
the "/scriptname?key=value form" . I believe the relevant Apache module is
called mod_rewrite.

-- Gerhard


From rsrchstr at msn.com  Sun Sep 29 19:40:24 2002
From: rsrchstr at msn.com (mike henley)
Date: Mon, 30 Sep 2002 00:40:24 +0100
Subject: idea for a much needed application
References: <576c1752.0209291326.56dcdf31@posting.google.com> <pasepuc0r73mh5vlcsu1l7erl4e2b17qvi@4ax.com>
Message-ID: <an831g$bfrqn$1@ID-157863.news.dfncis.de>

"TGOS" <tgos at invalid.invalid> wrote in message
news:pasepuc0r73mh5vlcsu1l7erl4e2b17qvi at 4ax.com...
>
> On 29 Sep 2002 14:26:05 -0700 rsrchstr at hotmail.com (mike henley) wrote
> in comp.lang.java.programmer:
>
> > I remember a while ago reading about tim burners-lee and how his
> > vision of the WorldWideWeb as a medium where all can freely and easily
> > publish information [...]
> > Hosts are either expensive or limit you in terms of what
> > languages, modules, applications, and OSs you can use.
>
> To publish information, all you need is writing it into a HTML file and
> a server that hands out the HTML file when getting asked for it.
> You
> don't need any special languages, modules or applications on the server
> to publish your information, other than the ones already present on a
> web-server.

"other than the ones already present on a web-server", exactly! i want to be
ably to use whatever language, modules or applications i choose, and over
the year use others, freely. if i choose perl, php, or even python, ruby or
tcl, then i want to be able to do that. if i want to use mysql or something
else that i prefer, then i want to have the freedom to do that. i do not
wish to be restricted by the "ones already present on a web-server". i do
not want to end up in a catch 22 situation, should i choose a good deal on a
host, and then choose applications within the limits of what it allows me,
or choose applications, and try to find a host that'll provide a good deal.


>
> Despite that, he has not said that you can publish information without
> having banners on the page placed their by your webhost. It's very easy
> to get free webspace if all you want to do is to keep some text online.

Agree, if that's all you want, but beyond keeping some text online, it's
difficult.

>
> Despite that, my webhost does not limit me in language, module or
> application, as long as it's not a permanent background process and does
> not keep any ports permanently open or permanently stays connect to a
> server (IRC bot and the like). In every other case, as long as I can get
> it running on the server and it doesn't swallow so much CPU time, that
> other customers suffer, I'm free to use it. This means I can even
> compile my own script interpreter on the server and use if I like.
> Therefor it's not free of course.
>
> > They also limit you in terms of space,
>
> He said you can "freely and easily publish information", he didn't say
> "infinite information". Limit yourself to information not already
> provided somewhere else. If it is already provided somewhere else, link
> there. That's how the WWW works, my friend.

my post is not about tim-burners lee or what he said.

>
> How are they going to finance their service, if everyone can eat up all
> space alone, leaving nothing for the rest?

my post is also not about the economics or politics of hosting; this sounds
much like the open source vs proprietary software arguments.

>
> > bandwidth used per period,
>
> Same as above.
>
> > number of executions of CGIs...
>
> Never heard of anything like that.

i never heard of it before i carefully read the terms and conditions of some
popular hosts.

>
> > I wish i can just use my own computer, with whatever technologies I
> > choose, keeping my content that i build up over time safe here on my
> > own machine,
>
> Actually my webhost makes a backup on a weekly basis of all data on the
> server. I have never made any backup of 90% of all my data so far. So
> it's a lot less safe on my own PC. Leaving aside that my own PC runs
> Windows, which is certainly a less safe OS than Solaris IMHO.
>
> Also keep in mind that my PC is for working, and I can't use it for
> working if it permanently has to fulfill HTTP requests of some surfers,
> that not just eat up plenty of my RAM, waste plenty of my CPU time, but
> also eat up nearly all my Internet bandwidth, so I can't even take a
> look at the latest news online anymore.
>
> > but unless you have a broadband connection this is not
> > possible. It may be possible using a dynamic DNS provider,
>
> Later below you wrote that you are on dial-up. How does a dynamic DNS
> provider compensate the fact, that your connection is way too slow,
> because you have no broadband? That makes no sense.
>
> Also I think there are still free DynDNS provider, they only ask you to
> "donate" some money and if nobody donates, they will go off-line again.
> But you are not forced to pay them.
>
> > most importantly, there's no way for others to know when you get
> > online.
>
> IRC, ICQ, AIM... the list is endless.

Yes, that is what my post is about, combining instant messaging presence
with dynamic DNS.

>
>  <snip>
>
> > So instead  of giving others a URL I can just give them an IM
> > username they can add to their client, and when i go online,
> > through my humble dial-up connection that changes IP whenever i
> > log in, they'll be aware of my presence, and be able to connect
> > to a server i host on my machine.
>
> Allow users to see your IP address when you are connected to an IRC
> network or to ICQ for example, then they can enter your IP address into
> their browser address bar and will get a webpage if you run a webserver
> and are not behind a firewall that blocks port 80.
>

Perhaps it can be possible to do a little application to automate this.

>  <snip>
>
> > Perhaps it can use existing open-source platforms, such as
> > jabber or gnutella,
>
> > The more automated it is, and the easier it is, the better.
>
> Use an IRC bot on your system,

I'll try that. I still think it can probably be easy to do an application
that combines dynamic DNS with instant messaging presence, say on popular
networks such asw AIM and MSN, as an add-on to say trillian or jabber, so it
would be much easier than IRC. (IRC is messy)




From mark at mceahern.com  Fri Sep 20 13:06:01 2002
From: mark at mceahern.com (Mark McEahern)
Date: Fri, 20 Sep 2002 12:06:01 -0500
Subject: average?
Message-ID: <JHEOKEOOLIGLDHCMAHMOGEDIEAAA.mark@mceahern.com>

I know I can average a sequence like this:

  def average(seq):
    return reduce(operator.add(seq) / len(seq))

Yes, I see that will raise ZeroDivisionError if seq == [].  Aside from that,

1.  Is there anything else dumb about the above?
2.  Am I missing a builtin cousin of min, max that would do this for me?

Oops, maybe the answer to 2 is:

  import Numeric
  Numeric.average(seq)

I guess I just answered my own question.

Cheers,

// m




From psage at ncaustin.com  Mon Sep 23 23:57:06 2002
From: psage at ncaustin.com (Paul Sage)
Date: Mon, 23 Sep 2002 22:57:06 -0500
Subject: Outlook 2002 and Python
Message-ID: <03F53405C6ED434A986221919785A6A78C9339@nca-postal.ncaustin.com>

What:  I saw an idea for a way to create a good SPAM filter for email.  

How: I would like to try to implement the algorithm in Python.

Why:  I am trying to get familiar with Python, as I will be using it at
work quite a bit, and would like to keep my skills honed.

Where:  Windows XP and Outlook 2002

Okay, so smart people would say, "just go use Visual Studio or Visual
Basic," I suppose.  I don't know those two languages either, and they
might be the right tools for the job.  However, that would defeat the
point of trying something in Python.  Doing something else probably
wouldn't be as useful to me personally.

Does anyone know how and where in the world I would even start to try to
write some kind of addin for Outlook 2002 in Python?   I don't even know
how to start the project currently, and I am turning here first. 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020923/16a6d403/attachment.html>

From bobx at linuxmail.org  Sun Sep 15 15:16:22 2002
From: bobx at linuxmail.org (Bob X)
Date: Sun, 15 Sep 2002 19:16:22 GMT
Subject: [ANN] PySQLite 0.3.0 released
References: <mailman.1031924237.12603.python-list@python.org>
Message-ID: <a05h9.6196$Lg2.1046166@news2.news.adelphia.net>

sismex01 at hebmex.com wrote:
<snip>
> 
> 
> Yeesh! You had to quote the whole message just to
> add two words?
> 
> :-(
> 
> -gus

My bust on that one. I usually do a <snip> like I did above. Sorry bout dat.

On that note...your two paragraph sig is okay right?  ; )

Bob



From phr-n2002b at NOSPAMnightsong.com  Mon Sep 16 03:22:18 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 16 Sep 2002 00:22:18 -0700
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <mailman.1032152058.11043.python-list@python.org> <m3admizdzq.fsf@mira.informatik.hu-berlin.de> <37fh9.112911$pX1.4044383@news2.tin.it>
Message-ID: <7xk7lml7px.fsf@ruckus.brouhaha.com>

Alex Martelli <aleax at aleax.it> writes:
> > You'll have to give your customers a sealed computer and no permission
> > to install additional software, or communicate with the internet, if
> > you want to reliably prevent them from analysing your program.
> 
> Incidentally, the latter option can be feasible in certain situations: some 
> key parts of your program's functionality might reside on a computer you
> control, accessible to the rest of the application (installed more 
> traditionally on users-controlled computers) only as a black box via the 
> net.  XML-RPC, SOAP, and other distributed-computing approaches
> such as Corba, make implementing this particularly easy these days.

It doesn't have to be via the net: you can literally deliver a sealed
computer to the customer, with the sealing enforced by contracts and
inspections (this is quite normal, e.g. some financial applications
are delivered as a rack-mount box that you install in your machine
room) or by tamper resistant hardware (for higher security
applications).  For VERY high security, of course you can't let
hostile parties touch the hardware.  Nonetheless, the pay TV industry
deploys millions of sealed computers (smart cards) every year to
customers, every single one of which is considered an attacker.  While
that industry had spectacular early security failures, these days it's
a fairly solved problem and they limit losses to acceptable levels.

It doesn't sound like the OP was asking for this type of solution though.


From Padraig at Linux.ie  Wed Sep 11 08:25:17 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Wed, 11 Sep 2002 12:25:17 GMT
Subject: regular expression for space seperated quoted string
References: <PqFf9.3627$cP3.7118@news.iol.ie> <alnbpb$ce5$1@wanadoo.fr>
Message-ID: <NCGf9.3631$cP3.7121@news.iol.ie>

Eric Brunel wrote:
> Padraig Brady wrote:
> 
> 
>>Hi, I'm trying to split a string that is seperated
>>by spaces and also contains double quoted words which
>>can contain spaces:
>>
>>For e.g. using: re.split('[ ]*"?([^"]*)"?[ ]*', s1)
>>on this string: s1='1 "2" "thre e"'
>>gives:          ['', '1 ', '', '2', '', 'thre e', '']
>>
>>Problem with this is the '' entries, but this isn't too bad.
>>
>>However using the above re with: s2='1 2 "th ree"'
>>I get:                           ['', '1 2 ', '', 'th ree', '']
>>
>>any ideas?
> 
> 
> What about:
> 
>>>>p = r'[^ \t\n\v\f"]+|"[^"]*"'
>>>>re.findall(p, '1 2 3')
>>>
> ['1', '2', '3']
> 
>>>>re.findall(p, '1 2 "three"')
>>>
> ['1', '2', '"three"']
> 
>>>>re.findall(p, '1 "2" "thr ee"')
>>>
> ['1', '"2"', '"thr ee"']
> 
>>>>re.findall(p, '"yeah it seems to be working!" yeah it seems to...')
>>>
> ['"yeah it seems to be working!"', 'yeah', 'it', 'seems', 'to...']
> 
> It leaves the double-quotes around the values, but they can be removed 
> quite simply...
> 
> I also tried to use look-ahead and look-behind features in re's, but it 
> doesn't have the expected result:
> 
> 
>>>>re.findall(r'(?<=")[^"]*(?=")|[^ \t\n\v\f"]+',
>>>
>         '"yeah it seems to be working!" yeah it seems to...')
> ['yeah it seems to be working!', 'yeah', 'it', 'seems', 'to...']
> 
>>>>re.findall(r'(?<=")[^"]*(?=")|[^ \t\n\v\f"]+', '1 2 3')
>>>
> ['1', '2', '3']
> 
>>>>re.findall(r'(?<=")[^"]*(?=")|[^ \t\n\v\f"]+', '1 "2" "thr ee"')
>>>
> ['1', '2', ' ', 'thr ee']
> 
> The " " between 2 and three is also matched...
> 
> HTH

thanks. Yes findall is probably more appropriate:
Also there is a canonical perl soultion for this,
search for split in: http://www.perldoc.com/perl5.6/faq/perlfaq4.html

P?draig.



From buzzard at urubu.freeserve.co.uk  Thu Sep 26 20:58:58 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Fri, 27 Sep 2002 01:58:58 +0100
Subject: This is driving me insane!
References: <mailman.1033084697.25598.python-list@python.org> <an09rk$p9e$1@newsg4.svr.pol.co.uk>
Message-ID: <an0ai3$dug$1@news6.svr.pol.co.uk>

> It isn't.  self.errorlog is a list containing self.grouptotals.

I should have said it contains a reference to the same list referenced by
self.grouptotals, to be more accurate.  Creating the copy means the
reference contained in self.errorlog is to a different list than that
referenced by self.grouptotals.

Duncan




From jubafre at brturbo.com  Fri Sep 13 15:34:10 2002
From: jubafre at brturbo.com (jubafre at brturbo.com)
Date: Fri, 13 Sep 2002 16:34:10 -0300 (GMT-03:00)
Subject: transform list of int in a list of string??
Message-ID: <-2032200682.1031945650320.JavaMail.nobody@webmail1>

my program have other modules imported, like re and sre and i put a convertion int to str but doesn?t work
t=str(x[i])
Traceback (most recent call last):
  File "C:\Documents and Settings\jubafre\Desktop\montador\teste.py", line 78, in ?
    t=str(x[i])
TypeError: 'list' object is not callable

Why i can?t transform a int number in string in my program???
I tryed with import string, and can?t too, 
but if i start a new python shell it works, why??

python shell
>>> x=[1,2,3,4,]
>>> s=[]
>>> for i in range(len(x)):
	t=str(x[i])
	s.append(t)
>>> print s
['1', '2', '3', '4']



Juliano Freitas
www.gebrasil.hpg.com.br

From wlfraed at ix.netcom.com  Mon Sep 23 20:35:34 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Mon, 23 Sep 2002 17:35:34 -0700
Subject: writing in a file as binary mode????
References: <mailman.1032810110.30498.python-list@python.org>
Message-ID: <n0coma.38e.ln@ix.netcom.com>

jubafre at brturbo.com fed this fish to the penguins on Monday 23 
September 2002 12:40 pm:

> i open the a file for writing in a binary mode, but doesnt work
> 
> final3=['0020 0000x9', '0030 0000xa', '00B4 0000x8', '0030 0000xb',
> '00F0', '0001', '0003', '0005'] destino=raw_input("Digite o nome do

        Pardon the KNode garbage wrapping...

        All those are just character strings, not binary data.

        First step would be to make that data binary...

        ... = [ 0x0020, 0x0000, 0x0030, ... 0x0005 ]


> arquivo .mem: ") f = open(destino,'wb')
> f.write('\x03AHM ')
> for i in final3:
>     f.write(i+' ')

        You won't be able to concatenate a space to what is now integer binary.



--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From wurmy at earthlink.net  Tue Sep 10 00:22:10 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Tue, 10 Sep 2002 04:22:10 GMT
Subject: Python 2.3
References: <3d7cfc19_5@news.newsgroups.com>
Message-ID: <3D7D7410.4070000@earthlink.net>

JB wrote:
> Will rational numbers be implemented in Python 2.3? And when 
> is the first beta due?

There's a PEP, but it doesn't look like this will be in 2.3:

   http://www.python.org/peps/pep-0239.html

I don't know what the status of this PEP is, BTW... it's not deferred or 
rejected, which suggests this is still under consideration.

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html



From sholden at holdenweb.com  Mon Sep 16 08:12:26 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 16 Sep 2002 08:12:26 -0400
Subject: Problem with HTTPS request
References: <uo2fveepf88175@corp.supernews.com> <b0f083db.0209130558.2aa60398@posting.google.com> <uo45ara0l62cdc@corp.supernews.com> <slrnao4dkn.288.gerhard.haering@lilith.my-fqdn.de> <mYqg9.8349$WV.1529@fe03> <slrnao4h5c.288.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <GWjh9.2108$Cf.1661@fe03>

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote ...
> Steve Holden wrote in comp.lang.python:
> > "Gerhard H?ring" <gerhard.haering at gmx.de> wrote ...
> > [complaints about crappy looking posts because of people using OE]
> > Gerhard:
> >
> > You're showing distinct signs of having too much time on your hands :-)
>
> Unlikely.
>
:-)
> > Can't you fix your brain instead so it will let you ignore other
> > people's mailer problems?
>
> Problems don't go away by ignoring them. People are so used to the
> postings in crappy form (I'm only talking about form, not content)
> offloaded by people like you that they need to be alerted to this from
> time to time.
>
"People like me". Aha, so now I've become a member of an abused minority?

> > an-unreconstructed-OE-user-myself-ly y'rs  - steve
>
> I wonder why some programmers who insist on proper form of source code
> don't care at all about the email/news messages they send to the
> public, even though they know better.
>
> And sorry, the Microsoft way of defining crap as standard won't work
> with me.
>

Well, I don't think we disagree on the importance of presentation and
compliance with standards. I simply think you have to allow some leeway to
people who either a) may not know any better, or b) may not have a choice
about the mailer they use at work, for example.

Basically you're pissing in the wind. Microsoft *like* crap. Especially
*expensive* crap.

but-feel-free-to-continue-ly y'rs  -s teve
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From edream at tds.net  Mon Sep  2 13:45:35 2002
From: edream at tds.net (Edward K. Ream)
Date: Mon, 02 Sep 2002 17:45:35 GMT
Subject: Does IDLE handle unicode?
References: <QdIc9.2412$Ak.1132837@kent.svc.tds.net>
Message-ID: <3tNc9.2631$Ak.1173546@kent.svc.tds.net>

Thanks very much to all who so helpfully replied.  Sorry I didn't find this
on the FAQ.

Edward




From skip at pobox.com  Wed Sep 11 13:23:16 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 11 Sep 2002 12:23:16 -0500
Subject: Large file support and stat vs stat64 in extension modules
In-Reply-To: <pan.2002.09.11.10.00.02.210522.16129@remove.Stanford.EDU>
References: <pan.2002.09.10.16.03.39.756929.4320@remove.Stanford.EDU>
        <mailman.1031750594.11844.python-list@python.org>
        <pan.2002.09.11.10.00.02.210522.16129@remove.Stanford.EDU>
Message-ID: <15743.31748.559786.593812@12-248-11-90.client.attbi.com>

    >> Did you #include <sys/stat.h>?

    Ben> Yes, but perhaps I'm not including something else.  Here is a short
    Ben> program (test.c):

    ...

Dunno what else might be causing it.  I modified your program slightly:

    #define __USE_FILE_OFFSET64
    #include <sys/types.h>
    #include <sys/stat.h>
    #include <unistd.h>

    main()
    {
      struct stat st;
      if (stat("/home/ben/work/big/bigfile", &st) != 0)
          perror("Error");
      else printf("OK\n");
    }

and ran it through gcc -E.  I expected the stat() call to be transformed,
but it wasn't.  Apparently, I don't have large file support builtin.  In
fact, I saw

    extern int stat (__const char *__restrict __file,
                     struct stat *__restrict __buf) ;

    extern int fstat (int __fd, struct stat *__buf) ;

in the output, suggesting that by the time that little bit of code in
sys/stat.h was executed __USE_FILE_OFFSET64 had been #undef'd.

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From skip at pobox.com  Mon Sep 30 11:55:04 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 30 Sep 2002 10:55:04 -0500
Subject: Python PAM module
In-Reply-To: <mxKl9.6230$1F3.3753@nwrddc02.gnilink.net>
References: <mxKl9.6230$1F3.3753@nwrddc02.gnilink.net>
Message-ID: <15768.29656.796965.23059@12-248-11-90.client.attbi.com>

    Larry> Hello, I am wondering if anyone can point me to the location of
    Larry> the Python PAM module. My google searches turn up references to
    Larry> www/ftp.pangalactic.org, but this site seems down.

You might get lucky with a little effort and locate it at Alexa's Wayback
Machine:

    http://www.archive.org/

-- 
Skip Montanaro - skip at pobox.com
"Airplanes don't fly until the paperwork equals the weight of the
aircraft. Same with i18N." - from the "Perl, Unicode and i18N FAQ"



From andrew.thompson at ashecastle.com  Thu Sep 19 05:11:13 2002
From: andrew.thompson at ashecastle.com (Andrew Thompson)
Date: Thu, 19 Sep 2002 10:11:13 +0100
Subject: lists in dictonaries ?
In-Reply-To: <20020919074722.GB77021@partagas.as.de.cw.net>
Message-ID: <007401c25fbc$81ed2d10$4802a8c0@blair>

You can put any python objects into a python dictionary, even the
dictionary itself...

For example :

aList=[1,2,3,4]
anotherList=[4,5,6,7]

aDictionary={}

aDictionary['foo']=[aList,anotherList]

Then...

aDictionary['foo'][0] answers aList

aDictionary['foo'][1] answers anotherList


Andrew

-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]
On Behalf Of Manuel Hendel
Sent: 19 September 2002 08:47
To: Python List
Subject: lists in dictonaries ?

Is it possible to put lists in dictionaries? And is it also possible
to put multiple lists in a dictionary for one key?

Thanks,
    Manuel

-- 
I married beneath me. All women do. 
-Lady Nancy Astor, first woman member of the British Parliment  

-- 
http://mail.python.org/mailman/listinfo/python-list




From thoa0025 at mail.usyd.edu.au  Fri Sep 13 19:24:55 2002
From: thoa0025 at mail.usyd.edu.au (:B nerdy)
Date: Fri, 13 Sep 2002 23:24:55 GMT
Subject: html template system?
Message-ID: <btug9.688$Ee4.1090@news-server.bigpond.net.au>

what are good templating systems avaliable for python?

ive been at http://zebra.sourceforge.org and the documentation is kinda sus.
zope DocumentTemplate is included with zope. but i want a smaller footprint
system.

what are my options? cheers




From gerson.kurz at t-online.de  Wed Sep 25 12:36:35 2002
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Wed, 25 Sep 2002 16:36:35 GMT
Subject: Outlook 2002 and Python
References: <mailman.1032839910.12131.python-list@python.org> <A_Rj9.13335$6g7.40581@news-server.bigpond.net.au>
Message-ID: <3d91e570.13288093@news.t-online.de>

On Tue, 24 Sep 2002 04:37:52 GMT, Mark Hammond
<mhammond at skippinet.com.au> wrote:
>You would need to use the new Universal Gateway code (see 
>win32com\test\test_pycomtest.py) and create a server that implements 
>IDTExtensibility2.  From there, you should be able to use any of the VB 
>samples to move forward - all the params passed are standard COM 
>objects, so you should be able to do whatever you want.  You would also 
>need to get your head around COM events and how to connect to them from 
>Python, as addins are very much event based.

Maybe take a look at my 

http://p-nand-q.com/pynospam.htm

Its a C++ extension for outlook that embedds python (well, a fixed
python script anyway). Source included.

I like my pop3filter approach better: 

http://p-nand-q.com/e/pop3filter.html


From peter at engcorp.com  Sun Sep  1 01:28:49 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 01 Sep 2002 01:28:49 -0400
Subject: urllib bug?
References: <mailman.1030848662.20745.python-list@python.org>
Message-ID: <3D71A591.13B87F4A@engcorp.com>

Steve Clift wrote:
> 
> Is this *really* a urllib bug, or am I missing something blindingly obvious?
[snip failing example]
> Fails the same way with 2.1. Works OK with 1.5.2 and (IIRC) 2.0.

Works fine here, under Windows 98 SE:

Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import urllib
>>> urllib.urlopen('http://www.python.org')
<addinfourl at 8596352 whose fp = <socket._fileobject instance at 0x00832AB0>>
>>> f = urllib.urlopen('http://www.python.org')
>>> f
<addinfourl at 8604976 whose fp = <socket._fileobject instance at 0x00834FA0>>

-Peter


From nas at python.ca  Thu Sep 26 11:37:13 2002
From: nas at python.ca (Neil Schemenauer)
Date: Thu, 26 Sep 2002 08:37:13 -0700
Subject: off topic -- spam assassin
In-Reply-To: <200209261407.g8QE76Wo028975@ratthing-b246.strakt.com>
References: <200209261407.g8QE76Wo028975@ratthing-b246.strakt.com>
Message-ID: <20020926153713.GA16671@glacier.arctrix.com>

Laura Creighton wrote:
> What's new for our side of the arms race?  

There is a lot of interest in Bayesian analysis right now.  The error
rates are impressively low.  Search for "bogofilter" or "spambayes".
I'm currently using "neilfilter.py" from the spambayes project with good
results.

  Neil



From franke at ableton.com  Tue Sep  3 16:52:20 2002
From: franke at ableton.com (Stefan Franke)
Date: Tue, 03 Sep 2002 22:52:20 +0200
Subject: Better tracebacks for exec and execfile
References: <3i99nusjbgpiepva9pkuv3mtl7equ7h88h@4ax.com> <87elcbm1ze.fsf@andreasen.org>
Message-ID: <u38anu4g63ih3hucdo6pgpu77cm2e1r9pr@4ax.com>

On 03 Sep 2002 19:12:21 +0200, "Erwin S. Andreasen" <erwin at andreasen.com> wrote:

>compile won't let you specify a start line number: for that I simply
>added "\n" * start_line to the start of the evaluated string :)

Now this is clever! And thought much too straight-forward
to have it figured out by myself <grin>.

Thanks,
Stefan



From gerrit.muller at philips.com  Wed Sep 11 07:06:56 2002
From: gerrit.muller at philips.com (Gerrit Muller)
Date: Wed, 11 Sep 2002 13:06:56 +0200
Subject: function pointer
References: <ffcedbc9.0209110215.644c1788@posting.google.com> <mailman.1031740094.15171.python-list@python.org>
Message-ID: <3D7F23D0.F9E4E61A@philips.com>

<...snip...>

> and even better python has bound methods:
> 
> >>> class a:
>          def f(self, arg):
>              print arg
> 
> >>> a1=a()
> >>> a1.f
> <bound method a.f of <__main__.a instance at 0x813ee74>>
> >>>
> 
> you can pass these functions and methods around and apply them
> as you wish. Very powerful.
> 
>     holger
Thank you for this hint. I did not know this. Although I don't see any
immediate applications it looks quite cool!

regards Gerrit
-- 
--------------------------------------------------------------------
Gerrit Muller                Philips Research Laboratories Eindhoven
Building:   WDC 2 - 007              Prof. Holstlaan 4
Phone: +31 40 27 45242               5656 AA  Eindhoven
Fax:   +31 40 27 45033               The Netherlands
mailto:gerrit.muller at philips.com
http://www.extra.research.philips.com/natlab/sysarch/index.html


From a-steinhoff at web.de  Sun Sep  1 14:56:13 2002
From: a-steinhoff at web.de (Armin Steinhoff)
Date: Sun, 01 Sep 2002 20:56:13 +0200
Subject: Psyco works with QNX6 ...
Message-ID: <3D7262CD.332A5D9A@web.de>

Hello,

 'Psyco' works also with QNX6!! See the performance tests below ....

# Python program without Psyco:
def hanoi(n):
        han_move("a", "c", "b", n)

def han_move(frm, to, also, n):
        if (n < 2):
                return frm, to
        else:
                return han_move(frm, also, to, n - 1), \
                                han_move(frm, to, also, 1), \
                                han_move(also, to, frm, n - 1)

i = 0
while i<10000:
        hanoi (8)
        i = i+1

# time python hanoi.py
   19.10s real    18.92s user     0.00s system
                  ^^^^^
----------------------------------------------------------------------------------------------------
# Python program with PSYCO:

#import the specialization compiler
import psyco   

def hanoi(n):
        han_move("a", "c", "b", n)

def han_move(frm, to, also, n):
        if (n < 2):
                return frm, to
        else:
                return han_move(frm, also, to, n - 1), \
                                han_move(frm, to, also, 1), \
                                han_move(also, to, frm, n - 1)

# specialize han_move
psyco.bind(han_move)

i = 0
while i<10000:
        hanoi (8)
        i = i+1

RESULT:

# time python psy_hanoi.py
    2.17s real     1.87s user     0.02s system
                   ^^^^
Python + PSYCO is 10 times FASTER !  Its nearly C speed ...


Regards

Armin


From fperez528 at yahoo.com  Mon Sep 30 21:14:19 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 30 Sep 2002 19:14:19 -0600
Subject: Reloading a module
References: <bondpaper-648404.19153230092002@nnrp04.earthlink.net>
Message-ID: <anastb$h2s$1@peabody.colorado.edu>

Tom wrote:

> 
> I'm developing some python mini applets on a Mac (using os 9 and the
> 2.1.1 Python IDE), and I've found it impossible to delete and/or reload
> a module after making changes to the code during the
> debugging/verification process. Is there an easy and/or effective way to
> do this? If I execute:
> 
> del mymodule
> 
> and then execute a print dir(mymodule), it shows up as nonexistent.
> Then, when I execute:
> 
> import mymodule
> 
> followed by a print dir(mymodule), the changes do not show up.

In [1]: reload?
Type:           builtin_function_or_method
Base Class:     <type 'builtin_function_or_method'>
String Form:    <built-in function reload>
Namespace:      Python builtin
Docstring:
    reload(module) -> module

    Reload the module.  The module must have been successfully imported 
before.

cheers,

f.


From claird at starbase.neosoft.com  Tue Sep 17 15:01:58 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 17 Sep 2002 14:01:58 -0500
Subject: Python GUI/graphics package recommendation?
References: <Xns9288738D9E9EEnospamanytime@216.148.227.77> <T6vg9.137$NS3.67606@news.uswest.net>
Message-ID: <B3651D1FCD3CE406.D89DC515647B33CB.0A4590C788AF3DCB@lp.airnews.net>

In article <T6vg9.137$NS3.67606 at news.uswest.net>,
Kevin Altis <altis at semi-retired.com> wrote:
>I don't think it is very nice to make your aunt do your programming for you
><wink>, but she can see if the instructions below are easy enough to give
>PythonCard a try.
>
>http://pythoncard.sourceforge.net/installation.html
>
>There are some walkthroughs to help you along once you have everything up
>and running.
>
>ka
>
>"David Ehrens" <no.spam at any.time> wrote in message
>news:Xns9288738D9E9EEnospamanytime at 216.148.227.77...
>> I am using Python 2.2 on Win32 for an AI class and would like to
>> be able to display a graphical representation of an agent-based
>> game I am developing. I have seen a number of GUI and graphics
>> packages for Python, and would appreciate some recommendations
>> since I really don't have the time to install, learn, and try all
>> of them.
>>
>> My main criteria are: (1) my great aunt should be able to install
>> it and immediately use it; (2) it should work with Win32; (3) it
>> should do menus and graphics.
>>
>> If it happens to be the #1 tool of its kind, that would be nice
>> too.
			.
			.
			.
Start with Tkinter, I'd think; it's part of the standard
Python installation, and it certainly "do[es] menus and
graphics", for at least a couple of senses of that phrase.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From bobx at linuxmail.org  Mon Sep 16 21:14:22 2002
From: bobx at linuxmail.org (Bob X)
Date: Tue, 17 Sep 2002 01:14:22 GMT
Subject: Python equivalent to perldoc?
Message-ID: <Olvh9.7898$Lg2.1389903@news2.news.adelphia.net>

In Perl you can run "perldoc warnings" and get the definition of what 
"warnings" does. Is there a equivalent in Python?



From lpgray at uop.com  Wed Sep 18 15:28:26 2002
From: lpgray at uop.com (Lee Gray)
Date: 18 Sep 2002 12:28:26 -0700
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <Xns928CED692541cliechtigmxnet@62.2.16.82>
Message-ID: <96904b50.0209181128.63cd71f9@posting.google.com>

Chris Liechti <cliechti at gmx.net> wrote in message news:<Xns928CED692541cliechtigmxnet at 62.2.16.82>...
> lpgray at uop.com (Lee Gray) wrote in 
> news:96904b50.0209171305.6d65f574 at posting.google.com:
> > My boss is pretty open-minded, but also needs to see something working
> > to be convinced Python is even a viable platform (he had never heard
> > of it).  Otherwise, .Net is a given, whether it's any good or not,
> > since corporate is MS all the way.
> 
> with python you have to choice of diffrent GUI libs. if you consider 
> wxWindows you can play around with the wxPython demo application that comes 
> with the download. on the other hand when you support win32 only the 
> win32all package has support for all the windows APIs and it has some 
> samples that get installed.
> 
> chris

I've been playing with wxWindows and Boa Constructor. wxWindows is
pretty darn cool and will make a good demo for him.  Boa looks good
but is still a bit rough (naturally, being alpha), but I'm looking
forward to future releases.

Thanks,
Lee


From huaiyu at gauss.almadan.ibm.com  Thu Sep 12 18:01:37 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Thu, 12 Sep 2002 22:01:37 +0000 (UTC)
Subject: Restricting methods in derived classes
References: <mailman.1031783414.10800.python-list@python.org>
Message-ID: <slrnao23m1.2ks.huaiyu@gauss.almadan.ibm.com>

Mark McEahern <marklists at mceahern.com> wrote:
>[Huaiyu Zhu]
>> What I'm looking for is a way to say which methods from dict are inherited
>> in MyClass at the time the class is defined, without incuring a cost when
>> the methods are called.  Is this possible?
>
>Would something like this work?
>
>$ python
>Python 2.2.1 (#1, Jun 25 2002, 10:55:46)
>[GCC 2.95.3-5 (cygwin special)] on cygwin
>Type "help", "copyright", "credits" or "license" for more information.
>>>> class foo(dict):
>...     def not_defined(self, *args, **kwargs):
>...             raise NotImplementedError
>...     update = not_defined
>...
>>>> f = foo()
>>>> f['a'] = 1
>>>> f
>{'a': 1}
>>>> d = {'a': 2}
>>>> f.update(d)
>Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
>  File "<stdin>", line 3, in not_defined
>NotImplementedError

I was hoping for a way to confirm wanted methods, because there are more
methods that I don't want.  But the above is good enough for preventing
obvious mistakes.  Thank you.

Huaiyu


From just at xs4all.nl  Thu Sep 12 03:02:34 2002
From: just at xs4all.nl (Just)
Date: Thu, 12 Sep 2002 09:02:34 +0200
Subject: iterators and generators, is this the Python Way???
References: <3D7FF921.2040305@one.net>
Message-ID: <just-DB3861.09023412092002@news1.xs4all.nl>

In article <3D7FF921.2040305 at one.net>,
 Michael Schneider <michaels at one.net> wrote:

> class FastFile:
>     def __init__(self, filename):
>         """ Read in entire file into memory"""
>         f = open(filename, 'r')
>         fileContents = f.read();
>         self.lines = fileContents.split('\n')
>         f.close()
>    
>     def  __iter__(self):
>         return self.fastFileGen()
> 
>     def fastFileGen(self):
>         for line in self.lines:
>                 yield line

I would write the __iter__ like so:

    def  __iter__(self):
        for line in self.lines:
                yield line

Same effect, less code...

Just


From kathuw at swbell.net  Thu Sep 26 01:27:31 2002
From: kathuw at swbell.net (Kat & Huw)
Date: Thu, 26 Sep 2002 00:27:31 -0500
Subject: curses troubles
Message-ID: <002a01c2651d$69dc56e0$0100a8c0@mshome.net>

Ok, problem.

I'm trying to take user entered data input and store it as a var.

inside an __init__ I have curses set up a window, and display some 
options. Then I use:
while 1:
    self.k = self.stdscr.ch()
    if self.k == ord('G'): do_this()
    elif self.k == ord('P'): do_another()
    elif self.k == ord('Q'): break

Now, when running this, if I hit either G or P, the applications closes as if I'd hit Q

any idea what I'm doing wrong here?

yours,
Huw
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020926/42f6f973/attachment.html>

From loewis at informatik.hu-berlin.de  Thu Sep 19 13:27:53 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 19 Sep 2002 19:27:53 +0200
Subject: debugging "Too many open files"
References: <mailman.1032378623.20448.python-list@python.org> <m3y99zj8kr.fsf@mira.informatik.hu-berlin.de> <mailman.1032454586.3117.python-list@python.org>
Message-ID: <j4znudx52e.fsf@informatik.hu-berlin.de>

Alex Polite <m2 at plusseven.com> writes:

> > I recommend you list /proc/<pid of python program>/fd/
> Thanks. This tells me that some of my processes have a LOT of files
> open, most of them pointing to /dev/hdc. Is this normal?

Depends on what you mean by "pointing to". It is extremely rare that a
process opens /dev/hdc, since that requires superuser permissions; I
know of no case where a program opens /dev/hdc more than once.

So by "pointing to", you probably meant something else. Care to
provide precise data? What did you do, how did the computer respond,
how do you interpret this response?

Regards,
Martin


From loewis at informatik.hu-berlin.de  Tue Sep 10 05:37:38 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 10 Sep 2002 11:37:38 +0200
Subject: Computing directory sizes problem
References: <d32d294c.0209061035.3e6fa400@posting.google.com> <j4fzwjbjza.fsf@informatik.hu-berlin.de> <d32d294c.0209091019.320433fc@posting.google.com>
Message-ID: <j4lm6a6v8d.fsf@informatik.hu-berlin.de>

nmichael at yahoo.com (Neophytos Michael) writes:

> Python version is 2.2.1.  Windows localized version I am not sure. 
> The laptop was purchased in the US so it must be the defaul US
> version.  The filenames in question are mp3 files whose names are made
> up from greek characters (at least I see them as greek when listing
> the files with file explorer).  Any advice on how to adjust my system
> locale would be appreciated.

I can see that this makes problems. Unfortunately, you did not report
the Windows version:
- On XP, navigate to
   Control Panel, Regional And Language Options,
   Advanced, Language for Non-Unicode Programs.
  Set language to Greek. You may also set your location to Greek, but
  that might be irrelevant.
- On W2k, navigate to
   Control Panel, Regional And Language Options,
   General.
  Choose Greek.

In the long run, Python should support Unicode file names independent
of the system locale; this will happen in Python 2.3. If you want to
experiment, you can use the patch in

http://python.org/sf/594001

You'll need CVS Python for that, and MSVC. With those changes,
invoking os.path.walk with a Unicode directory name should give you
Unicode file names again, which you then can pass to os.stat.

Regards,
Martin


From imbosol at vt.edu  Mon Sep  2 15:08:47 2002
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 2 Sep 2002 15:08:47 -0400
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <mailman.1030907703.26266.python-list@python.org> <akuifb$h0e$2@solaris.cc.vt.edu> <1030962064.628232@ampungk.ozonline.com.au>
Message-ID: <al0djh$a33$1@solaris.cc.vt.edu>

John La Rooy wrote:
> Carl Banks wrote:
>> Gerhard H?ring wrote:
>> 
>>>>which means the real time is not spent in the compile() function, but
>>>>in the match or find function. So basically, couldn't one come up with
>>>>a *human readable* syntax for re, and compile that instead?
>>>
>>>That's equally powerful? Most probably not.
>> 
>> Why not?  It won't be as fast, but it should be able to do anything a
>> regexp can do, and would be much more versatile.
> 
> I think the main problem is that *human readable* doesn't map really 
> well onto regular expressions.

Ridiculous.  If you can map human readable code into machine language,
then you can map human readable code into regular expressions.

Cryptic as they are, regular expressions are still systematic; thus it
is possible to systematically convert the cryptic regexp syntax into
more readable and consistent syntax.


> What would the equivalent of r"(.)(.)(.)\3\2\1"
> This means a "palindrome of 6 characters"
> But it is unlikely that the human readable processor would understand 
> that (isn't it??)

Nope.  You don't appear to appreciate the power of computers to
translate human readable text into comlicated internal data, and are
evidently forgetting that interpretters such as Python that do a much
more difficult translation of readable text.


> It would be more likely to look like this (I haven't put too much 
> thought into this)

No kidding.


> "anything,anything,anything,same_as_3rd,same_as_2nd,same_as_1st"
> or would you like to suggest something else?

How about:

pattern = Group(Any()) + Group(Any()) + Group(Any()) \
          + GroupRef(3) + GroupRef(2) + GroupRef(1)

There's no reason it has to be re.compile with a string.


[snip]
> 
> Sure there are some cases where the re is loaded with meta characters...

That's the idea, chief.  For a simple regexp like you gave above, it
would be overkill to use a human readable syntax.  And it would still
be overkill for many regexps more complicated that that.

But eventually, the regexps will become complicated enough that a more
human readable syntax is preferrable.  Not to mention that a human
readable syntax will be more versatile, when that is needed.


> hmmm
> OK is this about writing maintainable code or people not wanting to 
> learn all the ins and outs of re's?

Nope.  For me, this is about understanding that complicated regexps
could benefit from a more readable and consistent syntax, and that the
more consistent syntax could add a lot of power and versatility to
regexps.


-- 
CARL BANKS
http://www.aerojockey.com


From nospam at horvath.com  Sat Sep 28 16:32:24 2002
From: nospam at horvath.com (Bob Horvath)
Date: Sat, 28 Sep 2002 20:32:24 GMT
Subject: combining mailbox and email modules
References: <NPck9.581867$UU1.104088@sccrnsc03> <slrnap41re.1mh.matt@happy-hour.t-island.com> <LAxk9.79207$gA4.27843@sccrnsc02> <1fj6jb2.62mx08f47amsN%fgranger@alussinan.org> <mailman.1033200523.20530.python-list@python.org>
Message-ID: <slol9.484110$_91.709647@rwcrnsc51.ops.asp.att.net>

James J. Besemer wrote:
> 
> Fran?ois Granger wrote:
> 
>> Bob Horvath <nospam at horvath.com> wrote:
>>
>>  
>>
>>> The problem is that it looks good to me.  I've played around with it, 
>>> and if I take out the MIME gunk (It actually is a Klez virus), it 
>>> parses ok.
>>>   
>>
>>
>> I have had difficulties with the handling of multipart message with the
>> email package. I did not found a solution yet.
>>
> My experience is that a lot of Spam violates Mime rules.  Often they 
> omit the end marker, I guess relying on EOF to terminate the message. 
> Then the python packages can't parse any of the remaining messages.
> 
> --jb
> 

I was assuming the mailbox package handled splitting things in to 
messages correctly.  The HeaderParser class seemed to solve it for 
me, but then I wasn't interested inparsing the body.



From kingprad at mail.com  Fri Sep 13 09:07:27 2002
From: kingprad at mail.com (Carl)
Date: 13 Sep 2002 06:07:27 -0700
Subject: [newbie] Buying an introductory/reference text
References: <Mu5g9.1702$U_.1021@news04.bloor.is.net.cable.rogers.com>
Message-ID: <668b76ce.0209130507.2a6ea225@posting.google.com>

There are quite a few online books to teach the basic syntax /
structure of python coding, such as How To Think Like a Computer
Scientist (just search for it).

The absolute best reference manual and guide is the Python Bible. I'm
at work and don't remember the authors' names. It's a big book that
introduces you to every major module with example code
(non-trivial...stuff like a threaded web crawler in about 20 lines).
Awesome reference manual - I use it every single time I sit down to
code, because the ease of learning something new makes me keep trying
new things in my code. It's $40, and worth every dollar.

Carl


From duncan at NOSPAMrcp.co.uk  Fri Sep  6 08:42:40 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Fri, 6 Sep 2002 12:42:40 +0000 (UTC)
Subject: Dictionary...
References: <7ir8g7xqqk.fsf@enark.csis.hku.hk>
Message-ID: <Xns928189FB14BA5duncanrcpcouk@127.0.0.1>

Isaac To <kkto at csis.hku.hk> wrote in
news:7ir8g7xqqk.fsf at enark.csis.hku.hk: 

> After all, Python dictionaries are implemented by height balancing
> tree, so internally the needed data can be retrieved efficiently. 

I think you must be using a different implementation of Python than the 
rest of us. :-)

Python dictionaries are implemented using hash tables and have no implied 
order. Try creating a dictionary using complex numbers as keys if you don't 
believe me, then try sorting the keys of that dictionary.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From jb at cascade-sys.com  Sun Sep 29 03:35:38 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sun, 29 Sep 2002 00:35:38 -0700
Subject: Getting started
References: <GCEDKONBLEFPPADDJCOEKEPPFDAA.whisper@oz.net>
Message-ID: <3D96AD4A.8080202@cascade-sys.com>

David LeBlanc wrote:

>In fact, the names of some Lisp funtions came from the machine registers of
>the same name: car (current address register) and cdr (current data
>register) are the only ones I recall offhand.
>

Yup.  It was from IBM's 709/7094 computer.

Some Lisps had a 3rd, "CSR" but I don't recall the name of the register. 
 Status?  Subroutine?

Being able to branch in yet a third direction did not add as much 
utility as the first two.  And it didn't translate as well to most other 
architectures.  

It was nevertheless implemented in one CDC Lisp I am aware of, where 60 
bit words and 18 bit addresses left plenty of room for 3 addresses (with 
bits left over) in a cell.

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From martin at v.loewis.de  Tue Sep 10 02:58:47 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 10 Sep 2002 08:58:47 +0200
Subject: how to build wxPython w/ Sun's C++ compiler?
References: <mailman.1031607011.28525.python-list@python.org> <m3k7luojpq.fsf@mira.informatik.hu-berlin.de> <mailman.1031621231.16841.python-list@python.org>
Message-ID: <m33csimiu0.fsf@mira.informatik.hu-berlin.de>

Skip Montanaro <skip at pobox.com> writes:

> As for patches, it seems to me that distutils needs a generic way to
> override such settings (CC, CFLAGS, LDFLAGS, etc) via command line
> arguments, environment variables or ConfigParser files.  

This is true.

> It's not clear which way would be most Pythonic.

It would be Pythonic to honor the system conventions transparently. So
on Unix, environment variables should be considered; I don't know what
the conventions on other systems are (or whether the problem exists in
the first place). On Windows, you want a choice of using MSVC or gcc
or Borland C or ... This is already supported in a command line
switch, selecting the appropriate compiler class.

Regards,
Martin


From aleax at aleax.it  Fri Sep 20 05:34:31 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 09:34:31 GMT
Subject: monetary applications [was: Python GUI app to impress the bos s?]
References: <31575A892FF6D1118F5800600846864DCBD817@intrepid> <7xk7lhb943.fsf@ruckus.brouhaha.com> <9Qzi9.117104$ub2.2497933@news1.tin.it> <7xd6r9uj2i.fsf@ruckus.brouhaha.com>
Message-ID: <HYBi9.117876$ub2.2506835@news1.tin.it>

Paul Rubin wrote:

> Alex Martelli <aleax at aleax.it> writes:
>> Using double binary floating-point, the result of 0.70 x 1.05 is
>> 0.73499999999999998667732370449812151491641998291015625; the result
>> should have been 0.735 (which would be rounded up to $0.74) but instead
>> the rounded result would be $0.73.
> 
> Interesting.
> 
>> Should your business get tax-audited, inspectors may take a dim view
>> of your shaving a cent off the sales tax due on each and every item
>> you sell to the public at $0.70 -- "I found it hard to imagine" how
>> rounding effects might affect the result (when the law specifies
>> rounding to the nearest cent and always rounding half cents up) may
>> not be an acceptable defense.
> 
> I don't think they ask or care whether you round up or down on
> individual 70 cent items.  They want to know the total value of the
> stuff you sold, and then the tax you pay is X percent of that total,
> for whatever your local value of X is.

And they have no problems with letting you claim to your customers
that their sales tax is M cents when it's really N cents, N != M ...?

Wow -- your jurisdiction must be far laxer than ones I've dealt with.

Around here, if you assert the tax is M, and thus have the customer
pay an extra M, it HAD BETTER be M -- doing otherwise could be
constituted as fraud (depending on circumstances).

Say that you publish the erroneous information that each phone
call is $0.70 plus $0.03 sales tax for a total of $0.73.  Sales
tax is actually $0.04.  Now, a customer makes 100 calls.  Do
you charge them $74 explaining that 100 times 73 cents is 74
dollars?  Or do you charge $73 and eat the dollar loss yourself?
And even then, your tax books won't balance without fiddling --
even if the unbalance is against your interests, as in this case,
it's _still_ illegal (at least around here)... double-entry tax
book-keeping is supposed to balance to the last cent.

But what you said were that you had a hard time imagining any
commercially significant computation being off by a whole penny
(which, I imagine, means a whole cent) by using IEEE double
(binary floating point).  Surely this one example IS enough to
jog your imagination?  Surely it would not be a _mistake_ for
the rest of the program to sum up net-of-tax amounts and taxes
separately, item by item, rather than summing up just the
net-of-tax amounts and then recomputing overall tax based on
that -- then, the overall amount of tax computed and paid to
the authorities might be off by 1 cent per phone call billed,
which can sum up to quite a pretty sum (smacks of the "salami
slice fraud" to me, in fact).


Alex



From mfranklin1 at gatwick.westerngeco.slb.com  Wed Sep 18 12:20:02 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Wed, 18 Sep 2002 16:20:02 +0000
Subject: Tkinter:select directory dialog?
In-Reply-To: <20020918105332.00910.00007029@mb-ch.aol.com>
References: <20020918105332.00910.00007029@mb-ch.aol.com>
Message-ID: <200209181620.02942.mfranklin1@gatwick.westerngeco.slb.com>

On Wednesday 18 Sep 2002 2:53 pm, Park997 wrote:
> Hi,
> I have written an application that reads in several data files located in a
> single directory. Currently the directory is hard coded in a module as a
> string variable. On other computers, the data may be located in other
> directories. Is there a Tkinter dialog for choosing a directory, but not
> picking a file to open or write to? An entry field could be OK, but that
> would involve typing in a path, and probably validity checking, etc. A nice
> graphical "choose a directory"  Dialog would help. Does such a thing exist?
>
> Thanks,
>
> Wendell Cropper

from tkCommonDialog import Dialog
class Chooser(Dialog):
    command = "tk_chooseDirectory"
    def _fixresult(self, widget, result):
        if result:
            self.options["initialdir"] = result
        self.directory = result # compatibility
        return result

def askdirectory(**options):
    return apply(Chooser, (), options).show()






-- 
### Python Powered Signature
I started writting this email on 
Wed Sep 18 16:18:44 2002




From member at dbforums.com  Mon Sep 16 21:22:23 2002
From: member at dbforums.com (hpyhpy)
Date: Tue, 17 Sep 2002 01:22:23 +0000
Subject: Anyone can tell me why?
References: <1815200.1031973626@dbforums.com> <1817360.1032087013@dbforums.com> <am5ik7$ldv$1@usenet.Stanford.EDU>
Message-ID: <1822798.1032225743@dbforums.com>

Originally posted by Andy Salnikov
> "hpyhpy"  wrote in message
> news:1817360.1032087013 at dbforums.com"]news:1817360.1032087013 at dbforum-
> s.co-
> m[/url]...
> > But I try,change the filename as "example.so","_example.so",
> > "examplemodule.so",#_#,same error......have other advise?^_^
>
> 1. You run python from directory different from where your .so
>    lives in. Python finds everything in the well-defined places, which
>    includes local directory. From python prompt type:
>
> >>> import sys
> >>> print sys.path
>
>   to see the list of directories where python can find a module. Do
>   not
>   put your module outside of any directory in this path.
>
> 2. Your python was built without support for dynamic loading. Don't
>    know how to check this. Ask the person who installed/configured
>    your python.
>
> 3. Shared library was not build correctly. I'm not familiar with SCO
>    but usually with gcc I do 'gcc -shared object.o -o whatever.so' and
>    that
>    works.
>
> 4. Whatever else....
>
>   Andy.


Yeah,thank you for these advise.

I suppose it may associate with the second point you said,I download
python 2.xx,as GNU,./configure,./make,it can't install on sco,so I use
python 1.5 .

And swig is same thing,last I use v1.1.Somebody tell me I shall check
the compiler and python and swig should be same----Oh,I can't believe :

When compile python ,system use GCC,and compile swig,it use CC(sco
with it self).

I change it and ok^_^

--
Posted via http://dbforums.com



From sismex01 at hebmex.com  Thu Sep 19 10:09:19 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Thu, 19 Sep 2002 09:09:19 -0500
Subject: strong/weak - dynamic/static [Was: Getting started]
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233024CC847@mail.hebmex.com>

> 
> Sorry to go off at a tangent ...
> 
> "Lance" <lbrannma at cablespeed.com> writes:
> 
> > Hi All,
> > 
> > Last night I attended a Weak typing, C++ Templates, and 
> > Python talk by Bruce Eckel. It was great, my introduction
> > to Python.
> > 
> > I'm sold on Python and weak typing. I want to write a 
> > graphics module that will link to a C application,
> > permitting scatter plots, bar charts, log scales, etc.
> > I suppose the Python Imaging Library will be used for this
> > purpose.
> 

Sorry pal, someone must have sold you some ranch land
in Florida also. ;-)

Python is definitely NOT weakly typed, it's dynamicly-strong
typed. A variable, being only a "reference to *any* object"
may refer to any object, but that object (see the difference?)
may not change it's functionality or type during runtime.

So, being strong typed, you cannot "cast" an object to another
type of object, like in C, you can do some strange contortions
to get what you want (forcing the compiler to do what you need,
when doing it in ... assembler, for example, would be quite
direct. Obtain the first byte from a double number,
for example:

	unsigned char c = *(char*)(&somefloat);

Now, look at this variable "unsigned char c". It can never,
ever, during the execution of your program, become anything else.
Why? Because it's storage was staticly allocated during
runtime, it doesn't hold a "reference to an unsigned_char object",
it actually *is* the address of the byte where an unsigned-char
is held;  although it obtained it's value through some ugly
type contortions.

So, you see, in C (a strong, static typed language) I obtained
a char from a floating-point number; not the string representation,
just the first byte.

On the other hand, in Python, we have:

>>> num = 3.1315

"num" contains a reference to the floating point object 3.1415,
which exists "somewhere in memory" (although you mustn't care
about that).

If you do the following:

>>> num = "3.1415"

Now, "num" has a reference to an entirely different object,
a string which contains "3.1415", which is only a string,
it doesn't matter if it "looks" like a number.

If you should do something like:

>>> num += 10
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
TypeError: cannot concatenate 'str' and 'int' objects

See? Python won't let you do operations on an object which
correspond to a different object type. This is what it means
for "strong typing". But, since you can change a variable's
binding at runtime, it's dynamicly typed.

On the other hand, say, VBScript (yuck), you can do the
following:

DIM a,b
a = "Howdy!"
b = 3.1415

You have two variables, "a" which is a string, and "b" which
is a floating-point number. You can do operations such as:

a = a & b

"&" is a string concatenation operator, it's used to concatenate
strings; but since you're passing a floating-point number,
it automagically converts it to string and then concatenates
it. This is what weakly-typed languages to, they convert
objects to different types depending on the operation you're
trying to do with them.

VBScript is weak typed; Perl is notoriously weak typed
(everything's a string to Perl). Other scripting languages
are weak typed.


This question comes up every now and then, I think I'll
save this post, and repost it when the need comes. :-)

> A quick google search suggests that Bruce Eckel really claims that
> python is weakly typed (at least more weakly typed than C++).

This is Phalse (i.e., "very false"). An object in C++ can be
cast to a different type, same as with Java; in Python you
can do no such contortions (there's no need though).

The need in C++ and Java to have type casts comes from the
fact that variable's are staticly typed, so you can't bind
them to a different type of object than what they were
declared with, so you need to fix it up, by saying to the
program "you're receiving a X pointer, but it's not really
pointing to an X, but to a Y, and you're Y, so everythin's
OK."


HTH

-gus





Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From skip at pobox.com  Wed Sep 11 16:40:51 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 11 Sep 2002 15:40:51 -0500
Subject: "str.contains(part)" or alternatives?
In-Reply-To: <alo8jr$l04$1@news.online.de>
References: <alo8jr$l04$1@news.online.de>
Message-ID: <15743.43603.971831.920711@12-248-11-90.client.attbi.com>

    >>> try:
    ...     s.index('ll')
    ...     do_something_if_found()
    ... except ValueError:
    ...     do_something_if_not_found()

    Stefan> This has the problem that a ValueError raised in
    Stefan> do_something_if_found() may give the false impression that the
    Stefan> substring isn't contained in the string s. 

I try to minimize the amount of code executed in try blocks.  Use this
instead:

    try:
        s.index('ll')
    except ValueError:
        do_something_if_not_found()
    else:
        do_something_if_found()

Also note that in 2.3 the semantics of the 'in' operator will change to make
this sort of test more natural:

    % python
    Python 2.3a0 (#86, Sep  4 2002, 21:13:00) 
    [GCC 2.96 20000731 (Mandrake Linux 8.1 2.96-0.62mdk)] on linux2
    Type "help", "copyright", "credits" or "license" for more information.
    >>> 'll' in 'hello'
    True

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From whisper at oz.net  Mon Sep 16 01:47:03 2002
From: whisper at oz.net (David LeBlanc)
Date: Sun, 15 Sep 2002 22:47:03 -0700
Subject: couldnt find __dic__
In-Reply-To: <7153651c.0209151958.4099ac3e@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEAEOBEPAA.whisper@oz.net>

try __dict__ maybe?

More information might be useful. New style classes with slots don't have
__dicts__ as far as I know.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of black
> Sent: Sunday, September 15, 2002 20:59
> To: python-list at python.org
> Subject: couldnt find __dic__
>
>
> my book told me all dictionaries have __dic__ method but I searched
> each dictionaries and find nothing, where is it ???
> --
> http://mail.python.org/mailman/listinfo/python-list




From nevwill at altavista.net  Wed Sep  4 19:27:35 2002
From: nevwill at altavista.net (neville)
Date: 4 Sep 2002 16:27:35 -0700
Subject: Please help..
Message-ID: <bfe409d5.0209041527.6d51ab06@posting.google.com>

Hi,
I'm new to programming and python and I'm having trouble defining
functions.I have worked through 2 books and many tutorials and still
don't understand it.As this appears to be fundamental to python could
some-one please offer some suggestion.I have included a small program
that works, but how do I write it correctly.
from Tkinter import *
from time import *
t1=localtime()
n=asctime(t1)
t2=(2002,6,13,0,0,0,3,163,0)#13/6/2002
t3=(2002,6,14,0,0,0,4,164,0)#14/6/2002
ts=mktime(t1)-mktime(t2) #non smoking 
tr=mktime(t1)-mktime(t3) #for rent
d1=int(tr/((60*60)*24)) #convert to days
dr=d1%14 #rent due fortnigtly
if dr<7:
    c="The rent is due this Friday !"
elif dr==7:
    c="The rent is due today !!"
elif dr>7:
    c="The rent is due next week !"

d=int(ts/((60*60)*24))
a="Non-smoking for... %d" %d
m=(d/2)*50 #$50 saved every 2 days
b="Cash is... $%d" %m
root=Tk()
frame=Frame(root,bg="white")
frame.pack()
n=Label(frame,bg="white",fg="deeppink",text=n)
w=Label(frame,bg="white",fg="blue",text=a)
s=Label(frame,bg="white",fg="blue",text=c)
t=Label(frame,bg="white",fg="blue",text=b)
n.pack(padx=50,pady=10)
s.pack()
w.pack()
t.pack()
root.title("No Smoking!!!!")
root.mainloop()
I hope you can understand this and help me,
thanks, Neville


From eugene at ulka.com  Tue Sep 17 20:11:00 2002
From: eugene at ulka.com (Eugene Pervago)
Date: 17 Sep 2002 17:11:00 -0700
Subject: Loading modules from several directories
Message-ID: <a1247c5b.0209171610.243f3c02@posting.google.com>

Trying to reuse the code I found a problem (with Python or me) where
if I have for example, a foo package in two separate directories
(where foo/ contains common files and foo/project_a/ contains
project-specific files) I can't load from both.

If I put the common path first in the sys.path it tries to load the
project-specific stuff from there too. If I try putting the
project-specific one first it can't find the common stuff.

Anyone know of a good solution for this problem?

Thanks,
Eugene Pervago


From skip at pobox.com  Thu Sep 12 15:55:05 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 12 Sep 2002 14:55:05 -0500
Subject: Cursed newlines and readline()
In-Reply-To: <6u5g9.11312$Da3.1133980@kent.svc.tds.net>
References: <6u5g9.11312$Da3.1133980@kent.svc.tds.net>
Message-ID: <15744.61721.37284.447554@12-248-11-90.client.attbi.com>

    Edward> According to one of Leo's Linux users, the Python's readline()
    Edward> routine on Linux delivers "\r\n" line end strings verbatim,
    Edward> while the windows versions force the string to use the Unix
    Edward> convention of using only "\n". 

In 2.3 you can open files with the "U" flag and get "universal newline"
support: 

% python
Python 2.3a0 (#86, Sep  4 2002, 21:13:00) 
[GCC 2.96 20000731 (Mandrake Linux 8.1 2.96-0.62mdk)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> f = open("crlf.txt")
>>> line = f.readline()
>>> line
'This is an example of what I have come to call the "cursed newline" problem,\r\n'
>>> f = open("crlf.txt", "rU")
>>> line = f.readline()
>>> line
'This is an example of what I have come to call the "cursed newline" problem,\n'

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From robin at jessikat.fsnet.co.uk  Wed Sep  4 10:50:00 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 4 Sep 2002 15:50:00 +0100
Subject: jython memory usage
Message-ID: <hVm5ydAY2hd9Eww$@jessikat.demon.co.uk>

A colleague has been running tests of reportlab under jython-2.1 several
of the tests use prodigious amounts of memory (one even failed). We're
using win2k+j2sdk1.4.0_01 as the jython engine.

Anyone know what sorts of things cause x10 memory usage? I find one
particular test needs more that 1Gb and then falls over for lack of
memory. 
-- 
Robin Becker


From markus.vonehr at ipm.fhg.de  Tue Sep  3 10:11:58 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Tue, 03 Sep 2002 16:11:58 +0200
Subject: hackicon, tkicon
References: <3D749F85.962D8D39@ipm.fhg.de> <al2avp$7jp$1@wanadoo.fr>
Message-ID: <3D74C32E.9FB8BD47@ipm.fhg.de>

Hi Eric,

I just had a look at my python2.2 installation:
there is the tk83.dll in. Can I change it to tk84.dll?

Thanks,

Markus


Eric Brunel schrieb:
> 
> Markus von Ehr wrote:
> 
> > Hi,
> >
> > is there already a hackicon or tkicon module for
> > python 2.2 ?
> > The modules are useful for changing the window icon on win32.
> >
> > Thanks for any hints,
> >
> > Markus
> 
> IIRC, this is no more needed with Tcl/Tk 8.4: the standard methods
> iconbitmap and/or iconwindow should work on Windows. See:
> 
> http://www.tcl.tk/man/tcl8.4/TkCmd/wm.htm#M18
> 
> But I haven't actually tested it...
> 
> HTH
> --
> - Eric Brunel <eric.brunel at pragmadev.com> -
> PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From claird at starbase.neosoft.com  Tue Sep 24 17:56:11 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 24 Sep 2002 16:56:11 -0500
Subject: Reloading code without shutting down the app
References: <3nipma.itu.ln@cvs>
Message-ID: <93C6452FB23609A5.1A11EB70FD313CA7.B0E6371E22A16155@lp.airnews.net>

In article <3nipma.itu.ln at cvs>,
Max Ischenko  <max at malva.com.uaREMOVE.IT> wrote:
			.
			.
			.
>Does python have some ways to reload source code of already running
>modules? For example, when consumer recv. a SIGHUP it would delete
>currently used objects/classes/modules, import all again, creates objects
>and continues to process the data from the Queue.
>
>Is it possible? And how?
			.
			.
			.
The easy part is importing new stuff; read up on
eval and exec.

It takes more subtlety to "delete currently used
objects/classes/modules".  Easiest is to rewrite
your existing work slightly so it has enough self-
knowledge either to allow itself to be overwritten,
or to announce that it should be deleted.  Can you
make those changes?
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From klamidad at wanadoo.es  Tue Sep 10 20:58:48 2002
From: klamidad at wanadoo.es (klamidad)
Date: Wed, 11 Sep 2002 02:58:48 +0200
Subject: Lambda?!
References: <mailman.1029852811.13661.python-list@python.org>
Message-ID: <ag4mla-mcm.ln@micasa.localnet>

Bo M. Maryniuck wrote:

> Python 2.0 (#1, Jan 19 2001, 17:54:27)
> [GCC 2.95.2 19991024 (release)] on linux2
> Type "copyright", "credits" or "license" for more information.
>>>> def foo(a):
> ...     return "<%s>" % a
> ...
>>>> map(lambda bar: foo(bar).upper(), ['test', 'other test'])
> ['<TEST>', '<OTHER TEST>']
> 
> 
> Hura-a!
> 
> 
> Python 2.0 (#1, Jan 19 2001, 17:54:27)
> [GCC 2.95.2 19991024 (release)] on linux2
> Type "copyright", "credits" or "license" for more information.
>>>> class foo:
> ...     def bar(self, a):
> ...             return "<%s>" % a
> ...     def spam(self):
> ...             return map(lambda rope: self.bar(rope).upper(),
> ['test', 'other test']) ...
>>>> f = foo()
>>>> f.spam()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "<stdin>", line 5, in spam
>   File "<stdin>", line 5, in <lambda>
> NameError: There is no variable named 'self'
> 
> 
> Huh?!
> 
> 

Python 2.1.3 (#1, Jul 29 2002, 22:34:51) 
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> class foo:
...     def bar(self, a):
...             return "<%s>" % a
...     def spam(self):
...             return map(lambda rope: self.bar(rope).upper(), 
['test', 'other test'])
... 
<stdin>:4: SyntaxWarning: local name 'self' in 'spam' shadows use of 
'self' as global in nested scope 'lambda'
>>> f = foo()
>>> f.spam()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 5, in spam
  File "<stdin>", line 5, in <lambda>
NameError: global name 'self' is not defined

'self' local in 'spam' <---> 'self' global in 'lambda'

Python 2.1.3 (#1, Jul 29 2002, 22:34:51) 
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> class foo:
...     def bar(self, a):
...             return "<%s>" % a
...     def spam(self):
...             return map(lambda rope,slf=self: slf.bar(rope).upper(), 
['test', 'other test'])
... 
>>> f = foo()
>>> f.spam()
['<TEST>', '<OTHER TEST>']

Ahh!

 


From bokr at oz.net  Tue Sep  3 20:55:57 2002
From: bokr at oz.net (Bengt Richter)
Date: 4 Sep 2002 00:55:57 GMT
Subject: winsound.Beep() frequencies
References: <3ab719f7.0209031229.33667461@posting.google.com>
Message-ID: <al3lmt$uq6$0@216.39.172.122>

On 3 Sep 2002 13:29:23 -0700, megabytemonster at hotmail.com (Stuart) wrote:

>While looking at the python documentation, I noticed that there are no
>frequency/pitch tables for using with the winsound module.  Does
>anyone know whether this documentation exists, and/or what the
>corresponding frequencies to pitches are?
>
Well, 440 hz is an A, I believe. And an octave is a factor of 2.
But I believe the best sounding scale will not have 12 equal ratios
of 2**(1/12).

Google is your friend...

    http://ccrma-www.stanford.edu/CCRMA/Courses/220a/unjung/wellScale.html

(this table does appear to be based on equal ratios).

... which is easy to compute with Python:

 >>> notes = 'A Bb B C Db D Eb E F Gb G Ab'.split()
 >>> notes
 ['A', 'Bb', 'B', 'C', 'Db', 'D', 'Eb', 'E', 'F', 'Gb', 'G', 'Ab']
 >>> def noteFreq(name, oct):
 ...     return (27.5*2**oct)*2.0**(notes.index(name)/12.)
 ...
 >>> noteFreq('A',4)
 440.0
 >>> noteFreq('F',2)
 174.61411571650194
 >>> noteFreq('C',4)
 523.25113060119725
 >>> noteFreq('A',0)
 27.5
 >>> noteFreq('Ab',7)
 6644.875161279122


Here's more. Careful, it's easy to get interested ;-)

    http://www.organicdesign.org/peterson/tuning/

Regards,
Bengt Richter


From idjhh at hanmail.net  Sat Sep 14 23:52:50 2002
From: idjhh at hanmail.net (idjhh)
Date: 14 Sep 2002 20:52:50 -0700
Subject: registry question (_winreg)
Message-ID: <bcb4afaa.0209141952.6a1c53@posting.google.com>

hi ^^
i write this code today , but it is not running -_-;
i don`w find error position
see my code and then please advise me ^^
purpose of this program is to show window startprogram list ^^


=================================================================
import _winreg
root_key=_winreg.HKEY_LOCAL_MACHINE
sub_key=_winreg.OpenKey(root_key,"Software\\Microsoft\\Windows\\CurrentVersion\\Run")
number_value=_winreg.QueryInfoKey(sub_key)[1]
i=1
reg_name_list=[]
reg_data_list=[]
while i <= number_value :
    value_name=_winreg.EnumValue(sub_key,i)[0]
    value_data=_winreg.EnumValue(sub_key,i)[1]
    reg_name_list.append(value_name)
    reg_data_list.append(value_data)
    i=i+1

_winreg.CloseKey(sub_key)

==================================================================


From peter at engcorp.com  Tue Sep  3 23:29:15 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 03 Sep 2002 23:29:15 -0400
Subject: Design question.... parent/child class
References: <3d74b021$0$234$4d4ebb8e@news.nl.uu.net> <3d74b283$1@news.sentex.net> <3d74d0c9$0$225$4d4ebb8e@news.nl.uu.net>
Message-ID: <3d757def@news.sentex.net>

Gumuz wrote:
> I'm sorry, I don't think I understand it correctly.
> 
> Do you mean i should create a third Interface class just to communicate to
> the other class? Isn't that then just the same but with an extra class?

That's basically what I meant, but I wasn't really suggesting that
you needed to do that right now, just pointing it out as a possibility 
to consider.  It's not something you'd do if your code is still simple,
but *depending* on what SessionManager does in addition to this message
passing, it might be a necessary step to reduce complexity in the future.

As for "isn't it just the same?" I would say no, not if you end up
significantly increasing "cohesion" in the SessionManager class as
a result.  Learn about "coupling" and "cohesion" (probably many many
pages on the web that give tutorials on these terms) and you'll
understand where I was coming from.  (But I repeat, I'm *not* suggesting
you need to do that right now, if ever.)

-Peter



From malan at club-internet.fr  Thu Sep 19 18:09:53 2002
From: malan at club-internet.fr (Alan Marchand)
Date: Fri, 20 Sep 2002 00:09:53 +0200
Subject: sorry for this stupid question about struct module
Message-ID: <3d8a4ae7$0$280$7a628cd7@news.club-internet.fr>

Hi all,

I tryed this :

e:\temp>python.exe
Python 2.2.1 ( #34, Apr 9 2002, ...................)
...........
>>> import struct
>>> x = struct.pack( '>I', 10 )
>>> print repr( x )
'\x00\x00\x00\n'                       # why not !

How can I obtain the result a expect : '\x00\x00\x00\0A' ?

Remark :
This, to send a message to a IBM Mainframe in order to start an IMS
transaction ( via IMS CONNECT ).





From borcis at removethis.infomaniak.andthis.ch  Thu Sep 19 11:31:23 2002
From: borcis at removethis.infomaniak.andthis.ch (Boris)
Date: Thu, 19 Sep 2002 17:31:23 +0200
Subject: Iraq's offer to readmit arms inspectors
References: <d8f8142b.0209161913.12cce045@posting.google.com> <vWBh9.26065$Gp1.687517@weber.videotron.net> <3D86FF01.FE0CECC5@geneva-link.ch> <3d87069c.34652637@news.rcn.com> <3D871316.AA1E8FFC@geneva-link.ch> <3d882cd4.9181792@news.rcn.com> <3f74f974.0209180713.27889f74@posting.google.com> <l9ehouci1ga88ub2csdam1qroiik0ogd4r@4ax.com> <3d88c258$0$707$5402220f@news.sunrise.ch> <v4lhouolv1c0ijmihfom94j0lor2lhntli@4ax.com> <3d89be92$0$707$5402220f@news.sunrise.ch> <s6mjous6o5abg3icq12u156v5tijhcnk7h@4ax.com>
Message-ID: <3d89ee85$0$712$5402220f@news.sunrise.ch>

Malev : [after accusing me of an opaque vice on the basis of an
              incompetent interpretation of the headers of his copy
             of my post]
>
> >by this channel. This really has nothing to do with me.  You should
> >abstain from reading technical small print if it's beyond your ability
> >to interpret it adequately.
>
> Pompous patronizing fool.

I prefer that to being the unrepenting "innocent" liar that you are, further
compounding his crime by hiding wholly unjustified conclusions under
allusive
language; just like the American president, if I may say so.

I'd also like to warn you that citing my prose by cutting it the way you do,
is inviting supernatural catastrophes on yourself and your environment :)

OTOH, I can also withdraw the last sentence of the part that you find
offending,
provided you'll display enough standards to give a similar attention to a
similar
message of my own choosing :

python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")

What do you think it means ?

Boris




From DennisR at dair.com  Wed Sep 18 00:48:27 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Wed, 18 Sep 2002 04:48:27 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <Xns928CED692541cliechtigmxnet@62.2.16.82> <C05DA40F31B09329.0165201CAF094801.13AF9332EB64347A@lp.airnews.net> <s4Qh9.479726$q53.16320265@twister.austin.rr.com>
Message-ID: <vATh9.666$oY5.33977741@newssvr14.news.prodigy.com>

> > is in serious production use at tens of THOUSANDS of
> > sites.

> I read that all the time (and believe it), but that's not going to impress
> my boss, much. He's still going to want something demonstrated up close
and
> personal.

Would he be impressed by a user interface that powers 2.1 BILLION web
documents? (ref: alltheweb.com)

I am about 2 weeks away from releasing HTUIL, an integrated client for
Windows, hosting Python, Perl, and/or native PLD2 displaying in a standard
web browser.  See www.dair.com/hellopy.exe for a "hello" program written in
Python and distributed as a standalone EXE on Windows.

--

Dennis Reinhardt

http://www.dair.com




From bokr at oz.net  Mon Sep 16 18:10:26 2002
From: bokr at oz.net (Bengt Richter)
Date: 16 Sep 2002 22:10:26 GMT
Subject: "str.contains(part)" or alternatives?
References: <alo8jr$l04$1@news.online.de> <eWNf9.165073$l_4.88540@atlpnn01.usenetserver.com> <aloav7$mdc$1@news.online.de> <3d8013e8$0$309$39cecf19@nnrp1.twtelecom.net>
Message-ID: <am5ksi$lvc$0@216.39.172.122>

On Wed, 11 Sep 2002 21:18:23 -0700, Greg Fortune <lists at gregfortune.com> wrote:

><snip>> 
>>> I think the current proposal is to move to
>>> 
>>>     if ll in s:
>>> 
>>> for this construct. Happy?
>> 
>> I would like .contains more, but at least the proposal is _much_ better
>> than "if s.find(part) != -1". :-)
>> 
>> "if part in s" disturbs the pattern "if item in sequence" somewhat but
>> that has already been weakened by "if key in dict" (that I like,
>> nonetheless).
>> 
>> Stefan
>
>
>Why not simply use .count(part) > 0 ?  I didn't catch the first part of the 
>conversation, but if I'm looking for a contains, I just do.
>
>if(search_me.count(substr) > 0):
>        print 'found'
>else:
>        print 'not found'
>
>
Mostly it won't matter, but in cases like

    search_me = 'a'*10000000
    substr = 'a'

it will 'way suboptimal in comparison with find or index.

Regards,
Bengt Richter


From franke at ableton.com  Mon Sep  2 16:26:18 2002
From: franke at ableton.com (Stefan Franke)
Date: Mon, 02 Sep 2002 22:26:18 +0200
Subject: Regex for String Literals
Message-ID: <47i7nu8q5b3e5merhe7s19qvgsf244nkfq@4ax.com>

Does someone know a regular expression that matches all
kinds of Python string literals (along with their  finer points 
WRT line breaks, unicode..)?

Stefan



From chris_mk at hotmail.com  Fri Sep 13 19:29:07 2002
From: chris_mk at hotmail.com (Christopher)
Date: 13 Sep 2002 16:29:07 -0700
Subject: transform list of int in a list of string??
References: <-1483717605.1031940885180.JavaMail.nobody@webmail2.brturbo.com> <mailman.1031941396.19665.python-list@python.org>
Message-ID: <cc14131d.0209131529.6eac766c@posting.google.com>

I think you meant:
y = [str(n) for n in x]

Chris

Skip Montanaro <skip at pobox.com> wrote in message news:<mailman.1031941396.19665.python-list at python.org>...
> juliano> How can i transform a list of integer numbers in a list of
>     juliano> string numbers, like that: 
> 
>     juliano> x=[1,2,3,4] to y=['1', '2'...]????????
> 
> Try 
> 
>     y = map(str, x)
> 
> or
> 
>     y = [int(n) for n in x]


From rajarshi at presidency.com  Wed Sep 18 11:40:51 2002
From: rajarshi at presidency.com (Rajarshi Guha)
Date: Wed, 18 Sep 2002 11:40:51 -0400
Subject: isFloat: Without Exception-Handling
References: <ama6nt$n6i$06$1@news.t-online.com>
Message-ID: <pan.2002.09.18.11.40.50.766802.29524@presidency.com>

On Wed, 18 Sep 2002 11:41:47 -0400, Thomas Guettler wrote:

> Hi!
> 
> Is there a way to write the following method without using exceptions?
> 
> def isFloat(string):
>      is_float=1
>      try:
>          float(string)
>      except:
>          is_float=0
>      return is_float
> 
> print isFloat("asdf") # --> 0
> print isFloat("0.1")  # --> 1


def isFloat(string):
  if type(string) == type(1.0):
    return 1
  else
    return 0

This should do what you want without exceptions


From aleax at aleax.it  Fri Sep 20 05:15:26 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 09:15:26 GMT
Subject: Destructor never called ???
References: <mailman.1032506132.22396.python-list@python.org>
Message-ID: <OGBi9.117782$ub2.2505532@news1.tin.it>

Tim Peters wrote:
        ...
> __del__s will be invoked.  It's only the cases where Python can't possibly
> guess a safe order in which to call destructors where both __del__ and
> collection are inhibited.

I think it's a bit more than that: Python *COULD" "possibly guess a
safe order" when there is only exactly ONE object with a __del__ that
is part of the cycle -- namely, destroy that one object first.  It
doesn't seem to even TRY (note: I'm not saying it SHOULD -- but
"can't possibly" might be a bit too strong as a description of what
is actually going on).


Alex



From abuse at nospam.riot.com.au  Wed Sep 25 23:58:48 2002
From: abuse at nospam.riot.com.au (Glen Murphy)
Date: Thu, 26 Sep 2002 13:58:48 +1000
Subject: object changing itself to another object
Message-ID: <amu0lv$6ac$1@otis.netspace.net.au>

I'm creating a program that takes user input, and passes it to an object -
however, I want that object itself, and not the parent to change where that
input goes. In the highly simplified example below, I would like the output
to be "Obj1","Obj2","Obj1" - but not suprisingly, I just get Obj1, Obj1,
Obj1


### code begin  ###

class Obj1:
    def key1(self):
        # do obj1 specific stuff
        self = Obj2()
        print "Obj1"

class Obj2:
    def key1(self):
        # do obj1 specific stuff
        self = Obj1()
        print "Obj2"

a = Obj1()

# simulate user keypresses
a.key1()
a.key1()
a.key1()

### code end ###

I know I could achieve the result I want by doing horribly complicated trees
of ifs and such, but I'd like my code to be nice and expandable (in this
example, to easily add new Objs). I've looked through the Python FAQ, Google
Groups and Various O'Reilly books, but I don't really know the terminology
for what I'm looking for, so I haven't found anything so far, so does anyone
have any pointers?

Any help would be greatly appreciated,
Cheers,
Glen

--
Glen Murphy, http://glenmurphy.com/




From roy at panix.com  Wed Sep  4 07:30:57 2002
From: roy at panix.com (Roy Smith)
Date: Wed, 04 Sep 2002 07:30:57 -0400
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com> <unbp765rbsn809@corp.supernews.com>
Message-ID: <roy-815DD1.07305604092002@reader1.panix.com>

"Benjamin" <phncontact.libero.it@> wrote
>> i talked to a friend of mine, he told me that i should try php..

Learning new stuff is never a bad thing, but I'm not sure php is where I 
would go next.  While I've never used it myself, my impression is that 
it's really just a niche tool for creating dynamic web pages rather than 
a real programming language.  There is a lot to programming beyond the 
web, and I think you'd do better learning some more general-purpose 
tools and languages.

"RPM1" <rpm1deletethis at nospamfrontiernet.net> wrote:
> I think you should try coding in C.  It's used everywhere for
> everything.  It's faster than about any other language.
> Most other languages are written in C.  C is highly marketable.
> 
> But best of all...
> 
> You'll learn to *really* appreciate other languages like Python,
> Java, ...

I agree that learning C is a good next step if your plan is to do 
programming for a living.  For better or worse, it is the lingua franca 
of the programming industry today, and has been for the past 15 or 20 
years.  And, yes, it will give you a taste of a very different 
programming environment from Python.

After that, my guess would be at least two of C++, Perl, and Java.  
Exactly what you want to do will determine which.  If you goal is to get 
into e-commerce and the web, Java is probably a must.
 
Python is an cool language, but the unfortunate truth is that there just 
aren't that many paying jobs for people who know Python and nothing 
else.  Hopefully that will change over the next 5 years, but it's not 
assured to happen, and in any case will be a slow process.  In the 
meantime, hedge your bets.


From syver-en+usenet at online.no  Fri Sep 13 09:45:34 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Fri, 13 Sep 2002 13:45:34 GMT
Subject: Calling Python from C# - please help
References: <6de16ac9.0209130508.5d1b56c6@posting.google.com> <j4elbyt3yw.fsf@informatik.hu-berlin.de>
Message-ID: <ur8fyxau0.fsf@online.no>

loewis at informatik.hu-berlin.de (Martin v. L?wis) writes:

> adila001 at yahoo.com (AdrianC) writes:
> 
> > I am using Python 2.1 and I am tring to use my Python COM component.
> 
> > From VB6 everything is OK, but in .NET i don't know how to set the
> > reference to my COM.
> 
> I've never used the COM bridge in .NET, but I believe you need to
> create a type library first.

Although you don't have to make a typelibrary, your C# code will be
nicer if you do so (else you will have to use the dispatch interface
to work with the object).

-- 

Vennlig hilsen 

Syver Enstad


From bfarwick at juno.com  Thu Sep 26 05:32:41 2002
From: bfarwick at juno.com (brrrent)
Date: 26 Sep 2002 02:32:41 -0700
Subject: cursor positioning and color
References: <slrnaoqv2q.hn4.TuxTrax@fortress.tuxnet> <1b6dca50.0209231022.6fae7f6d@posting.google.com> <30770aa4.0209232242.1e91d9d5@posting.google.com>
Message-ID: <1b6dca50.0209260132.6d1d333a@posting.google.com>

bogusdrop at myself.com (TuxTrax) wrote in message news:<30770aa4.0209232242.1e91d9d5 at posting.google.com>...
> bfarwick at juno.com (brrrent) wrote in message news:<1b6dca50.0209231022.6fae7f6d at posting.google.com>...
> > TuxTrax at fortress.tuxnet.net (TuxTrax) wrote in message news:<slrnaoqv2q.hn4.TuxTrax at fortress.tuxnet>...
> > > Being new to python, I would like to know how one goes about such 
> > > tasks as precise cursor positioning and output to screen in color.
> > > 
> > 
> > Hi Mathew,
> > 
> > Curses is the way to go when using Linux/Unix.
> > 
> > Wconio is the way to go when using Microsoft.
> > You can find in the Vaults of Parnassus.  If you have any trouble
> > locating it, let me know, I can email it to you.  Small library, works
> > well.
> > 
> > All the best,
> > 
> > brrrent
> 
> Thanks Brent. Appreciate it. Am I to understand that there are no
> python libraries that do this as a standard part of the python
> distribution?

Mathew,

A linux or unix distribution without curses is just about unheard of.
About the only chance of finding a linux installation without curses
would be in a standalone firewall box.

WConio is a "tiny" download and very worth it.  Works like a champ. 
One heck of a lot easier to grasp than curses.  Many thanks to Mr.
Gonnerman for such a useful library.  If you write a python WConio
app, why not just send WConio along with the python code.  As I said,
it's "tiny".

brent

 
> That would be preferable, as it would eliminate the requirement of
> making
> the user have curses on their system before they can use the code.
> 
> Cheers,
> 
> Mathew


From jb at cascade-sys.com  Fri Sep 20 20:15:46 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Fri, 20 Sep 2002 17:15:46 -0700
Subject: Getting started
References: <uojinmcfapjf45@corp.supernews.com> <mailman.1032442884.18747.python-list@python.org> <H3mi9.133635$pX1.4751676@news2.tin.it>
Message-ID: <3D8BBA32.1070700@cascade-sys.com>

Alex Martelli wrote:

>$a = "2.3";
>$b = 45;
>print $a+$b;
>
>emitting 47.3 is pretty close to what I consider "weak typing".  
>

I don't see why you call this "weak typing".  In Perl, the "+" operator 
is defined to work with strings or numbers.  It may seem more confusing 
at first if you expect "+" also to be the string concatenation operator, 
which it is not.  In the above example,

    print $a . $b;

will print the string "2.345".

This all is no different in principle from Python's:

    a = 2.3
    b = 45L
    print a + b

and getting 47.3.  The "+" operator is defined for floats and also for 
longs.  Since they're all "numbers" it would seem really strange if they 
did not interact naturally like this.  But fundamentally they're 
distinct types as "type( 2.3 )" and "type( 45L )" illustrate.

Perhaps an even better example is the "*" operator:

        print  2  * 5     # integer x integer -> integer
        print [2] * 5     # list x integer -> list
        print "2" * 5     # string x integer -> string

This all makes sense even though radically different results and result 
types are produced each time.  It's not "weakly typed" but rather 
strong, dynamic typing.  The outcome is completely determined by the 
operator and the types of the operands and there is no way to break the 
rules.  Since we know Python is strongly typed then it follows that your 
Perl example must be also.

Anyway, why should "2" + 5 == "7" be all that much stranger than "2" * 5 
== "22222"?  If the operation is unambiguous and useful what rule would 
it break?  Personally, I think Python would be improved slightly by 
adding this conversion.  It's convenient and there's ample precedent in 
other languages (Awk, Snobol, Icon, Perl, etc.).  But I am sure that 
notion will provoke howls of protest from the True Believers.

There are a lot of reasons why Python is better than Perl but strongly 
vs. weakly typed is not one of them.  I don't see a material difference 
in the "strength," per se, of typing between Perl and Python.  They're 
both "strong" and "dynamic", though Python's type model is better 
thought out, particularly since object/type unification.  

Getting back to the original question, the quintessential example of 
"weak typing" is in old C or Fortran where you can define a function to 
take, say, a string argument but are allowed instead pass in a real or a 
pointer to a struct and get garbled results or even corrupt the program:

    printf( "%s\n", 2.5 );        // possible program fault

Another example is all too familiar to MFC programmers:

    CEdit* edit = (CEdit*)GetDialogItem( ID_FOO );

GetDialogItem() essentially returns a raw pointer.  The programmer is 
required to explicitly cast the pointer into the proper type, the same 
type as the dialog item identified by the numeric ID.  However, if the 
dialog item itself changes, say, from an Edit box to a List box then the 
cast becomes wrong, some operations on edit may fail catestrophically, 
and -- worst of all -- the compiler cannot detect and diagnose this 
error for you.
  
These types of errors are not possible in Python or Perl, which 
illustrates that they're both "strong".  

Curiously, in Python

    print "%s" % 2.5

prints the string "2.5" instead of, say, raising an exception.  So 
there's some small precedent for treating numbers as strings in a string 
"context" even in Python.

I'll go out on a limb and say that most "dynamic typed" languages are 
also "strong" by necessity.  If the user isn't keeping track of the 
types then the system HAS to.  An exception I can think of would be an 
ancient dialect of Lisp where integers also were machine addresses and 
thus callable without restriction.  E.g., if a function name evaluated 
to a number the interpreter performed a subroutine call to that address 
instead of, say, applying a lambda.  It was the mechanism through which 
system builtins were accessed.  However,  nothing prevented a user from 
"calling" an arbitrary integer and leaping off into never never land.  

"Strong typing" (the "static" kind) is also the norm in Pascal and C++. 
 I like to think of strong typing as the norm in ANSI C but there is 
still optional and thus may be considered "weak".  Furthermore, C 
programmers are on the "honor" system across module boundaries, as 
there's no link time checking.  There even are ways to cheat in C++, so 
some might still class it as a "weakly typed" language.  Of course, with 
discipline, it's possible to avoid the pitfalls and always use the 
strong typing facilities of C and with effort enjoy most of the benefit.

>I
>find it very hard to produce substantial, large programs using
>weakly-typed languages. 
>
Given my definition of "weakly typed," I agree it's more difficult than 
in strongly typed languages.

On the other hand, isn't it the case that much of Python is implemented 
in C, which is weakly typed?  So "weakly typed" makes it harder but not 
impossible, at least not for some.

> Python is strongly, albeit dynamically, typed,   
>
FWIW, so is Perl.

>Unfortunately, some seminal material of a few
>decades ago confused terminology (I vaguely remember something
>about it in some early article about Scheme, for example).
>
I think you're right.  I originally learned a different taxonomy for 
many of these concepts.  

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From donn at drizzle.com  Fri Sep  6 23:02:37 2002
From: donn at drizzle.com (Donn Cave)
Date: Sat, 07 Sep 2002 03:02:37 -0000
Subject: Multiple Python Apps on Linux
References: None <gade9.25681$w51.7126863@news2.news.adelphia.net>
Message-ID: <1031367756.99553@yasure>

Quoth "Robert Oschler" <Oschler at earthlink.net>:
| If I run several single-threaded Python apps simultaneously, each launched
| as their own process running on Linux, are they truly independent or will
| Python's "global library lock flag" come into play like it does for a single
| Python app that is multi-threaded?

They're independent, even if forked from a Python parent process, on
any platform.  The lock applies to threads only, that is, threads that
share a common address space.

	Donn Cave, donn at drizzle.com


From sismex01 at hebmex.com  Tue Sep 24 09:49:27 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Tue, 24 Sep 2002 08:49:27 -0500
Subject: confused with lambda~
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233025AA489@mail.hebmex.com>

> that describe in reference documentation is too short,
> I couldnt even understand what it trying to say.
> any explanation please ?

Lambdas are complicated to understand, but it's easier
if you think of them as a single-expression inline
function (not inline in the code-generation sense
of the word, but of the definition sense).

See, "def" is a statement, and defines a word in the
local namespace to be a function, and "lambda" returns
a reference to a function. Kinda hard to understand :-)

So, the equivalence would be something like this:

Using def:
>>> def square(a): return a*a

Using lambda:
>>> square = lambda a:a*a

lambda's syntax is "lambda <params>:<expression>", quite
simple really, but what makes them hard is the concept
of an "unbound" function. Another example:

Using def:
>>> def SumOverProduct(a,b): return (a+b)/(a*b)

Using lambda:
>>> SumOverProduct = lambda a,b: (a+b)/(a*b)

You may use anything Python in the argument list, including
the *args and **kw forms. One thing you have to look out
for is the form where local variables are propagated to
the lambda's namespace (it's a function, remember?) using
default arguments:

def some_func(...):
	... code ...
	L = lambda a, b=SomeLocalVar, c=Another: operations(a, b, c)
	... more code ...

In previous versions of Python, there was no way to access
variables outside the lambda's namespace unless it was
by propagating them via default arguments, in more recent
versions it's possible to do away with this practice,
if you know what you're doing :-)

Well... what more can I say. It's a complex subject,
I really recommend you practice them if you want to use
them, until you get the hang of using them to make your
programs more clear.

Notice: USE them, don't ABUSE them; they either make a
program much more clearer, or they make them a program
hideously obtuse and obscure, and difficult to read
and understand (we all know who the culprits might be
so I won't say names). ;-)

Good luck, practice, play around with them, and then
store them somewhere safe where they won't do much
harm. :-)

-gus

Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From tundra at tundraware.com  Sun Sep  1 21:00:02 2002
From: tundra at tundraware.com (Tim Daneliuk)
Date: Mon, 02 Sep 2002 01:00:02 GMT
Subject: [ANN] 'tdir' version 1.67 Released
Message-ID: <02duka.pbc2.ln@boundary.tundraware.com>

'tdir' 1.67 is released and available at:

        http://www.tundraware.com/Software/tdir/

'tdir' is an advanced command-line directory and file display utility
written in Python, which runs on both *nix systems like FreeBSD and
Linux, as well as Win32.

'tdir' displays directories and files distinctly and separately.
By default, 'tdir' sorts files by their 'extension' and the
character which delimits those extensions is user-selectable.
'tdir' can also be used to display just the directory tree rooted
anywhere in the file system.

Output and column width are user-selectable as well and the
program writes to stdout, thereby allowing redirection to files
and devices.

Update to the FreeBSD port has also been submitted and should be available
shortly.

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com



From gerhard.haering at opus-gmbh.net  Wed Sep  4 12:01:25 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 4 Sep 2002 16:01:25 GMT
Subject: embeddable ide?
References: <Ltpd9.6669$yt3.3168986@newssrv26.news.prodigy.com>
Message-ID: <slrnancbfs.270.gerhard.haering@haering.opus-gmbh.net>

Stevek wrote:
> We're planning on making extensive use of Python to provide our users with a
> powerful scripting mechanism.  I've successfully embedded and extended
> Python but now I want to give the user a decent development environment for
> writing their scripts.
> 
> I know there are several IDEs available for Python development, but I'm
> looking for one that can be used from within another application (embedded).
> 
> Any suggestions would be appreciated.

If you're not using the same toolkit as the Python IDE, then I think this
will get very difficult, if not impossible.

If you're only targetting Windows, then I think PythonWin will be possible.

Anything that uses Scintilla will be a candidate. Perhaps it's wise to
outsource this task, like the producers of WingIDE?

-- Gerhard


From jetman516 at hotmail.com  Thu Sep 12 18:13:11 2002
From: jetman516 at hotmail.com (The Jetman)
Date: 12 Sep 2002 15:13:11 -0700
Subject: [newbie] Buying an introductory/reference text
References: <Mu5g9.1702$U_.1021@news04.bloor.is.net.cable.rogers.com>
Message-ID: <a38d43b0.0209121413.1bf274c9@posting.google.com>

Brad Fonseca <linuxbrad at rogers.com> wrote in message news:<Mu5g9.1702$U_.1021 at news04.bloor.is.net.cable.rogers.com>...
> Greetings!
> 
> I'm looking for a few opinions on the above subject.  I'm a computer 
> science student who has had some introductory lectures on Python and I've 
> done some very basic mucking around with.  I'm looking for a clear 
> reference and/or introductory text.  Any suggestions would be nice.  I have 
> looked at http://www.python.org/cgi-bin/moinmoin/IntroductoryBooks but 
> there appears to be a number of good books to choose from there.
> 
> Regards,

Can't go wrong w/ Programming Python 2ndEd by Lutz from O'Reilly.  Here's 
a sample chapter.  If you can only afford a single book, this mite be it....Jet

http://www.oreilly.com/catalog/python2/chapter/ch15.html


From max at alcyone.com  Fri Sep  6 16:09:03 2002
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 06 Sep 2002 13:09:03 -0700
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com>
Message-ID: <3D790B5F.BDD8810C@alcyone.com>

Rod Stephenson wrote:

> "Python is cool to look at small bits of, but I think the "outline"
> syntax breaks down with larger chunks of code. I'm with Aristotle on
> the structure of discourse--a story should have a beginning, and
> middle, and an end. So should blocks"
> 
> I'm not quite sure what he's trying to get at here - I guess that for
> a long heavily indented chunk of code, you could lose track of the
> overall structure, but I don't write code this way.
> 
> Any comments?

I'm not really sure what he's getting at, either.  It sounds like he's
suggesting that using indentation for setting off blocks is confusing,
but this is a standard objection to Python that most people have
(especially those with experience in other languages), and my experience
is most people immediately drop this objection after using it for a
little while.

But at some level, what would you expect him to say?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ There is nothing so subject to the inconstancy of fortune as war.
\__/ Miguel de Cervantes
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From johnroth at ameritech.net  Thu Sep 26 23:20:13 2002
From: johnroth at ameritech.net (John Roth)
Date: Thu, 26 Sep 2002 23:20:13 -0400
Subject: Tab stops. was: Re: Larry Wall's comment on python...
References: <mailman.1033049075.18932.python-list@python.org>
Message-ID: <up7jarpdreiq28@news.supernews.com>

<sismex01 at hebmex.com> wrote in message
news:mailman.1033049075.18932.python-list at python.org...
> >
> > sismex01 at hebmex.com wrote:
> >
> > >>Ironically, Guido originally ruled (in the style guide, IIRC)
> > >>that thou shalt use tabs and tabs are 8 spaces, "as god
> > >>intentended them to be."
> > >
> > >Yeesh! Why in God's name would he say such a dumb thing?
> > >I mean, even ancient mechanical typewriters have movable
> > >tab settings. Rigid, 8-char tabs is just so... neolithical.
> > >
> >
> > Once upon a time, a large class of computers, and their attendant
> > peripherals, OSs and utility programs universally implemented tabs
> > hard-wired at 8 stops.  In some circles this heritage extended
> > back 30 or 40 years.  Unix originally followed this convention,
> > as did IIRC all Dec computers and that of most other "minicomputer"
> > vendors.
> > Ed and the original VI editors only had fixed tab stops consistent
> > with terminal drivers and all the unit record equipment at the time.
> > IIRC, even Emacs originally only had tab stops at every 8 columns.
> > The original PC also followed this convention, though not all
> > applications did.  I don't know for sure but I bet $1 that Linux
> > tty drivers today also are hard wired at 8 char tab stops.
> >
>
> Thanks for a beautiful reply.
>
> Well, I might not be an old fart :-), but I *was* kinda present
> at that time, I started work in highschool hacking on CP/M machines,
> writing drivers for diverse printers and terminals for some apps
> (ahh, customizable terminal control... we lost that during the
> PC era, everybody wanted CGA, MDA or Hercules).
>
> AND, I distinctly remember that most printers could customize
> their tab stops via escape codes, same as some (not all) terminals;
> I worked a lot on some Xoroc terminals, and lots on a Northstar
> Advantage (ahh, beautiful hardware... *snif*) and they also had
> customizable tabstops.

Well, you got in rather late. The entire tab stop thing began
with mechanical and electromechanical equipment, like TTYs.
All of these had customizable tab stops, set by actual hardware
tabs (little pieces of metal). The thing about them is that typists
set them when they were doing reports, and otherwise ignored
them. Once they started getting added to the earliest computers,
it was simply not feasible to manually set up the output typewriter
for each and every program.

So some genius simply set the stops every 8 characters
and left them there. The drivers used the fixed stops in
order to optimize their output stream, and all the utilities
simply assumed that the output was set up this way.

Time and technology moved on, but by then the assumptions
were built into way too many programs to fix, and there was
never a concensus about how to fix it, anyway.

> Hmmm... maybe, what I'm getting at is, why did terminal-designers
> everywhere toss out years of research in typewriter technology
> (at the time) and ignore customizable tab-stops?  It's simply not
> sane, I think.

As I said, it was very sane. Nobody wanted to change
manual tab stops. Program setable tab stops came later.
Much later, and too late for there to be a coherent solution.

As an interesting side note, much the same issue came up
with the old vacuum tube computers (1950s vintage.) They
used existing plugboard tab equipement as card readers,
punches and printers. People very quickly standardized
on one set of plug boards, stuck them in the machine and
forgot about them. Much easier than to create a special set
of plug boards for each job, although I suspect some analysts
tried.

John Roth




From markus68 at bigfoot.de  Wed Sep 11 15:25:32 2002
From: markus68 at bigfoot.de (Markus)
Date: Wed, 11 Sep 2002 21:25:32 +0200
Subject: deleting the first and the last character of a string
References: <20020911105818.GC2054@partagas.as.de.cw.net> <mailman.1031756234.22942.python-list@python.org>
Message-ID: <alo5df$hoa$03$1@news.t-online.com>

Sean 'Shaleh' Perry wrote:

>> Is there anything which says delete the first or last
>> character of a string? I checked the string module on the web, but I
>> couldn't find anything.
>>
> 
>>>> s = '|f|o|o|'
>>>> s.strip('|')

>>> s[1:-1]
'f|o|o'

-Markus


From jknapka at earthlink.net  Fri Sep  6 17:23:56 2002
From: jknapka at earthlink.net (Joseph A. Knapka)
Date: Fri, 06 Sep 2002 21:23:56 GMT
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com>
Message-ID: <3D791CED.807E744B@earthlink.net>

Rod Stephenson wrote:
> 
> Slashdot has a list of questions posed to Larry Wall (perl). When
> asked his thoughts on other scripting languages, he makes the
> following observation about python
> 
> "Python is cool to look at small bits of, but I think the "outline"
> syntax breaks down with larger chunks of code. I'm with Aristotle on
> the structure of discourse--a story should have a beginning, and
> middle, and an end. So should blocks"
> 
> I'm not quite sure what he's trying to get at here - I guess that for
> a long heavily indented chunk of code, you could lose track of the
> overall structure,

...whereas with Perl, you can lose track of the overall structure
in a single line, brackets or no.

-- Joe
  "I'd rather chew my leg off than maintain Java code, which
   sucks, 'cause I have a lot of Java code to maintain and
   the leg surgery is starting to get expensive." - Me


From a-steinhoff at web.de  Sat Sep 21 06:48:06 2002
From: a-steinhoff at web.de (Armin Steinhoff)
Date: 21 Sep 2002 03:48:06 -0700
Subject: Python threading?
References: <f9Pi9.10275$Lg2.2191294@news2.news.adelphia.net> <mailman.1032571709.25544.python-list@python.org>
Message-ID: <ddc19db7.0209210248.3f1eded9@posting.google.com>

Trent Mick <trentm at ActiveState.com> wrote in message news:<mailman.1032571709.25544.python-list at python.org>...
> [Robert Oschler wrote]
> > How can I tell if the version of Python I am running is thread-enabled?
> 
> Run Lib/test/test_thread.py with your Python build. If you are on any
> common platform then Python, unless specifically crippled, is
> thread-enabled.
> 
> > Also, what is a reasonable number of threads to expect to be able to run
> > before context switching overhead becomes a problem (I'm using a PIII 500
> > Mhz with 512MB ram if that helps).
> 
> I have no experience putting that many threads to use.

Check out Stackless Python which supports micro threads (very small
context switching overhead ..)

Armin

> 
> 
> Trent


From toto2001 at hananet.net  Tue Sep 10 09:22:58 2002
From: toto2001 at hananet.net (konghyeogjun)
Date: Tue, 10 Sep 2002 22:22:58 +0900
Subject: default variable value and cmd module
Message-ID: <hmmf9.201$GH5.711@news.hananet.net>

This sample using 'cmd' module do not prints default variable value, 5.

import cmd
class x(cmd.Cmd):
    def do_y(self, value = 5):
        print value
z = x()
z.cmdloop()

but, the following sample prints '5'. why don't the sample above prints?

class x:
    def do_y(self, value = 5):
        print value
z = x()
z.do_y()














From aleax at aleax.it  Mon Sep 16 02:46:23 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 16 Sep 2002 06:46:23 GMT
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <mailman.1032152058.11043.python-list@python.org> <m3admizdzq.fsf@mira.informatik.hu-berlin.de>
Message-ID: <37fh9.112911$pX1.4044383@news2.tin.it>

Martin v. Loewis wrote:

> Gerhard H?ring <gerhard.haering at gmx.de> writes:
> 
>> You should perhaps reconsider if and why you need to hide your source
>> code.
> 
> You should also recognize that, if your computer can execute and
> perform a certain algorithm, users can always observe simulate the
> computer operations, thus breaking any protection scheme.
> 
> You'll have to give your customers a sealed computer and no permission
> to install additional software, or communicate with the internet, if
> you want to reliably prevent them from analysing your program.

Incidentally, the latter option can be feasible in certain situations: some 
key parts of your program's functionality might reside on a computer you
control, accessible to the rest of the application (installed more 
traditionally on users-controlled computers) only as a black box via the 
net.  XML-RPC, SOAP, and other distributed-computing approaches
such as Corba, make implementing this particularly easy these days.

Feasibility depends on how troublesome it is to demand that the
users' computers be net-connected whenever the application may
run, and how time-critical is the round-trip time whenever the
code that runs on the users' computers must delegate some part
of an operation to code running on the computer you control.  In
some application areas these feasibility issues may be killers, in
others they're no trouble at all.  When the latter conditions prevail,
this approach does yield something like a protection scheme that
you can, at least in theory, make "unbreakable" (by satisfying other
conditions, such as statefulness and other ways to make the part
running under your control arbitrarily hard to reverse-engineer).


Alex



From marklists at mceahern.com  Tue Sep 24 12:31:31 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 24 Sep 2002 11:31:31 -0500
Subject: What is the mechanism behide?
In-Reply-To: <e895ad50.0209240740.6b59f426@posting.google.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOEECAEBAA.marklists@mceahern.com>

> I'm studying the singleton pattern, there are something I don't
> understand in the sample code below:
[snip]
>     I'm puzzled that what is behide the statement "print x" ?

This is a more straightforward sample to cut your teeth on, imho:

>>> class foo:
...     def __str__(self):
...             print "calling __str__"
...             return "whatever we like"
...     def __repr__(self):
...             print "calling __repr__"
...             return "something different"
...
>>> f = foo()
>>> print f
 calling __str__
whatever we like
>>> print `f`
calling __repr__
something different
>>>

So __str__ simply allows you to override what happens when str() is called
on an instance of your class.  str() is what print calls, I believe.

__repr__ allows you to override what happens when repr() is called.  The
backticks (e.g., ``) are equivalent to repr().

You might want to lookup print, __str__, __repr__, str(), and repr() in the
docs.

I don't have the patience to decode the OnlyOne stuff.  Perhaps someone else
does.

Cheers,

// m




From mcfletch at rogers.com  Thu Sep  5 11:11:03 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Thu, 05 Sep 2002 11:11:03 -0400
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <20020905.072837.159259470.2038@cybermesa.com> <j4y9ag3377.fsf@informatik.hu-berlin.de>
Message-ID: <3D777407.4060203@rogers.com>

 >>Why not simply disallow the using of a name already defined in an
 >>enclosing scope? Similar to Smalltalk?
 >
 >
 > That might break compatibility with existing code.

I'll go one further:
	That _would_ break large quantities of existing code.

#consider
def blah( somefile ):
	string = somefile.read()
	return string[1:-1]

is now dependent on string not being imported at a higher level? 
Perfectly predictable code, obviously no relationship to "string" in a 
higher scope, but suddenly raising errors.

#consider
class window:
	pass
def whah( window ):
	window.whah( someglobal, somenumber, something )

Which is now raising errors because the temp variable/argument matches 
the class-name (yes, I know, normally there's a case-diff, but the 
pattern happens with types and the like).

If, for some reason this functionality (assigning to an enclosing scope 
which is not global (does this really occur much? I've never felt such a 
need that I can recall)), figure out a better way, such as:

	global(2) x

(Yes, ugly, but doesn't kill lots of code).  Or:

	global existing x

where the existing form says that an existing x record must be used.

But please sirs, don't kill our code :) ,
Mike

Martin v. L?wis wrote:
> "Jay O'Connor" <joconnor at cybermesa.com> writes:
> 
> 
>>Why not simply disallow the using of a name already defined in an
>>enclosing scope? Similar to Smalltalk?
> 
> 
> That might break compatibility with existing code.
> 
> Regards,
> Martin
_______________________________________
   Mike C. Fletcher
   Designer, VR Plumber, Coder
   http://members.rogers.com/mcfletch/





From jggramlich at yahoo.com  Thu Sep  5 14:14:42 2002
From: jggramlich at yahoo.com (Joshua Gramlich)
Date: 5 Sep 2002 11:14:42 -0700
Subject: PyListObject & C Modules
References: <mailman.1031085786.12546.python-list@python.org> <m3k7m2tf0y.fsf@mira.informatik.hu-berlin.de> <bdade4c9.0209040823.3624d164@posting.google.com> <j41y89emce.fsf@informatik.hu-berlin.de>
Message-ID: <bdade4c9.0209051014.5913ef9e@posting.google.com>

Okay...what I have now:

##############################code
block#################################

#include "Python.h"
#include <stdio.h>

static PyObject *ErrorObject;
#define onError(message) \
    { PyErr_SetString(ErrorObject, message); return NULL; }

PyObject*
get_first_elem(PyObject *unused, PyObject *args)
{
  PyObject *list;
  PyObject *first;
  PyObject *result;

  if(!PyArg_ParseTuple(args, "!O", &list))
    return NULL;

  if(!PyList_Check(list)) {
    PyErr_SetString(PyExc_TypeError, "get_first_elem expects a list");
    return NULL;
  }

  if(PyList_Size(list) == 0){
    PyErr_SetString(PyExc_ValueError, "empty lists not allowed");
    return NULL;
  }

  first = PyList_GetItem(list, 0);
  result = PyList_New(1);
  if (!result) return NULL;
  Py_INCREF(first);

  if (PyList_SetItem(result, 0, first) != 0) {
    Py_DECREF(first);
    Py_DECREF(result);
    result = NULL;
  }
  return result;
}

static struct PyMethodDef slist_methods[] = {
 {"getfstelm",     get_first_elem,     1},
 {NULL,                NULL}
};

void initslist()
{
    PyObject *m, *d;

    m = Py_InitModule("slist", slist_methods);

    d = PyModule_GetDict(m);
    ErrorObject = Py_BuildValue("s", "slist.error");
    PyDict_SetItemString(d, "error", ErrorObject);

    if (PyErr_Occurred())
        Py_FatalError("cannot initialize module slist");
}

##########################end code block############################

I have been reading the tutorial (amongst other things), and yes, I
should
probably have eventually caught the ADDREF bit.  The above compiles
fine and what-not, but when I run it via this:

############### hello.py #####################

stuff = 7,8,8,23,222,12
print(stuff)
pidge = slist.getfstelm(stuff)
print(pidge)

##############################################

I get this:

[d0ora3 ~/work/grid_query/test]$ python hello.py 

(7, 8, 8, 23, 222, 12)
Traceback (most recent call last):
  File "hello.py", line 25, in ?
    pidge = slist.getfstelm(stuff)
TypeError: argument 1 must be impossible<bad format char>, not tuple
[d0ora3 ~/work/grid_query/test]$ 



Now, if "stuff" is changed to "[7,8,8,23,222,12]" then the TypeError
ends with "not list" instead of "not tuple" and so on for ints and
strings.

Thoughts?


From loewis at informatik.hu-berlin.de  Wed Sep  4 12:44:17 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 04 Sep 2002 18:44:17 +0200
Subject: PyListObject & C Modules
References: <mailman.1031085786.12546.python-list@python.org> <m3k7m2tf0y.fsf@mira.informatik.hu-berlin.de> <bdade4c9.0209040823.3624d164@posting.google.com>
Message-ID: <j41y89emce.fsf@informatik.hu-berlin.de>

jggramlich at yahoo.com (Joshua Gramlich) writes:

> get_first_elem(PyObject *unused, PyObject *args)
> {
>   if(!PyArg_ParseTuple("!0", &list))
>     return NULL;
> slist.c:10: warning: passing arg 1 of `PyArg_ParseTuple' from
> incompatible pointer type

Right. This should be
   if(!PyArg_ParseTuple(args,"!0", &list))
     return NULL;

Reading the Extending and Embedding tutorial would have helped you to
fix this error on your own.

>   Py_ADDREF(first);
> ./slist.so: symbol Py_ADDREF: referenced symbol not found

Right. It is INCREF, not ADDREF. Reading the tutorial really helps.

You'll also learn that you lead additional boilerplate code to make an
extension module.

Regards,
Martin


From paoloinvernizzi at dmsware.com  Thu Sep  5 03:59:26 2002
From: paoloinvernizzi at dmsware.com (Paolo Invernizzi)
Date: Thu, 05 Sep 2002 07:59:26 GMT
Subject: Servizi Internet
References: <mailman.1031056925.7383.python-list@python.org> <f9fbnu8dghga0tbpvf7nsv64jo3csrlsm1@4ax.com>
Message-ID: <y9Ed9.2904$%z1.72415@twister1.libero.it>

Benjamin wrote:
> by the way, how many people understand italian in
> this nesgroup? ^_^

** a little hand is raised! **



From see_reply_address at something.invalid  Wed Sep 18 19:43:44 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Thu, 19 Sep 2002 11:43:44 +1200
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <mailman.1032329366.6813.python-list@python.org> <j4ptvbaapm.fsf@informatik.hu-berlin.de>
Message-ID: <3D890FB0.4030408@something.invalid>

Martin v. L?wis wrote:

> Depends on your notion of "danger"; I usually associate it with fear
> and uncertainty :-)


Well, I *do* experience fear and uncertainty when
I find that the normally secure and predictable
environment of Python suddenly starts crashing
on me. It takes all the fun out of things.


> If you know it can crash when you do certain things, and you document
> that limitation, there is no danger. There are already ways to crash
> Python, the dlmodule being an example in the core distribution.


The problem in this case is that it's not something
*I* do which is going to cause the crash, it's
something done by the internals of some piece of
C code that I know nothing about. Something which
is not going to be documented, because it's a
perfectly legitimate and unremarkable thing for
C code to do.

So, the only way to find out whether a particular
extension is safe to use in Stackless is to try
it. And that only gives a definite answer when
the answer is "no, it's not safe". If it doesn't
crash, I still can't be confident that it won't
crash at some time in the future.

That's why I say that I would be nervous about
using Stackless. Not to say that would stop me
from using it, but I'd feel nervous while I was
doing it! :-)


> Also, readers should notice that your example requires Stackless to be
> used:


Yes, I understand that, and I acknowledge that it
won't cause any harm to have Stackless in the core
if the Stackless stuff is switched off by default.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From 2002 at weholt.org  Wed Sep 25 09:47:49 2002
From: 2002 at weholt.org (Thomas Weholt)
Date: Wed, 25 Sep 2002 13:47:49 GMT
Subject: What function called this function ??
Message-ID: <98jk9.22340$0p1.391334@news2.ulv.nextra.no>

Is it possible, by hacking the traceback/stack/etc, to get information about
the function, that functions class and in what module that function and
class are defined, from inside a function located elsewhere? Something like
a way of answering the question of  "who called me?" without sending the
answer along as a parameter.

Best regards,
Thomas




From notvalid at hotmail.com  Tue Sep 24 13:45:03 2002
From: notvalid at hotmail.com (jano)
Date: Tue, 24 Sep 2002 10:45:03 -0700
Subject: cgi post problem
References: <FVRj9.192186$Jo.71059@rwcrnsc53> <9aWj9.39314$V7.10439005@twister.socal.rr.com>
Message-ID: <amq8av$2cum$1@agate.berkeley.edu>

"Jeff Davis" <jdavis at empires.org> wrote in message
news:9aWj9.39314$V7.10439005 at twister.socal.rr.com...
> Hi,
>
> It looks as though you have a lot of code. It might be helpful to try to
> narrow the problem further by removing a lot of the code unrelated to your
> problem, which appears to be that "keys" does not contain any items. After
> you work that out, then put the other code back in.
>
> My first impression is that you have some "weird" form element names
> because they contain the "/" character. That might be perfectly legal, but
> perhaps the cgi.FieldStorage class does not properly handle such element
> names, or perhaps it considers such names erroneous.
>
> Next, it seems that the line "keys = form.keys()" might be out of
alignment
> with the rest of the code in it's block. Perhaps it's just the way it
> looks after you copied your code into the post.
>
> Regards,
>         Jeff
>
>



Yes, the problem is that keys does not contain any items.  The 'weird' names
are not the problem, because i tried it also with regular 1-word names (no
slashes), with the same results.  Also, the keys=form.keys() line is
properly indented in the code -- it was just a copy/paste problem.  Here is
the code again, with the extraneous stuff stripped out:

#!/usr/local/bin/python

import cgitb; cgitb.enable()
import sys, traceback, cgi

def main():
    try:
        form = cgi.FieldStorage()
        input = form.keys()

        print "Content-type: text/html\n\n"
        print("""<?xml version='1.0' encoding='UTF-8'?>
        <!DOCTYPE html PUBLIC '-//W3C//DTD XHTML 1.1//EN'
'http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd'>
        <html xmlns='http://www.w3.org/1999/xhtml' xml:lang='en'>
            <head>
                <title>test</title>
            </head>
            <body>""")
        print "<p>",keys,"</p>"
        print "</body></html>"
   except:
        pass

if __name__ == "__main__":
     main()

And the form is:

<form action="http://someurl.com/cgi-bin/updatexml.py" method="post">
     <input type="text" name="Name" size="20" />
     <input type="text" name="PartNumber" size="20" />
     <input type="text" name="Inventory" size="20" />
     <input type="text" name="Cost" size="20" />
     <input type="text" name="WholesalePrice" size="20" />
    <input type="submit" value="Update Records" />
</form>


The script, again, is currently just supposed to return the keys, which I am
expecting to be Name, PartNumber, etc., but it returns only an empty list.
Any ideas.

thanks,

jano





From rond at ixsys.com  Thu Sep  5 11:30:11 2002
From: rond at ixsys.com (Ron Darling)
Date: Thu, 5 Sep 2002 10:30:11 -0500
Subject: What database should I use
In-Reply-To: <CC2C254F4F467345A0A5754CA6E25F0B2B9426@PRTRRA06-XS03.telkom.co.za>
Message-ID: <BAEILLEBFGFMOHOGFDACKEIDCJAA.rond@ixsys.com>

>> (though it's it considered obsolete now-a-days).

This is only a consideration for the uni-formed or those that have never
really used it.  MS is still dumping lots of money into it.  Just shipped
v7.0 will be going into Beta with 8.0 in a couple of months.

Amazing String capabilties.  As a matter of fact there is now a VFP style
.net class that moves a lot of the wonderful commands internal to VFP to
.NET.  I think the Fox Dev team created.

Cursor Technology, Rushmore Speed and many other things have moved there way
into not only Access but SQL Server and ADO.

Due to the fact that each part of a table is its own file.  .DBF is the
table.  .FPT is the Memo field  .CDX is the index.  The file size limitation
of the OS of 2 gig is at the individual file level.  In Access everything is
stored in the same file (you could break up but not really meant to) so the
2 gig file size is limited to the app and data which is more than likely in
the same file.  I have dealt with lots of situations where the .DBF was 1.8
gig, the index was 1.2 gig and the FPT was another 1+ gig size.

Indexes are stored in a .CDX file.  A .CDX file can have as many as 255
indexes (take into account 2 gig file size).  Each open table can have 10
open CDX's so you can have 2550 indexes on a table.

Very Robust SQL pass Through for back end databases.

There is nothing out there as fast as VFP for pure raw speed.  SQL 2000 gets
close but not quite there, the advantage to SQL as far speed goes is it does
not bog down under load.  Load means 25 + heavy users.

As far as speed goes I have seen the return of queries on tables as large as
21 million records within seconds and that was pulling data across a 10 meg
network with Foxpro For DOS.  United We Stand (remember Ross Perot)
member/prospect database.  Sounds a little dated but the point is its been
that way in Fox for a long time.

Fox and VFP managed/tracked the military assets to the Persian Gulf and is
till in use.
Manages the flow of info for Chunnel.
It is far more prevalent than most realize.

Very good OOP.  It seems that VFP is becoming known for the best language
that can really implement the "Abstract Factory Pattern"

Very good interface. Rich Native Controls (No OCX) and does very well with
OCX's.

Easy Distribution.  (Drop a couple of DLLS into system directory not reg)

Rich Debugger.

Can access any data ODBC, ADO or ADO.Net can access.

Internal XML.  Select Result set from SQL into a local cursor then use
CURSORTOXML().  With the XMLTOCCURSOR() and
CURSORTOXML() commands it is very easy to build thin wire apps.  Bind the
interface to the cursor.

I can try any and all commands from the command prompt and even from within
a debugging suspend.  Pretty much interrogate the entire system at hand very
easily from a suspend.

Many options for hanging VFP behind IIS and ISAPI.

You can create com objects and run behind other languages even from within
SQL Server or visa versa.

One of the most robust SOUNDEX/Fuzzy Search add-ons available with the
PHDbase Library.  You can index memo fields and do 2 - 3 second searches on
huge amounts of memo data.

Best Table Repair Libraries that I am aware of are available.

Error trapping has been far superior to VB and a lot of languages since the
Fox 2.0 days (1992).

Last but not least http://www.universalthread.com.  Great support from all
over the world.


obsolete...hmmmmm maybe lack of exposure or conflicting marketing problems
with MS owning Access, SQL and Fox but certainly not obsolete.

-----Original Message-----
From: activepython-admin at listserv.ActiveState.com
[mailto:activepython-admin at listserv.ActiveState.com]On Behalf Of Sathish
Chandrasekaran (S)
Sent: Thursday, September 05, 2002 9:24 AM
To: printers at sendme.cz; python-list at python.org; tutor at python.org;
activepython at listserv.activestate.com; python-help at python.org
Cc: db-sig at python.org; wx-users at lists.wxwindows.org
Subject: RE: What database should I use


In that case MS-access will be better on the size issue. But on the
other hand if u want ur application to be flexible and ease of use I
would rather suggest FoxPro (though it's it considered obsolete
now-a-days).


Here Is A Test To Know Whether The Mission Of Your Life Is Over .
If You Are Alive It Isn't .

Sathish C
Infra-Amdocs


-----Original Message-----
From: A [mailto:printers at sendme.cz]
Sent: 05 September 2002 15:21
To: python-list at python.org; tutor at python.org;
activepython at listserv.ActiveState.com; python-help at python.org
Cc: db-sig at python.org; wx-users at lists.wxwindows.org
Subject: What database should I use


Hello,
I am going to program a small application (in Python and wxPython)
which consists about
2000 - 3000 records . Can you please recommend which database is the
best for that. I
use Windows.
Maybe MS Access  or FoxPro or MySQL or better?
I want to keep the program as small as possible.
Thanks
Ladislav

_______________________________________________
ActivePython mailing list
ActivePython at listserv.ActiveState.com
To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
Other options:
http://listserv.ActiveState.com/mailman/listinfo/ActivePython

_______________________________________________
ActivePython mailing list
ActivePython at listserv.ActiveState.com
To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
Other options: http://listserv.ActiveState.com/mailman/listinfo/ActivePython





From tdelaney at avaya.com  Sun Sep 29 20:31:39 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 30 Sep 2002 10:31:39 +1000
Subject: comparing all values of a list to regex
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5F3@natasha.auslabs.avaya.com>

> From: Alex Martelli [mailto:aleax at aleax.it]
> 
> Delaney, Timothy wrote:
>         ...
> >> > elif matches != len(alist):
> >> >     match_list = match_some
> >> > 
> >> > match_list.append(alist)
> >> 
> >> Yes, choosing a list first and then appending to the chosen
> >> list is nicer, but you cannot do that within the original
> >> specs (for all I know the match_some list might grow
> >> unbearably large over time, in this case).
> > 
> > Hmm - why not? Ignoring that the original specs didn't 
> actually *have*
> > that case, 
> 
> Not ignoring it, one would think the specs must be to do nothing
> special in that case.
> 
> If your program was specified to "append to list1 if the time is
> exactly 11:22:33, to list2 if the time is exactly 17:26:35", would
> you consider it reasonable to make your program appent to list3
> if it's any other time of day?

Well, of course. I thought you were talking about the implementation (since
you yourself mentioned the unspecified case) - not the conformance to the
specs.

Not having that case in the specs is not a cause for simply ignoring it.
What you should instead to is get an explicit statement of what is to occur
in that case into the specs (e.g. if more than one match is found, but not
everything, the line should be ignored).

It was an *incomplete* specification, which would require querying by the
(usually) system engineer in order to produce complete specifications. If
such specs made it to a developer I should bloody hope that they would say
"and what should I do in this case?".

Tim Delaney



From itamar at itamarst.org  Mon Sep 30 18:06:33 2002
From: itamar at itamarst.org (Itamar Shtull-Trauring)
Date: Mon, 30 Sep 2002 18:06:33 -0400
Subject: ANN: Bannerfish 0.1.1 - banner ad server
Message-ID: <20020930180633.2cd106cc.itamar@itamarst.org>

Apologies if this appears twice, my ISP's NNTP server doesn't seem to like me.

 Bannerfish banner ad server
=============================

This is the first public announcement of Bannerfish, v.0.1.1, a banner ad
server for small to medium sized websites.

The latest version of Bannerfish should always be available from:

  http://itamarst.org/software/bannerfish/


 Features
----------
- Supports images and Flash
- Embeds ads using javascript, so it should work with any web platform
- Weighting and multiple ads per page
- Lets customers view banner stats (clicks and views)
- Automatic deactivation of ads

The goal of Bannerfish is to be a small, simple banner ad server
sufficient for the needs of small websites.


 Requirements
--------------
Twisted 0.99.2 or later
Python 2.2

Bannerfish has been tested under Debian GNU/Linux, but should work
fine under other Unix-like systems and on Windows as well.


 Author and Licensing
----------------------
Bannerfish was written by Itamar Shtull-Trauring (itamar at itamarst-DOT-org).

Bannerfish is licensed under the GPL, but licensing under other terms
can be arranged - send me an email.


-- 
Itamar Shtull-Trauring    http://itamarst.org/
Available for Python, Twisted, Zope and Java consulting



From llothar at web.de  Wed Sep 18 09:16:54 2002
From: llothar at web.de (Lothar Scholz)
Date: Wed, 18 Sep 2002 15:16:54 +0200
Subject: Will "Python in a Nutshell" be too short?
References: <mailman.1032288261.15607.python-list@python.org>
Message-ID: <fc7fougmsppuleug0br2napq7k4i219bvs@4ax.com>

On Tue, 17 Sep 2002 19:46:21 +0100, Hamish Lawson
<hbl at st-andrews.ac.uk> wrote:

>Many of us were hoping that "Python in a Nutshell" would not only be a 
>comprehensive reference of pretty much all of the modules in the standard 

I never understood why people wants another reference to Python other
then the standard online documentation. We should concentrate on this
rather then on everything else. But currently they seem to be written
in stone. 

The annotated PHP manual is a complete other dimension of
documentation with active development. You don't have to write a PEP
before changing a sentence.

>library that aren't obsolescent or platform-specific, but would also cover 
>some of the most popular third-party libraries - e.g. DB-API, PIL, 
>mxDateTime, Numeric, win32all, wxPython, mod_python, ReportLab.

Here comes another problem. You can't write a simple 40 page
introduction. What we really need are not books for beginners but
books for advanced programmers. There is too much garbage in the
bookshelfs (even in the python bookshelf) of my favorite shop.

Perhaps i'm to frustrated but is education really so low in the python
community that you must explain simple method calls or thinks like
this ? I would like to see a detailed description about the actions
that must be done on every method call by the interpreter with
references to the c source code.





From see_reply_address at something.invalid  Mon Sep  2 00:38:58 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Mon, 02 Sep 2002 16:38:58 +1200
Subject: ANN: Pyrex 0.4.4
Message-ID: <mailman.1031050990.31309.clpa-moderators@python.org>

Pyrex 0.4.4 is now available:

    http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/

New Features:

  * Tracebacks: Pyrex modules now produce tracebacks that
    show you where the error occurred in the Pyrex code.

  * ",".join(x) and the like now work.

  * Float literals don't require digits on both sides of
    the point.

There are also some other minor improvements and bug
fixes. See the CHANGES file in the distribution for a
full list.


What is Pyrex?
--------------

Pyrex is a new language for writing Python extension modules.
It lets you freely mix operations on Python and C data, with
all Python reference counting and error checking handled
automatically.


-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg




From leif at crysberg.dk  Fri Sep  6 10:29:54 2002
From: leif at crysberg.dk (Leif Jensen)
Date: Fri, 6 Sep 2002 16:29:54 +0200 (CEST)
Subject: What database should I use
In-Reply-To: <3D77763F.23080.BFAFA3@localhost>
Message-ID: <Pine.LNX.4.21.0209061629050.19270-100000@samba.crysberg.dk>

    PostgreSQL, no doubt, if you use Linux, WinNT or possible WinXP !

 Leif



On Thu, 5 Sep 2002, A wrote:

> Hello,
> I am going to program a small application (in Python and wxPython) which consists about 
> 2000 - 3000 records . Can you please recommend which database is the best for that. I 
> use Windows.
> Maybe MS Access  or FoxPro or MySQL or better?
> I want to keep the program as small as possible.
> Thanks
> Ladislav
> 
> 
> _______________________________________________
> wx-users mailing list
> wx-users at lists.wxwindows.org
> http://lists.wxwindows.org/mailman/listinfo/wx-users
> 




From peter at engcorp.com  Tue Sep 24 21:19:36 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 24 Sep 2002 21:19:36 -0400
Subject: Non-interpreted Python?
References: <3D8F65A4.38124945@engcorp.com> <y7Kj9.400$UC2.36205@read2.inet.fi>
Message-ID: <3d910f20$1@news.sentex.net>

Tapio Kelloniemi wrote:
> Peter Hansen <peter at engcorp.com> wrote:
> 
>>There are other approaches, but there's no point describing them until
>>you explain *why* you want to do this.  What are your requirements?
> 
> I'm using Linux and searching for a good high-level language to use in
> my projects. I don't have any special ideas, it depends on language. I
> don't like the idea of using any interpreted language, but have heard so
> much good about Python. I have used mostly C before and so it would be
> best to give up my speed requirements.

Python is definitely a "good high-level language" for a vast range of
projects.  We use it in my company for many things including internal 
utilities, a factory automated testing system for wireless telecom 
products, our Intranet (via Zope), and as the primary language (98%
of the code) in several of our main products.

I think you would be well advised to let go your fear of using an
"interpreted" language, since unless you are writing device drivers
for a living you are unlikely to experience performance bottlenecks
in most situations you'll encounter.  On the other hand, there are
*significant* advantages which you will NOT find using non-interpreted
languages, and these in many cases outweigh the few disadvantages.

A common expression around here is that developer time is worth far
more than CPU time, and in this respect Python has shown itself
capable of significantly increasing productivity in many instances.

>>>Or better, if it is possible to compile code into object format and then
>>>link with C/C++/whatever code.
>>
>>Python has excellent facilities for embedding in applications written
>>in C/C++.  Check the documentation for the relevant sections.
> 
> I have thought of writing some parts of programs in Python (some kind of
> library) and then be able to use it in C++/Ruby/anything else, if it is
> possible.

Excellent idea.  Why not take the approach many do, of writing the
whole application in Python and then (once it's working) rewriting
only those parts which profiling shows have performance problems?

-Peter



From see_reply_address at something.invalid  Tue Sep 24 22:30:57 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Wed, 25 Sep 2002 14:30:57 +1200
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <%8ci9.138477$5r1.5541622@bin5.nnrp.aus1.giganews.com> <amltd3$3k$1@glue.ucr.edu>
Message-ID: <3D911FE1.1050801@something.invalid>

thp at cs.ucr.edu wrote:

> But why use grotesque hacks in the first place?  "a ? b : c" is a
> well-established syntax for conditional expressions.  Why doesn't
> Python simply adopt it?


I believe because Guido (a) thinks it's ugly and (b) doesn't
think it would add enough to the language to be worth
including.

Personally, I rather like

    a if b else c

but Guido doesn't like that either, unfortunately.

Perhaps I'll slip it into Pyrex when no-one's looking...:-)

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From jepler at unpythonic.net  Tue Sep 10 16:53:58 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 10 Sep 2002 15:53:58 -0500
Subject: run-mailcap in python?
In-Reply-To: <3D7E016D.2646AD9F@engcorp.com>
References: <alkcfb$9ql$07$1@news.t-online.com> <alkco4$j0k$06$1@news.t-online.com> <3d7de60b$1@news.sentex.net> <alksap$ekd$02$1@news.t-online.com> <3D7E016D.2646AD9F@engcorp.com>
Message-ID: <20020910205355.GD15936@unpythonic.net>

On Tue, Sep 10, 2002 at 10:27:57AM -0400, Peter Hansen wrote:
> Thomas Guettler wrote:
> > 
> > Peter Hansen schrieb:
> > > import  os
> > > if os.name == 'posix':
> > >     def startfile(name):
> > >         os.system('run-mailcap "%s"' % name)
> > >     os.startfile = startfile
> > 
> > Yes, ugly but it works. Thank you!
> 
> In Pythonia, "ugly" is sometimes spelled "pragmatic".  <grin>

Of course, what if it's arranged for name to equal
    '"; rm -rf / & echo you're a loser"'
?  If an attacker controls the string (eg by supplying it from e-mail or
a web page), you're sunk.

this should be safe, at least as long as run-mailcap doesn't commit the
same kind of error:
    os.spawnvp(os.P_WAIT, "run-mailcap", ["run-mailcap", name])

Jeff



From pinard at iro.umontreal.ca  Thu Sep 26 11:36:58 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Thu, 26 Sep 2002 11:36:58 -0400
Subject: RELEASE: Recodec 0.0
Message-ID: <oqu1kc92zp.fsf@carouge.sram.qc.ca>

Hello to all `recode' pretesters and Python friends.

Here is the initial throw for the Recodec program and library.  Recodec
is the start of a Python prototype for Free Recode.  See `Recodec and
Free Recode' in the `README' file to know how they relate to one another.

The Recodec program and its library, as Free Recode already do,
convert files between character sets and usages.  They recognise or
produce more than 240 different character sets and transliterate files
between almost any pair.  When exact transliteration are not possible,
they get rid of offending characters or fall back on approximations.

The `README' is repeated as `http://www.iro.umontreal.ca/~pinard/recodec/',
and `http://www.iro.umontreal.ca/~pinard/recodec/Recodec.tar.gz' holds
the latest release.  This is more a snapshot than a finished release,
actually.  This is still alpha quality software, which should not be used
without caution in a production environment.  Pretesters are welcome.

Please gently report problems, suggestions or other comments to
`mailto:pinard at iro.umontreal.ca'.  Recode-related discussions might
be held on `mailto:recode-forum at iro.umontreal.ca', this list is
opened to its members only.  Python specific discussions might go to
`mailto:python-list at python.org'.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From gerhard.haering at opus-gmbh.net  Wed Sep  4 12:09:33 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 4 Sep 2002 16:09:33 GMT
Subject: Please help in  calling python fucntion from 'c'
References: <mailman.1031142606.11252.python-list@python.org> <zOpd9.291180$Aw4.12264386@bin2.nnrp.aus1.giganews.com>
Message-ID: <slrnancbv4.270.gerhard.haering@haering.opus-gmbh.net>

Terry Reedy wrote:
>> [ The information contained in this e-mail is confidential and is
> intended for the named recipient only. If you are not the named
> recipient, please notify us by telephone on +44 (0)1249 442 430
> immediately, destroy the message and delete it from your computer.
> 
> I will gladly delete it, but expensive call? no.  Pretty whacko.

It obviously makes certain people feel safer. But if it really has any
legally binding effect in any country, I'd like to stay away from that
country.

Btw. your header "X-DMCA-Complaints-To" makes me hope that the tries to get
similar laws passed in the EU will fail.

-- Gerhard


From anton at vredegoor.doge.nl  Fri Sep 27 12:26:07 2002
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Fri, 27 Sep 2002 18:26:07 +0200
Subject: watching mutables?
References: <mailman.1033137396.14482.python-list@python.org>
Message-ID: <an211j$oi$1@news.hccnet.nl>

On Fri, 27 Sep 2002 09:34:55 -0500, "Mark McEahern"
<marklists at mceahern.com> wrote:

>l = Watched()
>l.append(1)

Thanks for reacting to my question. However, I am looking for
something that calls a defineable function if *any* mutable variable
in the current scope changes. Variables should be 'marked' to install
this function for them. For example 'watch(m,f)' indicates that
function f is called if mutable variable m is changed. I do not want
function f to be a method of object m since I want object m to be
*any* mutable python object and furthermore I do not want to change
the orginal object. The call of function f would be triggered by *any*
change of object m. This would make debugging (and some other things I
am tinkering with) a lot easier. Is it possible?

Regards,

Anton.


From walterm at parque.homelinux.net  Mon Sep  9 23:38:26 2002
From: walterm at parque.homelinux.net (Walter Moreira)
Date: Tue, 10 Sep 2002 00:38:26 -0300
Subject: RPy / Win2k
In-Reply-To: <aljd9u$k9n$1@newsg4.svr.pol.co.uk>
References: <albgmp$cgv$1@newsg2.svr.pol.co.uk> <mailman.1031613673.1670.python-list@python.org> <aljd9u$k9n$1@newsg4.svr.pol.co.uk>
Message-ID: <20020910033826.GB21801@parque.homelinux.net>

On Tue, Sep 10, 2002 at 01:09:18AM +0100, Duncan Smith wrote:
> ... [About a Windows release of RPy] ...
> 
> Thanks Walter, that's good news.  When I spoke to Brian he said that R could
> not be built as a shared library under Windows.  Does this mean I won't need
> to reinstall R?  Cheers.

Yes, you don't need to compile neither R nor Python. The binary
distributions already provide the dll's. I'm trying to make the
installer to find automatically the location of the R library.

Regards:
Walter



From jdavis at empires.org  Thu Sep 19 02:36:34 2002
From: jdavis at empires.org (Jeff Davis)
Date: Thu, 19 Sep 2002 06:36:34 GMT
Subject: [Help]: mailbox classes
References: <jylimd-DAD12F.17305418092002@skates.gsfc.nasa.gov>
Message-ID: <Sfei9.18259$V7.4642808@twister.socal.rr.com>

Ok, I'll assume your using mbox format, but this is easily adapted for 
maildir, etc.

import mailbox
mbox_file = '/path/to/mbox'

mbox_fp = open(mbox_file)
mbox = mailbox.UnixMailbox(mbox_fp)
msg = mbox.next()

there, now you have a message object (a rfc822.Message object, to be 
exact). Now just import the rfc822 module and do whatever you need. You'll 
probably want to use a loop to process all of the messages.

Regards,
        Jeff
        

> Hi,
> 
> I am trying to extract a message from a unix mailbox file. Python does
> provide few classes for easy and uniform access. So it says! After
> having read the libiary references, I went nowhere.
> 
> Can anyone give me an example as how to use these mailbox classes and
> what is it relationship with email.message classes.
> 
> thanks



From Oschler at earthlink.net  Fri Sep  6 16:14:08 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Fri, 06 Sep 2002 20:14:08 GMT
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <3D7958DB.50903@nospam.free.fr>
Message-ID: <k08e9.25533$w51.7064239@news2.news.adelphia.net>

"laotseu" <bdesth at nospam.free.fr> wrote in message
news:3D7958DB.50903 at nospam.free.fr...
> > Any comments?
>
> def fun(args):
> #BEGIN
>      if ceci or cela:
>      #BEGIN
>          do_this(with_that)
>          [snip a lot of code]
>      #END
> #END
>
>
laotseu,

Although new to Python, I've read many posts complaining about the indenting
issue,  usually from "other language" proponents.  But for the longest time
I wondered why not do exactly what you have done with comment lines?  In
essence I'm saying, I'm glad you made this point. :)

In Pascal, C/C++ and VB (no rock-throwing please, it was a paycheck) I
frequently did things like the following, to use a C example:

if (something == something_else)
{
} // if (something == something_else)

As you can see, using comments to label my end blocks explicitly.  For me
this issue was never an issue.

thx




From chris.walsh at real-sense.com  Wed Sep 11 19:58:21 2002
From: chris.walsh at real-sense.com (Chris Walsh)
Date: 11 Sep 2002 16:58:21 -0700
Subject: Getting __builtins__ into my local dictionary
References: <1691b372.0209110828.31501d82@posting.google.com>
Message-ID: <1691b372.0209111558.33c2990@posting.google.com>

chris.walsh at real-sense.com (Chris Walsh) wrote in message news:<1691b372.0209110828.31501d82 at posting.google.com>...
> Hi All,
> 
> I know this may sound like I am repeating a much discussed topic but I
> want to create a multiple environments within the one Python
> interpreter so that each envronment contains it's own local defs and
> attributes but also understands functions such as str() (e.g. contains
> __builtin__ module).
> 
> This is how I think I should do it:
> 
> ================START=OF=CODE================
> 	PyObject *pMyMod  = PyModule_New("unique000");
> 	PyObject *pMyDict = PyModule_GetDict(pMyMod);
> 
> 	PyDict_SetItemString(pMyDict, "__builtins__", PyEval_GetBuiltins());
> 
> 	PyObject *pResultT;
> 	pResultT = PyRun_String("dir()\n",
> 		Py_eval_input, pMyDict, pMyDict);
> 	DumpAndRelease("dir()", pResultT);
> 
> 	pResultT = PyRun_String("dir(__builtins__)\n",
> 		Py_eval_input, pMyDict, pMyDict);
> 	DumpAndRelease("dir(__builtins__)", pResultT);
> =================END==OF=CODE================
> 
> The results I get from my Dump module (which just throws out the
> contents of pResultT) is:
> 
> ================START=OF=DUMP================
> dir()
> 	List of 3 items: ['__builtins__', '__doc__', '__name__']
> 
> dir(__builtins__)
> 	List of 37 items: ['__class__', '__cmp__', '__contains__',
> '__delattr__', '__delitem__', '__doc__', '__eq__', '__ge__',
> '__getattribute__', '__getitem__', '__gt__', '__hash__', '__init__',
> '__iter__', '__le__', '__len__', '__lt__', '__ne__', '__new__',
> '__reduce__', '__repr__', '__setattr__', '__setitem__', '__str__',
> 'clear', 'copy', 'get', 'has_key', 'items', 'iteritems', 'iterkeys',
> 'itervalues', 'keys', 'popitem', 'setdefault', 'update', 'values']
> =================END==OF=DUMP================
> 
> But... if I do a "dir(__builtins__)" from the Python interpreter there
> are 118 items (not 37) including amongst other stuff, the str()
> function.
> 
> How do I create my own module/dictionary which contains the
> __builtin__ module which contains all the handy functions that
> "dir(__builtins__)" shows up from the interpeter?
> 
> Kind Regards,
> 
> Chris.

PS: My Python dir(__builtins__) lists the following on startup:

================START=OF=PYTHON===============
Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on
win32
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help
>>> dir(__builtins__)
['ArithmeticError', 'AssertionError', 'AttributeError',
'DeprecationWarning', 'EOFError', 'Ellipsis', 'EnvironmentError',
'Exception', 'False', 'FloatingPointError', 'IOError', 'ImportError',
'IndentationError', 'IndexError', 'KeyError', 'KeyboardInterrupt',
'LookupError', 'MemoryError', 'NameError', 'None', 'NotImplemented',
'NotImplementedError', 'OSError', 'OverflowError', 'OverflowWarning',
'ReferenceError', 'RuntimeError', 'RuntimeWarning', 'StandardError',
'StopIteration', 'SyntaxError', 'SyntaxWarning', 'SystemError',
'SystemExit', 'TabError', 'True', 'TypeError', 'UnboundLocalError',
'UnicodeError', 'UserWarning', 'ValueError', 'Warning',
'WindowsError', 'ZeroDivisionError', '__debug__', '__doc__',
'__import__', '__name__', 'abs', 'apply', 'bool', 'buffer',
'callable', 'chr', 'classmethod', 'cmp', 'coerce', 'compile',
'complex', 'copyright', 'credits', 'delattr', 'dict', 'dir', 'divmod',
'eval', 'execfile', 'exit', 'file', 'filter', 'float', 'getattr',
'globals', 'hasattr', 'hash', 'help', 'hex', 'id', 'input', 'int',
'intern', 'isinstance', 'issubclass', 'iter', 'len', 'license',
'list', 'locals', 'long', 'map', 'max', 'min', 'object', 'oct',
'open', 'ord', 'pow', 'property', 'quit', 'range', 'raw_input',
'reduce', 'reload', 'repr', 'round', 'setattr', 'slice',
'staticmethod', 'str', 'super', 'tuple', 'type', 'unichr', 'unicode',
'vars', 'xrange', 'zip']
=================END==OF=PYTHON===============


From wlfraed at ix.netcom.com  Sun Sep 29 19:31:56 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Sun, 29 Sep 2002 16:31:56 -0700
Subject: tkinter & serial port
References: <IhKl9.45766$bX.8904@sccrnsc02>
Message-ID: <ch28na.7vj.ln@ix.netcom.com>

 user at domain.invalid fed this fish to the penguins on Sunday 29 
September 2002 02:30 pm:

> 
> How do I do this or its moral equivalent under Python so that it
> runs without modification under both BSD and Windows? The closest

        Jump on Bill Gates a few times so he changes windows to allow any data 
stream in a select() <G>

> that I've been able to figure out is to poll the serial device
> a few times per second. This is ugly, involves overhead and generally
> not up to the standard of Python -- clearly I don't yet 'see the
> light.'
>
        Possibly put the serial device I/O into a separate thread, package up 
the lines, and send them using a socket protocol which you can then use 
with select?

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From gq437x at yahoo.de  Wed Sep 25 04:05:42 2002
From: gq437x at yahoo.de (Volker Apelt)
Date: 25 Sep 2002 10:05:42 +0200
Subject: summary:  (was: callback + __call__ + TypeError: unbound method; on non method)
References: <lg8z1rtvzi.fsf@mu.biosolveit.local> <JiZj9.157856$pX1.5687433@news2.tin.it> <lgwupbs84w.fsf@mu.biosolveit.local>
Message-ID: <lg1y7ize7d.fsf_-_@mu.biosolveit.local>

## Thank you, Alex Martelli for your help on this problem. 

Problem: 
  Python 2.1.1. took a reference to a function
  stored in a class attribute as a class method.
  Calling that function through an object instance 
  was not possible. 

Solution for 2.2 + Workaround for python < 2.2:
  In Python 2.2 use staticmethod(). 
  For Python < 2.2 (eg: tested with 2.1.1) use a 
  class replacement for staticmethod and add a second
  group of () to the call. 

Example:

import time

def call_me():
    return time.time

## workaround for python < 2.2   
class staticmethod:
    def __init__(self, func): self.func = func
    def __call__(self, *args, **kwds): return self.func(*args, **kwds)
   
class Callback_good:
    """
    """
    _default_Func = staticmethod(call_me) #class wide default callback function
    
    def __init__(self): 
        """ """
        self._m_func = Callback_good._default_Func
        
    def __call__(self): 
        """ """
        # uncomment one of these
        #return self._m_func()() # for python < 2.2
        #return self._m_func()   # for ptyton >= 2.2

x = Callback_good()

## both prints are equal
print time.time 
print x()


-- 
Volker Apelt                   





From kkto at csis.hku.hk  Fri Sep  6 09:22:57 2002
From: kkto at csis.hku.hk (Isaac To)
Date: 06 Sep 2002 21:22:57 +0800
Subject: Dictionary...
References: <7ir8g7xqqk.fsf@enark.csis.hku.hk> <ala7qe$aog$1@bob.news.rcn.net>
Message-ID: <7iheh3qmku.fsf@enark.csis.hku.hk>

>>>>> "Raymond" == Raymond Hettinger <python at rcn.com> writes:

    Raymond> Hello Issac, Try looking at seqdict in the Vaults of Parnassus:
    Raymond> http://home.arcor.de/wolfgang.grafen/Python/Modules/Modules.html

Ah... after I read this it makes sense again now, since Python map is not
comparison based.  Perhaps I've got the wrong idea from the beginning...

    Raymond> Your other ideas were good too.  The auxilliary map of
    Raymond> successors and precessors is likely to be the fastest.  To
    Raymond> efficiently update a sorted python list, try the bisect module.

But difficult to code.  I've to maintain 3 data structures instead of 1.
Well... perhaps time to write a class. =)

Regards,
Isaac.


From phlip_cpp at yahoo.com  Tue Sep  3 00:38:24 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: Tue, 03 Sep 2002 04:38:24 GMT
Subject: opening 'file open' dialog from console
References: <WFWc9.6430$yt3.2976494@newssrv26.news.prodigy.com>
Message-ID: <41Xc9.1255$Qz2.231397557@newssvr16.news.prodigy.com>

dsavitsk wrote:

> I would like to open a file open dialog from a console on windows (2000).
> I do this now using the line
> 
>>>> from tkFileDialog import *
>>>> fileName = askopenfilename(filetypes=[(".txt", ".txt")], title="Select
> the file")
> 
> but this opens a second window and leaves it there.  what is the best way
> (tkinter or MFC) to do this w/o the second window?

Fetch the root singleton (or whatever it is) out of Tk land:

        root = Tk()

Withdraw it:

        root.withdraw()

Warning: Sometimes this withdraws the wrong thing. >sigh<

Alternately, pass it into 'askopenfilename', if possible.

-- 
  Phlip
          http://www.greencheese.org/ParodyMode
  --  Have a :-) day  --



From martin at v.loewis.de  Thu Sep  5 03:41:07 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 05 Sep 2002 09:41:07 +0200
Subject: tk8.4 in python22 ?
References: <3D76FFB4.FE5EF48A@ipm.fhg.de>
Message-ID: <m37ki0ubn0.fsf@mira.informatik.hu-berlin.de>

Markus von Ehr <markus.vonehr at ipm.fhg.de> writes:

> I'd like to use tk8.4 with python22, in fact the windows installer
> installs tk8.3.  How can I replace tk8.3 to 8.4? Do I have to
> compile any python files?

You will need to recompile _tkinter.pyd, using the Tcl 8.4 headers and
import libraries.

Regards,
Martin


From dbasch at yahoo.com  Sun Sep 29 22:22:47 2002
From: dbasch at yahoo.com (Derek Basch)
Date: Sun, 29 Sep 2002 19:22:47 -0700 (PDT)
Subject: Catching user defined exceptions from foreign modules
Message-ID: <20020930022247.86781.qmail@web20810.mail.yahoo.com>

Hello, 
I have defined 2 simple modules, test1 and test2.
test1 calls the run function of test 2 which returns a
user defined TestError function. I then catch the
exception in test1 and perform a print. Heres the
code:
http://www.geocities.com/dbasch/test1.txt
http://www.geocities.com/dbasch/test2.txt

So, this code works fine for me. However, I have 2
much larger modules that have the same code structure
but in these larger modules the test1 refuses to catch
the raised TestError unless I change the except clause
to:

except test2.TestError:
    print 'fudge'

I know that TestError has been imported because I can
put a "print TestError" before the except clause in
test1 and I get "test2.TestError". Can anyone tell me
why I am having to put the test2.TestError instead of
just TestError? It confuses me even further that my
small test modules work like I would expect and it
doesn't work on the larger modules.

Thanks for the help,
Derek Basch





 

__________________________________________________
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
http://sbc.yahoo.com



From aleax at aleax.it  Sun Sep 15 03:39:38 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 15 Sep 2002 07:39:38 GMT
Subject: Tkinter in PyRun_SimpleString(char* s)
References: <am12be$4j0$1@schleim.qwe.de>
Message-ID: <_OWg9.109539$pX1.3874894@news2.tin.it>

Torsten Mohr wrote:

> Hi,
> 
> it doesn't seem to be as easy as i thought to load a script
> from a C program and execute it, if i import Tkinter, it fails:
        ...
>   File "/usr/lib/python2.1/lib-tk/Tkinter.py", line 1476, in __init__
>     baseName = os.path.basename(sys.argv[0])
> AttributeError: 'sys' module has no attribute 'argv'

As you can see, the problem is with sys.argv, which you have
not set.  You can set it up from the Python code you execute,
before you import Tkinter, or with C-API call PySys_SetArgv.


Alex



From maarten at tccn.info  Wed Sep 18 07:10:20 2002
From: maarten at tccn.info (Maarten van Hulsentop)
Date: 18 Sep 2002 04:10:20 -0700
Subject: 'Real' global (write var to different namespace)
Message-ID: <29242dfa.0209180310.94c915d@posting.google.com>

Hi y'all 

I'm sorry if it's a FAQ but i honestly can't find to solution to my
problem; i've created a 'store' (a class) to store vars using the
(really cool) pickle classes. Now i can store a var doing:
"mystore['varname'] = myvar" and read it again by "myvar =
mystore['varname']". My question is if it's possible to create a
function like 'mystore.RestoreVars()' that sets ALL the stored vars in
the _calling_ namespace at once.

Hope it's clear enough, i'm still a python newbie so excuse me if i
used the wrong terms or if i didn't explain it enough.

Thanks

Maarten van Hulsentop


From totoNOSPAM at toto.fr  Tue Sep 24 05:52:06 2002
From: totoNOSPAM at toto.fr (toto)
Date: Tue, 24 Sep 2002 11:52:06 +0200
Subject: embedding Python in C
Message-ID: <3D9035C6.BAC23A90@toto.fr>

Hello,

I try to use a simple programme to try to run Python code from C by
calling the PyRun_SimpleString API function. But I have the following
problem when I compile the programme C :

------------------------------------------------
bash-2.04$ make -f usermod.mak
gcc embed-usermod.o
/local/edf-python-2.1.1/lib/python2.1/config/libpython2.1.a -L/usr/lib
-L/usr/X11R6/lib -lgdbm -ltk8.3 -ltcl8.3 -lX11 -lm -ldl -g
-export-dynamic  -o embed-usermod
/local/edf-python-2.1.1/lib/python2.1/config/libpython2.1.a(posixmodule.o):
In function `posix_tmpnam':
/usr/src/redhat/BUILD/Python-2.1.1/./Modules/posixmodule.c:4212: the use
of `tmpnam_r' is dangerous, better use `mkstemp'
/local/edf-python-2.1.1/lib/python2.1/config/libpython2.1.a(posixmodule.o):
In function `posix_tempnam':
/usr/src/redhat/BUILD/Python-2.1.1/./Modules/posixmodule.c:4168: the use
of `tempnam' is dangerous, better use `mkstemp'
/local/edf-python-2.1.1/lib/python2.1/config/libpython2.1.a(thread.o):
In function `PyThread_start_new_thread':
/usr/src/redhat/BUILD/Python-2.1.1/Python/thread_pthread.h:135:
undefined reference to `pthread_create'
/usr/src/redhat/BUILD/Python-2.1.1/Python/thread_pthread.h:159:
undefined reference to `pthread_detach'
/local/edf-python-2.1.1/lib/python2.1/config/libpython2.1.a(posixmodule.o):
In function `posix_openpty':
/usr/src/redhat/BUILD/Python-2.1.1/./Modules/posixmodule.c:1754:
undefined reference to `openpty'
/local/edf-python-2.1.1/lib/python2.1/config/libpython2.1.a(posixmodule.o):
In function `posix_forkpty':
/usr/src/redhat/BUILD/Python-2.1.1/./Modules/posixmodule.c:1785:
undefined reference to `forkpty'
collect2: ld returned 1 exit status
make: *** [embed-usermod] Erreur 1
------------------------------------------------

If some of you can give me some ideas of this problem ? Thank you  very
much.

Yang



From jhorneman at pobox.com  Mon Sep 30 02:51:00 2002
From: jhorneman at pobox.com (Jurie Horneman)
Date: Mon, 30 Sep 2002 08:51:00 +0200
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com> <3d976fbc$0$23666$91cee783@newsreader02.highway.telekom.at> <8428na.7vj.ln@ix.netcom.com> <3d979184$0$20332$79c14f64@nan-newsreader-03.noos.net> <8sf8na.86m.ln@ix.netcom.com>
Message-ID: <3d97f3f4$0$32596$91cee783@newsreader02.highway.telekom.at>

"Dennis Lee Bieber" <wlfraed at ix.netcom.com> wrote in message
news:8sf8na.86m.ln at ix.netcom.com...
> Guy Rabiller fed this fish to the penguins on Sunday 29 September 2002
> 04:49 pm:
>
> >
> > yes it works on MS-DOS too.
> >
>         Ah, good to know...

Dennis Reinhardt was right: it doesn't work in the Win98 DOS shell.

>         Still need -vv to get the details from import though; -v just says
> "not found"

Ah, I thought the -vv was a typo. 'python -h' does not list 'vv' as an
option. Thanks.

Jurie.





From jacek.generowicz at cern.ch  Fri Sep 20 04:03:18 2002
From: jacek.generowicz at cern.ch (Jacek Generowicz)
Date: 20 Sep 2002 10:03:18 +0200
Subject: strong/weak - dynamic/static [Was: Getting started]
References: <uojinmcfapjf45@corp.supernews.com> <tyf8z1y5cp5.fsf@pcitapi22.cern.ch> <5ami9.114674$ub2.2440492@news1.tin.it>
Message-ID: <tyffzw53x6h.fsf@pcitapi22.cern.ch>

Alex Martelli <aleax at aleax.it> writes:

> Jacek Generowicz wrote:
>         ...
> > Could someone provide a serious reference ...

> For example, Amer Diwan's lessons on Fundamentals of Programming
> Languages at the University of Colorado

...

> and the already-referenced URL gives some fundamental bibliography

Thank you, Alex.


From shalehperry at attbi.com  Mon Sep 23 19:55:00 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Mon, 23 Sep 2002 16:55:00 -0700
Subject: Elementary confusion
In-Reply-To: <jwbaxter-31BCDD.16364723092002@corp.supernews.com>
References: <878z1s1kqf.fsf@smtp.blueyonder.co.uk> <jwbaxter-31BCDD.16364723092002@corp.supernews.com>
Message-ID: <200209231655.00621.shalehperry@attbi.com>

On Monday 23 September 2002 16:36, John Baxter wrote:
> >>> a=0.1
> >>> b=0.2
> >>> c=a+b
> >>> c==0.3
>
> 0
>
> >>> c
>
> 0.30000000000000004
>
> >>> d=0.3
> >>> d
>
> 0.29999999999999999
>
>
> We "know" that c is 0.3.  The computer has added two inexact values
> which are "too big" and produced a result which is therefore both "too
> big" and not the nearest representation of 0.3, which is "too small".
>
>   --John

Every C programming book I have ever read makes the comment that == and 
floating point do not mix.  Just one the things you learn and accept.



From andreas.ulbrich at gmx.net  Mon Sep 23 10:11:04 2002
From: andreas.ulbrich at gmx.net (Andreas Ulbrich)
Date: Mon, 23 Sep 2002 16:11:04 +0200
Subject: Session handling
References: <c5b3549a.0209230511.5bbe6849@posting.google.com>
Message-ID: <amn7dp$6cn$1@mamenchi.zrz.TU-Berlin.DE>

beachboy wrote:
> Hi!
> Please can anyone help me.
> My problem:
> I'm trying to program a web application with python. This tool has to
> provide diffrent languages. The user has to choose the language at the
> beginning. So i have to save the information which language the user
> chooses. I think this is a perfect example for using sessions.
> The only problem is I've got no idea how to realise this in python.
> Please help me as soon as you can because I have to solve this problem
> very quickly.
> 
> hoping to hear from you soon
> beachboy

Whenever I need to write a web-application I do use jython and write 
servlets. This combines the power of Servlet/JSP with the worlds best 
programming language.

a jython servlet would look like this:

from javax.servlet.http import HttpServlet

class MyServlet(HttpServlet) :

	def doGet(self, request, response) :

		session = request.getSession(1)
		lang = session.getAttribute('language')
		if not lang :
			# initialise the session (i.e. choose language)

		# produce output suitable to lang
		if lang == 'DE' :
			print >> res.writer, "Hallo Welt!"
		else :
			print >> res.writer, "Hello World!"

There are other solutions. I believe that Zope offers session-management 
too.



From jubafre at brturbo.com  Mon Sep 23 17:53:44 2002
From: jubafre at brturbo.com (jubafre at brturbo.com)
Date: Mon, 23 Sep 2002 18:53:44 -0300 (GMT-03:00)
Subject: list problem
Message-ID: <-29615135.1032818024099.JavaMail.nobody@webmail2.brturbo.com>

i have:	
x=[['0020', '0000x9'], ['0030', '0000xa'], ['00B4', '0000x8'], ['0030', '0000xb'], ['00F0'], ['0001'], ['0003'], ['0005']]

and i want:
y=['0020', '0000x9', '0030', '0000xa', '00B4', '0000x8', '0030', '0000xb','00F0', '0001', 
'0003', '0005']

how to do this?

Juliano Freitas
www.gebrasil.hpg.com.br 

From just at xs4all.nl  Mon Sep  9 16:36:37 2002
From: just at xs4all.nl (Just)
Date: Mon, 09 Sep 2002 22:36:37 +0200
Subject: Windows/DOS: double clicking a .py file
References: <just-A936EC.17062909092002@news1.xs4all.nl> <alie4t$1q4jol$1@ID-59885.news.dfncis.de> <just-F18BA3.17425609092002@news1.xs4all.nl> <3h7f9.77472$l_4.28399@atlpnn01.usenetserver.com>
Message-ID: <just-C78A8A.22363709092002@news1.xs4all.nl>

In article <3h7f9.77472$l_4.28399 at atlpnn01.usenetserver.com>,
 "Steve Holden" <sholden at holdenweb.com> wrote:

> "Just" <just at xs4all.nl> wrote in message
> news:just-F18BA3.17425609092002 at news1.xs4all.nl...
> [ how top see messages ]
> >
> > Thanks for both suggestions. Another workaround just occured to me: do a
> > raw_input("type enter to exit") in an exception hook. That should work
> > just fine in a py2exe-fied app as well I suppose.
> >
> 
> python -i
> 
> will allow you to enter interactive statements after your execution
> finishes, which can be useful when you're debugging.

Actually, debugging on my side is not really the problem (I usually 
start the app from the command line anyway, it's a simple 
platform-neutral command line tool), but it's end users: they want to 
drop files onto the app. The DOS window can go away if everything went 
ok (it only displays some progress info), but not if an exception 
occurred: I want them to a) be able to tell something's wrong, and b) 
tell me _what_ went wrong ;-). The raw_input() call upon exception seems 
the easiest workaround for now.

Just


From usenet at wojas.vvtp.tudelft.nl  Wed Sep  4 09:50:26 2002
From: usenet at wojas.vvtp.tudelft.nl (Konrad Wojas)
Date: Wed, 04 Sep 2002 15:50:26 +0200
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com>
Message-ID: <al5332$fm$1@news.tudelft.nl>

Benjamin wrote:

> Hi guys (and girls)
> I just begun to learn how to program. And I choosed Python for
> doing that. surfing around the web i found an article about this
> language, they said it's the ideal language for beginners to the
> world of programming. and i admit it, the syntax is very clear making
> the code very readable, i learn easily the concepts of programming
> and i soon wrote some nice programs. but since i finished my
> newbie-tutorial, i am wondering about one question: is it worth to go
> deeper into python? i talked to a friend of mine, he told me that i
> should try php..
> what should i do?? ^_^ in other words: what is python good for?
> why should i continue with python?

I've been doing web development in PHP since 1998 and have written a content 
management system with a quite complex shopping cart system and a 
backoffice system to handle all sales. It's over 10000 lines of PHP code, 
excluding the html code.

I've really had it with PHP, it's unmaintainable, ugly and slow. Errors like 
'parse error on line 487' are not fun to work with. I really like python. 
The systax is a lot clearer, exceptions are great and I really love those 
stacktraces if an exception is not handles. Object oriented programming in 
PHP is a joke, it really sucks. And PHP is buggy. I've encountered numerous 
bugs, especially in the extension modules. XSLT in PHP? Forget it, the 
module will crash in 50% of the cases. And compiling PHP yourself with a 
lot of extension modules can be a real nightmare. And above all: python is 
a generic language, PHP is only useable for web development and small 
maintainance scripts. It doesn't have bindings for anything beyond web 
development. 

I'm still maintaining the PHP code I have written, but I use python for all 
new projects. Don't start big projects in PHP, you'll regret it ;-)

Regards,
-- 
Konrad Wojas


From GavinT at ResMed.com.au  Tue Sep 24 00:06:11 2002
From: GavinT at ResMed.com.au (Gavin Tomlins)
Date: Tue, 24 Sep 2002 14:06:11 +1000
Subject: Python and activescripting within a host environment
Message-ID: <362133AB12B2CE4EA3C01E12CD6CCC6E0254E0B7@osprey.corp.resmed.org>

Greetings,

We are still trying to solve one or two problems that we have with Python in
a ActiveScripting host. Currently we can access COM based objects
successfully from the python command line and perform normal operations.
However, I was curious to know if people have any examples of products with
ActiveScripting embedded within them that exposes a object model that can be
utilised via Pythonscript. eg. (http://www.atozedsoftware.com/
FinalBuilder). In particular, we are unable to access variables from the
host application. My understanding was that upon a request for a variable or
unknown type it would be fired to the ActiveScripting engine to return the
value or an error.
eg. Within the hostapplication, they expose WindowTitle, ProjectVar
Shouldn't I be able to do something simple like

MyWindowTitle = WindowTitle
MyProjectVar = ProjectVar

or 
ProjectVar = "A new value"

ie. above logic works with Vbscript etc.

is there any special notation needed to identify host variables, or does the
entire variable space of the host application become global to the python
instance.

Is there any documentation outlining the integrating or the absorption of
such operations with Python and ActiveScripting.

Regards

Gavin


Warning:  Copyright ResMed.  Where the contents of this email and/or attachment includes materials prepared by ResMed, the use of those materials is subject exclusively to the conditions of engagement between ResMed and the intended recipient.

This communication is confidential and may contain legally privileged information. By the use of email over the Internet or other communication systems, ResMed is not waiving either confidentiality of, or legal privilege in,the content of the email and of any attachments.
If the recipient of this message is not the intended addressee, please call ResMed immediately on  +61 2 9886 5000 Sydney, Australia.




From tismer at tismer.com  Wed Sep 18 11:42:45 2002
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 18 Sep 2002 17:42:45 +0200
Subject: Stackless Python, eventual merge?
References: <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <20020918060823.GB81158@hishome.net> <3D8880AC.7040902@tismer.com> <20020918181834.A27703@hishome.net>
Message-ID: <3D889EF5.1080605@tismer.com>

Oren Tirosh wrote:
...

>>Just-need-an-idea-how-to-determine-that - ly y'rs - chris
> 
> 
> Who says you have to determine that in advance? Just try. If the
> extension actually calls the interpreter recusively it means that it
> needs its own stack.  Let it have exclusive ownership of the current OS
> thread and its associated stack and allocate another thread from a
> dynamic thread pool to continue tasklet scheduling. When the extension
> returns back to nesting level 0 its thread will be released back into
> the pool. This way a relatively small pool of OS threads can be used to
> juggle a much larger pool of tasklets.

This is of course an interesting idea.
The reason why I thought I need to determine in advance
is that I don't know when a new extension shows up on
my stack. But when this happens, bazillions of tasklets
may have been created already, and they have live cstack
objects, which cannot be relocated (at least not in the near
future). So they do need to use this one stack again.

> This can also be a good way to handle extensions that make blocking I/O
> calls like database engines - a relatively small pool of OS threads can
> serve requests from a bigger pool of tasklets. This can be transparent:
> when the number of concurrent requests exceeds the number of threads any
> new request frames can be queued until a thread becomes available. The
> level of concurrency for any resource can be tuned dynamically to
> optimize the performance without affecting the overall structure of the
> program.

Sure, that's a thing I definately will do, if I can find
a good solution to the other problem above.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From see_reply_address at something.invalid  Thu Sep 26 19:28:28 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Fri, 27 Sep 2002 11:28:28 +1200
Subject: How do you pronounce "__" (double underscore)?
References: <525eaba.0209260956.72bf6786@posting.google.com>
Message-ID: <3D93981C.1030101@something.invalid>

Pat Notz wrote:

> Is there a short (one or two syllable) pronounciation for "__" (double
> underscore).


How about

    _ == "uh"

    __ == "uuuh"

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From llothar at web.de  Sun Sep 15 21:28:59 2002
From: llothar at web.de (Lothar Scholz)
Date: Mon, 16 Sep 2002 03:28:59 +0200
Subject: AspectP?
References: <jOQg9.6057$Lg2.882928@news2.news.adelphia.net> <mailman.1032095664.13848.python-list@python.org>
Message-ID: <7cl9oucm65a4qr2tolm74gltqn7morbbm5@4ax.com>

On Sun, 15 Sep 2002 08:13:34 -0500, "Mark McEahern"
<marklists at mceahern.com> wrote:

>[Robert Oschler]
>> Anybody seen a robust (fairly ready for primetime) implementation
>> of Aspect Oriented Programming for Python?
>
>Ignoring the criteria of robust...

I think from this posting he has no problems in finding instable 
implementations.



From eric.brunel at pragmadev.com  Wed Sep 18 04:46:36 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Wed, 18 Sep 2002 10:46:36 +0200
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <Xns928CED692541cliechtigmxnet@62.2.16.82> <6_Ph9.479711$q53.16315681@twister.austin.rr.com>
Message-ID: <am9jj9$c2m$1@wanadoo.fr>

Lee Gray wrote:
[snip]
> For a newbie, what's easier, Tk or wxWindows?

Matter of taste, really... I do prefer Tkinter, because I find it far 
simpler than other GUI toolkits. But then, I've been into Tk for some years 
now, and maybe I'm too used to it.

The first time I've looked at wxWindows and seen what's needed to open a 
simple window, it just made me run away screaming ;-). But seeing it again 
now, it's not that bad. See:

http://www.wxpython.org/tut-part1.php

In fact, example one in this tutorial would be much shorter in Tkinter than 
in wxWindows, but examples 2 & 3 would require much more code in pure 
Tkinter because of the status bar: it's apparently automatically managed by 
wxWindows, but not by Tkinter. Adding Pmw to Tkinter would make things 
easier ( see http://pmw.sourceforge.net ).

For the record, here is the equivalent (untested) Tkinter+Pmw code for 
examples 1 & 3 in the wxPython tutorial:

--- example 1 -------------------------
from Tkinter import *

topWdw = Tk()
topWdw.title("Hello from Tkinter")
topWdw.mainloop()
----------------------------------------

--- example 3 -------------------------
from Tkinter import *
import Pmw

## Main window
topWdw = Tk()
topWdw.title("Hello from Tkinter")
## Status bar
statusBar = Pmw.MessageBar(topWdw)
statusBar.pack(side=BOTTOM, fill=X)
## Helper for menu items
helper = Pmw.Balloon(topWdw)
helper.configure(statuscommand=statusBar.helpmessage)
## Function displaying about screen
def showAbout():
  showinfo('About Me', "This sample program shows off\n"
                       "frames, menus, statusbars, and this\n"
                       "message dialog.")
## Menus
menuBar = Pmw.MainMenuBar(topWdw, balloon=helper)
topWdw.configure(menu=menuBar)
menuBar.addmenu('File', '')
menuBar.addmenuitem('File', 'command',
                    'More information about this program',
                    label='About', command=showAbout)
menuBar.addmenuitem('File', 'separator')
menuBar.addmenuitem('File', 'command',
                    'Terminate the program',
                    label='Exit', command=topWdw.quit)
topWdw.mainloop()
---------------------------------------

Just see for yourself the one you find clearer.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From sismex01 at hebmex.com  Wed Sep 18 10:59:53 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Wed, 18 Sep 2002 09:59:53 -0500
Subject: minimizing Tk programs to the taskbar
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062330249BC4C@mail.hebmex.com>

> 
> Do properly-coded Tk extensions benefit Tkinter users 
> directly?  Is there a way to load them?  I doubt it, but
> I haven't thought long about it ...
>

Of course there is, only it's kinda roundabout,
and you have to declare a new type of widget,
and play around with some Tkinter stuff, not
very newbie friendly.

I made bindings for the tkhtml and tktable extensions,
and they more-or-less worked well; I didn't have enough
time to keep pursuing it, so I didn't go very far;
but they did work.

HTH

-gus
--




Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From steve at ferg.org  Sat Sep  7 08:01:05 2002
From: steve at ferg.org (Stephen Ferg)
Date: 7 Sep 2002 05:01:05 -0700
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net>
Message-ID: <b16e4ef7.0209070401.16087fd@posting.google.com>

"Joseph A. Knapka" <jknapka at earthlink.net> wrote in message news:<3D798461.B3FE0E06 at earthlink.net>...

> you probably shouldn't refer to the parent/child
> relationship between UI elements as "binding", since
> that will cause great confusion when you get around
> to discussing the binding of events to code (using
> the widget.bind() method).

You're probably right.  I'll see if I can't rewrite that.
 
> Also, you should be aware of the "command" option

I know about it, but I think that for beginners, it is very confusing.
 For one thing, it leads down the path to lambda, which is tricky
enough itself.  For another, it doesn't automatically pass an event
object, which makes it even more difficult to use.  So I think that
beginners should consistently use the more plodding technique of
explicit binding.

But I should probably add a note to that effect.  Otherwise, the
avoidance of the "command" option might be confusing.

Thanks for the feedback! :-)

-- Steve


From tjreedy at udel.edu  Mon Sep 23 21:22:01 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 24 Sep 2002 01:22:01 GMT
Subject: writing in a file as binary mode????
References: <mailman.1032810110.30498.python-list@python.org>
Message-ID: <Y6Pj9.241352$z91.10208055@bin3.nnrp.aus1.giganews.com>

<jubafre at brturbo.com> wrote in message
news:mailman.1032810110.30498.python-list at python.org...
>i open the a file for writing in a binary mode, but doesnt work

Depends what you mean by 'work'.  On *nix, *all* read/write is binary
mode.  On Win*, the *only* difference is conversion of \r\n to \n on
input and vice versa on output.

>just the "f.write('\x03AHM ')" work, i open the file in the notepad
and my file appears >like a text file and not as a binary file, i dont
know why??

*All* files are string of binary digits.  Notepad tries to interpret
any file as as extended-ASCII text file as best it can.

>>file.mem
>AHM 0020 0000x7 0030 0000x8 0010 0000x9 00F0 0003 0002 0000

>>i have the file like this, how??
>AHM  ? 0 ?  f ?

That is how Notepad displays what it cannot properly interpret.

Terry J. Reedy




From bruce_deletethis_dawson at cygnus-software.com  Sun Sep 22 01:19:25 2002
From: bruce_deletethis_dawson at cygnus-software.com (Bruce Dawson)
Date: Sun, 22 Sep 2002 05:19:25 GMT
Subject: Minimal debug/rep functionality
References: <J%ai9.893$1D.574@nwrddc01.gnilink.net>
Message-ID: <3D8D52D0.7050107@cygnus-software.com>

On a tangential note...

While doing some Python optimizing (for our about to ship Python based 
game) I tweaked the HAP Python debugger to count lines of Python code 
being executed. Then I displayed lines-of-code per frame in the game 
title bar. That turns out to be a very useful coarse indicator.

Then I added the expression HeDbg.GetLineCount()[1] to the watch window.
Since the GetLineCount() feature returns total lines executed and lines 
executed since the last call, and since the watch expression is 
evaluated each time control returns to the debugger, this simple 
expression tells me how many lines of code execute each time I step or go.

So, if I step over a trivial looking function and the watch window says 
700 then I investigate to find out whether 700 lines of Python executed 
for the function is reasonable.

For profiling of a game I found this feature set worked really well.

I don't think we've released an updated version of HAP that has this 
feature, but we certainly will - lots of other updates also.

Mark wrote:

> Hello all,
> 
> In a nutshell, what I'd like to do is develop a utility that would print out 
> every line in a python program as the program executes (I may even just 
> want to print line numbers).  I can see hints of how to do this in pdb (and 
> more directly in the underlying bdb class) and also in the code module.
> 
> So, 
> 
> 1) Should I try and inherit a class and override some methods or should I 
> just pull out the relevant code and hand spin a small interpreter.
> 
> 2) Anyone have a 20 liner solution for this?
> 
> Thanks.
> 
> Regards,
> Mark
> 



From SSchukat at dspace.de  Fri Sep 20 10:41:59 2002
From: SSchukat at dspace.de (Stefan Schukat)
Date: Fri, 20 Sep 2002 15:41:59 +0100
Subject: Python COM and variable number of arguments
Message-ID: <84257D042AA7D411B3F700D0B7DB9B7C13B2FC@PDC-DSPACE>

You just need the makepy generated wrapper and then you can 
just use named arguments:

Object.Add(Item,After=3)

Without the wrapper you are out of luck. 

	Stefan

> -----Original Message-----
> From: Greg Green [mailto:gregory.p.green at boeing.com]
> Sent: Thursday, September 19, 2002 10:36 PM
> To: python-list at python.org
> Subject: Python COM and variable number of arguments
> 
> 
> I am using python with a third-party closed source application. It has
> Visual Basic for Applications embedded in it for use as an extension
> language. My question is what to do about com calls that have a
> variable number of arguments in Visual Basic. It seems that I have to
> use all of the arguments when I call the method using python. But I
> don't know what to use for a value when I don't want that argument.
> 
> For instance, In VB i can call the method .Add with 1, 2, or 3
> arguments. The 2nd and 3rd arguments are optional. In this case the
> 2nd argument is called 'Before', and the 3rd is 'After'. So if I want
> to add an item after the 3rd item. I do: 
> 
>         .Add(Item,,3) 
> 
> In python I don't know what to use as the second argument:
> 
>         Object.Add(Item=Item,Before=????,After=3)
> 
> In case you are wondering, I have tried 'None', or leaving off the
> argument all together, but that doesn't work. :(
> 
> Thanks for any replies. This is causing me a lot of pain.
> 
> -- 
> 
> Greg Green
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 



From robin at alldunn.NOSPAM.com  Thu Sep 19 22:24:20 2002
From: robin at alldunn.NOSPAM.com (Robin Dunn)
Date: Fri, 20 Sep 2002 02:24:20 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com>
Message-ID: <oFvi9.113$FY2.82713@news.randori.com>

Lee Gray wrote:
[...]
> Can someone point me to a good Python GUI app to demonstrate (other
> than IDLE)?

There are some new screen shots of apps written in wxPython at 
http://wxpython.org/screenshots.php

-- 
Robin Dunn
Software Craftsman
http://wxPython.org  Java give you jitters?  Relax with wxPython!



From see_reply_address at something.invalid  Thu Sep 19 21:17:41 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Fri, 20 Sep 2002 13:17:41 +1200
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <mailman.1032329366.6813.python-list@python.org> <j4ptvbaapm.fsf@informatik.hu-berlin.de> <3D890FB0.4030408@something.invalid> <m3admeo4u1.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3D8A7735.3090403@something.invalid>

Martin v. Loewis wrote:

> That is not true. It is more reliable to inspect the source code.
> 
> In general, it is very easy to do so: if there are no callbacks to
> Python, the module is stackless-safe.


That's a fair point, and I guess it takes care of
the majority of extensions.

But the cases where there is doubt are typically
going to be big, complicated things like GUI
libraries. Making a line-by-line inspection of
one of those isn't exactly something you do over
a cup of tea.

And you don't just have to look at the code, you
have to understand what it's doing, because the
things which cause problems are dynamic rather
than static features of the code.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From quiteblack at yahoo.com  Wed Sep 25 01:05:32 2002
From: quiteblack at yahoo.com (black)
Date: Tue, 24 Sep 2002 22:05:32 -0700 (PDT)
Subject: about webbrowser.open~
Message-ID: <20020925050532.601.qmail@web21304.mail.yahoo.com>


Hi, all~

  I write something like this: webbrowser.open_new("http://www.python.org")

it works, but it open the page with an old one, not the new page at all, what's wrong ? my OS is window2000 professional, thanx in advance~



---------------------------------
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020924/0d2cc479/attachment.html>

From mfranklin1 at gatwick.westerngeco.slb.com  Mon Sep 16 04:46:16 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Mon, 16 Sep 2002 08:46:16 +0000
Subject: Tkinter listbox clicking problem
In-Reply-To: <sGog9.307183$aA.53945@sccrnsc02>
References: <sGog9.307183$aA.53945@sccrnsc02>
Message-ID: <200209160846.16854.mfranklin1@gatwick.westerngeco.slb.com>

Matthew,


On Friday 13 Sep 2002 4:49 pm, Matthew Bonig wrote:
> (lb.bind("<Button-1>", someFunc)) it calls function someFunc. Now,
> someFunc() simply prints the curselection(). The problem is it's lagged...
> if I first select list item 1 (index 0), and then 2, and then 3 (array


Simply change the binding to <ButtonRelease-1>



>>> root=Tk()
>>> lb=Listbox(root)
>>> lb.insert("end", 1)
>>> lb.insert("end", 2)
>>> lb.insert("end", 3)
>>> lb.insert("end", 4)
>>> lb.pack()
>>> def callback(event):
...     print event.widget.curselection()
...     print event.widget.get(event.widget.curselection()[0])
...
>>> lb.bind("<ButtonRelease-1>", callback)
'136346756callback'
>>> ('0',)
1
('1',)
2
('2',)
3
('3',)
4
('3',)
4
('2',)
3
('2',)
3
('1',)
2
('0',)
1
('1',)
2
('2',)
3
('3',)
4

HTH
Martin

-- 
### Python Powered Signature
I started writting this email on 
Mon Sep 16 08:40:42 2002




From marklists at mceahern.com  Tue Sep 24 00:20:09 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 23 Sep 2002 23:20:09 -0500
Subject: Outlook 2002 and Python
In-Reply-To: <03F53405C6ED434A986221919785A6A78C9339@nca-postal.ncaustin.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOCEPEEAAA.marklists@mceahern.com>

[Paul Sage]
>What:  I saw an idea for a way to create a good SPAM filter for email.
[snip]

You might want to checkout:

  http://spambayes.sf.net/

// m
-




From eric.brunel at pragmadev.com  Thu Sep 12 05:48:33 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Thu, 12 Sep 2002 11:48:33 +0200
Subject: class variable won't icrement!
References: <3d8054ef$0$234$4d4ebb8e@news.nl.uu.net>
Message-ID: <alpner$bm$1@wanadoo.fr>

Gumuz wrote:

> Hello everyone,
> 
> I have something very strange and i can not understand it. probably, it's
> not _that_ strange : )
> 
> I have this simple class which registers itself on a class variable in the
> following code:
> 
> 
----------------------------------------------------------------------------
> ------
> class test:
>     dict = {}
>     id = 1
>     def __init__(self, name):
>         self.dict[name] = self
>         self.id = self.id + 1
> 
>     def show(self):
>         print "class variable dict element count: ", str(len(self.dict))
>         print "class variable id: ", self.id
> 
----------------------------------------------------------------------------
> --------
> 
> now, look at the following:
> 
>>>> john = test("john")
>>>> frank = test("frank")
>>>> bill = test("bill")
>>>> bill.show()
> class variable dict element count:  3
> class variable id:  2
>>>> frank.show()
> class variable dict element count:  3
> class variable id:  2
> 
> The dict variable acts fine, maintaining it's content throughout every
> instance of the class.
> The id variable however, seems to be only incremented within the instance.
> In the end this variable should've been 4.
> 
> Could someone explain this to me? Is there something about class variables
> whta i should know?

Yep: to access them, you'd better always use the syntax <class 
name>.<attribute name>. Instances do in fact "inherit" the attributes of 
their class, but it leads to misunderstandings just as yours. In your 
example, here is what happens:

> class test:
>     dict = {}

Here, you create an attribute dict for class test. The value for this 
attribute is an *object*, which is *mutable*, that happens to be an empty 
dictionary.

>     id = 1

Here, you create an attribute id for class test. The value for this 
attribute is an *integer*, which is *not* an object, and is *immutable*.

>     def __init__(self, name):
>         self.dict[name] = self

Here, you use the attribute dict, inherited from the class, and you modify 
it. These modifications happens *in place*, which means that the pointed 
dictionary is actully modified. Since both test.dict and self.dict point to 
the same object, they'll both see the modification. This is exactly the 
same if you do:

a = {}
b = a
b['foo'] = 'bar'
print a

You'll see that a has been modified. In your example, here is roughly what 
happens:

test.dict = {}
self.dict = test.dict    # Implicit inheritance from class to instance
self.dict[name] = self

>         self.id = self.id + 1

This is a completely different story: here, you *recreate* an instance 
attribute named id, based on the id attribute inherited from the class. So, 
in fact, you're doing:

test.id = 1
self.id = test.id    # Again, implicit inheritance from class to instance
self.id = self.id + 1

just as in:

a = 1
b = a
b = b + 1
print a

Of course, a is not modified here...

I always found that inheritance of class attributes to instance attributes 
was really misleading, and I even think it would be better if it were 
simply removed from Python. IMHO, it would far more conform to the 
"explicit is better than implicit" rule that is one of the basis of Python.

To avoid its drawbacks:
- never declare as a class attribute something that's actually an instance 
attribute
- always use <class name>.<attribute name> to access class attributes, 
especially when you modify them

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From lu_gio at hotmail.com  Wed Sep  4 18:16:57 2002
From: lu_gio at hotmail.com (Max)
Date: Wed, 04 Sep 2002 22:16:57 GMT
Subject: Database BSD on Mac?
References: <3144be3c280c062e5e71a98ab191f664.98737@mygate.mailgate.org> <mailman.1031095985.2997.python-list@python.org> <_7vd9.57894$ub2.928923@news1.tin.it> <slrnancvnh.107.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <tDvd9.57994$ub2.932923@news1.tin.it>

> > I read shelve can be used for small amount of data... will it work with
5mb?
>
> Certainly. And probably for _lots_ of more data, if your problem fits
> an on-disk hashtable like shelve.

Thanks, I'll try it.
I was reading on my book it's possible to create encrypted modules, but
don't explain much how.
My idea is to create a exe file in C joust to execute the python encrypted
modules stored with shelve.
Is it possible? Can you point me to a tutorial/example code?





From shalehperry at attbi.com  Wed Sep 18 16:47:39 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 18 Sep 2002 13:47:39 -0700
Subject: linux python ideas
In-Reply-To: <3D88E578.6010001@diespammerdieuselesspython.com>
References: <3D88E578.6010001@diespammerdieuselesspython.com>
Message-ID: <200209181347.39562.shalehperry@attbi.com>

On Wednesday 18 September 2002 13:38, Rob Andrews wrote:
> I'll be giving a presentation on Python to my local LUG
> (http://lugoj.org) in a few weeks, and would like to point out at least
> a few things that would be of particular interest to linux users.
>
> The idea is not "how to program in Python", which would take more than
> one quick presentation, but a one-off demonstration of ways in which
> Python is particularly useful for such a group.
>
> Any suggestions would be appreciated, since my notes are pretty generic
> so far.
>
> regards,
> Rob Andrews

obvious ideas are things like zope and mailman as well as cheetah.



From jubafre at brturbo.com  Mon Sep 16 15:45:11 2002
From: jubafre at brturbo.com (jubafre at brturbo.com)
Date: Mon, 16 Sep 2002 16:45:11 -0300 (GMT-03:00)
Subject: Alex Martelli <aleax@aleax.it> -> search re and put in a list??
Message-ID: <1138212326.1032205511801.JavaMail.nobody@webmail1>

mem=['7','8', '9']
labels=['D1','D2','D3'] 
S=['20 D1', '30 D2', '10 D3', 'F0', ' 3', ' 2', ' 0']

i want final=['20 7', '30 8', '10 9', 'F0', ' 3', ' 2', ' 0']

#++++++++++++++++++++++++++++++++++++++++++++++
repdict = dict(zip(labels, mem))
big_re = re.compile('|'.join(labels))
def replacer(mo, repdict=repdict):
    return repdict[mo.group()]
final = [ big_re.sub(x, replacer)for x in S ]
print final 
#++++++++++++++++++++++++++++++++++++++++++++++

 Traceback (most recent call last):
  File "<string>", line 1, in ?
  File "C:\Documents and Settings\jubafre\Desktop\teste.py", line 27, in ?
    final = [ big_re.sub(x, replacer)for x in S ]
TypeError: expected string or buffer
	

what happened????



Juliano Freitas
www.gebrasil.hpg.com.br 

From martin at v.loewis.de  Sun Sep 29 17:18:32 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 29 Sep 2002 23:18:32 +0200
Subject: Python, unicode, Windows32 and NTFS...
References: <an1l2o$fvc$1@dackel.pdb.sbs.de>
Message-ID: <m38z1ko5p3.fsf@mira.informatik.hu-berlin.de>

"Werner Merkl" <werner.merkl at fujitsu-siemens.de> writes:

> Python -- at least since version 1.6 -- is supposed to understand
> and handle UNICODE. But my experience is, it doesn't support
> UNICODE for NTFS file names.

That's basically correct. Since Python 2.1, you can use non-ASCII
characters in a Unicode file name as long as they are included in the
"ANSI code page" (aka "mbcs encoding").

If you want to access arbitrary Unicode file names (within the
limitations of the file system), you need the facility described in

http://www.python.org/peps/pep-0277.html

which is implemented in

http://python.org/sf/594001

Regards,
Martin



From op73418 at mail.telepac.pt  Wed Sep  4 13:53:46 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Wed, 04 Sep 2002 18:53:46 +0100
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de>
Message-ID: <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com>

On 04 Sep 2002 18:47:34 +0200, loewis at informatik.hu-berlin.de (Martin v.
L?wis) wrote:

>Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:
>
>> So why hasn't this been extended when nested scopes were introduced,
>> e.g. by reusing the global directive or some other one? 
>
>I think the main reason is that nobody could propose an acceptable
>syntax. Reusing the global directive does not work.

Can you explain why it does not work?

>
>Regards,
>Martin

All the best,
Gon?alo Rodrigues


From ianb at colorstudy.com  Thu Sep 19 13:45:56 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 19 Sep 2002 12:45:56 -0500
Subject: Need advice on choosing the technology for intranet based
	application
In-Reply-To: <3d88fd1e@news.ivm.net>
References: <3d88fd1e@news.ivm.net>
Message-ID: <1032457557.611.1179.camel@dsl-65-184-205-5.telocity.com>

On Wed, 2002-09-18 at 17:30, Nenad Propadovic wrote:
> Hello pythonistas,
> I'm about to write an application related to injury prevention in big
> industrial companies. It's supposed to be intranet-based (because the admins
> are reluctant to install it on x (where x>>1) clients). So: the client is a
> web-browser. I still don't know whether I will get a server on my own, or
> I'll have to fit into an unknown web server environment, both options being
> realistic.
> By now it seems the logic of the application will be simple. It's going to
> be a relational database with several tables, a modest number of users with
> write priviledges, some more with read priviledges only. Performance is not
> a big issue.
> The GUI has a part I consider complex: I have to present some statistics in
> form of charts or graphs.

I would encourage you to consider the graph-creation process to be
separate from the web interface.  You can program and debug this outside
of a web environment (e.g., just creating graphics files in a
directory), and then plug it into whatever environment you want -- CGI,
Webware, Skunkweb, etc.  Skunkweb isn't Windows-compatible, however,
which may be a concern.  CGI is slow, but easy to install.  Webware
should install anywhere, but may take a bit more server setup -- if
possible I would recommend using Apache, but it doesn't rule out IIS. 
There's a bunch of other frameworks -- Quixote, Albatross, jonpy... I
can't remember them all, but they are all actually fairly similar in
scope.  I would *not* recommend Zope, which is a content management
application and framework, not a good web development framework.

CGI will be a little harder to program -- be sure to use cgitb if you go
that route -- but the application you describe sounds fairly
straight-forward.

For the graphics, you may want Reportlab for nice reports, or another
graphic library if you want to inline the graphs in HTML pages.  I don't
believe PIL does what you want -- there are wrappers for gnuplot, and
there's a *ton* of other graphing packages (I think the scientists tend
to use lots of graphs and lots of Python).  Search Parnasus.

  Ian





From martin at v.loewis.de  Wed Sep 18 00:47:43 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 18 Sep 2002 06:47:43 +0200
Subject: How to make a C extension module backwards compatible?
References: <mailman.1032319522.8608.python-list@python.org>
Message-ID: <m3sn07hpjk.fsf@mira.informatik.hu-berlin.de>

Andrew McNamara <andrewm at object-craft.com.au> writes:

> Is there any community wisdom on how to make a C extension module that
> implements a new-style class (type?) backwards compatible (albeit,
> at the cost of some functionality)?
> 
> Put another way, how *should* I make a class implemented in C behave
> like a new-style class when installed with python 2.2 and above, and
> like an old style class when installed with older pythons?

What properties of a new-style class are you interested in? If you
implement a plain type, it will work on all Python versions, no need
to change anything for Python 2.2.

Regards,
Martin



From peter at engcorp.com  Tue Sep  3 23:58:39 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 03 Sep 2002 23:58:39 -0400
Subject: default tuple unpacking?
References: <slrnanaknb.1gh.huaiyu@gauss.almadan.ibm.com> <3d758180$1@news.sentex.net>
Message-ID: <3d7584d2$1@news.sentex.net>

Peter Hansen wrote:
> Huaiyu Zhu wrote:
> 
>> It is often very convenient to have default arguments such as
>>
>>     def func(a, b, *c): ...
>>
>> Is there a way to use this in statements too?  Examples:
>>
>>     a, (b, *c), *d = e

Hmm... I think there might be an inconsistency or an unexpected
result in this.  What if e already contains a tuple?

   e = (1, 2, 3)
   *a = e        # a is ((1, 2, 3), )

That seems sensible, and it's just what happens if you pass
"e" in to a function defined as "def func(*a)".

Now what about this:

   a, *b = e     # a is 1, b is (2, 3)

That's one possible result, and presumably the one you want with
this syntax, but it's not what you'd get with the function syntax:

   def func(a, *b):
     print a
     print b

   func(e)

Here of course you get (1, 2, 3) in "a" and () in b.  To get the
same as the proposed syntax above, you can't pass in a tuple, but
have to pass the parameters separately:

   func(1, 2, 3)  # now a is 1, b is (2, 3)

The problem is that in function calling, you can pass in a single
tuple, which is one argument, or you can pass the elements separately
as multiple arguments, which is when the special * syntax kicks in.

When doing assignment with tuples you don't have the same two options:

   a, *b = 1, 2, 3

is the same as

   a, *b = (1, 2, 3)

So maybe this won't fly.  Comments?

-Peter



From jan.verstrepen at pandora.be  Thu Sep  5 17:06:08 2002
From: jan.verstrepen at pandora.be (jan verstrepen)
Date: 5 Sep 2002 14:06:08 -0700
Subject: deeply embedded python sources
Message-ID: <5f0846a1.0209051306.2ea5a986@posting.google.com>

Hi,

It seems that the link to the "deeply embedded python" site is dead.
Is it possible that someone has already downloaded those sources
before the link died.

I have already shrinked down a 2.0 version (so not pippy) by leaving
out floating, long, complex, parser/compiler, docs but now I'm very
interested in the modifications the deeply embedded project made to
the sources itself.

Or has anyone maybe one or two tips...

thanks
Jan


From wlfraed at ix.netcom.com  Mon Sep 23 20:46:08 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Mon, 23 Sep 2002 17:46:08 -0700
Subject: e vs exp()? / financial applications
References: <mailman.1032801271.6597.python-list@python.org>
Message-ID: <gkcoma.38e.ln@ix.netcom.com>

Andrew Thompson fed this fish to the penguins on Monday 23 September 
2002 10:15 am:

> 
> 1+w+w+w+w
> 
> and
> 
> w+w+w+w+1
> 
> which are both *different* according to the == operator, but for the
> purposes of numerical accuracy are identical.
> 
> This behaviour is quite mild, but it shows the point that rounding
> errors can easily creep in, and using == to test for equality is not
> always sensible.
>

        Don't they teach anything about floating point in college anymore?

        I'm pretty sure it came up by my second year "advanced FORTRAN" class 
that one should not rely upon testing for equality between any two 
floating point values, but instead should test for a difference less 
than some predetermined epsilon...

        abs(f1 - f2) < epsilon

-- 
--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From debl2nononospmmmy at bellatlantic.net  Sat Sep 21 19:36:34 2002
From: debl2nononospmmmy at bellatlantic.net (David Lees)
Date: Sat, 21 Sep 2002 23:36:34 GMT
Subject: Math optimization
References: <6T5j9.145247$pX1.5195840@news2.tin.it>
Message-ID: <3D8D02A0.FF572334@bellatlantic.net>

This depends on your hardware, OS and probably Python version. Further
you need to specify the data type.  Integer, floating point or infinite
precision integers.  It is quite easy to run benchmarks, which you
should do yourself.  

David Lees


Zed wrote:
> 
> Hi!
> What's faster in python? Divide or multiply?
> (a= b*c  is faster than   a=b/c)?
> 
> Math expressions with constants are calcolated during .pyc creation or real
> time?
> (a=b*5   is faster than    a=b*5*2/2 )?
> 
> Thanks


From marklists at mceahern.com  Fri Sep 13 11:35:26 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 13 Sep 2002 10:35:26 -0500
Subject: Restricting methods in derived classes
In-Reply-To: <JHEOKEOOLIGLDHCMAHMOGEENDOAA.marklists@mceahern.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOEEIFDOAA.marklists@mceahern.com>

Huaiyu,

The problem with my previously posted solution is that I was trying to
rewrite the class' methods at class __init__ time when I needed to be doing
that at __new__.  Here's a working model.  As usual, the names I pick are
probably plain goofy.  I'd be interested to know whehther this is what you
were looking for.

Cheers,

// mark

#!/usr/bin/env python

import inspect
import unittest

def make_unwanted(cls, method_name):
    def unwanted(self, *args, **kwargs):
        msg = "'%s' object has no attribute '%s'" % (cls.__name__,
method_name)
        raise AttributeError(msg)
    return unwanted

def is_unwanted(member, wanted_methods):
    return not member[0].startswith('__') and member[1] not in
wanted_methods

class Restricted(type):

    def __new__(cls, classname, bases, classdict):
        wanted_method_key = 'wanted_methods'
        # Don't do anything if we don't find the list of wanted methods.
        if wanted_method_key not in classdict:
            return type.__new__(cls, classname, bases, classdict)
        # Hide any method from bases not explicitly listed in
wanted_methods.
        wanted_methods = classdict[wanted_method_key]
        for base in bases:
            for method in inspect.getmembers(base,
inspect.ismethoddescriptor):
                if is_unwanted(method, wanted_methods):
                    classdict[method[0]] = make_unwanted(cls, method[0])
        return type.__new__(cls, classname, bases, classdict)

class RestrictedDict(dict):

    __metaclass__ = Restricted

    wanted_methods = [dict.update]

class tests(unittest.TestCase):

    def test_update(self):
        """Wanted methods should be accessible."""
        d = RestrictedDict()
        d['a'] = 1
        expected = {'a': 1}
        self.assertEquals(expected, d)
        d.update({'a': 2, 'b': 3})
        expected = {'a': 2, 'b': 3}
        self.assertEquals(expected, d)

    def test_error(self):
        """Unwanted methods should raise AttributeError."""
        d = RestrictedDict()
        d['a'] = 1
        self.assertRaises(AttributeError, d.copy)

    def test_using_base(self):
        """We should still be able to access base methods if we really want
to.

        <wink>
        """
        d = RestrictedDict()
        d['a'] = 1

        hidden_method_name = 'copy'
        for b in d.__class__.__bases__:
            if hasattr(b, hidden_method_name):
                copy_method = getattr(b, hidden_method_name)
                break
        d2 = copy_method(d)
        self.assertEquals(d2, d)

if __name__ == "__main__":
    unittest.main()




From writeson at earthlink.net  Mon Sep  2 13:32:51 2002
From: writeson at earthlink.net (Doug Farrell)
Date: 2 Sep 2002 10:32:51 -0700
Subject: RedHat 7.3 and Python2.2.1
Message-ID: <88bc63c6.0209020932.54723791@posting.google.com>

Hi,

I'm trying to install the latest version of Python
(python2-2.2.1-1.i386.rpm) from the www.python.org download page, and
I'm having problems. When I run

rpm -i python2-2.2.1-1.i386.rpm

I get all kinds of conflict messages like this:

file /usr/bin/pydoc2 from install of python2-2.2.1-1 conflicts with
file from package python2-2.2-16

probably hundreds of them in fact. Can anyone tell me what is going on
here, as the www.python.org site says the RPM I'm using is for RH7.3,
which is what I'm using. If you can tell me what I'm doing wrong and
how to address this, that would be great too! My ultimate goal is to
also install wxPython for the python2.2.1 installation, so if anyone
has some 'heads up' on that, I'd appreciate it.

Thanks in advance,
Doug


From syver-en+usenet at online.no  Sun Sep  8 11:52:09 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Sun, 08 Sep 2002 15:52:09 GMT
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com>
Message-ID: <uy9acbht9.fsf@online.no>

steve at ferg.org (Stephen Ferg) writes:

> "Joseph A. Knapka" <jknapka at earthlink.net> wrote in message
> news:<3D798461.B3FE0E06 at earthlink.net>...

> > Also, you should be aware of the "command" option
> 
> I know about it, but I think that for beginners, it is very
> confusing.

I think you are right here Stephen. Only learning the command option
is a bit like using a wizard to generate code, you can get quite far
with it but you haven't really understood what you are doing. bind is
much more general and let's the student know what he/she is doing.



-- 

Vennlig hilsen 

Syver Enstad


From danielnuriyev at yahoo.com  Tue Sep 10 19:32:11 2002
From: danielnuriyev at yahoo.com (Daniel Nuriyev)
Date: 10 Sep 2002 16:32:11 -0700
Subject: Python GUI
References: <3d8d3c0.0208262339.e7bc02c@posting.google.com> <uniomu8pdi7tm169b4g6225d5oqdrkj2go@4ax.com> <8_8b9.3412$pw.4534678@ruti.visi.com>
Message-ID: <3d8d3c0.0209101532.6fe25365@posting.google.com>

Gentlemen, some of you have misunderstood what I wrote against using
the native widgets.
What I meant was not to use instead of the native widgets the least
common denominator for all platforms
but to use one basic graphical object of no matter which origin by
means of which to simulate
look&feels you like. This is what I meant when I compared this method
with Java Swing.
Surely it envolves a lot of work for those who will develop such
packages. But in my opinion it is better
than to call directly the native widgets. I apologize for being lazy
to go into the explanation of why.
But if you're interested in the comparoson of the two approaches, you
can have a look at the advantages
of Java Swing vs Java AWT.
Another point I would like to make clear is that what I meant by my
message was the general issue
of a general idea of a general GUI package for Python. But of course
each one of us has a concrete package we like and
it is absolutely not compulsory to use something generic. I, for
instance, use Tk for quickies. I don't
like it but it's always there. When I need something nice I look at
the needs of a concrete project and
make choices not necessarily based on my personal taste.


From steven_shaw at adc.com  Mon Sep  9 23:23:55 2002
From: steven_shaw at adc.com (Steven Shaw)
Date: 9 Sep 2002 20:23:55 -0700
Subject: dynamism
Message-ID: <66715c02.0209091923.265fe77e@posting.google.com>

I've been away from Python for awhile and I'm wondering whether Python got
less "dynamic" while I was away.

I thought you could just extend an object's attributes at run-time like this:

>>> p = object()
>>> p.x = 1
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: 'object' object has no attribute 'x'

I thought that you could access dictionary values with object syntax:

>>> x = { 'a': 1, 'b':2 }
>>> x
{'a': 1, 'b': 2}
>>> x.a
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: 'dict' object has no attribute 'a'

Maybe you could never do this as shown but I'm sure there were ways of getting
this effect. Anyone care to remind me?

Cheers, Steve.


From shalehperry at attbi.com  Thu Sep 12 15:05:49 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 12 Sep 2002 12:05:49 -0700
Subject: [newbie] Buying an introductory/reference text
In-Reply-To: <Mu5g9.1702$U_.1021@news04.bloor.is.net.cable.rogers.com>
References: <Mu5g9.1702$U_.1021@news04.bloor.is.net.cable.rogers.com>
Message-ID: <200209121205.49267.shalehperry@attbi.com>

On Thursday 12 September 2002 11:59, Brad Fonseca wrote:
> Greetings!
>
> I'm looking for a few opinions on the above subject.  I'm a computer
> science student who has had some introductory lectures on Python and I've
> done some very basic mucking around with.  I'm looking for a clear
> reference and/or introductory text.  Any suggestions would be nice.  I have
> looked at http://www.python.org/cgi-bin/moinmoin/IntroductoryBooks but
> there appears to be a number of good books to choose from there.
>
> Regards,

I believe that "Python: Essential Reference" is about all you need if you have 
programmed in at least one other language.  Top notch book.



From erikprice at mac.com  Wed Sep  4 10:17:26 2002
From: erikprice at mac.com (Erik Price)
Date: Wed, 4 Sep 2002 10:17:26 -0400
Subject: Why Python?
In-Reply-To: <al5332$fm$1@news.tudelft.nl>
Message-ID: <09CB9911-C011-11D6-BB50-00039351FE6A@mac.com>

On Wednesday, September 4, 2002, at 09:50  AM, Konrad Wojas wrote:

> I've been doing web development in PHP since 1998 and have written a 
> content
> management system with a quite complex shopping cart system and a
> backoffice system to handle all sales. It's over 10000 lines of PHP 
> code,
> excluding the html code.

I learned to program from PHP, and wrote an application that was 
comparable (somewhere over 10,000 lines of code).  It didn't deal with 
e-commerce but was essentially an extensive database front-end, and 
included shopping cart-like "features".

> I've really had it with PHP, it's unmaintainable, ugly and slow. 
> Errors like
> 'parse error on line 487' are not fun to work with. I really like 
> python.
> The systax is a lot clearer, exceptions are great and I really love 
> those
> stacktraces if an exception is not handles. Object oriented 
> programming in
> PHP is a joke, it really sucks. And PHP is buggy. I've encountered 
> numerous
> bugs, especially in the extension modules. XSLT in PHP? Forget it, the
> module will crash in 50% of the cases. And compiling PHP yourself with 
> a
> lot of extension modules can be a real nightmare. And above all: 
> python is
> a generic language, PHP is only useable for web development and small
> maintainance scripts. It doesn't have bindings for anything beyond web
> development.

I think that PHP is prettier than Perl but not nearly as pretty as 
Python.  Its support for objects is adequate for some purposes, but for 
a fully object-oriented application it might not be quite up to snuff.  
If your code is mostly procedural and you use a few objects to make 
your life easier, though, you should be able to manage.

The error messages from PHP were always fine for me, but there's no 
elegant way to handle them like in Python where you can catch 
exceptions, etc.

PHP does have a few kinks, but the bugs are esoteric.  It is a pretty 
well-developed language IMO.  One of the biggest differences between 
PHP and Python is that in PHP most of the functions are built into the 
global namespace.  I know that some people in Python have a lot of 
problems with that, but PHP is not a general-purpose language and so 
doesn't use modules like Python and Perl -- having these functions 
available is actually a blessing.  But in a language like Python or 
Java I prefer the modular approach.

PHP also has extremely good interconnectivity with the MySQL database, 
so many people who use MySQL see PHP as the web-language of choice.  
PHP/MySQL tutorials and books abound, which I think is interesting 
because there are so few jobs (in New England, US at least) that 
actually want PHP programmers.

> I'm still maintaining the PHP code I have written, but I use python 
> for all
> new projects. Don't start big projects in PHP, you'll regret it ;-)

I would go back to PHP for some simple dynamic web page scripts, but 
not much more.  I agree with the above -- if you're going to write 
full-blown applications, you will probably like the OO facilities of 
Python (or Java) better.


Erik



--
Erik Price

email: erikprice at mac.com
jabber: erikprice at jabber.org




From jtk at yahoo.com  Fri Sep  6 23:21:27 2002
From: jtk at yahoo.com (Jeff Kowalczyk)
Date: Fri, 6 Sep 2002 23:21:27 -0400
Subject: Sequence or Identity in gadfly
Message-ID: <albrbp$rej$1@main.gmane.org>

I don't expect that it does (it would make more sense to do it in your
python code), but is there a Sequence type in gadfly so that I can make an
autonumber for ID columns similar this postgres code?:

CREATE SEQUENCE id
start 1000 increment 1 maxvalue 100000

SELECT nextval ('id');

Or some other construct for creating an identity value?








From gumuz at looze.net  Thu Sep  5 09:19:41 2002
From: gumuz at looze.net (Gumuz)
Date: Thu, 5 Sep 2002 15:19:41 +0200
Subject: encryption (passwords)
References: <3d75d115$0$221$4d4ebb8e@news.nl.uu.net> <7xd6ru2gfi.fsf@ruckus.brouhaha.com> <3d75f0f0$0$230$4d4ebb8e@news.nl.uu.net> <7xsn0pzcol.fsf@ruckus.brouhaha.com>
Message-ID: <3d775896$0$231$4d4ebb8e@news.nl.uu.net>

actually, i found a very good example for this at:
http://www.nyx.net/~awestrop/crypt/dh2.htm
it's a java applet

only thing i'm still puzzling with is how to apply this properly

"Paul Rubin" <phr-n2002b at NOSPAMnightsong.com> wrote in message
news:7xsn0pzcol.fsf at ruckus.brouhaha.com...
> "Gumuz" <gumuz at looze.net> writes:
> > although this is just a play-learn-project and security is not really an
> > issue, i'd like to include encryption in it as well to get experience in
> > this, too.
>
> If you want to learn programming cryptography, the book you want is
> "Applied Cryptography" by Bruce Schneier.  You might like to implement
> everything yourself rather than using a library.
>
> > Actually, I am trying to create a sort of simple instant messenger
> > server+client.
>
> I think this should use Diffie-Hellman key exchange to choose the
> decryption keys.  DH is a way for two people to agree on a key based
> on random parameters they each choose.  The traditional version goes:
>
>    Alice and Bob agree beforehand on public parameters g and P,
>    where P is a big prime number (like 300 digits) with certain
>    properties and g is a generator of Z//p.  If you don't understand
>    that, don't worry too much--there are standard values of g and P
>    that you can use.
>
>    To start talking to each other, Alice chooses a secret random number x,
>    and Bob chooses a secret random number y.  They do not reveal these
>    numbers to anyone.  Alice instead computes the number X = g**x mod P
>    and sends X to Bob.  Bob computes Y = g**y mod P and sends Y to Alice.
>    Computing these modular exponentials is trivial in python because
>    of Python's built-in long integers and its 3-argument pow function.
>    You can just say Y = pow(g,y,P).
>
>    Since Alice receives Y from Bob and knows x already, she can compute
>    K = Y**x mod P.  Notice Y**x mod P == (g**y)**x mod P = g**(yx) mod P.
>    Bob likewise can compute K = X**y mod P = (g**x)**y mod P = g**(xy) mod
P.
>    Since multiplication is commutative, xy==yx so both have found the same
K.
>    The coolness here is the computation required knowing at least one of
>    the secrets, x or y.  An eavesdropper knowing neither secret doesn't
>    have any easy way to find K.
>
>    At the end of the conversation, Alice and Bob should both erase their
>    secret values x, y, and K from computer memory.  That means the
>    conversation can never be recovered by a third party, even by forcing
>    Alice and Bob to reveal their passwords, turn over their computers,
etc.
>    The keys are gone forever, like burning a document.
>
> Note you will also have to protect against "man in the middle"
> attacks, and deal with some other subtleties, to make DH secure--the
> description above is just to sketch the process.  See Applied
> Cryptography for more info.




From imbosol at vt.edu  Mon Sep  2 19:34:29 2002
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 2 Sep 2002 19:34:29 -0400
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <mailman.1030907703.26266.python-list@python.org> <akuifb$h0e$2@solaris.cc.vt.edu> <1030962064.628232@ampungk.ozonline.com.au> <al0djh$a33$1@solaris.cc.vt.edu> <1031008966.343893@ampungk.ozonline.com.au>
Message-ID: <al0tkb$3nh$1@solaris.cc.vt.edu>

John La Rooy wrote:
> Carl Banks wrote:
> 
>>>"anything,anything,anything,same_as_3rd,same_as_2nd,same_as_1st"
>>>or would you like to suggest something else?
>> 
>> 
>> How about:
>> 
>> pattern = Group(Any()) + Group(Any()) + Group(Any()) \
>>           + GroupRef(3) + GroupRef(2) + GroupRef(1)
>> 
> Err symantically that's exactly the same as the re and my suggestion
> only the syntax is different. It's still nothing like saying
> 
> pattern = "6 character palindrome"


Oh.  Well, methinks you don't give humans enough credit.  I'm a human,
and I can read the verbose regexp.  Of course, humans can read regular
expressions, too.  Maybe I don't give humans enough credit, either.
:-)

The point is, I was arguing for less-cryptic, more-verbose regular
expressions as a way to make complicated patterns more transparent.  I
certainly wasn't arguing for "6 character palindrome".


Sorry-for-the-confusion-ly y'rs


-- 
CARL BANKS
http://www.aerojockey.com


From grayshade at libero.it  Wed Sep 18 11:37:18 2002
From: grayshade at libero.it (Domenico)
Date: 18 Sep 2002 08:37:18 -0700
Subject: Python doesn't suck!
Message-ID: <e34e8ce6.0209180737.49868d9f@posting.google.com>

http://www.theperlreview.com/at_a_glance.html

According to the most accurate algorithm, Python is the second
less-sucking language.


From theller at python.net  Mon Sep  9 12:12:30 2002
From: theller at python.net (Thomas Heller)
Date: Mon, 9 Sep 2002 18:12:30 +0200
Subject: win32com.client.gencache and py2exe
References: <uit1ftba1.fsf@online.no>
Message-ID: <alih9e$1qjt12$1@ID-59885.news.dfncis.de>

"Syver Enstad" <syver-en+usenet at online.no> wrote in message news:uit1ftba1.fsf at online.no...
>
> When using py2exe to make package a python program that uses gencache
> to generate wrapper files from COM type libraries, the resulting
> installation fails because the win32com system is unable to find the
> generated files. Is there any workaround to this problem?z
>

It's already fixed in CVS. It should be sufficient if you replace
the existing build_exe.py by the new one which you can grab
from the CVS repository:
http://cvs.sourceforge.net/cgi-in/viewcvs.cgi/py2exe/py2exe/py2exe/build_exe.py
(the server seems down at the moment).

There's a new command line flag, which you must use to specify
the progids you use, like "--progids Excel.Application".

HTH,

Thomas




From Padraig at Linux.ie  Wed Sep 11 07:35:11 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Wed, 11 Sep 2002 11:35:11 GMT
Subject: regular expression for space seperated quoted string
References: <PqFf9.3627$cP3.7118@news.iol.ie>
Message-ID: <PTFf9.3628$cP3.7117@news.iol.ie>

Padraig Brady wrote:
> Hi, I'm trying to split a string that is seperated
> by spaces and also contains double quoted words which
> can contain spaces:
> 
> For e.g. using: re.split('[ ]*"?([^"]*)"?[ ]*', s1)
> on this string: s1='1 "2" "thre e"'
> gives:          ['', '1 ', '', '2', '', 'thre e', '']
> 
> Problem with this is the '' entries, but this isn't too bad.
> 
> However using the above re with: s2='1 2 "th ree"'
> I get:                           ['', '1 2 ', '', 'th ree', '']
> 
> any ideas?

Well this one does better:
re.split('[ ]+|"([^"]*)"', string)

It does leave '' and None entries in the list,
but I can remove these easily.

I presume there is something more elegant.

P?draig.



From eric.brunel at pragmadev.com  Mon Sep 23 04:21:07 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 23 Sep 2002 10:21:07 +0200
Subject: Tk Getting a directory name
References: <amkbso$hre$1@newsg2.svr.pol.co.uk>
Message-ID: <ammieg$o6b$1@wanadoo.fr>

Ken Payne wrote:

> HI everyone -
> 
> I'm writing a Tk interface to an app and in it I want the user to choose a
> directory. There's an easy way to ask the user to pick a file -
> 
> fileWanted=askopenfilename()
> 
> but is there a similar way to ask for a directory?
> 
> - Ken

http://groups.google.com/groups?hl=en&lr=&ie=ISO-8859-1&q=tkinter+select+directory&meta=group%3Dcomp.lang.python.*
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From peter at engcorp.com  Thu Sep 12 00:27:43 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 12 Sep 2002 00:27:43 -0400
Subject: I must be an idiot, please pity me!
References: <mailman.1031782274.18676.python-list@python.org> <oPPf9.2866$XE6.173592@newsfep1-win.server.ntli.net>
Message-ID: <3d8017ac@news.sentex.net>

newbie wrote:
 > On Wed, 11 Sep 2002 18:10:42 +0100, Mark McEahern wrote:
 >
 >>[newbie]
 >>
 >>>But it doesn't seem to work like that.
 >>
 >>What are the symptoms of failure?
 >
> Perhaps I wasn't clear, but I don't believe I mentioned anything not
> functioning properly. 

I believe Mark read your above comment the same way I did, which
was "it doesn't seem to work [when configured] like that", as opposed
to what you appear to have meant which would be "it doesn't seem
to work in that manner".  A simple misunderstanding perhaps. :)

-Peter



From robin.siebler at corp.palm.com  Mon Sep 16 16:08:44 2002
From: robin.siebler at corp.palm.com (Robin Siebler)
Date: 16 Sep 2002 13:08:44 -0700
Subject: Difflib question
Message-ID: <886c5e4b.0209161208.25012958@posting.google.com>

Is there a way to get difflib to output *only* the lines that are
different (instead of all of the lines)?


From sholden at holdenweb.com  Thu Sep  5 14:04:21 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 5 Sep 2002 14:04:21 -0400
Subject: Maybe it's a mod_python issue I'm hitting?
References: <3D6E6BDA.7337426C@bullfrog-tech.com> <3D6E7022.3DE649D4@bullfrog-tech.com> <3D6E82E3.866CBA49@bullfrog-tech.com>
Message-ID: <KxRd9.5584$%M3.3847@atlpnn01.usenetserver.com>

"Manus Hand" <manus at bullfrog-tech.com> wrote in message
news:3D6E82E3.866CBA49 at bullfrog-tech.com...
> Another follow-up....
>
> It seems that all my code of the form:
>     vars(self).update(someDict)
> still DOES work fine under 2.2.  (Yay!  Whew!)
>
It's not guaranteed to work in later versions, though, I suspect.

> The one thing that STOPPED working reliably is this:
>
> locals()[varName] = someValue
>
Don't do that (see below).

> My initial thought was that PEP 227 had made locals() read-only, hence my
post
> at the start of this thread.  However, I soon realized that this line
worked
> fine
> under 2.2 while it sat at global scope.  The line started failing when I
put it
>
> into a function (specifically, a handler() function, as part of a port to
a
> mod_python installation).
>
Erm, you should NOT treat the locals dictionary as modifiable, whether
running under mod_python or not. The locals() function returns a dictionary
(mapping, anyway), but for a long time now the documentation has said you
should not (necessarily) expect changes to this mapping object to result in
changes to the local name space.

[other confusions]

>From the library manual (section 2.1, Built-In Functions):
"""
locals()
Return a dictionary representing the current local symbol table. Warning:
The contents of this dictionary should not be modified; changes may not
affect the values of local variables used by the interpreter.
"""

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                        pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From kseehof at neuralintegrator.com  Wed Sep  4 15:40:47 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Wed, 4 Sep 2002 12:40:47 -0700
Subject: Could Python supplant Java?
In-Reply-To: <3D64C3C5.7000502@softhome.net>
Message-ID: <EFEFIHAINIINNELLHFAICEHCDFAA.kseehof@neuralintegrator.com>

Nick Ruisi wrote:

> .net compiled code can run on linux with the mono CLR
> (http://www.go-mono.com). One can use the mono tool kit to write /
> compile .Net code on a linux box, without ever touching a microsoft
> product.  The compiler and SDK for C# and VB.Net is a free (albeit
> large) download. Although I don't like VB.Net as a language, C# is fine.
> It's just like java.

C# and .Net are transparent ploys by Microsoft.  This isn't paranoia, or
fashionable "Microsoft bashing", it's just plain experience.  The only
possible motivation for Microsoft to market a product that increases
compatibility across platforms, languages and applications is to be in
a position to severely damage that compatibility at some point in the
future.  Anyone who has been in the computer industry for any length of
time should see that this is obvious.  Open Source, open architecture,
and open anything, are extremely harmful to Microsoft's well being, and
they know it.  They aren't evil; they are just trying to survive.

I expect .Net to provide what it promises in the short run, at least
to some degree, in order to gain some degree of credibility.  Hence,
it works fine on Linux for now.  Once people are hooked, you can expect
microsoftisms to creep in for no reason other than to lock out competition.

If this isn't obvious to you, please read Microsoft's own internal
documentation as leaked in the "Halloween Documents".  And don't waste
space on this thread arguing with me until you have at least read part 1.
Especially be sure to read about "decommoditizing protocols".

"OSS [open source software] projects have been able to gain a foothold in
many server application because of the wide utility of highly commoditized,
simple protocols. By extending these protocols and developing new protocols,
we can deny OSS projects entry into the market." - Microsoft internal memo

  http://www.opensource.org/halloween/

- Ken Seehof

> Rex Ballard wrote:
>
> > netvegetable wrote:
> >
> >
> >>Not offering any personal opinion on this conjecture, but a lot of
> >>people are saying that java has failed to become a widely accepted
> >>cross platform language for applications.
> >>
> >
> > Java is still quite popular, and has established a pretty
> > substantial market share.  Microsoft is fighting Java and trying to
> > Undermine Java 2 by bundling an older version of Microsoft's
> > insecure JVM, but getting JRE is very easy, and even JDK is pretty
> > easy to get if you want it.  And since JDK 1.3 have some execellent
> > IDEs available, on both Linux and Windows, it's pretty easy to
> > implement applications that work on Both Linux and Windows.
> >
> >
> >>These leaves the world defenseless against the Microsoft Borg.
> >>
> >
> > PLEASE!!!  - Microsoft has no really good general purpose scripting
> > languages, and no low-cost development languages.  PERL, PYTHON,
> > JAVA, and TCL can all be downloaded for free.  Microsoft obviously
> > doesn't pay $4 billion/year to promote these Open Source based
> > languages which can create products and projects that can run on
> > both Linux and Windows.
> >
> > Of COURSE Microsoft is going to spend their $Billions promoting C#
> > and Visual Basic, which run Exclusively on Windows and do not
> > support comprehensive implementations on Linux.
> >
> > Of course, if you want some heavy-duty portability, you have
> > cygwin, which give you the ability to run Linux programs on
> > Windows, including Windows NT, 2000, and XP.  Cygwin isn't as
> > secure, stable, and fast as Linux (after all you are ultimately
> > dependent on Windows infrastructure), but it's a nice way to make
> > the transition.
> >
> >
> >>But could Python do the trick? The python interpreter is smaller
> >>than the JRE, and it's certainly a nicely structured language,
> >>with nearly all the coding features of Java.
> >>
> >
> > Unfortunately, Python, PERL, TCL/TK, Ruby, and the other fine
> > languages all have to be loaded AFTER the machines are shipped.
> > Obviously, it would be nice if OEMs could install PERL, Java 2,
> > Python, and others as part of their OEM installation, when it can
> > be put on the master disk.  It would cost nothing, or extremely
> > little, and would provide new capabilities not currently available
> > to Windows.  It would also make hundreds of applications available
> > as well.
> >
> > Unfortunately, Microsoft has insisted on complete control of the
> > preinstalled suite and is doing everything they possibly can to
> > maintain that control despite the unanimous ruling of the Appeals
> > court full panel that such exclusion of competitors is explicitly
> > illegal.
> >
> > Sun is likely to be the strongest candidate to be prepared to fight
> > this or have Microsoft cited for contempt.  Sun could offer Java
> > JRE and/or JDK, Star Office, and other applications which could be
> > installed by OEMs for prices that are very competitive with
> > Microsoft's office, JDK, and IE/VBScript.
> >
> > Red Hat could also press for the inclusion of Cygwin, in this case
> > possibly even at no cost to OEMs.  Corporate customers could also
> > begin using Cygwin - including Python, Perl, and TCL/TK.
> >
> > These are "first steps" - steps that would create an application
> > base that could lead to a marketplace in which both Windows and
> > Linux compete in an open and equally accessible market.
> >
> > For this reason, Microsoft will do everything it can to fight this
> > technology, since it cannot be controlled.  Microsoft has even
> > tried to force Corporations to agree to NOT use Open Source
> > technology such as Python on their Windows desktops.  The DOJ has
> > ignored this, the Courts have ignored this, and Microsoft has tried
> > to divert attention by releasing a support pack that allows users
> > to choose which web browser they want to use.
> >
> > As long as Microsoft controls access to the OEM distribution
> > channel, and has the ability to control what OS, Applications, and
> > systems can be installed on PCs, they will do everything in their
> > power to "Lock out" any technology which can be supported by both
> > Linux and Windows.
> >
> >
> >>--
> >>netvegetable at excite.com
> >>
> >
>
>
>
> -----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
> http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
> -----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
> --
> http://mail.python.org/mailman/listinfo/python-list
>




From kmp at NO2SPAMatrium.fsnet.co.uk  Mon Sep 23 13:42:22 2002
From: kmp at NO2SPAMatrium.fsnet.co.uk (Ken Payne)
Date: Mon, 23 Sep 2002 18:42:22 +0100
Subject: Tk Getting a directory name
References: <amkbso$hre$1@newsg2.svr.pol.co.uk> <ammieg$o6b$1@wanadoo.fr>
Message-ID: <amnjgh$578$1@newsg2.svr.pol.co.uk>

Eric -

Thanks. Just what I needed.

Best wishes

- Ken
Eric Brunel wrote:

> Ken Payne wrote:
> 
>> HI everyone -
>> 
>> I'm writing a Tk interface to an app and in it I want the user to choose
>> a directory. There's an easy way to ask the user to pick a file -
>> 
>> fileWanted=askopenfilename()
>> 
>> but is there a similar way to ask for a directory?
>> 
>> - Ken
> 
> 
http://groups.google.com/groups?hl=en&lr=&ie=ISO-8859-1&q=tkinter+select+directory&meta=group%3Dcomp.lang.python.*



From sholden at holdenweb.com  Fri Sep 27 17:16:01 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 27 Sep 2002 17:16:01 -0400
Subject: Local Python Server Installation
References: <3d8b47c6.277669@news.jaring.my> <slrnaomn6m.39s.gerhard.haering@haering.opus-gmbh.net> <3d8c993c.2190426@news.jaring.my> <mailman.1032625889.13783.python-list@python.org> <3d8d4009.375344@news.jaring.my> <bQYk9.1630$6g7.1090@fe03> <3d947c26.2656654@news.jaring.my>
Message-ID: <JR3l9.2341$79.998@fe03>

"Kenneth Gomez" <kennethg at pd.jaring.my> wrote ...
> Steve,
>
> I stumbled onto this problem while going through the first exmaple in
> the book Python:How to Progra, by Deitel, Deitel, Liperi and
> Wiedermann.
>
> My little code sniipet, copied exactly from the book except for the
> path to my python exe is :
>
> ________________________________________________
>
> #!C:\prog\Python22\python.exe
>
> import time
>
> def printheader(title):
> print """Content-type: text/html
>
> <?xml version = "1.0" encoding = "UTF-8"?>
> <!DOCTYPE html PUBLIC
> "-//W3C//DTD XHTML 1.0 Strict//EN"
> "DTD/xhtml1-strict.dtd">
> <html xmlns = "http://www.w3.org/1999/xhtml">
> <head><title>%s</title></head>
>
> <body>""" % title
>
> printheader("Current Date & Time")
> print time.ctime(time.time())
> print "</body></html>"
> ___________________________________________________
>
This looks OK given the mailer mangling that's (presumably) lost some tabs
from the code.

>
> I get the following error in IE6 :
>
> The XML page cannot be displayed
> Cannot view XML input using style sheet. Please correct the error and
> then click the Refresh button, or try again later.
>
Interesting that it even recognises XML given the Content-Type header and
the fact that the program doesn't seem to be running.
>
> --------------------------------------------------------------------------
------
>
> Invalid at the top level of the document. Error processing resource
> 'http://localhost/ken.py'. Line 1, Position 1
>
> #!C:\prog\Python22\python.exe
> ^
>
So, basically, IE6 is also seeing a listing of the script. This isn't
surprising: the server behavior is unlikely to change with a change of
client.
>
> and when using Mozilla 1.0, all I get is the listing of the script.
>
> Bummer.
>
> Any advise?
>
Don't know. I presume you have already made

    http://httpd.apache.org/docs-2.0/platform/windows.html

your bible? It does tell you almost all of what you need to know.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From ken at hotmail.com  Fri Sep 20 23:07:28 2002
From: ken at hotmail.com (Ken)
Date: Sat, 21 Sep 2002 13:07:28 +1000
Subject: Python about file I/O
Message-ID: <amgnnl$5qc9n$1@ID-49758.news.dfncis.de>

Hi, I have 2 questions on IO, not necessary related to one another:

1.) How do you list the filenames within a given directory so user can
choose which file to open?

2.) How do you ensure the program doesn't crash if I specify a filename to
open but the file doesn't exist? (I.E. how to check if file exist before
opening the file?)

Thanks




From pyth at devel.trillke.net  Sun Sep  1 18:56:27 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 2 Sep 2002 00:56:27 +0200
Subject: Python IPC options
In-Reply-To: <3D727AFA.1FE19151@bio.gla.ac.uk>; from F.L.Burton@bio.gla.ac.uk on Sun, Sep 01, 2002 at 09:39:22PM +0100
References: <1030829572.943495@irys.nyx.net> <ddc19db7.0209010836.50242309@posting.google.com> <3D727AFA.1FE19151@bio.gla.ac.uk>
Message-ID: <20020902005627.F3893@prim.han.de>

Francis Burton wrote:
> Armin Steinhoff wrote:
> > SPREAD -> http://www.python.org/other/spread/
> 
> This seems to be very close to what I am looking for.
> Thanks for the pointer!
> 
> > Hope this is enough :)
> 
> More than enough. :-) I'll investigate Spread first,
> then maybe delve into some of the others.

Check back if you get stuck.  It took me some
hours to get something meaningful running.

One thing i didn't solve (and this might be a design-problem)
is to dynamically add/remove a participating host/server. 
I had to restart *all* servers with updated configuration files
to get a new server in.  That's a bit 90'

regards,

    holger



From martin at v.loewis.de  Wed Sep 18 00:44:16 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 18 Sep 2002 06:44:16 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid>
Message-ID: <m31y7rj49r.fsf@mira.informatik.hu-berlin.de>

Greg Ewing <see_reply_address at something.invalid> writes:

> Also, I believe that, even on the supported
> architectures, it plays dangerous tricks with
> the C stack. I'd prefer to see that fixed
> before making it a standard part of the
> distribution.

They aren't really dangerous, just unconventional. And that cannot be
fixed, as it is essential to operation of Stackless.

Regards,
Martin



From pyth at devel.trillke.net  Wed Sep 11 19:29:14 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 12 Sep 2002 01:29:14 +0200
Subject: Destructor never called ???
In-Reply-To: <76710511.0209111450.40999dd8@posting.google.com>; from muk_pr@yahoo.com on Wed, Sep 11, 2002 at 03:50:24PM -0700
References: <76710511.0209111450.40999dd8@posting.google.com>
Message-ID: <20020912012914.W3893@prim.han.de>

I presume that you are using CPython (not jpython).

vb wrote:
> Hi, 
>   I have the following code my problem is when I call any methods of  
>        TCVector_Object the destructor never executes.
> 
> I mean
> 
> v = module.CreateTCVector()
> del v;

or maybe you are using JPython :-)

> //destructor runs. message shown .OK
> 
> but 
> in
> 
> v = module.CreateTCVector()
> v.Append(123);
> del v;//or closing python interpreter

At least you are coming from java :-)

getting ontopic, what does your 'Append' method do? It's not the
usual list.append method, right?

> destructor never gets called. Any ideas?

You need to provide *example* code, showing unexpected behaviour.
Tossing in some information about the python version and the operating
system might help, too.  Especially when it's about low-level behaviour.

regards,

    holger



From StRanGy at dijana.vest.hr  Fri Sep  6 12:52:49 2002
From: StRanGy at dijana.vest.hr (Ivica)
Date: 6 Sep 2002 16:52:49 GMT
Subject: Getting a proccesses ID
References: <Qu3e9.188987$_91.214523@rwcrnsc51.ops.asp.att.net>
Message-ID: <slrnanhnb1.ikp.StRanGy@dijana.vest.hr>

On Fri, 06 Sep 2002 15:05:20 GMT, Micah Mayo <astrophels at yahoo.com> wrote:
> Hi,
> 
> I'm writing a little tool for SysAdmins on FreeBSD - I'm trying to make
> it as portable as possible(as far as the unices go), and have come upon
> a problem - one of the things this tool does is restarts various deamons
> on the system, in order to do that I have to send a killsignal to the
> pid. The problem is getting the process id - right now I've got a hack
> using ps, grep, and awk, which I doubt will translate well to other
> flavors of *nix because they all seem to have their own method of
> delivering info via ps. Can anyone tell me a cross platform way of
> grabbing a particular processes ID
> 
> Here's what I'm doing now, by the way:
> 
> pid = os.popen("ps -x | grep sshd | grep -v grep | awk '{ print $1 }'")
> pid = pid.atoi()
> 

Maybe this will help you:

def GetPids(list, procname):
        pidlist = []
        for i in range(len(list)):
                if string.find(list[i], procname) != -1:
                        str = list[i].strip()
                        sp = string.find(str, ' ')
                        pid = str[:sp].strip()
                        pidlist.append(pid)
        mypid = os.getpid()
        if mypid in pidlist:
        pidlist.remove(mypid)
        return pidlist
	
This I use for killing a proccess ny his name and not killing myself if the
process is python

the list argument is a list os proccesses which i got using ...

pidlist = os.popen('ps').read()
pidlist = pidlist.split('\n')


-- 
---------------------------------
# RealName: Ivica Munitic
# E-Mail: imunitic at vest.hr
---------------------------------


From gumuz at looze.net  Tue Sep 24 10:42:11 2002
From: gumuz at looze.net (Gumuz)
Date: Tue, 24 Sep 2002 16:42:11 +0200
Subject: how to build up a guest book with python ?
References: <mailman.1032854050.26863.python-list@python.org> <6FYj9.157588$pX1.5677010@news2.tin.it>
Message-ID: <3d90784e$0$231$4d4ebb8e@news.nl.uu.net>

i might add that this is an exellent book!



"Alex Martelli" <aleax at aleax.it> wrote in message
news:6FYj9.157588$pX1.5677010 at news2.tin.it...
> black wrote:
>
> >     I'm learning python for almost 1 month and decide to build up a
guest
> >     book with it, but I really know nothing about web programming, any
> >     tutorials for me please ?
>
> Steve Holden's "Python Web Programming" (New Riders) should be
> perfect for your needs -- it touches on all the technologies you
> need to be acquainted with for the purpose, and gives a good
> tutorial grounding in each of them.  (I'm biased: Steve is a
> friend and I was a tech editor for the book -- but I have some
> counter-bias to compensate, as I could be recommending my own
> books instead of his, so it probably balances out:-).
>
>
> Alex
>




From dance_code at hotmail.com  Sat Sep 28 20:31:53 2002
From: dance_code at hotmail.com (lee lion)
Date: Sun, 29 Sep 2002 00:31:53 +0000
Subject: A simpler singleton pattern?
Message-ID: <F119SRjCftClBbwt0Rv00003274@hotmail.com>

     singleton pattern is a way to provide one and only one object of a 
particular type. There are many ways to implement a singleton pattern, is 
the below one?
class OnlyOne:
    instance = None
    def __init__(self, arg):
        OnlyOne.instance = arg
    def __getattr__(self, name):
        return getattr(OnlyOne.instance, name)

if __name__ == "__main__":
    x = OnlyOne("x")
    print x
    y = OnlyOne("y")
    print y
    print x
    print `x`
    print `y`

The output is:
x
y
y
'y'
'y'



"The best way of living is to earn a little more than you cost and to cost 
just as you need to."
                                      Li Hang

_________________________________________________________________
?????????????? MSN Messenger: 
http://messenger.microsoft.com/cn/




From davegaramond at icqmail.com  Sat Sep 21 11:28:08 2002
From: davegaramond at icqmail.com (David Garamond)
Date: Sat, 21 Sep 2002 22:28:08 +0700
Subject: if...else on the same line?
Message-ID: <3D8C9008.70805@icqmail.com>

i'm sorry for this terrorizingly basic question, but is it possible at 
all to write an if ... else statement on a single line in python?

--
dave




From ADATE at kc.rr.com  Tue Sep  3 18:37:36 2002
From: ADATE at kc.rr.com (Shashank Date)
Date: Tue, 03 Sep 2002 22:37:36 GMT
Subject: Threads
References: <kIBb9.85641$mj7.1268320@twister.rdc-kc.rr.com> <al0jd5$f5p$1@panix1.panix.com>
Message-ID: <QQad9.100528$mj7.1509619@twister.rdc-kc.rr.com>

Splendid !

Thanks a lot ...

-- Shanko

"Aahz" <aahz at pythoncraft.com> wrote in message
news:al0jd5$f5p$1 at panix1.panix.com...
> In article <kIBb9.85641$mj7.1268320 at twister.rdc-kc.rr.com>,
> Shashank Date <ADATE at kc.rr.com> wrote:
> >
> >I am trying to write a simple threading app in Python 2.2 (on Windows)
which
> >will spawn mutiple threads.
> >Each thread will go out on the web (various web sites) and fetch certain
> >documents. The main thread will wait for all the threads to terminate OR
> >some external event (like deletion of a file). If the external event
fires
> >before termination of child-threads it kills all the spawned threads and
> >terminates.
> >
> >Any ideas how to go about doing this.
>
> See my web page.
> --
> Aahz (aahz at pythoncraft.com)           <*>
http://www.pythoncraft.com/
>
> Project Vote Smart: http://www.vote-smart.org/




From cnetzer at mail.arc.nasa.gov  Mon Sep 30 18:17:43 2002
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Mon, 30 Sep 2002 15:17:43 -0700
Subject: Why doesn't this work?
In-Reply-To: <an66jo$cb6$0@216.39.145.102>
References: <an66jo$cb6$0@216.39.145.102>
Message-ID: <200209302217.PAA07614@mail.arc.nasa.gov>

On Saturday 28 September 2002 23:29, Michael Bakkemo wrote:
> Why doesn't the rectangle start somewhere in the middle of the canvas
> (where one would expect (40,40) to be, instead of all the way over on
> the left?

Because of the way Tk works.  You are setting up a lot of things before 
entering the mainloop(), and not giving Tk a chance to execute and 
update it's internal state (not your fault, really).

But to fix things, add the line:

sc.update()

Just before:

rect = MovingRect(root, sc, 40,40, 10, 10)

(ie. right near the end of the program, but before the mainloop).  This 
let's the canvas get information from the window manager about how big 
the window that opened actually was, so it can position things like you 
expect it to.

-- 

Chad Netzer
cnetzer at mail.arc.nasa.gov



From jdhunter at ace.bsd.uchicago.edu  Wed Sep 25 10:53:30 2002
From: jdhunter at ace.bsd.uchicago.edu (John Hunter)
Date: Wed, 25 Sep 2002 09:53:30 -0500
Subject: Removing all occurences of a character from a string.
In-Reply-To: <amrav1$4rk$1@driftwood.ccs.carleton.ca> ("Sean Ross"'s message
 of "25 Sep 2002 03:36:01 GMT")
References: <amr8nh$rd8$1@driftwood.ccs.carleton.ca>
	<DL9k9.201596$216.7001230@bin4.nnrp.aus1.giganews.com>
	<amrav1$4rk$1@driftwood.ccs.carleton.ca>
Message-ID: <m2u1kece8l.fsf@mother.paradise.lost>

>From the example you gave, I was wondering if a dictionary might be a
more suitable data structure:

attr = { "gill-attachment" : ["a", "d", "f", "n"], 
         "gill-spacing"    : ["c", "d", "w"],
         "gill-size"       : ["b", "n"]
         }

print attr["gill-size"]

John Hunter



From aleax at aleax.it  Tue Sep 24 08:57:13 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 24 Sep 2002 12:57:13 GMT
Subject: callback + __call__ + TypeError: unbound method; on non method
References: <lg8z1rtvzi.fsf@mu.biosolveit.local>
Message-ID: <JiZj9.157856$pX1.5687433@news2.tin.it>

Volker Apelt wrote:


> But I can't remove the .__call__ method from my
> Callback class. The method is essential for

The problem is not with method __call__, as it gets called
with an instance.  Rather, it's right here:

> class Callback:
>     """
>     """
>     _default_Func = call_me # class wide default callback function
>     
>     def __init__(self):
>         """ """
>         self._m_func = Callback._default_Func

Here, self._m_func is NOT a function -- it's an unbound method,
thus it's not callable unless the first argument is an instance
of the class (or any subclass thereof).

When you fetch from a class an attribute that is a Python coded
function object, what you get is an unbound method object.  This
behavior is intrinsic to Python's OO mechanisms.

Remedy: have Callback._default_Func be just about ANY callable
except a Python-coded function object:-).  In Python 2.2, the
built-in staticmethod is made just for the purpose:

class Callback:
     """
     """
     _default_Func = staticmethod(call_me) # class wide default callback 

and the rest can stay unchanged (haven't examined it in detail, but
this is the surely-buggy spot in your code and explains the traceback
that you report).


Alex



From amk at amk.ca  Thu Sep 19 20:08:03 2002
From: amk at amk.ca (A.M. Kuchling)
Date: Fri, 20 Sep 2002 00:08:03 -0000
Subject: Need advice on choosing the technology for intranet based application 
References: <3d88fd1e@news.ivm.net>
Message-ID: <uokpn32h8dhqfa@corp.supernews.com>

In article <3d88fd1e at news.ivm.net>, Nenad Propadovic wrote:
> -I assume the simplest way to present a graph/chart kind of graphic is to
> create a jpg/giff/tiff on the server side, and just show it on the client?

Yes.  PIL, the Python Imaging Library, would be one possibility; so is
the gd module.  For example, the pricing charts on the pages at
http://www.mems-exchange.org/catalog/ are generated from Quixote using
gd.  PDF likely isn't helpful, because browsers won't display PDF
images inlined in a web page, though it's certainly well-suited for 
multi-page reports that need to be printed.

--amk


From marklists at mceahern.com  Thu Sep 26 20:40:03 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 26 Sep 2002 19:40:03 -0500
Subject: Help! This is driving me insane!
In-Reply-To: <FYMk9.12763$wH.1166@sccrnsc01>
Message-ID: <JHEOKEOOLIGLDHCMAHMOGEPBEBAA.marklists@mceahern.com>

[kara.klosterman]
> I created a very simple class that totals some data
> but I am not getting the output from it that I would
> expect at all.
> Here is the class:
> http://www.geocities.com/dbasch/insane.txt
>
> I would expect to get:
> [0, 0]
> [0, 0]
> [[0, 0]]
>
> but I am getting:
> [0, 0]
> [0, 0]
> [[1, 1]]
>
> I dont understand how the 'grouptotals' list is
> getting the incremented values prior to the
> 'grouptotalcounter' function being called.
> furthermore, I can place a 'print grouptotals' right
> after the errorlog append and it isnt the incremented
> values! Please tell me this isnt a normal thing!

I glanced at your code and tried to decipher it.  My overall reaction is
that the intention behind your code is obscure to me.  Note that you don't
even really tell us how to run it and how you got your output.  I'm guessing
you did this:

  t = Test()
  x = t.run()
  print x

// m




From bwucew at yahoo.com  Thu Sep 12 13:12:05 2002
From: bwucew at yahoo.com (Bwuce_Wee)
Date: 12 Sep 2002 10:12:05 -0700
Subject: Extending a C++ App
Message-ID: <996eade8.0209120912.3a653d7@posting.google.com>

There are lots of utilities that allow a person to bind a DLL's
classes, variables, etc to use as a python module.

I can find incredibly little information though on how to run python
INSIDE a C++ EXE, and still have that binding of classes,etc.
For instance, I want to make application MyApp.EXE that can load in x
number of python scripts that are called inside the app.
However, there is a application object model that I want the scripting
to be able to interact with (from both C++ code and from the script)

If you found my explanation too vague - I apologise
Please mention it so I can ellaborate

Any help is appreciated - thanks in advance for the replies


From max at alcyone.com  Sun Sep  8 02:28:08 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 07 Sep 2002 23:28:08 -0700
Subject: ANN: empy 1.4 -- Embed Python in template text as markup
Message-ID: <3D7AEDF8.767B27DC@alcyone.com>

Summary

    empy is a system for embedding Python expressions and statements
    in template text; it takes empy source files, processes them, and
    produces output.  This is accomplished via expansions, which are
    special signals to the empy system and are set off by a special
    prefix (by default the at sign, '@').  empy can expand arbitrary
    Python expressions and statements in this way, as well as a
    variety of special forms.  Textual data not explicitly delimited
    in this way is sent unaffected to the output, allowing Python to
    be used in effect as a markup language.  Also supported are
    recording and playback via diversions and dynamic, chainable
    filters.  The system is highly configurable via command line
    options.

    Expressions are embedded in text with the '@(...)' notation; as a
    shortcut, simple variables and expressions can be abbreviated as
    '@variable', '@object.attribute', '@function(arguments)',
    '@sequence[index]', and combinations.  Full-fledged statements
    are embedded with '@{...}'.  A '@' followed by a whitespace
    character (including a newline) expands to nothing, allowing
    string concatenations and line continuations.  Comments are
    indicated with '@#' and consume the rest of the line, up to and
    including the trailing newline.  '@%' indicate optional
    "significators," which are special forms of variable assignment
    intended to specify per-file identification information in a
    format which is easy to parse externally.  '@(...?...:...)' allows
    for conditional expressions, and '@(...$...)' allows handling of
    exceptions thrown from expressions.  Finally, a '@@' sequence
    expands to a single literal at sign.


Getting the software

    The current version of empy is 1.4.

    The latest version of the software is available in a tarball here:
    http://www.alcyone.com/pyos/empy/empy-latest.tar.gz.

    The official URL for this Web site is
    http://www.alcyone.com/pyos/empy/.


Requirements

    empy should work with any version of Python from 1.5.x onward.


License

    This code is released under the GPL.


	...


Release history [since 1.3]

    - 1.4; 2002 Sep 7.  Fix bug with triple quotes; collapse
      conditional and protected expression syntaxes into the single
      generalized '@(...)' notation; 'empy.setName' and 'empy.setLine'
      functions; true support for multiple concurrent interpreters
      with improved sys.stdout proxy; proper support for 'empy.expand'
      to return a string evaluated in a subinterpreter as intended;
      merged Context and Parser classes together, and separated out
      Scanner functionality.


-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ The work of many cannot be done alone.
\__/ (a Bemba proverb)
    Python chess module / http://www.alcyone.com/pyos/chess/
 A chess game adjudicator in Python.


From billgates at msn.com  Sun Sep 22 09:08:29 2002
From: billgates at msn.com (El Jesuita)
Date: Sun, 22 Sep 2002 14:08:29 +0100
Subject: wxTreeCtrl and Drag&Drop
Message-ID: <amkf9j$6i34b$1@ID-162186.news.dfncis.de>

Hi, 

how can I use drag and drop any wxTreeCtrl node?

Thank you
-- 
fmarcelo(AT)airtel(DOT)net


From gustav at morpheus.demon.co.uk  Mon Sep  9 17:05:50 2002
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: 09 Sep 2002 22:05:50 +0100
Subject: Pysol Solitaire for Windows wanted
References: <algahc$e8q$1@news7.svr.pol.co.uk> <2MPe9.47925$Jo.5395@rwcrnsc53> <alh591$sqa$1@brokaw.wa.com> <alh9r7$vni$1@brokaw.wa.com>
Message-ID: <fzwirhzl.fsf@morpheus.demon.co.uk>

"Alan Winston" <awinston at scn.org> writes:
> > I have done the Google exercise several times over recent months, and have
> > never gotten a working Windows installable, only the pyc files, which
> > briefly open a Python window, display some messages (presumably errors)
> > then close again before I can read the messages.

There are bugs in the Windows code for the latest version
(4.80). IIRC, it also depended on a version of Tk which isn't the
standard. None of the problems were impossible to solve, but they were
a mild pain, and the result is *not* the same as the original, as the
supplied .pyc file is somehow built as a single file from multiple
source files.

The author is also effectively uncontactable, which makes getting
fixes included in the source pretty much impossible.

It's a real pain, as PySol is a really good program. If anyone could
package it up again, I'm sure the result would be of value to a lot of
people...

Paul.


From never at mind.info  Wed Sep 18 07:43:16 2002
From: never at mind.info (G. Willoughby)
Date: Wed, 18 Sep 2002 12:43:16 +0100
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com>
Message-ID: <am9os4$95b$1@newsg3.svr.pol.co.uk>

> Can someone point me to a good Python GUI app to demonstrate (other
> than IDLE)?

I wrote a GUI interface to another python console app to create maps for an
online game by reading the installed in-game graphics, 3D model and terrain
data, see it here:

http://www.thecalm.fsnet.co.uk/MapperGUI

--G. Willoughby




From maxm at mxm.dk  Tue Sep  3 05:05:23 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 03 Sep 2002 11:05:23 +0200
Subject: Why it is so dramatical?
References: <mailman.1030957743.4021.python-list@python.org> <mailman.1030962662.4274.python-list@python.org> <20020902.104707.1789376348.1554@cybermesa.com> <mailman.1031039705.9065.python-list@python.org>
Message-ID: <3D747B53.6000307@mxm.dk>

Bo M. Maryniuck wrote:
> On Monday 02 September 2002 05:47, Jay O'Connor wrote:
> 
>>As a result Perl is much faster at some kinds of operations than
>>Python, and vice-versa.
> 
> 
> Yeah, actually, but I just would like to find some optimal Python solution :)


Actually it is possible to make it as readable as concatenation with:

result = []
a = result.append

a('some string')
a('another string')
a('third string')
for letter in 'abcdefghijklmnop':
     a(letter)

result = ''.join(result)


rather than:

result = ''
result += 'some string'
result += 'another string'
result += 'third string'
for letter in 'abcdefghijklmnop':
     result += letter


regards Max M



From peter at engcorp.com  Sun Sep 15 23:09:07 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 15 Sep 2002 23:09:07 -0400
Subject: can wxpython do this?
References: <737aefee.0209151702.3f735252@posting.google.com> <spcaou804m5oo6gah31oqaeuqsuj60h0ip@4ax.com>
Message-ID: <3d854b44@news.sentex.net>

Lothar Scholz wrote:
> On 15 Sep 2002 18:02:17 -0700, erick_papadakis at yahoo.com (Erick
> Papadakis) wrote:
> 
> There is no problem to setup an installer easily that takes
> Python,WXPysthon and whatever you need. It simply blowes up the
> installer package size. 

A better option, although I haven't tried it myself, would be
to use py2exe to build an executable, then package that up
using one of the regular installers.  This should come out
substantially smaller than a full Python+wxPython+app
installation.

-Peter



From shalehperry at attbi.com  Fri Sep  6 13:06:13 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 6 Sep 2002 10:06:13 -0700
Subject: question: iso 14977 ebnf grammar for Python?
In-Reply-To: <77977a28.0209060856.6974e6eb@posting.google.com>
References: <77977a28.0209060856.6974e6eb@posting.google.com>
Message-ID: <200209061006.13290.shalehperry@attbi.com>

On Friday 06 September 2002 09:56, Michael Weigend wrote:
> Hello!
>
> Is there a ebnf grammar for Python written according to the ISO 14977
> standard somewhere in the www?
>
> Michael asks

not familiar with that standard but in the source under Grammar/ is a 
definition of the grammar in something very close to ebnf.



From op73418 at mail.telepac.pt  Thu Sep 26 13:28:10 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Thu, 26 Sep 2002 18:28:10 +0100
Subject: design question
References: <1ha6pukqdqell8f31q1k7rtor2rv5ogjs0@4ax.com> <mailman.1033059455.32230.python-list@python.org>
Message-ID: <r5g6pug88d7uq5jm8kbckom2l30n6ec9jn@4ax.com>

On Thu, 26 Sep 2002 11:56:05 -0500, "Mark McEahern"
<marklists at mceahern.com> wrote:

>[Gon?alo Rodrigues]
>> I know I am being a little vague in my question, but does anyone have
>> any ideas? Open-source code where this "same problem" has been tackled?
>
>Have you considered metaclasses?  Do you need a concrete Wrapper for each
>class?  Why not a generic wrapper that wraps any given class?
>

Well, I have thought that somewhere further down the way some (though
not all) of the wrapping code could be factored out in some kind of
metaclass, and then I would just set __metaclass__ at module level. But
since I am still at the entrance, I am kinda polling for ideas.

Assuming that my description of the problem is enough, how would you go
about writing a metaclass as a general wrapper?

>// m
>

All the best,
Gon?alo Rodrigues


From max at alcyone.com  Thu Sep  5 15:11:49 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 05 Sep 2002 12:11:49 -0700
Subject: Graham's spam filter
References: <mailman.1029992314.27908.python-list@python.org> <3D647B4B.28D9BF04@alcyone.com> <mailman.1030050253.11330.python-list@python.org> <3D655AE6.A097AA25@alcyone.com> <ak44j1$1f2ft4$2@ID-125932.news.dfncis.de> <c2ef5940.0209041912.16e102fb@posting.google.com> <3D770156.2E55864C@alcyone.com> <al7tg4$1o5a9k$5@ID-125932.news.dfncis.de>
Message-ID: <3D77AC75.B9B7F27@alcyone.com>

Christopher Browne wrote:

> Have you considered simply replacing strings that appear to be
> base64-encoded with a token like "base64-text"?
> 
> That allows the database to at least be aware that the spam commonly
> contains base64 data.

Well, that really depends on what your goal is.  Again, if you're one of
those people that has a very tight circle of email buddies and so
essentially any unsolicited email is by definition spam, then you can
tighten down your spam filter in all kinds of very powerful ways.

I, as I've mentioned before, receive unsolicited email from my Web sites
and various projects, and so unfortunately don't have the luxury of
doing this.  So I need to support receiving email from faraway lands and
unknown email addresses as well as trying to vigorously filter spam.

Fact is, unfortunately, lots of people send legitimate email that is
MIME encoded.

> -> Supposing there is interesting text encoded (such as source code
>    for a virus) inside the base64 stuff, it _would_ be useful to
>    decode it;
> 
> -> Supposing the base64 stuff is basically just a GIF/JPEG/PNG, or
>    something else that doesn't contain "interesting text," you'll
>    have not much of value from the decoding process.
> 
> Making the "tokenizing" step a tad smarter (e.g. - recognizing "this
> is likely base 64" and collecting stats on numbers of lines of base64
> material) requires minimal added effort, and I expect it would buy you
> _most_ of the benefits of decoding.

Spammers are hitting upon the strategy, though, of sending emails in
which the body consists of nothing but a completely encoded base64 MIME
part.  So in that case, the entire body of your message would consist
solely of your "base64encoded" token.  So in the general case of any
kind of spam filter (not just limited to a Graham filter), it's
questionable how useful this will be, unless you plant to always filter
against that token, presuming it to always indicate spam.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ There is nothing so subject to the inconstancy of fortune as war.
\__/ Miguel de Cervantes
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From lienard.bruno at free.fr  Sat Sep 21 05:01:54 2002
From: lienard.bruno at free.fr (Bruno Liénard)
Date: Sat, 21 Sep 2002 11:01:54 +0200
Subject: Htmlxtract modules
Message-ID: <3d8c3668$0$225$626a54ce@news.free.fr>

I am looking for the HtmlXtract packages quoted in the "vaylts of Parnassus"
repository. The quoted link seems to be dead. So can somebody send it to me
?

Thanking you by advance,

Bruno Lienard




From ulbi at ivs.tu-berlin.de  Tue Sep 24 02:56:32 2002
From: ulbi at ivs.tu-berlin.de (Andreas Ulbrich)
Date: Tue, 24 Sep 2002 08:56:32 +0200
Subject: Jython file manipulation?!
References: <2yej9.32454$gA4.4053@sccrnsc02> <amn87j$73a$1@mamenchi.zrz.TU-Berlin.DE> <CaIj9.414430$_91.558962@rwcrnsc51.ops.asp.att.net>
Message-ID: <amp2b4$qhm$1@mamenchi.zrz.TU-Berlin.DE>

Matthew Bonig wrote:
> There are no exceptions raised...
> If you look at the source Jython uses for the os.rename function it just
> uses Java's File.renameTo()...
> that function returns false... which then the os.rename function throws a
> OSError exception, but it's not helpful in any way.
> The only thing I can think is happening is that I am doing this renaming on
> files that are on a netword drive on a win32 box, but I have double
> checked... the new filename is NOT there... (since win32 won't overwrite the
> new file if it exists, and therefore return an error, but that's not
> happening). I've also double checked to make sure I have write permissions
> on the network drive, which I most definetly do (I usually write 3-5 files
> onto that drive a day).
> This is why I thought there might be some sort of security issue with
> Jython... I'm pretty sure it's in the Jython side since when I tried exactly
> the same thing (os.rename(oldfilename, newFilename)) in python it worked
> perfectly.

Sounds weird. Have you tried to write a java program and check the 
results? Maybe it's an java issue ... Which java you are using?



From mstenner at phy.duke.edu  Tue Sep 10 08:59:28 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Tue, 10 Sep 2002 08:59:28 -0400
Subject: Pickle problem
In-Reply-To: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220D10@hendrix.empolisuk.com>; from hst@empolis.co.uk on Tue, Sep 10, 2002 at 01:44:57PM +0100
References: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220D10@hendrix.empolisuk.com>
Message-ID: <20020910085928.B4329@phy.duke.edu>

On Tue, Sep 10, 2002 at 01:44:57PM +0100, Harvey Thomas wrote:
> Erik Price wrote:
> > 
> > On Tuesday, September 10, 2002, at 06:13  AM, Harvey Thomas wrote:
> > 
> > > I'm trying to pickle a complicated data structure which includes a 
> > > dictionary whose entries are created by (simplified):
> > >
> > > a[x] = re.compile(y).match
> > >
> > > The pickle fails, however, with the error:
> > >
> > > PicklingError: Can't pickle <built-in method match of 
> > _sre.SRE_Pattern 
> > > object at
> > >  0x00838700>: it's not found as __main__.match
> > >
> > > I'd be grateful if someone could tell me if it is possible 
> > to pickle 
> > > the data structure and what I'm doing wrongly.
> > 
> > Are you trying to store the MatchObject returned by a call to 
> > a regular 
> > expression object's match() method?  If so, it looks like you're 
> > forgetting the parens at the end of "re.compile(y).match".
> > 
> No, I haven't forgotten the parentheses. I'd like to pickle the match method of the compiled regular expression.

My bad... I just reread this.  I had misread it before.  Sorry for the
misunderstanding.  I am bit curious about what you get out of pickling
the method that you can't get from the compiled re object itself.
You can always call the match method after unpickling, right?

-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From zopestoller at thomas-guettler.de  Thu Sep 26 06:37:18 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Thu, 26 Sep 2002 12:37:18 +0200
Subject: Length of file extension Was: convention for Pickler file extension
References: <918bc22f.0209230208.4eaca702@posting.google.com> <svEj9.226013$z91.9600123@bin3.nnrp.aus1.giganews.com> <ampfbe$rq0$07$2@news.t-online.com> <smm0pus5t41l783ffo8ms6uur2umob0ftf@4ax.com>
Message-ID: <amuns8$cc2$03$1@news.t-online.com>

Richie Hindle schrieb:
>>I call files *.pickle
> 
> 
> <aol>Me too.</aol>
> 
> Are there any non-medieval platforms that don't let you have arbitrary
> length file extensions?

I think it works for Win98, WinNT, Win2000 and of corse Unix.

Don't know about Mac or MacOSX

  thomas



From tchur at optushome.com.au  Sat Sep  7 19:01:16 2002
From: tchur at optushome.com.au (Tim Churches)
Date: Sun, 08 Sep 2002 09:01:16 +1000
Subject: Smoothing a discrete set of data
References: <3csmdso5.fsf@morpheus.demon.co.uk> <B99F9C2F.1238E%pereira@cis.upenn.edu> <mailman.1031432711.19023.python-list@python.org> <aldsas$b8$1@peabody.colorado.edu>
Message-ID: <3D7A853C.5E5BD96B@optushome.com.au>

Fernando P?rez wrote:
> 
> Tim Churches wrote:
> 
> > AFAIK, there is no native Python package which
> > implements a
> > comprehensive range of regression and smoothing techniques, but I can
> > highly
> > recommend the R package for statistics and statistical graphics - this
> > mature, free, open
> > source package has more statistical facilities than you are ever likely
> > to need.
> 
> While not as comprehensive as R in its statistical coverage, SciPy probably
> has enough in it for what the OP wants (which sounded fairly basic). Quick
> linear (in the parameters) fits are easily done with scipy, plotting, simple
> smoothing (if he wants to), etc. I know R is industrial-strength statistics,
> but it sounds like more than what the OP needed.

Yup, although I suggested R because a) some very gentle companion
introductions 
to basic statistical analysis are available for R (see for example
http://cran.r-project.org/doc/manuals/R-intro.pdf or the various
documents listed
at http://cran.r-project.org/other-docs.html ); and b) R will not run
out of steam
when the OP realises that there might be a wee bit more to the analysis
of his
data than he first thought eg some of his data are almost certainly time
series 
and hence he will need access to techniques which take auto-correlation
etc into
account. 

Of course, the biggest hurdle for Pythonistas who want to use R is the
fact that 
R uses one-based indexing, not zero-based indexing as <insert deity or
demi-god 
of your choice> intended.

Tim C



From gerhard.haering at opus-gmbh.net  Tue Sep 10 10:05:14 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 10 Sep 2002 14:05:14 GMT
Subject: os.tempnam() warning
References: <pan.2002.09.10.09.56.50.41586.29853@presidency.com>
Message-ID: <slrnanruqh.24s.gerhard.haering@haering.opus-gmbh.net>

Rajarshi Guha wrote:
> The C library advises against using tempnam() and suggests using
> mkstemp().  Is there any Python equivilant?

tempfile.mktemp()

-- Gerhard


From nospam_t.roessler at v-d.com  Thu Sep 26 07:16:52 2002
From: nospam_t.roessler at v-d.com (Thilo =?ISO-8859-15?Q?R=F6=DFler?=)
Date: Thu, 26 Sep 2002 13:16:52 +0200
Subject: Zope, MySQL and pictures in blob-fields ...
Message-ID: <amuqck$e0r$1@nathan.regio.net>

Hello to everybody!

I'd like to provide users with a possibility to upload pictures into a 
MySQL-Database. Therefore I created a web-form (using method="post" and 
enctype="multipart/form-data") with an input for uploading files (<input 
type="file" name="VorschlagBild">). The action of the form calls a 
SQL-Method that stores the data received from the into a MySQL-table 
(insert into Vorschlag values([...]<dtml-sqlvar VorschlagBild 
type="string">)). The picture should be put into a blob-field. I used 
'type="string"' because "file" or something similar does not exist.
Finally, the record gets stored into the MySQL-database, but the picture 
isn't saved the way I expected. Taking a look at the content of the table 
using webmin, the appropriate field just shows 
"<ZPublisher.HTTPRequest.FileUpload instance at 0x893a7ec>".
Can anybody tell he how I manage to save/upload the picture and possibly 
even show it again in a website afterwards?

Thanks in advance, and sorry if it is a simple question ... I tried to find 
a solution but couldn't find one.

Thilo


From cliechti at gmx.net  Fri Sep 27 13:24:53 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 27 Sep 2002 19:24:53 +0200
Subject: watching mutables?
References: <mailman.1033137396.14482.python-list@python.org> <an211j$oi$1@news.hccnet.nl>
Message-ID: <Xns9296C58C9ADFCcliechtigmxnet@62.2.16.82>

anton at vredegoor.doge.nl (Anton Vredegoor) wrote in 
news:an211j$oi$1 at news.hccnet.nl:

> On Fri, 27 Sep 2002 09:34:55 -0500, "Mark McEahern"
> <marklists at mceahern.com> wrote:
> 
>>l = Watched()
>>l.append(1)
> 
> Thanks for reacting to my question. However, I am looking for
> something that calls a defineable function if *any* mutable variable
> in the current scope changes. Variables should be 'marked' to install
> this function for them. For example 'watch(m,f)' indicates that
> function f is called if mutable variable m is changed. I do not want
> function f to be a method of object m since I want object m to be
> *any* mutable python object and furthermore I do not want to change
> the orginal object. The call of function f would be triggered by *any*
> change of object m. This would make debugging (and some other things I
> am tinkering with) a lot easier. Is it possible?

see EventObj:
http://oomadness.tuxfamily.org/p-pyobjtools.php

chris
-- 
Chris <cliechti at gmx.net>



From whisper at oz.net  Mon Sep 16 23:26:06 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 16 Sep 2002 20:26:06 -0700
Subject: More on Protecting Source Code
Message-ID: <GCEDKONBLEFPPADDJCOEAEDDFAAA.whisper@oz.net>

I have read what I consider to be the narrow minded and short sighted views
wrt to protecting IP in the form of Python source code, and I would like to
make a few points.

* Not everyone believes that all or most software ought to be free. I for
one, don't.

* Owners of property have the right to control it. If you disagree and own a
nice car or house, I'll be right over :->

* Why invest a substantial amount of time and money developing in a language
that makes it trivial to gain access to the work product?

Python may be easy to use, but if it's also easy to steal software written
in it, that's an impediment to it's being widely used in commercial
products, especially shrink-wrapped products. Perhaps that's a reason why
there seems to be so few commercial products in Python, beyond those that
are fairly closely licensed and/or have substantial parts of the app written
in C/C++.

I suppose one solution is to modify the Python interpreter with different
op-codes and that ought to make it somewhat painful for the average hacker.
A better solution is to make a .pyc file approximately as hard as a binary
.exe file to decompile - however that could be done.

David LeBlanc
Seattle, WA USA




From simon at callan.demon.co.uk  Tue Sep 17 14:43:56 2002
From: simon at callan.demon.co.uk (Simon Callan)
Date: Tue, 17 Sep 2002 19:43:56 +0100
Subject: using python as cgi on apache unix
References: <am6fm301477@enews2.newsguy.com>
Message-ID: <664284774b.simon@satin.callan.demon.co.uk>

In message <am6fm301477 at enews2.newsguy.com>
          <josh_py at sautez.com> wrote:

> All,
> 
> I have developed a python solution to run and connect to a mysql
> database. I did all my development under windows apache 1.3.26 and
> python 2.2.1.  I know want to move this to my unix server. 
> Unfortunately, no matter what I do I can not get apache on my linux
> box to execute any python scripts.  I get the following errors in
> apache:
> 
> Premature end of script headers: /var/www/html/cgi-bin/test.py

Check the access permissions on the file. On my setup, the file has to
have the "other" access permissions set to read and execute. Ie try
typing "chmod o+rx test.py", and trying again.

It took me some time to spot this problem, when I had it.

Simon



From aleax at aleax.it  Fri Sep 20 10:01:54 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 14:01:54 GMT
Subject: Newbia alert .. working my way through Easy Python Montana Edu page, need help..
References: <MPG.17f540bc75626c66989683@news.worldonline.nl>
Message-ID: <mTFi9.119296$ub2.2532256@news1.tin.it>

Highwaykind wrote:
        ...
> def print_options() :

You define this function, but never call it anywhere.

> choice = 'p'
> while choice != 'q':
>     if choice == 'r':
>         a=input ("width")
>         b=input ("height")
>     print "result",rect (width,height)

Note that the print is NOT aligned with the input calls, but
rather with the if statement.  Thus, the print (and its call
to rect) is not conditional -- it happens even when choice
is NOT 'r'.  This is unlikely to be what you want, so, indent
this print statement 4 more spaces rightwards.

However, the crux of your problem is another:

> But the thing keeps telling me width is not defined, also I can;t use

Inddeed, width is not a defined variable: what you obtained
from the call to input("width") you did not assign to the
name width, but rather to the name a.  Therefore, the variable
that is defined once this statement has executed is named a.

Therefore, change your call to rect  to rect(a,b) -- and
similarly for the following passages in your code.

> the ELIF command .. THink I could do with some help :)

The reason you can't use elif (not a 'command' -- a clause
in an if statement) is probably the indentation problem I
mentioned earlier -- once you fix that you can probably
also change the following if's to elif's (not the first one
of course).  When indentation returns to the same level as
the word 'if', it means the if statement is over: therefore
you cannot "reprise" it later, continuing it, so to speak,
with an elif clause.  Once you correct the indentation of
the print statements, this will change.


Alex



From bokr at oz.net  Mon Sep 30 00:42:39 2002
From: bokr at oz.net (Bengt Richter)
Date: 30 Sep 2002 04:42:39 GMT
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05> <an7bvd$burft$1@ID-125932.news.dfncis.de> <an7p7t$cpg$0@216.39.172.122> <mailman.1033338759.6775.python-list@python.org>
Message-ID: <an8knv$mhr$0@216.39.172.122>

On Sun, 29 Sep 2002 17:33:57 -0500, "Chris Gonnerman" <chris.gonnerman at newcenturycomputers.net> wrote:

>----- Original Message ----- 
>From: "Bengt Richter" <bokr at oz.net>
>
>
>> On 29 Sep 2002 17:06:53 GMT, Christopher Browne <cbbrowne at acm.org> wrote:
>> 
>> >The problem with FP is that FP involves the use of binary
>> >approximations, which commonly cannot exactly represent 
>> >decimal values.
>> >
>> >Notably, the decimal fraction 0.1, which is /exactly/ 1/10, 
>> >does not have any exact encoding in binary.  Its 
>> >representation in binary is a repeating binary fraction.
>>
>> How is that a significantly different problem from repeating 
>> decimal fractions? (E.g., 1/3. or 1/7. or wxyz/9999. which 
>> will give you 0.wxyzwxzy..?)
>
>Humans think in decimal.  To a human, 1.0/10.0 == 0.1, not
>0.10000000000000001.  OK, so this *rounds* to 0.10 at two
>decimal places, but what about:
>
>>>> .70 * .05
>0.034999999999999996
>
>Rounded to two decimals, that's 0.03; but done in proper
>decimal numbers, the answer should be 0.035, rounded to 0.04
>using either "classic" or "banker's" decimal rounding.
>
>If you are calculating sales tax, for instance, this is 
>unacceptable.
>
>(I borrowed this example from memory from a previous poster,
>and if I could remember who it was I'd give credit...)
>
>You ask how it is a different problem.  Simple.  A human,
I hope that includes me ;-)
>training in decimal math in grade school, knows that 
>
>    1/3 = 0.3333333...
>
>and would *expect* it to round to 0.33 (still in two decimal
>places).  The problem with *binary* floating point is the
>"surprise factor."
>
Old programmers aren't surprised by anything. They expect the unexpected ;-)

>So, Bengt, how do I do this right using only floats?  I can't.
>I have to have decimal arithmetic to get the right answer.
>
It goes back to requirements. E.g.,

 >>> for x in xrange(100):
 ...     for y in xrange(100):
 ...         fx = float('.%02d' %x)
 ...         fy = float('.%02d' %y)
 ...         fxy = fx*fy
 ...         xy = x*y
 ...         sfxy = '%.2f' % (fxy*1.000000001)
 ...         sxy = '0.%02d' % ((xy+50)/100)
 ...         assert sfxy == sxy
 ...

finishes sans problem, as a quick check on a limited problem domain. Obviously
a proper test suite and some math would be required to show validity for a
realistic problem domain.

If you can specify exactly how your decimal virtual machine works,
(i.e., how many significant figures are carried, how it rounds/truncates
intermediate values etc., and how many figures are going to be in the inputs
and outputs, then we can determine how much of a problem it will be to do it
using floating point.

The example you cited is included in the above:
 >> '%.2f' % ( .05 * .70)
 0.03'
 >> '%.2f' % (-.05 * .70)
 -0.03'

changes to

 >> '%.2f' % (( .05 * .70) * 1.000000001)
 0.04'
 >> '%.2f' % ((-.05 * .70) * 1.000000001)
 -0.04'

There is not unlimited room in the double floating point space around the
true abstract real values corresponding to decimal representations, but there is
quite a bit, unless you are carrying a lot of figures. E.g., as an
approximate indication,

 >>> '%.2f' % ((.0349999999    ) * 1.0000000001)
 '0.03'
 >>> '%.2f' % ((.03499999999   ) * 1.0000000001)
 '0.03'
 >>> '%.2f' % ((.034999999999  ) * 1.0000000001)
 '0.04'

and the multiplier is probably bigger than it has to be (and better specified
in binary). But that can be optimized when you specify your decimal virtual machine.

If you need rounding that alternates direction on alternate exact trailing fives,
then you need an extra step.

If you visualize the true theoretical decimal values representable in your decimal virtual
machine as little red glowing dots on the real axis, and visualize as little glowing blue dots
all the exact theoretical numbers corresponding to all valid bit combinations in a floating
point double, then if every pair of successive red dots has thousands of blue dots in between,
and the error in floating point calculations can be held to within a few blue dots, then
you can see how to select the red dot that represents the true decimal answer, even if
there is no blue dot exactly coinciding with it. Twelve decimal digits would give about a thousand
blue dots between the red ones. And a thousand times that while still in the FPU, before
storing in memory. Rounding an exact decimal is moving from one red dot to another,
perhaps skipping hundreds or thousands of dots, depending on fractional decimals carried. But
we must round by moving from some blue dot to another blue dot. By shifting the blue
uncertainty set reliably to one side of its true red dot, we can move from blue dot to
blue dot where the new blue dot will reliably be in an uncertainty set corresponding
to the proper red dot. If more blue dots are needed, they can be arranged.
If you are doing very few operations, so that the range of blue dot uncertainty that can be
accumulated is small, then you can have more red dots.

Regards,
Bengt Richter


From sadams123 at optushome.com.au  Tue Sep 17 00:50:39 2002
From: sadams123 at optushome.com.au (Steven)
Date: Tue, 17 Sep 2002 14:50:39 +1000
Subject: More on Protecting Source Code
References: <mailman.1032232999.22661.python-list@python.org>
Message-ID: <3d86b4a0$0$23173$afc38c87@news.optusnet.com.au>

"David LeBlanc" <whisper at oz.net> wrote in message
news:mailman.1032232999.22661.python-list at python.org...
> I have read what I consider to be the narrow minded and short sighted
views
> wrt to protecting IP in the form of Python source code, and I would like
to
> make a few points.
>
> * Not everyone believes that all or most software ought to be free. I for
> one, don't.
>
> * Owners of property have the right to control it. If you disagree and own
a
> nice car or house, I'll be right over :->
>
> * Why invest a substantial amount of time and money developing in a
language
> that makes it trivial to gain access to the work product?
>
> Python may be easy to use, but if it's also easy to steal software written
> in it, that's an impediment to it's being widely used in commercial
> products, especially shrink-wrapped products. Perhaps that's a reason why
> there seems to be so few commercial products in Python, beyond those that
> are fairly closely licensed and/or have substantial parts of the app
written
> in C/C++.

but if you're worried about protecting your IP, then wouldn't your package
be under a closed license?

> I suppose one solution is to modify the Python interpreter with different
> op-codes and that ought to make it somewhat painful for the average
hacker.
> A better solution is to make a .pyc file approximately as hard as a binary
> .exe file to decompile - however that could be done.

>From what I've read of that thread, the gist of it was

"once its released, a determined person can, to some extent get your code"

and

"Security through obscurity is no security against a determined attacker"

So, someone who is willing to hacka nd hack and hack away at your code, can
get at it, and this is true for things coded in assembler, C, Python or
Befunge. Most people aren't determined attackers, they're simply Joe Average
who needs to get a  job done.

An operating system is a complex beast, compiled into machine code its even
more complex, how many people out there are using an unlicensed copy of
Windows or Office? they didn't decompile, that would take skill, they just
copied it bit for bit and posted the registration key.

Do a very very simple risk analysis. Who is your target audience, what is
the price of your software, will people bother going through your bytecode,
or will they just post the registration key to a website? Will stolen
software lead to a loss in revenue - or will those pirate users be people
who would never have bought it in the first place? (yes, the pirate users
are annoying, and you would feel ripped off, but OTOH you haven't lost
money, and you've gained a bunch of users)

Steven





From martian_bob at portalofevil.com  Thu Sep 26 18:17:28 2002
From: martian_bob at portalofevil.com (Bob Arens)
Date: Thu, 26 Sep 2002 22:17:28 GMT
Subject: Newb - web portal for Unix tool
Message-ID: <Xns9295AFF64AE99martianbobportalofev@204.127.202.16>

Hiyo, total newbie to Python though not HTML, etc. I'm writing a web-based 
portal allowing users to use Expat by entering their XML into a form.

The HTML is trivial - just a text area and submit button. I'm using a 
pretty good reference for the CGI bit (Internet Programming with Python by 
Watters, van Rossum and Ahlstrom), but their examples don't seem to be 
working for me. So here are my big questions:

1. How do I write out input from a text area to a file?
2. How do I make a command line call from within a CGI script?
3. How can I debug my script?

Here (more or less) is what I have for the script (where METHOD="POST" and 
TEXTAREA NAME="XMLin" in the HTML, and testme.xml is the file I'm writing 
to)
----------------------------
import genCGI, string

class ExPortal(genCGI.ParseMixin, genCGI.Verbose_Mixin, genCGI.cgiBase):
    	MultiValuedFields = ["XMLin"]

def success(self):
    	items = self.formdict.items()
    	items = map(str, items)
    	items = string.joinfields(items)
    	outfile = open("testme.xml")
    	outfile.write(items)
    	outfile.close()

if __name__=="__main__":
    	ExPortal().GO()
-----------------------------

Any help is much appreciated!


From fedor at mailandnews.com  Wed Sep 18 09:21:42 2002
From: fedor at mailandnews.com (Fedor Baart)
Date: Wed, 18 Sep 2002 15:21:42 +0200
Subject: running external programs
References: <p%_h9.507714$UU1.83878@sccrnsc03>
Message-ID: <am9ukp$dtj$1@news.hccnet.nl>

You might want to try:

import webbrowser
webbrowser.open('url')

or

import os
os.system('application')


"Jonathan Beebe" <jonbeebe85 at yahoo.com> wrote in message
news:p%_h9.507714$UU1.83878 at sccrnsc03...
> Is there a function or library i can import that i can use to run external
> programs from within my program?
>
> example:
>
> runprogram("iexplore.exe")
>
> thanx,
> -JONATHAN BEEBE
>
>
>




From loewis at informatik.hu-berlin.de  Tue Sep  3 06:56:29 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 03 Sep 2002 12:56:29 +0200
Subject: Which XML parser under Python can handle Unicode?
References: <9ac02e81.0209030214.515d6c0@posting.google.com>
Message-ID: <j4vg5ngx42.fsf@informatik.hu-berlin.de>

jhorneman at pobox.com (Jurie Horneman) writes:

> Could someone please recommend an XML parser under Python which can
> handle Unicode? 

The "standard" parser of Python 2.x, ie. xml.parser.expat, supports
Unicode quite well, as do all libraries on top of this parser
(xml.sax, xml.dom.minidom). Since PyXML 0.7, xmlproc supports Unicode as well.

> Specifically, most European languages and Asian
> languages such as Chinese (traditional, simplified), Japanese, Korean
> and Thai? (Please don't make me look up the actual codetables.)

All of those languages are supported by all Python XML parsers, in
various encodings - but best in UTF-8.

> Ideally the parser would use Python's codec system.

pyexpat/Expat has builtin codecs for UTF-8 and Latin-1, and falls back
to Python codecs - but only for byte-oriented encodings. xmlproc uses
the Python codec system.

> I'd prefer an event-based parser, although tree-based would not be a
> huge problem.

Both pyexpat and xmlproc are event-based. I'd recommend to use SAX on
top of this, which is also event-based.

Regards,
Martin



From pedronis at bluewin.ch  Sun Sep 29 11:23:35 2002
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Sun, 29 Sep 2002 17:23:35 +0200
Subject: FYI (was: watching mutables?)
References: <an1p3c$jrd$1@news.hccnet.nl> <3d96f25a$0$498$7a628cd7@news.club-internet.fr>
Message-ID: <3d971af8$1_3@news.bluewin.ch>

"Jiba" <jiba at tuxfamily.org> ha scritto nel messaggio
news:3d96f25a$0$498$7a628cd7 at news.club-internet.fr...
>
> Try EditObj (http://oomadness.tuxfamily.org/en/editobj) and look at the
> module editobj.eventobj.
>
> It does exactely what you need, but it is quite a hack...

FYI in Python 2.3: __class__ will be immutable for builtin types, and only
mutable for (user) subtypes.

See:

http://mail.python.org/pipermail/python-checkins/2002-August/028681.html

regards.




From jjl at pobox.com  Sat Sep 28 13:12:24 2002
From: jjl at pobox.com (John J. Lee)
Date: Sat, 28 Sep 2002 18:12:24 +0100
Subject: Strange behavior of ClientCookie and site-packages
In-Reply-To: <6cb82943.0209191238.4b116988@posting.google.com>
References: <6cb82943.0209191238.4b116988@posting.google.com>
Message-ID: <Pine.SOL.4.44.0209281803360.5154-100000@mimosa.csv.warwick.ac.uk>

On 19 Sep 2002, Brian Ibbotson wrote:
[...ClientCookie works from command line, but...]
> However, Idle for Python 2.2 cannot do the same. With an identical set
> of paths reported by sys.path, including the ClientCookie reference
> above, Idle reports:
>
>    >>> import ClientCookie
>    Traceback (most recent call last):
>      File "<pyshell#0>", line 1, in ?
>        import ClientCookie
>    ImportError: No module named ClientCookie
>    >>>
>
> Can anyone suggest an explanation for this behavior? Other third-party
> modules installed to site-packages (such as ChartDirector and MySQLdb)
> have had no problems once they are included in sys.path

That's odd.  I don't use IDLE -- have you checked the IDLE docs for any
information about import peculiarities?  All ClientCookie does, as you've
no doubt seen, is to import the 'public' names from various modules into
__init__.py, so that they can be imported using

from ClientCookie import foo

rather than

from ClientCookie._some_obscure_module import foo

So you may be able to work around the problem by doing the latter (look in
__init__.py to see what import statements you need).

Perhaps you'd have better luck asking directly about IDLE's behaviour in
this case.  Let me know if it's something that can/should be fixed in
ClientCookie.

BTW, you'll have a much better chance of getting a prompt response from me
if you mail me questions about ClientCookie rather than (or in addition
to) posting them here -- I don't read c.l.py very often now.


John



From justnotworking at yahoo.com  Wed Sep 18 09:51:42 2002
From: justnotworking at yahoo.com (mowestusa)
Date: Wed, 18 Sep 2002 08:51:42 -0500
Subject: (Newbie) Menu Not Functioning
Message-ID: <uoh10cjh4sct0e@corp.supernews.com>

I have created the following simple program.

import os

def menu():
    print "1. Copy Python Files from Office Computer to Floppy"
    print "2. Copy Python Files from Home Computer to Floppy"
    print "3. Copy Python Files from Floppy to Office Computer"
    print "4. Copy Python Files from Floppy to Home Computer"
    print "5. Quit"
    print

def officetoflop():
    os.system("xcopy C:\MyDocu~1\*.py a:")

def hometoflop():
    os.system("xcopy E:\MyDocu~1\*.py a:")

def floptooffice():
    os.system("xcopy a:\*.py C:\MyDocu~1")

def floptohome():
    os.system("xcopy a:\*.py E:\MyDocu~1")

menu_choice = 0

while menu_choice < 5:
    menu()
    menu_choice = input("Enter your selection:")
    if menu_choice == 1:
        officetoflop()
    elif menu_choice == 2:
        hometoflop()
    elif menu_choice == 3:
        floptooffice()
    elif menu_choice == 4:
        floptohome()
    else:
        print "Goodbye"

As you can see it is designed to simply keep all of my python files current
between home and work.  I know that there are probably better ways to write
this program by not using "os.system", but using "shutil" instead.  I tried
that first but I couldn't figure out how you could get just the ".py" files
from a directory, and then put into "shutil.copyfile".

The problem I have with my script is that I can only run it once.  For
example it lets me choose "1".  It copies all of the files onto the floppy
without problem.  Then the menu appears and the line "Enter your
selection:", but it doesn't allow me to make another choice.  I can't type
in any key from the keyboard at all.  The program seems to freeze.  Any
ideas why?  I tried using the same technique shown in the menu examples
given in the "Non-programmers Tutorial."

Thanks for the help.

mowestusa




From gupsmail at ihatespam.shaw.ca  Wed Sep 25 11:02:06 2002
From: gupsmail at ihatespam.shaw.ca (Grant Hallam)
Date: Wed, 25 Sep 2002 15:02:06 GMT
Subject: problem with dictionaries within lists
References: <pan.2002.09.25.05.21.07.477231.167@ihatespam.shaw.ca> <S4fk9.163219$pX1.5840444@news2.tin.it> <pan.2002.09.25.14.28.41.435139.181@ihatespam.shaw.ca> <J3kk9.142193$ub2.3103834@news1.tin.it>
Message-ID: <pan.2002.09.25.14.59.50.755604.181@ihatespam.shaw.ca>

Thnks for the clarification...I must have gotten confused on that one.
:), I didn't realize that dict was a builtin...changed to be more
representative of the actual contents.  Using the .copy() worked great.
Thanks.

Grant (da gup)

On Wed, 25 Sep 2002 07:51:21 -0700, Alex Martelli wrote:

> Grant Hallam wrote:
> 
>> The problem with blanking out the dictionary with dict{} is that it
>> wipes the keys out.  As the loop goes through it assigns the matching
>> second half of the list to a matching key.  If you insert a print
>> statement on the dict you can see the values in the dict are changing.
>> Thus while it is the same dict the key values within the dict are
>> different.  Thats why I'm somewhat confused as to why its the same dic
>> value.  As for the tabs,
> 
> It's the same dict object because you didn't copy it.  You just appended
> many references to it to the list.  So, just use:
> 
>>> webdatabaselist.append(dict.copy())
> 
> appending *copies* of the dict to the list, and you should be OK.
> 
> BTW, avoid naming your own variables like builtins -- list, dict, file,
> int, str, &c -- you'll come to grief eventually when you do that and
> then some other piece of your code tries to use the builtin in the
> normal way.  It's always possible to find a more significant name,
> anyway, so avoiding the types' names is no burden. (This has nothing to
> do with your specific problem -- it's just general advice).
> 
> 
> Alex


From oliver.eichler at dspsolutions.de  Mon Sep  9 03:37:22 2002
From: oliver.eichler at dspsolutions.de (Oliver Eichler)
Date: Mon, 09 Sep 2002 09:37:22 +0200
Subject: Howto debug self?
Message-ID: <alhivr$fom$04$1@news.t-online.com>

Hi 

Is it possible to debug self with pdb? I mean it in the sense of getting a 
list of vars and their values. I know I can do: 

pdb> print self.xyz

but is there a way to dump a complete list of all vars within self. The same 
applies for a class instance. I would like to dump all of it vars at a 
time. Without knowing the class structure like the print command in pdb 
assumes.

Pretty much stuck here...

Oliver


From ws at mystrobl.de  Sun Sep 22 16:33:31 2002
From: ws at mystrobl.de (Wolfgang Strobl)
Date: Sun, 22 Sep 2002 22:33:31 +0200
Subject: How to get memory usage on Windows?
References: <amilhk$bi8$1@reader1.tiscali.nl> <amkpo5$kur$1@reader1.tiscali.nl> <f51sou8o4j9pkilhc3g6jevkqmkk7jr3pl@4ax.com> <aml3th$qu3$1@reader1.tiscali.nl>
Message-ID: <ap9sou4crm9qjiilnqcov2m7uu2ktlqiiq@4ax.com>

"GerritM" <gmuller at worldonline.nl>:

>"Wolfgang Strobl" <ws at mystrobl.de> schreef in bericht
>news:f51sou8o4j9pkilhc3g6jevkqmkk7jr3pl at 4ax.com...
>> "GerritM" <gmuller at worldonline.nl>:
>>
>> >Searching in books , documentation (a.o. win32api) and net did not yet
>help
>> >I found this article on the net:
>> >http://www.developeriq.com/Magazinestories/02jul29memory.php3
>> >but how to call this function from Python?
>>
>> I guess thanks to Mark Hammond you already have everything you need.
>>
>> Have a look at  \Python22\Lib\site-packages\win32\Lib\win32pdhutil.py,
>> uncomment the test function, run it.
>>
>> I snarfed the  following snippet straight from win32pdhutil.py:
><...snip...>
>> --
>> Wir danken f?r die Beachtung aller Sicherheitsbestimmungen
>
>Wolfgang,
>
>thanks. I think this will work only on NT. Here at home I have "only"
>windows98, which gives the following message:
>"RuntimeError: The pdh.dll entry point functions could not be loaded."

Well, I haven't used windows 98 anymore for years, but I guess you're
right.

>
>I will have to try this at work, on a windows2000 machine.

That's what I'd do.

Another option would be using calldll 
(see http://www.nightmare.com/software.html). This is more dangerous,
but will most probably work on any win32 system.



-- 
Wir danken f?r die Beachtung aller Sicherheitsbestimmungen


From pete at shinners.org  Mon Sep 23 11:41:02 2002
From: pete at shinners.org (Pete Shinners)
Date: Mon, 23 Sep 2002 15:41:02 GMT
Subject: PyGame + PyOpenGl + ZBuffer?
References: <3D8EDCEF.70F827F6@hotmail.com>
Message-ID: <iCGj9.186982$Jo.66084@rwcrnsc53>

Max wrote:
> I found my code is not working well in all computers: with a SiS 300/200
> graphic card, the Z buffer seems disabled.
> 
> The OpenGL initialization code is:
> 
> pygame.init()
> pygame.display.set_mode((640,480), OPENGL|DOUBLEBUF|HWSURFACE,32 )
> glEnable(GL_DEPTH_TEST)
> glDepthFunc(GL_LESS)
> glDepthRange(0,1)
> 
> How can I force Opengl to utilize ZBuffer?

pygame has a function to control extra setting for opengl. it must be
called before display.set_mode(). you'd probably want to pass something
like "pygame.display.gl_set_attribute(GL_DEPTH_SIZE, 16)". perhaps try
different sized depths?

http://pygame.org/docs/ref/pygame_display.html#gl_set_attribute


you might also try testing with GLUT, which usually builds with pyopengl.



From thoa0025 at mail.usyd.edu.au  Tue Sep 17 08:11:55 2002
From: thoa0025 at mail.usyd.edu.au (:B nerdy)
Date: Tue, 17 Sep 2002 12:11:55 GMT
Subject: raw post data
References: <Wt8h9.9583$Ee4.22928@news-server.bigpond.net.au> <w_fh9.113093$pX1.4054147@news2.tin.it> <gOxh9.13698$Ee4.40115@news-server.bigpond.net.au> <siAh9.118475$pX1.4269024@news2.tin.it>
Message-ID: <f_Eh9.15761$Ee4.46455@news-server.bigpond.net.au>

yeah good guess! and thats something that looks like it will werk =B

thanks

"Alex Martelli" <aleax at aleax.it> wrote in message
news:siAh9.118475$pX1.4269024 at news2.tin.it...
> :B nerdy wrote:
>
> > how could i make it so when i read the stdin it still is there.. cause i
> > use cgi.FormVariables() also
>
> When you instantiate class cgi.FieldStorage (there's no such
> thing as cgi.FormVariables, so I'm just guessing at what it IS
> that you want!) you can pass it an optional first parameter
> that is a file object open for reading -- it uses sys.stdin
> as the default if you don't pass the first parameter.
>
> Thus, the following should work (untested!):
>
> import cgi, sys, cStringIO
>
> copyInput = cStringIO.StringIO(sys.stdin.read())
> fieldStorage = cgi.FieldStorage(copyInput)
>
>
> The raw posted data are now available as the string returned
> by calling copyInput.getvalue(), and are also parsed by cgi to
> obtain fieldStorage.
>
>
> Alex
>




From robin at jessikat.fsnet.co.uk  Thu Sep  5 10:57:13 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 5 Sep 2002 15:57:13 +0100
Subject: jython memory usage
References: <hVm5ydAY2hd9Eww$@jessikat.demon.co.uk>
 <3d76366a$1_5@news.bluewin.ch> <rnLbuKAv5jd9Ewzz@jessikat.demon.co.uk>
 <3d76494d$1_5@news.bluewin.ch> <YR13DVAUNld9EwXH@jessikat.demon.co.uk>
 <3d765ce6_4@news.bluewin.ch> <zzKd7KAqhnd9EwxT@jessikat.fsnet.co.uk>
 <3d767c4a$1_1@news.bluewin.ch> <YWyHyIAJKyd9EwHA@jessikat.demon.co.uk>
 <3d774f40$1_1@news.bluewin.ch>
Message-ID: <RTiVzIAJD3d9Ew2t@jessikat.demon.co.uk>

In article <3d774f40$1_1 at news.bluewin.ch>, Samuele Pedroni <pedronis at bluewin.ch> writes
>Ok, I have discovered an undocumented fact and maybe relevant fact.
.......
>Now one can log class loading, you can edit jython.bat and
>put -verbose:class in there:
>
>"..java.exe" -verbose:class "-Dpython.home=..."
>
>then running jython.bat <the-mem-overflowing-code>
>
>you should get an initial burst of [Loaded <classname>]
>and then normally nothing or a pattern of
>


OK I get over 600 lines at the initial burst which ends with
.......
[Loaded sun.reflect.GeneratedMethodAccessor18]
[Loaded sun.reflect.GeneratedMethodAccessor19]
[Loaded org.python.modules.MD5Object]
[Loaded reportlab.lib.fonts$py]
[Loaded java.io.RandomAccessFile from C:\Program Files\Java\j2re1.4.0_01\lib\rt.jar]
[Loaded sun.reflect.GeneratedMethodAccessor20]

memory usage climbs continuously and then
Exception in thread "CompileThread0" java.lang.OutOfMemoryError: requested 32756 bytes
C:\jython-2.1\reportlab\test>

>[Loaded ...]
>...
>[Unloading class ...]
>
>If you get an increasing sequence of [Loaded ...] without corresponding
>[Unloading class ...] then we have reduced
>the possible suspects.
>
>regards.

-- 
Robin Becker


From edream at tds.net  Fri Sep 20 09:36:23 2002
From: edream at tds.net (Edward K. Ream)
Date: Fri, 20 Sep 2002 13:36:23 GMT
Subject: Setting cursor colors in tkinter
Message-ID: <rvFi9.926$fO.166520@kent.svc.tds.net>

Hi,

I would like to set the foreground an background colors of a cursor.

Following:

http://www.tcl.tk/man/tcl8.3/TkLib/GetCursor.htm#M4

I am synthesizing a cursor description with

cursor="xterm " + fg + " " + bg

yielding, e.g.,

cursor = "xterm red green"

I then call:

body.configure(cursor=cursor)

However, Tcl complains:

    self.tk.call((self._w, 'configure')
TclError: bad cursor spec "xterm red green"

Can anyone say how this should properly be done?  Put another way, I may be
asking what the proper format of

"name [fgColor [bgColor]]"

is in the URL above.

Thanks.

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From marklists at mceahern.com  Thu Sep 12 13:42:22 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 12 Sep 2002 12:42:22 -0500
Subject: dynamic includes from classes...
In-Reply-To: <Vz3g9.246505$kp.845705@rwcrnsc52.ops.asp.att.net>
Message-ID: <JHEOKEOOLIGLDHCMAHMOKEDGDOAA.marklists@mceahern.com>

[mike]
> First, is there a better way than exec to dynamically include a module?
> i.e.
> importvar = "foo"
> exec("from %s import %s" % (importvar, importvar))

s = "sys"
sys = __import__(s)
print sys.path

// m
-




From shagshag13 at yahoo.fr  Fri Sep 20 10:40:18 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Fri, 20 Sep 2002 16:40:18 +0200
Subject: bit operations ?
Message-ID: <amfbuq$56f28$1@ID-146704.news.dfncis.de>

hello,

i'm looking for any libs which could handle bit operations on integer.
(i must be able to get 0 to n bits from any int., only bit j, etc.)

thanks in advance,
s13.







From aunueco at veritas.com  Sat Sep 21 06:10:16 2002
From: aunueco at veritas.com (Allen Unueco)
Date: 21 Sep 2002 03:10:16 -0700
Subject: Creating a standalone embedded Python app.
Message-ID: <f0eefec9.0209210210.56067a8f@posting.google.com>

I'm looking for a way to somehow bundle the Python modules that my
embedded application uses.  I would like to be able to distribute an
executable file that would work without having Python installed on the
system.

I would rather have everything in one file but having the modules in a
separate bundle (pickle?) would be workable.

Is there any clever way to do this?

Any help would be very appreciated.

-allen


From kseehof at neuralintegrator.com  Tue Sep 10 19:32:02 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Tue, 10 Sep 2002 16:32:02 -0700
Subject: COM, Outlook, and constants
Message-ID: <4.3.1.0.20020910162432.00c10c80@pop.hurrah.com>

I just upgraded my windows box to XP, and resumed working on a Outlook COM 
project
that I haven't touched for a few weeks.  For some reason it doesn't work 
anymore.
Anyway, the symptom I am trying to resolve now is that 
win32com.client.constants
doesn't seem to work anymore.  It's supposed to acquire attributes for all 
constants
defined in loaded COM servers through its __dicts__ attribute.

The following is typical of python COM client code.  Any hints?

 >>> import win32com.client
 >>> ol = win32com.client.Dispatch("Outlook.Application")
 >>> ns = ol.GetNamespace("MAPI")
 >>> oc = win32com.client.constants
 >>> oc.olFolderInbox
Traceback (most recent call last):
   File "<interactive input>", line 1, in ?
   File "C:\Python22\lib\site-packages\win32com\client\__init__.py", line 
131, in __getattr__
     raise AttributeError, a
AttributeError: olFolderInbox
 >>> oc.__dicts__
[]
 >>>

Thanks,
- Ken Seehof





From wlfraed at ix.netcom.com  Thu Sep 19 17:04:44 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Thu, 19 Sep 2002 14:04:44 -0700
Subject: execfile syntax & zope error
References: <a65110ed.0209190107.54b4ae46@posting.google.com>
Message-ID: <c5edma.2p3.ln@ix.netcom.com>

dd fed this fish to the penguins on Thursday 19 September 2002 02:07 am:

> Hi,
> 
> Im having problem using execfile.
> 
> Using execfile('filename.exe') on zope for web purpose, generate an
> error:
> 
> An error has occurred. The Site Admistrator has been notified.
> global name 'execfile' is not defined and Traceback (innermost last)
>

        Doesn't Zope run python in bastion (?) mode -- IE, restricted 
execution to prevent access to non-Zope file space?


--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From sismex01 at hebmex.com  Wed Sep 25 11:35:39 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Wed, 25 Sep 2002 10:35:39 -0500
Subject: parsing from a position found with a regexp
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233025D46A6@mail.hebmex.com>

> 
> Hi,
> 
> I would like to parse only a portion of a particular
> file. The portion I am interested in starts after the
> last occurrence of a particular string in the file.
> Ideally, I was hoping there might be a function
> (perhaps like the string rfind function, but for
> files) that would make this quick and simple. Any
> suggestions?
> 
> Thanks,
> Mark
> 

Dealing with text files, in a random-access fashion, is
never "quick and simple" :-(

But, something like the following might help?

def file_rfind(fp, string):
   last_pos = -1
   pos = fp.tell()
   for line in fp:
      if line.find(string) > -1: last_pos = pos
      pos += len(line)
   return last_pos


I haven't tested it though, but the final solution won't
be too far away, conceptually. :-/  AND, remember that
text files in M$ operating systems are 2-chars, and on
unix and mac-os based systems are 1-char, and this can
cause problems when opening files in text mode.

Good luck :-)

-gustavo





Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From gmuller at worldonline.nl  Wed Sep 18 14:23:31 2002
From: gmuller at worldonline.nl (GerritM)
Date: Wed, 18 Sep 2002 20:23:31 +0200
Subject: MarketProductLifecycle
References: <am818c$pam$1@reader1.tiscali.nl> <am81ff$pdh$1@reader1.tiscali.nl> <cVNh9.15168$T4.10611@fe04>
Message-ID: <amagf0$2h2$1@reader1.tiscali.nl>

"Steve Holden" <sholden at holdenweb.com> schreef in bericht
news:cVNh9.15168$T4.10611 at fe04...
> "GerritM" <gmuller at worldonline.nl> wrote in message
> news:am81ff$pdh$1 at reader1.tiscali.nl...
> >
> > Oops pressed the send key before verifying the destination...
> >
> > This does not belong here, but how to cancel from OE5?
> >
>
> Select your message (you have to be able to see it on your server...) and
> select "cancel message" from the "Message" menu.
>
> regards
> -----------------------------------------------------------------------
> Steve Holden                                  http://www.holdenweb.com/
> Python Web Programming                 http://pydish.holdenweb.com/pwp/
> Previous .sig file retired to                    www.homeforoldsigs.com
> -----------------------------------------------------------------------
>
thanks, I already succeeded in cancelling. The file which was attached is
the source version of the (very preliminary) presentation at:
http://www.extra.research.philips.com/natlab/sysarch/MarketProductLifecycleS
lides.pdf
which has no relation with Python, except that generation of documents and
webpages are automated via Python.

regards Gerrit

--
www.extra.research.philips.com/natlab/sysarch/





From no at mail.com  Mon Sep 16 19:24:05 2002
From: no at mail.com (BT)
Date: Mon, 16 Sep 2002 23:24:05 GMT
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de><pUdh9.9377$yt3.4604908@newssrv26.news.prodigy.com><mailman.1032184821.5180.python-list@python.org> <c%qh9.102794$ub2.2120685@news1.tin.it> <mailman.1032212180.11973.python-list@python.org>
Message-ID: <pKth9.118182$pX1.4250378@news2.tin.it>

> > I really don't understand posts like these.
> > You think I'm fool to protect my work?
>
> No, only a bit uninformed if you think you can get "security through
> obscurity".  Obscuring the code will dissuade only the most casual or
> incompetent from decompiling/disassembling your code.  Sort of like
> hiding a spare key to your house under the "Welcome" mat.

I was talking to people who suggest to "seek psychiatric help" if you want
to protect your code.
I know that no protection is unattacable but, I think it's better to protect
something instead of
release source code and write in a .txt file "please, don't copy it".

> > I can give you the phone of my clinic, but stop it NOW, join the ideal
> > world.
>
> You mean the "real" world.

No, I meant "ideal": I was persuading Id people to join open source
ideology.

> This question ("how can I obscure Python byte-code?") comes up at least
> once a month on this list.  It's fairly clear that any scheme you come
> up with will take longer to implement than it will to break by a
> competent hacker.

True.. that's why I'm planning to obfuscate the code then encrypt it.
A hacker (not every programmer) is specialized in modify jmp instructions
and could find the password
to decrypt the .pyc in less than 20 minutes, but maybe 5% of them knows
about python.
Then obfuscated source should persuade everyone to steal the code: sometimes
it's difficoult to modify
even a source with comments.

> Anyway, good luck!  If your software is any better than your protection
> scheme I'll download it from http://www.warez.com when you release it ;)

I'd like it: it would means my software it's worth to crack.




From cben at tx.technion.ac.il  Thu Sep 12 14:50:36 2002
From: cben at tx.technion.ac.il (Beni Cherniavsky)
Date: 12 Sep 2002 11:50:36 -0700
Subject: "str"(expr)"str" interpolation syntax
References: <mailman.1031688793.11971.python-list@python.org>
Message-ID: <f7930f23.0209121050.5c1870e3@posting.google.com>

"Mark McEahern" <marklists at mceahern.com> wrote on 2002-09-10:
> Beni Cherniavsky wrote:
> > I was searching for a new interpolation syntax ...
> 
> <applying the brakes>
> 
> You might be interested to:
> 
> 1.  Checkout the PEPs for string interpolation and simple string
> substitution.
> 2.  Read this thread:
> 
> http://groups.google.com/
> groups?selm=ouit4e63fr.fsf%40andlx-anamika.cisco.com
> 
> 3.  Read the archives from Python-Dev regarding the relevant PEPs.
> 
Read them all before (except py-dev but I dind't find much new there now).

> After you do all that, you may reconsider or reframe your suggestion in
> light of what has been discussed.
> 
To tell the truth my idea looked nice while I tossed it in my head but when
I actually wrote the post I noticed that something doesn't look well...
The "Bleagh" reply said it best though perhaps a bit strongly :-)
Anyway I wanted to get feedback on how it looks to others.

As for the motivation [to answer the other posts]: I'm a happy user of
"%d" % x but I was targeting here the cases of long strings and other
reasons why PEPs 215 and 292 came to be.  I didn't like (and still don't)
the idea that I need to scan string literals for _code_.  Especially if I
might want to nest strings inside that code...  It can confuse editors but
that's a sign it can also confuse me.  I thought for a moment that I found
a nice solution but it turns out not to be nice...  Still I don't like
interpolations as in the PEPs.  [Never liked double quotes in perl/shell
either].  I'd rather stick to %s...

Maybe something very simple along the lines of:

subst("A template string - here: % it goes\n", foo,
      "Also multiple values: %, %\n", bar, quux,
      "Quoting percent (%) sings probably not needed...\n", '%')

could be useful...  Yes it smells like perl formats but the
intelacing of pictures/values is really near optimal IMHO.  I've just
read perl 6's plans for formats and they have hyphenation!  Beat me
if I want that in a language core ;-)


From bobx at linuxmail.org  Mon Sep 16 16:42:52 2002
From: bobx at linuxmail.org (Bob X)
Date: Mon, 16 Sep 2002 20:42:52 GMT
Subject: [ANN] PySQLite 0.3.0 released
References: <mailman.1031924237.12603.python-list@python.org> <a05h9.6196$Lg2.1046166@news2.news.adelphia.net> <Wyjh9.1922$Cf.1588@fe03>
Message-ID: <gnrh9.7074$Lg2.1331402@news2.news.adelphia.net>

Steve Holden wrote:
<snip>
> 
> No, his two-paragraph "sig" is a PITA, but probably it's not aa sig but
> something that his company's mailer sticks on the end of all message quite
> gratuitously, just so some corporate lawyer can sleep better (fancying his
> ass is covered).
> 
> regards
> -----------------------------------------------------------------------
> Steve Holden                                  http://www.holdenweb.com/
> Python Web Programming                 http://pydish.holdenweb.com/pwp/
> Previous .sig file retired to                    www.homeforoldsigs.com
> -----------------------------------------------------------------------
>  
And here I thought a PITA was a sandwich!  :-)

Bob




From brueckd at tbye.com  Fri Sep  6 18:36:14 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Fri, 6 Sep 2002 15:36:14 -0700 (PDT)
Subject: Larry Wall's comment on python...
In-Reply-To: <pc5bla.off.ln@boundary.tundraware.com>
Message-ID: <Pine.LNX.4.44.0209061534290.1578-100000@ns0.tbye.com>

On 6 Sep 2002, Tim Daneliuk wrote:

> > The very idea of Larry Wall lecturing on code readability is hilarious.
> > 
> > -Dave

> Well, that's true, but for all of its similarity to line noise, perl has
> been an enormous contribution to our profession/community. I would never want
> to diminish Wall's stature, therefore.

Me neither, and I didn't intend it that way. There's just something very 
pot-calling-the-kettle-black about it, that's all. 

-Dave




From mstenner at phy.duke.edu  Fri Sep 13 16:46:12 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Fri, 13 Sep 2002 16:46:12 -0400
Subject: Coding Style: Quotes
In-Reply-To: <altgev$10u79$1@ID-77524.news.dfncis.de>; from thorsten@thorstenkampe.de on Fri, Sep 13, 2002 at 10:05:45PM +0200
References: <bt1g9.17429$Pf7.590136@news1.west.cox.net> <altgev$10u79$1@ID-77524.news.dfncis.de>
Message-ID: <20020913164612.A14152@phy.duke.edu>

On Fri, Sep 13, 2002 at 10:05:45PM +0200, Thorsten Kampe wrote:
> * John Waycott
> > I'm curious if others have adopted any standards for choice of single vs.
> > double quotation marks to delimit strings. A look through the standard
> > library reveals the choice of one over the other is rather arbitrary. I
> > suspect it really makes no difference, but the question has come up during a
> > code review.
> 
> Python itself seems to "prefer" single quotes slightly:
> 1: >>> "String"
> 1:     'String'
> 2: >>>

I hate to be so picky, but it _is_ after all, my nature... :)

Just because it uses the single quote doesn't mean it prefers it.  It
(or the authors, rather) had to choose.  They couldn't just not pick
one.  Also, the method you use to input the string IS NOT part of the
string, and so should be dropped after it's role (encoding the string)
is complete.  Therefore, your example does not support your statement.

Of course, I suspect it is for just this reason that you quoted
"prefer". :)
					-Michael
-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From phoebe_1 at att.net  Tue Sep 24 20:43:40 2002
From: phoebe_1 at att.net (Holden Caulfield)
Date: 24 Sep 2002 17:43:40 -0700
Subject: Solaris 64 bit compilation
References: <c2595393.0209201844.2ed62265@posting.google.com> <m3sn03k44m.fsf@mira.informatik.hu-berlin.de> <c2595393.0209210627.44f672a8@posting.google.com> <m3d6r7i9o0.fsf@mira.informatik.hu-berlin.de> <c2595393.0209240941.5f3fc565@posting.google.com> <j4fzvzclfb.fsf@informatik.hu-berlin.de>
Message-ID: <c2595393.0209241643.5b276788@posting.google.com>

loewis at informatik.hu-berlin.de (Martin v. L?wis wrote in message news:<j4fzvzclfb.fsf at informatik.hu-berlin.de>...
> phoebe_1 at att.net (Holden Caulfield) writes:
> 
> > > > "./Include/pyport.h", line 480: #error: "LONG_BIT definition appears
> > > > wrong for platform (bad gcc/glibc config?)."
> > > 
> > > So what value has SIZEOF_LONG in pyconfig.h?
>  [...]
> > To answer your questions:
> >   - pyconfig.h does say 8 as SIZEOF_LONG
> 
> That raises the question then why the error in pyport.h is
> produced. LONG_BIT comes from limits.h, and, if _LP64 is defined, is
> 64. So I'm still uncertain what you did to provoke this error.
> 
> >   - 'curses' modules does not compile. 
> >   - 'dbm' modules compiles fine but core dumps during test.
> > So, I disabled these modules in 'setup.py' and built it and ran the
> > make test.
> 
> curses fails for me as well, since it finds a 32-bit ncurses
> installation; dbm also core dumps here.
> 
> > test test_time failed -- Traceback (most recent call last):
> >   File "./Lib/test/test_time.py", line 46, in test_mktime
> >     time.mktime, (999999, 999999, 999999, 999999,
> >   File "/apps/python/build/Python-2.2.1/Lib/unittest.py", line 279, in
> > failUnlessRaises
> >     raise self.failureException, excName
> > AssertionError: OverflowError
> 
> This is a known bug (#460357), which has been fixed in CVS.
> 
> > 2)
> > test_gettext
> > test test_gettext crashed -- exceptions.OSError: [Errno 20] Not a
> > directory: 'xx/LC_MESSAGES/gettext.mo'
> 
> I believe this also has been fixed in CVS.
> 
> > Are there any plans to support 64-bit python executable/module on
> > solaris platforms for future versions?
> 
> Building python in 64-bit mode on Solaris works already. Some modules
> may still not work; patches are welcome.
> 
> > Is anyone even interested in the configure.in patches?
> 
> I don't really see a need, since it works fine without the patches.
> 
  Well, I will have to disagree here. As currently at lease in V2.2.1 vanilla
  source there is *no* option to built in 64 bit mode, unless you hack the
  CFLAGS, LDSHARED etc.
  Unless, you know of way to pass the appropriate flags? 
  Because my expectation was that I could build 64bit excutable/object/module
  by running "configure", currently out of the box I could *not* figure out a
  way to do this. 

> Regards,
> Martin


From emile at fenx.com  Sun Sep  1 21:17:25 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 1 Sep 2002 18:17:25 -0700
Subject: Add methods to a class at runtime?
References: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com> <slrnan4dl3.18h.gerhard.haering@lilith.my-fqdn.de> <wmvc9.21757$FJ1.3181934@e3500-atl1.usenetserver.com> <mailman.1030916762.5578.python-list@python.org> <Eaxc9.33161$vY2.738637@e3500-atl2.usenetserver.com>
Message-ID: <akue62$v7m$1@main.gmane.org>

Robert:
> (wow what a great password this link would make):
>
>
http://groups.google.com/groups?hl=en&lr=lang_en&ie=UTF-8&oe=UTF-8&safe=acti
>
ve&threadm=mailman.997727540.19628.python-list%40python.org&rnum=1&prev=/gro
>
ups%3Fhl%3Den%26lr%3Dlang_en%26ie%3DUTF-8%26oe%3DUTF-8%26safe%3Dactive%26q%3
> DRe%253A%2BCopy%2Bconstructors%2Bpython
>

OT

A trick to shortening a groups.google link is to identify the thread or
message identifer and the specific message id and just pass that part.
These are the segments that start 'selm=' or 'threadm='.  In the reference
above, it's this part:

threadm=mailman.997727540.19628.python-list%40python.org

then put http://groups.google.com/groups? in the front to get:

http://groups.google.com/groups?threadm=mailman.997727540.19628.python-list%
40python.org


Emile van Sebille
emile at fenx.com







From gmcm at hypernet.com  Wed Sep 11 18:37:16 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 11 Sep 2002 22:37:16 GMT
Subject: Building a python executable archive
References: <6c20157a.0209110826.3e98a77@posting.google.com>
Message-ID: <Xns9286BD7ED6CE1gmcmhypernetcom@199.171.54.216>

[posted and mailed]

Rich Browne wrote:

> I'm trying to build python as an archive bound executable.  
> 
> This sounds dumb, Python is inherently dynamic, so why am I doing
> this?
> Because the python executable will have to execute without any dynamic
> dependencies on shared libs outside the build environment that I plan
> to install it in.
> 
> I have succeeded in statically linking most of the shared objects...

snip

> ... This works, but I'm looking for an
> easier(automated)way to do this.  


Have you looked at whether Installer
 http://www.mcmillan-inc.com/install1.html
can be made to suit your needs?

It's not statically linked, so you have to control the
shared libs along with the exe, but it's designed to operate
from only it's own directory tree, (so you can ship an app
built from one version of Python without fear of conflicts
with an installed version).

> Also, what is the difference between the *.pyo and *.pyc files.

At this point, not much (the pyo's omit line numbers and maybe
doc strings?). The goal is to have pyo's be the product of an
optimizing compiler.

-- Gordon
http://www.mcmillan-inc.com/




From jdhunter at ace.bsd.uchicago.edu  Thu Sep 26 19:07:32 2002
From: jdhunter at ace.bsd.uchicago.edu (John Hunter)
Date: Thu, 26 Sep 2002 18:07:32 -0500
Subject: indexing and searching pdf files
In-Reply-To: <pan.2002.09.26.18.05.37.70342.4621@presidency.com> (Rajarshi
 Guha's message of "Thu, 26 Sep 2002 18:05:37 -0400")
References: <pan.2002.09.26.18.05.37.70342.4621@presidency.com>
Message-ID: <m27kh88i4r.fsf@mother.paradise.lost>

>>>>> "Rajarshi" == Rajarshi Guha <rajarshi at presidency.com> writes:

    Rajarshi> Hi, I have a load of pdf files and I would like to index
    Rajarshi> them so that I can serach them for keywords. I was
    Rajarshi> thinking of using pdftotext to generate the textfile and
    Rajarshi> then create the index from that.

As far as I have been able to tell, the pdftotext approach is about as
good as any.  Darrell Gallion was kind enough to forward a Plex parser
to me that he had been working on which I'm sure he would send to you
if you asked: dgallion1 at yahoo.com.  I haven't had a chance to look at
it yet.

You could also consider exposing your pdf dirs to a web server, and
induce google to crawl it.  You could then use their advanced pdf
searching, cacheing, html-izing and indexing capabilities by doing an
advanced search restricted to your server.  You could use pygoogle and
do the same if you wanted to stay within the fold.

JDH



From paul at boddie.net  Mon Sep 30 09:54:34 2002
From: paul at boddie.net (Paul Boddie)
Date: 30 Sep 2002 06:54:34 -0700
Subject: DBI cursor behaviour with multiple statements?
References: <7e964d1d.0209270137.568a3f3@posting.google.com>
Message-ID: <23891c90.0209300554.5b37c195@posting.google.com>

Henrik.Weber at sys.aok.de (Henrik Weber) wrote in message news:<7e964d1d.0209270137.568a3f3 at posting.google.com>...
> Hello,
> 
> I'm trying to write a DBI2 compliant module for Microsoft Access
> databases. Now I'm not quite sure how to interpret the DBI2
> specification when it comes to the execution of multiple data
> returning statements with the same cursor.
> 
> For example if c is my cursor and I do:
> c.executemany("SELECT * FROM table WHERE column = '%s'", [["parm1"],
> ["parm2"], ["parm3"]])

Note that you probably don't want the quoting inside the SQL
statement. In other words, it should be...

  "SELECT * FROM table WHERE column = %s"

Personally, I'd recommend supporting different parameter notations,
since this issue always confuses people.

> should the result be the union of all three statements or just the
> result of the first or the last statement? Or should the user switch
> from one resultset to the next with nextset?

An interesting, related issue is the treatment of the IN operator, and
this doesn't necessarily yield uniform treatment across database
modules, even for the same database system (compare pyPgSQL and
psycopg on PostgreSQL). For some applications of your query (where you
want the union), the IN operator is probably what is desired:

  "SELECT * FROM table WHERE column IN %s"

This appears highly counter-intuitive, of course, since the
Python-style %s parameter marker actually represents a list in this
case - another reason for choosing a better parameter notation.

Paul


From jlselsewhere at my-deja.com  Thu Sep  5 16:24:04 2002
From: jlselsewhere at my-deja.com (J. L. Schilling)
Date: 5 Sep 2002 13:24:04 -0700
Subject: Memory fault - core dumped   while compiling python 2.2.1 on  SCO
References: <3D757EFA.9050401@nospam.yahoo.com> <yu99sn0qqw1r.fsf@europa.research.att.com> <m3bs7etem9.fsf@mira.informatik.hu-berlin.de> <3D765A05.6030004@nospam.yahoo.com> <j48z2h61j9.fsf@informatik.hu-berlin.de> <3D768753.8030302@nospam.yahoo.com> <m3n0qx8j3w.fsf@mira.informatik.hu-berlin.de> <3D76A139.1030304@nospam.yahoo.com> <al69pn$sg$0@216.39.172.122>
Message-ID: <ff3c0649.0209051224.3da59b27@posting.google.com>

bokr at oz.net (Bengt Richter) wrote in message 
> > [...]
>
> How about
>     cc -v
> ?

Just for the record, cc -V is the option on the SCO compiler
to report its version.  It should show something like this
on an OSR 5.0.5 system:

217$ /bin/cc -V
SCO UNIX Development System  Release 5.1.1A 27Jul98
Usage: cc [ options ] files ...

The cc -v option enables lint-like warnings.

Jonathan Schilling


From python at rcn.com  Mon Sep  9 23:45:35 2002
From: python at rcn.com (Raymond Hettinger)
Date: Mon, 9 Sep 2002 23:45:35 -0400
Subject: dynamism
References: <66715c02.0209091923.265fe77e@posting.google.com>
Message-ID: <aljpvd$5qk$1@bob.news.rcn.net>

"Steven Shaw" <steven_shaw at adc.com> wrote in message
> I thought that you could access dictionary values with object syntax:
>
> >>> x = { 'a': 1, 'b':2 }
> >>> x
> {'a': 1, 'b': 2}
> >>> x.a
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> AttributeError: 'dict' object has no attribute 'a'
>
> Maybe you could never do this as shown but I'm sure there were ways of
getting
> this effect. Anyone care to remind me?

>>> class AttrDict: pass

>>> ad = AttrDict()
>>> ad.__dict__ = {'a': 1, 'b':2}
>>> ad.a
1
>>> ad.b
2

>>> class BetterAttrDict(dict):
 def __getattr__(self, key):
  return dict.__getitem__(self, key)


>>> bad = BetterAttrDict({'a': 1, 'b':2})
>>> bad.a
1
>>> bad.b
2


Raymond Hettinger




From donn at drizzle.com  Wed Sep 25 00:31:40 2002
From: donn at drizzle.com (Donn Cave)
Date: Wed, 25 Sep 2002 04:31:40 -0000
Subject: popen2() gives broken pipe for large files - is it fixed yet?
References: None <mailman.1032879997.20241.python-list@python.org>
Message-ID: <1032928299.406088@yasure>

Quoth Frank Gibbons <fgibbons at hms.harvard.edu>:
...
| It worked fine for a small test file, but I get a broken pipe for realistic 
| sizes (1MB). I've written a little test script that shows the same behavior 
| (see below). I've searched the newsgroups, and find that popen2() has a 
| history of breaking the pipe on large files. Does anyone know if this has 
| been fixed? Is popen3 any better? I've also read that the os module now has 
| a popen function (as of Python 2.x, forget which x). Is this any better?
...
| data = f.readlines()
...
| zip_output, zip_input = popen2.popen2("%s -c  "% GZIP)
| data = string.join(data, "")
...
| zip_input.write(data)
...
| zipfile = open("popen.gz", "w+")
| zipfile.write(zip_output.read())

It isn't going to be fixed, either, because there are just some
axiomatic truths in operation here, like "I/O buffers aren't
infinitely elastic." Since your example doesn't do anything with
the output, you don't have to shoot yourself in the foot this way.
Try something like this -

  zip_input = os.popen('gzip -c > popen.gz', 'w')
  while 1:
      data = f.read(16000)
      if not data:
          break
      zip_input.write(data)
  zip_input.close()

The main point is that this allows gzip to write directly to disk.
If your actual application needs to process the data further at this
point, you may read it back from disk and write to a new file if you
like.   Otherwise, if you really want a second pipe, you have to figure
out how to read some, write some, etc., to keep gzip's output from
filling up while you're trying to write all its input.

	Donn Cave, donn at drizzle.com


From pjdurai at hotmail.com  Mon Sep 30 14:12:21 2002
From: pjdurai at hotmail.com (pjd)
Date: 30 Sep 2002 18:12:21 GMT
Subject: embedding question: cant import some modules
Message-ID: <ana465$rsk$0@198.102.102.57>

Hi

I am embedding Python 22 in a C++ application.

Everything works fine except when I import a module like win32api, which is
installed in C:\Python22\Lib\site-packages\win32, it complains that it cant
find the module.

The sample script I am executing works fine when executed from the shell
with Python.exe.

But failes with module not found, when executed from my C++ app.

I have tried to set PYTHONPATH( = C:\Python22\Lib\site-packages\win32), and
PYTHONHOME to various values . No avail.

Is there something I need to do inside my app, to make it find the installed
modules ?

Any ideas ?

Appreciate your time.

cheers
pj





From sross at connectmail.carleton.ca  Wed Sep 18 09:07:11 2002
From: sross at connectmail.carleton.ca (Sean Ross)
Date: 18 Sep 2002 13:07:11 GMT
Subject: Splitting a list of strings
References: <mailman.1032309381.3626.python-list@python.org> <am8pvo$dcn$1@driftwood.ccs.carleton.ca> <w0Uh9.116763$5r1.4544705@bin5.nnrp.aus1.giganews.com>
Message-ID: <am9tpv$9gp$1@driftwood.ccs.carleton.ca>

Terry:
Thank you. However, I do not decide what format will be applied to the data,
I operate on what is provided. Moreover, the data really is not what I'm
interested in. (Mark asked to see a sample, so I provided it.). Your
suggestion is certainly true and I would follow it, if I could.

What I was interested in was finding a way to do a three pronged filter of a
list of strings using a single list comprehension. Sort of like a change
filter, where you drop handfuls of change into the top of the device and
they come out the other side sorted into pennies, nickels, dimes, quarters.
In this case I had comments, attributes, and data. The comments could be
redirected to a chaff outlet, i.e., ignored, so long as the attributes and
data where filtered properly.
The suggestions people have provided are fine (using an attribute and data
list, marching through the list of lines from the file and sticking an
attribute string into the attribute list and a data string into the data
list). They do what I need to accomplish, but they are not quite what I was
looking for. Think of it this way, if I have a pile of pennies, nickels, and
dimes, and pull out all the nickels, then I have a pile of nickels and a
pile of pennies and dimes. Using list comprehensions, I've been able to
create the pile of nickels, but the original pile is unchanged. That's as it
should be, I suppose. However, I guess what I was looking for was a
side-effect, or some such mechanism where, by extracting the nickels, I get
the two piles: one of nickels, one of pennies and dimes.
I want my cake, and to eat it too, in other words.
You can probably just chalk this all up to an infatuation with list
comprehensions.

Thanks again, to all, for your suggestions
Sean





From claird at starbase.neosoft.com  Wed Sep 18 09:25:05 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 18 Sep 2002 08:25:05 -0500
Subject: minimizing Tk programs to the taskbar
References: <3D8872F1.87850856@optushome.com.au> <Vz_h9.126208$pX1.4528221@news2.tin.it>
Message-ID: <656913E3D19D40D2.0831BD5165F06A87.FDA870F87AEF2207@lp.airnews.net>

In article <Vz_h9.126208$pX1.4528221 at news2.tin.it>,
Alex Martelli  <aleax at aleax.it> wrote:
>Richard Tardif wrote:
>
>> Is there any way to minimize programs to the taskbar? I'm talking about
>> windows here. By taskbar I mean the area down in the bottom right
>> corner, near the clock, that shows icons for what proggies you've got
>> running. Thanks for any help.
>
>I don't think Tkinter offers such an intrinsically non-cross-platform 
>functionality. I suspect therefore that you'd need to work this part 
>through the win32all add-ons, win32gui.Shell_NotifyIcon &tc.
			.
			.
			.
?  Alex and Richard, I think you'll want to experiment with
  root = Tkinter.Tk()
    ...
  Tkinter.Wm.iconify(root)

Alex' intuition that a lot of the operations people say they
want are platform- or window-manager-specific is certainly
apt.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From shalehperry at attbi.com  Wed Sep 25 03:40:12 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 25 Sep 2002 00:40:12 -0700
Subject: problems defining an empty list
In-Reply-To: <20020925070654.GB57911@partagas.as.de.cw.net>
References: <20020925070654.GB57911@partagas.as.de.cw.net>
Message-ID: <200209250040.12209.shalehperry@attbi.com>

On Wednesday 25 September 2002 00:06, Manuel Hendel wrote:
> This is what I'm trying to do:
>
>     # empty list for lines per domain
>     linesperdomain = []
>     linesperdomain = [domain, domainlines]
>     print linesperdomain
>
> domainlines is also a list.
>
> Error:
>
> linesperdomain = []
>                   ^
> IndentationError: unindent does not match any outer indentation level
>

it sounds like you have a hidden piece of whitespace.  Perhaps the code is 
indented when it should not be?  If the code is not in a loop, class, if 
structure, etc then it should not be indented.



From fb at ultranet.com  Mon Sep  9 20:47:31 2002
From: fb at ultranet.com (basegmez)
Date: Mon, 9 Sep 2002 20:47:31 -0400
Subject: returning values from FORTRAN with f2py
References: <c3835e5f.0209071122.338a787d@posting.google.com>
Message-ID: <aljg4r$osb$1@bob.news.rcn.net>

Chris Fonnesbeck wrote in message ...
>I am wondering how to persuade some fortran code that I have compiled
>with f2py to return its calculated values to me in python.  I have
>some code that does gaussian elimination (mostly from Numerical
>Reciopes), and is supposed to return an inverted matrix so that I can
>solve a system of linear equations.  However, when run in python, the
>function runs without error but returns no values:
>
>>>> a = [[2,1,3],[1,-1,1],[3,6,-1]]
>>>> b = [7,2,10]
>>>> import gauss
>>>> gauss.gauss(a,3,b,1)
>>>>
>
>Is there something that I have to change in my fortran code to make
>this work?
>
>Many thanks,
>cjf

I just used Numeric to do something similar.  I used Newton-Raphson to solve
a set of non-linear equations.

from __future__ import nested_scopes
from Numeric import array
from Numeric import matrixmultiply as mm
from Matrix import inverse
g = 386.08

''' Given Ix, Iy, Iz, and weight, calculates X, Y and Z dims of a
rectangular box
    uses Newton-Raphson to solve three simultaneous nonlinear equations'''
def dim_box(w, Ix, Iy, Iz, tol = .0001):
    m = w/g
    mc = m/12
    def f(x, y, z):   # Function
        return array([(y**2 + z**2)*mc - Ix, (x**2 + z**2)*mc - Iy, (x**2 +
y**2)*mc - Iz])
    def J(x, y, z):    # Jacobian
        return array([[0, 2*y*mc, 2*z*mc], [2*x*mc, 0, 2*z*mc], [2*x*mc,
2*y*mc, 0]])
    x, y, z = 1.8, 4, 5
    for i in range(200):
        res = mm(inverse(J(x, y, z)), -f(x, y, z)) + array([x, y, z])
        x, y, z = res[0], res[1], res[2]
        check = f(x, y, z)
        if abs(check[0]**2 + check[1]**2 + check[2]**2) <= tol:
            return x, y, z
    print 'No solution!', '\n', f(x, y, z), '\n', x, y, z

# Example:
# >>> dim_box(1000, 1600, 700, 1800)
# (45.6599607331 79.0853715939 34.0329252342)


Alternatively, you can use the solve_linear_equations from
LinearAlgebraModule.

from Numeric import array
from LinearAlgebra import solve_linear_equations as solve

#Example:
# a + b = 5
# 2a - b = 7
#should give
# [4., 1.]

LHS = array([[1, 1], [2, -1]])
RHS = array([5, 7])
print solve(LHS, RHS)


Fahri Basegmez








From yaipa at aol.com  Wed Sep 18 13:30:53 2002
From: yaipa at aol.com (yaipa)
Date: 18 Sep 2002 10:30:53 -0700
Subject: Is Python 2.1 for Solairs 8 broke on SunFreeware?
References: <8d148763.0209131005.65de04a8@posting.google.com> <j4k7lnrvd4.fsf@informatik.hu-berlin.de>
Message-ID: <8d148763.0209180930.56cd3f92@posting.google.com>

loewis at informatik.hu-berlin.de (Martin v. L?wis wrote in message news:<j4k7lnrvd4.fsf at informatik.hu-berlin.de>...
> yaipa at aol.com (yaipa) writes:
> 
> > % /usr/local/bin/python CronTar.py -c cronTar.cfg
> >   Traceback (most recent call last):
> >     File "/export/fruid/bin/CronTar.py", line 26, in ?
> >       import time
> >   ImportError: ld.so.1: /usr/local/bin/python: fatal: 
> >     libgcc_s.so.1: open failed: No such file or directory
> > %
> 
> That is somewhat unfortunate: gcc  3.2 links another shared library to
> every binary.
> 
> > Am I required to have gcc installed also?
>  [...]
> > Any ideas?
> 
> No need to do that. Just install the libgcc package from Sunfreeware
> also. This is much less than the full GCC.
> 
> Regards,
> Martin


Thanks Martin. I'll give that a try this weekend. One of those times when
I should have know, and if not, should have asked sooner. 

Thanks again.

Yaipa


From pedro_rodriguez at club-internet.fr  Mon Sep 30 09:02:59 2002
From: pedro_rodriguez at club-internet.fr (Pedro RODRIGUEZ)
Date: Mon, 30 Sep 2002 15:02:59 +0200
Subject: Find out link location
References: <7uf9na-vd9.ln@lbdb01.gsi.de>
Message-ID: <an9hpi$q4u$1@s1.read.news.oleane.net>

On Mon, 30 Sep 2002 14:26:47 +0200, Klaus Hoeppner wrote:

> with the os.path module I can use the islink function to check, wether a
> file is symbolic link. But how can I find out, where the link points to.
> E.g. if I have readme.txt --> ../packages/mypackage/readme.txt how do I
> get the "../packages/mypackage/readme.txt" with python?
> 

in http://www.python.org/doc/current/lib/os-file-dir.html :

os.readlink("readme.txt")

Pedro


From fgibbons at hms.harvard.edu  Tue Sep 24 11:04:03 2002
From: fgibbons at hms.harvard.edu (Frank Gibbons)
Date: Tue, 24 Sep 2002 11:04:03 -0400
Subject: popen2() gives broken pipe for large files - is it fixed yet?
Message-ID: <5.0.0.25.2.20020924105700.00ad1a08@hms.harvard.edu>

Hi,

I'm trying to allow users of my CGI script to download rather large data 
files by clicking a button. The solution which worked for small test files 
was this:

- Create a button whose action is to print the formatted data with the 
appropriate content-type header, so that the desired application will open.
- Create another button whose action is to open a pipe to gzip (using 
popen2()), write the same data into the pipe, then write the output from 
the pipe to the browser, again with appropriate content-type, to say that 
it's a zipped file.

It worked fine for a small test file, but I get a broken pipe for realistic 
sizes (1MB). I've written a little test script that shows the same behavior 
(see below). I've searched the newsgroups, and find that popen2() has a 
history of breaking the pipe on large files. Does anyone know if this has 
been fixed? Is popen3 any better? I've also read that the os module now has 
a popen function (as of Python 2.x, forget which x). Is this any better?

TIA,

Frank Gibbons

#!/usr/bin/python

import os
import popen2
import string
GZIP = "/bin/gzip"
filename = "ChIP_Squared_Template_FR2.txt"
f = open(filename)
data = f.readlines()
f.close()

print "Read %d lines" % len(data)
zip_output, zip_input = popen2.popen2("%s -c  "% GZIP)
data = string.join(data, "")
print "LEN(data) = %d" % len(data)
zip_input.write(data)
#zip_input.close()

zipfile = open("popen.gz", "w+")
zipfile.write(zip_output.read())
zipfile.flush()
zipfile.close()

PhD, Computational Biologist,
Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA.
Tel: 617-432-3555       Fax: 
617-432-3557       http://llama.med.harvard.edu/~fgibbons




From gerhard.haering at gmx.de  Fri Sep  6 02:45:29 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Fri, 6 Sep 2002 08:45:29 +0200
Subject: How to install informixdb-1.3 for python2.2 on RH7.3?
In-Reply-To: <c345ff6f.0209052222.37221ef1@posting.google.com>
References: <c345ff6f.0209042255.65dd5824@posting.google.com> <c345ff6f.0209052222.37221ef1@posting.google.com>
Message-ID: <20020906064529.GB614@lilith.ghaering.test>

* J Dubal <dubal at khimjiramdas.com> [2002-09-05 23:22 -0700]:
> Hello again!
> 
> I managed to compile and install informixdb module manually for py221.
> However, when I import it in python2, it gives following error:
> ImportError: Cannot locate C core module for informix db interface.

Normally, I write a setup.py using Python's distutils in such cases,
however informixdb seems to use some sort of SQL preprocessor for the
ext/_informixdb.ec file.

So, I'd try to perform the step with the SQL preprocessor manually, then
invoke "python setup.py build".

> If I try import _informixdb, it gives following error: ImportError:
> /opt/informix/lib/esql/libifgen.so: undefined symbol: stat

You need to link against a additional libraries. If the Informix
documentation is silent about this, I'd recommend to ldd on some client
utility of the database and add these libraries in the Makefile.

-- Gerhard



From marco at boo.it  Wed Sep  4 12:18:24 2002
From: marco at boo.it (kyro)
Date: Wed, 04 Sep 2002 16:18:24 GMT
Subject: Program for ... HELP!!!
Message-ID: <knqd9.25$%z1.943@twister1.libero.it>

Hello everybody, I need a little piece of program thath calculate the
determinant of a square matrix.
Can somebady help me?!??!?!?!

Thanks a lot

>>kyro<<

P.S.: I think to use the recursion, but I don't use thr recursion in Python,
I'm a newby....  :oP




From bkc at Murkworks.com  Mon Sep  9 18:15:22 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Mon, 9 Sep 2002 18:15:22 -0400
Subject: Larry Wall's comment on python...
References: <mailman.1031369470.18741.python-list@python.org>
Message-ID: <3d7d1ae6$1_1@goliath.newsgroups.com>

 "All is fair if you predeclare."

I declare, Larry is some kind of anti-Guido!






-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From tdelaney at avaya.com  Thu Sep 26 01:09:03 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 26 Sep 2002 15:09:03 +1000
Subject: revenge of the python puzzle
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5E3@natasha.auslabs.avaya.com>

> From: Michal Wallace [mailto:sabren at manifestation.com]
> 
> Okay, so you guys think you're so smart. Try this one!  No
> contest this time, I've learned my lesson.  (I suppose if
> someone really wants some free hosting bad enough to solve
> this puzzle, you could talk me into it, but I'm just posting
> this for revenge... Um, I mean "fun".)

Hmm - a little bit harder. The fake perl actually made it easier though
(gave while and s).

def decrypt(yourstring):
    dedict = {
        'a': 'm',
        'b': 'r',
        'c': 'y',
        'd': 'j',
        'e': 'b',
        'f': 'd',
        'g': 'l',
        'h': 'i',
        'i': 'e',
        'j': 'u',
        'k': 'w',
        'l': 'v',
        'm': 'o',
        'o': 'k',
        'p': 't',
        'q': 'g',
        's': 's',
        't': 'x',
        'u': 'p',
        'v': 'f',
        'w': 'c',
        'x': 'h',
        'y': 'n',
        'z': 'a',
    }
    return "".join([dedict.get(i,i) for i in yourstring])

data = r'''wbjig_ucpxmy, kxzp_xzli_h_fmyi \
    ='ilhg.zuuiyf(ghsp(izwxmyi))',\
"zkvjg wmfi hs sm ajwx vjy"
hv 'h za sazbp': 'hp djsp ahqxp bjy'
"hp's sm ajwx kmbsi pxzy 'xiggm kmbgf'"
"kxc, pxisi yitp ghyis zgamsp gmmo ghoi uibg!"

'kxhgi(<>){wxmau;s/uibg/ucpxmy;ubhyp}';mbzyqis='vbjhp';
ezf="'iyhpgzlm bjmc oyhbf'=qsa;qsa gzemgqtqsa pyhbu";
iaupc="";qgmezg izwxmyi;ilhg=[];

'ejp z siwbip xibi hs hy cmjb bizwx'
zuugis=mbzyqis, "izwxmyi=izwx"
fiv fm(pzso="fm kxzp h zso"):
     qgmezg izwxmyi; itiw pzso
"zyf djsp hy wzsi cmj'bi ymp uibugitif,"
kmbsi,ymk = (ezf.sughp('t'),"!")
vmb izwx hy kmbsi:
    itiw zuugis[1]
    hv izwx hy kmbsi:
        fm(wbjig_ucpxmy)

kmbsi = ilhg; "ilhg".sughp().bilibsi()
[izwx.bilibsi() vmb izwx hy kmbsi]
"zyf ymk ki'li bizwxif pxi vhyzg libsi";
azu(fm, azu(iaupc.dmhy, kmbsi))'''

print decrypt(data)



From johann.hoechtl at myrealbox.com  Mon Sep 30 07:28:46 2002
From: johann.hoechtl at myrealbox.com (Johann Höchtl)
Date: Mon, 30 Sep 2002 11:28:46 GMT
Subject: Stackless Python Howto
Message-ID: <1033385295.69247@charme.kabsi.at>

Hello!

Can someone give me a short briefing how to use stackless python? I thought
it's all about a replacement of python.dll, and as long as you don't use
stackless threading and continuations, at least arbitrary, lightweight
recursion will be available.

However, the famous ackermann function with parameters ack (3, 7) and
onwards still failes because of stack constraints.

What else do i have to specify to be go stackless?

Thank you,
    Johann




From edream at tds.net  Thu Sep 12 14:59:14 2002
From: edream at tds.net (Edward K. Ream)
Date: Thu, 12 Sep 2002 18:59:14 GMT
Subject: Cursed newlines and readline()
Message-ID: <6u5g9.11312$Da3.1133980@kent.svc.tds.net>

According to one of Leo's Linux users, the Python's readline() routine on
Linux delivers "\r\n" line end strings verbatim, while the windows versions
force the string to use the Unix convention of using only "\n". This seems
unfortunate.  Is there a justification for it?  If not, I would like to
suggest that readline() work the same on all platforms and that it deliver
only "\n" as a line terminator.

This is an example of what I have come to call the "cursed newline" problem,
a problem that seems trivial, and in fact has caused me, by far, the most
grief as a programmer trying to write cross-platform software.  BTW, TK
handles this problem superbly, so well that one might not even suspect that
it is a problem.

Any comments?

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From jggramlich at yahoo.com  Thu Sep  5 17:40:19 2002
From: jggramlich at yahoo.com (Joshua Gramlich)
Date: 5 Sep 2002 14:40:19 -0700
Subject: PyListObject & C Modules
References: <mailman.1031085786.12546.python-list@python.org> <m3k7m2tf0y.fsf@mira.informatik.hu-berlin.de> <bdade4c9.0209040823.3624d164@posting.google.com> <j41y89emce.fsf@informatik.hu-berlin.de>
Message-ID: <bdade4c9.0209051340.758a9ec1@posting.google.com>

Okay...I finally got it.

in:

if(!PyArg_ParseTuple(args, "O!", &list))
    return NULL;


The option "O!" doesn't work.  If you look at the
<sarcasm>tutorial</sarcasm>*  :D


http://www.python.org/doc/current/ext/parseTuple.html

...you'll see this:

"O" (object) [PyObject *] 
Store a Python object (without any conversion) in a C object pointer.
The C program thus receives the actual object that was passed. The
object's reference count is not increased. The pointer stored is not
NULL.

"O!" (object) [typeobject, PyObject *] 
Store a Python object in a C object pointer. This is similar to "O",
but takes two C arguments: the first is the address of a Python type
object, the second is the address of the C variable (of type PyObject
*) into which the object pointer is stored. If the Python object does
not have the required type, TypeError is raised.

Removing the '!' from the PyArg_ParseTuple solved the problem (before
it was returning TypeError), and when I ask for item 3 (really 4) I
get the following:


[d0ora3 ~/work/grid_query/test]$ python hello.py 
[0, 3, 8, 22, 9]
[22]
[d0ora3 ~/work/grid_query/test]$ 



THANKS FOR ALL THE HELP!!!


From martin at v.loewis.de  Tue Sep 10 14:38:34 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 10 Sep 2002 20:38:34 +0200
Subject: how to build wxPython w/ Sun's C++ compiler?
References: <mailman.1031607011.28525.python-list@python.org> <m3k7luojpq.fsf@mira.informatik.hu-berlin.de> <mailman.1031621231.16841.python-list@python.org> <23891c90.0209092359.11acb607@posting.google.com> <mailman.1031662034.25815.python-list@python.org>
Message-ID: <m3k7ltitat.fsf@mira.informatik.hu-berlin.de>

Skip Montanaro <skip at pobox.com> writes:

> Agreed.  However, disutils commands have this structure:
> 
>     python setup.py [ options ] command [ more options ] ...
> 
> which I think can get kind of confusing.

You forgot 

  python [ options ] setup.py [ options ] command [ more options ] ...

:-)

Regards,
Martin


From loewis at informatik.hu-berlin.de  Fri Sep 20 12:57:56 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 20 Sep 2002 18:57:56 +0200
Subject: Porting woes
References: <3a4b2539.0209200809.407d50d5@posting.google.com>
Message-ID: <j4sn041tuz.fsf@informatik.hu-berlin.de>

jstavnitzky at atsautomation.com (Jay) writes:

> This failure is occuring during the function classic_mro().  The
> stack at that point is
> classic_mro(),mro_implementation(),mro_internal(),PyType_Ready(),
> PyTypeReady(),Py_ReadyTypes(),Py_Initialize(),Py_Main().
> 
> Help!! What is happening at this point, and what could be causing this
> assertion to fail?

It checks whether the cls argument to classic_mro really is a classic
class. It is surprising that it calls into classic_mro. Can you tell
which line in Py_ReadyTypes comes from? (type, list, string, ...?)

None of these are classic classes, so this code should not have been
executed. It appears that this was executed because of the test in

		if (PyType_Check(base))
			parentMRO = PySequence_List(
				((PyTypeObject*)base)->tp_mro);
		else
			parentMRO = classic_mro(base);

i.e. base appears not to be a type. This, again, should not happen:
base is

		PyObject *base = PyTuple_GET_ITEM(bases, i);

which in turn comes from

	bases = type->tp_bases;

All base types of the types initialized in Py_ReadyTypes have only
types as bases, so it is surprising that one of them has a base which
is not a type.

It would be good to study the value of type, and tp_bases, at this
point. It might be that some initialization of global objects has
failed to happen, or has happened incorrectly.

Regards,
Martin


From praveen.patil at silver-software.com  Wed Sep  4 08:36:16 2002
From: praveen.patil at silver-software.com (Praveen Patil)
Date: Wed, 4 Sep 2002 13:36:16 +0100
Subject: Please help in  calling python fucntion from 'c'
Message-ID: <NFBBLJFBNMKMLGNLJMFCEELOCCAA.praveen.patil@silver-software.com>

Hi,

I have written 'C' dll(MY_DLL.DLL) . I am importing 'C' dll in python
file(example.py).
I want to call python function from 'c' function.
For your reference I have attached 'c' and python files to this mail.
In my pc:
python code is under the directory D:\test\example.py
dll is under the directory C:\Program Files\Python\DLLs\MY_DLL.pyd

Here are the steps I am following.

step(1): I am calling 'C' function(RECEIVE_FROM_IL_S) from python.
         This 'C' function is existing imported dll(MY_DLL).
step(2): I want to call python function(TestFunction) from 'C'
function(RECEIVE_FROM_IL_S).


Python code is(example.py)  :-
----------------------------
import MY_DLL

G_Logfile          = None

def TestFunction():
    G_Logfile = open('Pytestfile.txt', 'w')
    G_Logfile.write("%s \n"%'I am writing python created text file')
    G_Logfile.close
    G_Logfile = None
#end def TestFunction

if __name__ == "__main__":

   MY_DLL.RECEIVE_FROM_IL_S(10,50)


'C' code is (MY_DLL.c) :-
---------------------
#include <windows.h>
#include <stdio.h>
#include <Python.h>

PyObject* _wrap_RECEIVE_FROM_IL_S(PyObject *self, PyObject *args)
{
    FILE* fp;
    PyObject* _resultobj;
    int i,j;

    if( !(PyArg_ParseTuple(args, "ii",&i,&j)))
    {
       return NULL;
    }
    fp= fopen("RECEIVE_IL_S.txt", "w");
    fprintf(fp, "i=%d   j=%d" , i,j);
    fclose(fp);

    /* Here I want to call python function(TestFunction). Please suggest me
some solution*/

    _resultobj = Py_None;
    return _resultobj;
}


static PyMethodDef MY_DLL_methods[] = {
      { "RECEIVE_FROM_IL_S", _wrap_RECEIVE_FROM_IL_S, METH_VARARGS },
      { NULL , NULL}
      };

__declspec(dllexport) void __cdecl initMY_DLL(void)
  {
    Py_InitModule("MY_DLL",MY_DLL_methods);
  }


Please anybody help me solving the problem.


Cheers,

Praveen.
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: exampl.py
URL: <http://mail.python.org/pipermail/python-list/attachments/20020904/7d4538d3/attachment.ksh>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: MY_DLL.c
Type: application/octet-stream
Size: 788 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020904/7d4538d3/attachment.obj>
-------------- next part --------------
[ The information contained in this e-mail is confidential and is intended for the named recipient only. If you are not the named recipient, please notify us by telephone on +44 (0)1249 442 430 immediately, destroy the message and delete it from your computer. Silver Software has taken every reasonable precaution to ensure that any attachment to this e-mail has been checked for viruses. However, we cannot accept liability for any damage sustained as a result of any such software viruses and advise you to carry out your own virus check before opening any attachment. Furthermore, we do not accept responsibility for any change made to this message after it was sent by the sender.]

From marklists at mceahern.com  Wed Sep 18 13:24:06 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 18 Sep 2002 12:24:06 -0500
Subject: New RPMs at python.org
In-Reply-To: <20020918171900.GB13272@tummy.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOOEJHDPAA.marklists@mceahern.com>

[Sean Reifschneider]
> Find them at:
> 
>    http://python.org/2.2.1/rpms.html

Sean, thanks!  Is this a typo in the above?

<quote>Installs "python2.2" as well as "python2.2".</quote>
                       ^^^                    ^^^

They are the same, no?

// m

-




From jpmieville at bluewin.ch  Wed Sep 18 15:40:47 2002
From: jpmieville at bluewin.ch (Jean-Paul Mi=?ISO-8859-1?B?6Q==?=ville)
Date: Wed, 18 Sep 2002 21:40:47 +0200
Subject: Problem with urllib
Message-ID: <B9AEA35F.7FF7%jpmieville@bluewin.ch>

I don't understand why with some URL I have got [Errno 57] Socket is not
connected.

The code is :

import urllib2

page = urllib2.urlopen(' http://www.python.org/doc/NonEnglish.html')
print page.read()

If I do the same thing with python interactive, I can't reproduce the error.

Thanks in advance,
Jean-Paul



From marco at sferacarta.com  Thu Sep 26 18:17:14 2002
From: marco at sferacarta.com (Marco Mariani)
Date: Fri, 27 Sep 2002 00:17:14 +0200
Subject: Who knows somefunction?
In-Reply-To: <3D9382AF.5070800@aon.at>
References: <664e469667.69667664e4@icomcast.net> <3D9382AF.5070800@aon.at>
Message-ID: <20020926221714.GA11579@sferacarta.com>

On Thu, Sep 26, 2002 at 11:57:03PM +0200, Gregor Lingl wrote:

> How did you understand my question?
> (Errors should never pass silently)
> Gregor

Yeah, it's useless and a bad idea.

However, let's try:

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


// Pointer.c

#include <Python.h>
#include <ctype.h>

PyObject* pointer(PyObject* self, PyObject* args) {
        PyObject* p;

        if (!(PyArg_ParseTuple(args,"l",&p)))
                return NULL;

        Py_INCREF(p);
        return p;
}


static PyMethodDef pointer_methods[] = {
        {"pointer",     (PyCFunction) pointer,
                        METH_VARARGS, NULL},
        {NULL,NULL,0}
};


void initpointer(void) {
        Py_InitModule("pointer", pointer_methods);
};


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

## setup.py

from distutils.core import setup, Extension

module1 = Extension('pointer',
            sources = ['pointer.c'])

setup(name="Pointer",
      version = "0.1",
      author = "Anonymous Coward",
      author_email = "noway at example.com",
      description = "Blah",
      ext_modules = [module1]
      )


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



Build with "python setup.py build"


Now you can harm yourself. :-)

Tested with python 2.2.1



From gerhard.haering at gmx.de  Sun Sep  1 21:11:34 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 2 Sep 2002 01:11:34 GMT
Subject: curses: AttributeError: initscr
References: <un5dhs4f7qlte8@corp.supernews.com>
Message-ID: <slrnan5em7.4s9.gerhard.haering@lilith.my-fqdn.de>

Cesar wrote in comp.lang.python:
> I'm running python 1.5/2.2.1 under redhat linux 7.2.
> screen = curses.initscr()
> ==> AttributeError: initscr
> 
> any idea ??

The Python curses module is likely packaged in a separate RPM.


From cmay4 at yahoo.com  Tue Sep 24 15:56:08 2002
From: cmay4 at yahoo.com (Chuck May)
Date: Tue, 24 Sep 2002 19:56:08 GMT
Subject: Limit number of concurrent threads
References: <Xns929393554CAB0cmay4yahoocom@165.112.130.8> <uwupbdwnj.fsf@online.no>
Message-ID: <Xns9293A21B3DA0Acmay4yahoocom@165.112.130.8>

That sounds interesting.  I'll try it.

I was just hacking a much less elegant way of doing it.  Here is what I 
came up with:

from threading import Thread
import time, random

maxthreads =  5  # maximum number of concurrent threads
totalops   = 80  # hold total number of operations
all        = {}  # global holding the current running threads

class Worker(Thread):
    global all
    
    def __init__(self, id):
        Thread.__init__(self)
        self.all = all
        self.id = id

    def run(self):
        time.sleep(random.randint(10, 100) / 1000.0)
        print "id %d: " % self.id, all.keys()
        del all[self.id]

if __name__ == "__main__":
    for i in range(totalops):
        while (len(all) > maxthreads):
            time.sleep(.1)
            
        all[i] = 1
        w = Worker(i)
        w.start()

There may be problems with that, but it does seem to work.

Syver Enstad <syver-en+usenet at online.no> wrote in
news:uwupbdwnj.fsf at online.no: 

> Chuck May <cmay4 at yahoo.com> writes:
> 
>> I've searching for examples that fit my problem, and I haven't found 
>> any.  Most examples I've found look something like this:
>> 
>>     threads = []
>>     for i in range(number_of_tasks):
>>         thread.append(thread_for_task)
>>     
>>     # let the threads run - just wait for shutdown
>>     for t in threads:
>>         t.join() 
>> 
>> which whould work fine if I wanted to start all 80 threads at once.
>> 
>> I've read about the Queue, but I'm not sure if it's what I want, and
>> I 
>> 
>> can't seem to find any good examples on it.
> 
> I think Queue would be nice to use, Have a thread (the main thread
> perhaps) feed all the table names (or a table object if you please)
> into a Queue. 
> 
> Set up some worker threads with access to the Queue
> (global variable, member). The worker threads should loop something
> like this:
> 
> while 1:
>    queueItem = self._queue.get()
>    doSomethingWithQueueItemHere
> 
> Then you can start your workers.
> 
> (You'll have to find out some way to make the workers exit when the
> work is done, perhaps putting a number of None's equal to the number
> of workers last in the queue and testing for None in the working
> thread loop. 
> 
> 



-- 
Chuck May
IMS, Inc.
may at ims.nci.nih.gov


From oren-py-l at hishome.net  Thu Sep 12 10:47:10 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Thu, 12 Sep 2002 10:47:10 -0400
Subject: Coding Style: Quotes
In-Reply-To: <bt1g9.17429$Pf7.590136@news1.west.cox.net>
References: <bt1g9.17429$Pf7.590136@news1.west.cox.net>
Message-ID: <20020912144710.GA78280@hishome.net>

On Thu, Sep 12, 2002 at 02:25:11PM +0000, John Waycott wrote:
> I'm curious if others have adopted any standards for choice of single vs.
> double quotation marks to delimit strings. A look through the standard
> library reveals the choice of one over the other is rather arbitrary. I
> suspect it really makes no difference, but the question has come up during a
> code review.

The default repr of a string uses single quotes unless the string 
contains a single quote. Most of the time I use the same convention but 
I'm not 100% consistent.

For docstrings I always use triple double-quotes even if the docstring 
fits in a single line.

I use triple single-quotes to comment out blocks of code. This is very 
useful since "if 0:" would require re-indentation.

	Oren



From thorsten at thorstenkampe.de  Wed Sep 18 10:34:33 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Wed, 18 Sep 2002 16:34:33 +0200
Subject: Python equivalent to perldoc?
References: <Olvh9.7898$Lg2.1389903@news2.news.adelphia.net> <%Avh9.7933$Lg2.1392332@news2.news.adelphia.net> <am7sna$3emja$2@ID-77524.news.dfncis.de> <kaPh9.8886$Lg2.1572004@news2.news.adelphia.net>
Message-ID: <ama2tm$40up1$1@ID-77524.news.dfncis.de>

*  Bob X
> "Thorsten Kampe" <thorsten at thorstenkampe.de> wrote in message
>>> Shouldn't Python know its own libs?
>>
>> It does:
>> ,---
>> | Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
>> | Type "help", "copyright", "credits" or "license" for more information.
>> | >>> import pydoc
>> | >>>
>> `---
> It doesn't...you are executing from with Python...I am trying to do a ViM
> script that executes it.

So Vim doesn't doesn't know about Python's libs. No surprise.

>>> Or at least put the "libs" in the ENV variable?
>>
>> Windows XP doesn't have a ENV variable.
>
> Lets see...right click "My Computer"...go to the "Advanced" tab....OH LOOKY
> a button that says "Environment Variables". I could have been a wee bit more
> specific like...putting the the lib dir in the PATH.

It's not reasonable to put $PYTHONPATH in your $PATH, it's vice 
versa. And the Python executables (python, pydoc, etc.) should be put 
to where your path is: /bin or /usr/bin or /usr/local/bin.

Thorsten


From loewis at informatik.hu-berlin.de  Thu Sep 12 06:14:04 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 12 Sep 2002 12:14:04 +0200
Subject: ignoring dtd file in sax parser
References: <1010a26.0209112251.3efd57f4@posting.google.com>
Message-ID: <j4bs73347n.fsf@informatik.hu-berlin.de>

amichail at cse.unsw.edu.au (Amir Michail) writes:

> I tried:
> 
>         parser.setFeature(feature_namespaces, 0)
>         parser.setFeature(feature_validation, 0)
> 
>         parser.setContentHandler(ph)
> 
>         try:
>             parser.parse(f)
>     #        print "NO PARSE ERRORS"
>         except xml.sax._exceptions.SAXParseException:
>             print "XML parse errors in " + fn
> 
> 
> However, it still complains about not finding the dtd file.

You need to install an EntityResolver in the parser; that resolver can
then return an empty resource.

Regards,
Martin



From tjreedy at udel.edu  Wed Sep 25 10:51:35 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 25 Sep 2002 14:51:35 GMT
Subject: Newbie References Question
References: <3d91c72f$0$8558$79c14f64@nan-newsreader-02.noos.net>
Message-ID: <X3kk9.217211$216.7449571@bin4.nnrp.aus1.giganews.com>

"Guy Rabiller" <grabiller at 3dvf.net> wrote in message
news:3d91c72f$0$8558$79c14f64 at nan-newsreader-02.noos.net...

> let say I have:
> i1 = 1
> i2 = 2
> and
> p = [i1,i2]
>
> How can I have:
> p=[*i1,*i2]
> rather than
> p=[**i1,**i2] as it is curently ?
>
> ( Sorry for this nasty hybrid syntax )
>
> Textualy, how if I want that p[0] refere to the 'i1' container, and
not to
> the refererence it contains ?

Names, in themselves, are not Python objects and are certainly *not*
containers.  Once turned into Python string objects, they may be
contained within a dictionary and as key, associated with a reference.
But they still do not *contain* the reference; the dictionary, not the
name, is the container.

As a side note but still relevant: as a optimization in CPython at
least, local names within a function that are not used within nested
scopes are turned into array indexes.  They may be tossed after
compilation since they serve no further purpose.

> What I want is that if now I set:
> i1 = 4
> that automaticaly:
> p -> [4,2]
> and not keeping [1,2]

Too bad, since, it should not be clearer, this is not possible.  The
assignment means: associate 'i1' with [int object with value] 4 after
removing any previous association.

Terry J. Reedy




From david at no.westcontrol.spam.com  Fri Sep 27 08:33:43 2002
From: david at no.westcontrol.spam.com (David Brown)
Date: Fri, 27 Sep 2002 14:33:43 +0200
Subject: hex numbers??
References: <mailman.1032881550.14068.python-list@python.org>
Message-ID: <an1ipa$29v$1@news.netpower.no>

<jubafre at brturbo.com> wrote in message
news:mailman.1032881550.14068.python-list at python.org...
> >>> y=hex(x)
> >>> y
> '0xa'
> >>> type(y)
> <type 'str'>
>
> the hex() function put the hexadecimal number in a string type like this
'0xa', but the real hexadecimal number of 10 is 'a', why python puts a '0x',
and how i can get just the real hexadedcimal number, for example i want to
write in a file 10 in hexadecimal, but if i use hex(10) the result is '0xa'
as a string, is not the same as 'a'.
>
> '0xa' != 'a'
>
> x = [12,13,14]
> y = ['0xc', '0xd', '0xe']
>
> hex() create a list of string,  how i can write in a file just the real
value of hexadecimal number?
>

If you want to do the conversion over a list of numbers, try:
    map(lambda a : "%x" % a, x)






From missive at frontiernet.net  Sun Sep  1 16:54:14 2002
From: missive at frontiernet.net (Lee Harr)
Date: Sun, 01 Sep 2002 20:54:14 -0000
Subject: Installation of multiple versions of Python on Unix
References: <pAYb9.226358$v53.12275516@news3.calgary.shaw.ca> <un2mk4u8ukv17@corp.supernews.com> <slrnan2qoe.5hi.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <un4vjmcgvq4e81@corp.supernews.com>

In article <slrnan2qoe.5hi.gerhard.haering at lilith.my-fqdn.de>, Gerhard
H?ring wrote:
> Lee Harr wrote in comp.lang.python:
>> I have a very similar setup on a FreeBSD server (Zope still at 2.4.X tho
>> as I have not had a chance to upgrade yet)
>> 
>> I just leave /usr/local/bin/python as a hard link to
>> /usr/local/bin/python2.1 and run the other as /usr/local/bin/python2.2
> 
> Any reason not to use a symbolic link? Any reason to use a hard link
> for anything, ever? I never needed one.
> 

Not sure... that is the way the port does it, so I just
did the same thing. If it were me I would probably have
just used a symlink, but I figured there must be some reason.


>> One issue is building ports.
> 
> Erh. Really? <0.8 wink>
> 
>> Whenever I want to build a port to go with 2.2 (like I just added
>> twisted python, for instance) I will change the plain python to link
>> to python2.2 before I make and make install, then switch it back
>> when I am done. That has no impact on the running Zope at all.
> 
> For the ports you want to build, use:
> 
> $ make PYTHON_VERSION=python2.1
> 
> to use the FreeBSD ports system instead of working around it ;-)
> 
> Look into /usr/ports/Mk/bsd.python.mk for why this works :-)
> 

Thanks. That should be much easier :o)



From somebody at nojunk.invalid  Mon Sep 30 19:13:13 2002
From: somebody at nojunk.invalid (Russell E. Owen)
Date: Mon, 30 Sep 2002 17:13:13 -0600
Subject: [Mac] Editor on Macintosh?
References: <1fj7za2.1erbqps1uvzpj8N%news@agapow.net> <an7u1s$33de$1@nntp6.u.washington.edu> <300920020926012729%pecora@anvil.nrl.navy.mil>
Message-ID: <analq8$1d9e$1@nntp6.u.washington.edu>

In article <300920020926012729%pecora at anvil.nrl.navy.mil>, "Louis M. 
Pecora" <pecora at anvil.nrl.navy.mil> wrote:

>In article <an7u1s$33de$1 at nntp6.u.washington.edu>, Russell E. Owen
><somebody at nojunk.invalid> wrote:
>
>> Python has the Mac IDE is a bit limited (need an extra plug-in even to 
>> do syntax coloring)
>
>What Plug-in works for the Python IDE?

I know an add-on of some kind used to exist, but I just looked for it 
and couldn't find it. I've never used it myself so I may just have 
overlooked it. If you subscribe to the MacPython mailing list I suggest 
you ask there.

-- Russell


From cbbrowne at acm.org  Fri Sep 27 00:19:58 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: 27 Sep 2002 04:19:58 GMT
Subject: ANN: Tim Peters' FixedPoint on SourceForge
References: <mailman.1033085920.21837.python-list@python.org>
Message-ID: <an0m9e$a00fc$2@ID-125932.news.dfncis.de>

"dougfort" <dougfort at dougfort.net> wrote:
> What would it take to get you to use FixedPoint in a serious application....
>  benchmarks?
>  celebrity endorsements?
>  sample code?
>  other?

Photo opportunities with spokesmodels!  :-)
-- 
(reverse (concatenate 'string "moc.enworbbc@" "sirhc"))
http://www.ntlug.org/~cbbrowne/python.html
Whatever  you  do don't  mail  me at  pink-and-wobbly at asdkjlwelkj.com,
because then I'll know you're just an address-harvester, and blacklist
your IP until the end of time


From danb_83 at yahoo.com  Thu Sep 19 15:45:47 2002
From: danb_83 at yahoo.com (Dan Bishop)
Date: 19 Sep 2002 12:45:47 -0700
Subject: monetary applications (et al)
References: <mailman.1032383305.18515.python-list@python.org>
Message-ID: <ad052e5c.0209191145.7c059403@posting.google.com>

terry <tg5027 at citlink.net> wrote in message news:<mailman.1032383305.18515.python-list at python.org>...
> ...
> I do admit to being OO inhibited, so maybe you could take it from this 
> short piece code and execution results, and, provide a brief 
> description of what you would put into this money class that would 
> eliminate this error message, while maintaining the code simplicity.
> [code snipped]
> ...

def _roundToEven(x):
   if x < 0:
      return -_round(-x)
   iPart, fPart = divmod(x, 1.)
   iPart = int(iPart)
   if fPart == 0.5:
      if iPart & 1:
         return iPart + 1
      else:                
         return iPart      
   elif fPart > 0.5:
      return iPart + 1
   else:
      return iPart


class Money(object):
   def __init__(self, dollars):
      self.__cents = _roundToEven(dollars * 100)
   def __repr__(self):
      return "Money(%0.2f)" % float(self)
   def __str__(self):
      return "$%0.2f" % float(self)
   def __float__(self):
      return self.__cents / 100.
   def __pos__(self):
      return self
   def __neg__(self):
      return Money(-float(self))
   def __add__(self, other):
      assert isinstance(other, Money)
      return Money(float(self) + float(other))
   def __sub__(self, other):
      assert isinstance(other, Money)
      return Money(float(self) - float(other))
   def __mul__(self, other):
      return Money(float(self) * other)
   def __div__(self, other):
      if isinstance(other, Money):
         return float(self) / float(other)
      else:
         return Money(float(self) / other)
   def __cmp__(self, other):
      return (self.__cents).__cmp__(other.__cents)
   def __hash__(self):
      return hash(float(self))


From chris.gonnerman at newcenturycomputers.net  Sun Sep 29 18:59:21 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sun, 29 Sep 2002 17:59:21 -0500
Subject: Graphics library
Message-ID: <00af01c2680b$d9fa8c80$ba01010a@local>

----- Original Message ----- 
From: "David Siroky" <dsiroky at email.cz>

> Hi!
> 
> I'd like to know, which library for creating images (like 
> libgd) is "the best". I want to use it with Python.

This is a sore point with me.  I am the current maintainer
of the Python gdmodule:

    http://newcenturycomputers.net/projects/gdmodule.html

I took it over because I needed to annotate graphic images
using TrueType fonts.  The PIL (PythonWare Imaging Library)
is probably superior otherwise, but the font support stinks.

There are two other projects, claimed to be PIL-compatible,
one supporting FreeType (TrueType) and the other supporting
Postscript Type 1; but when I asked here previously for 
examples and/or support using either of them I got resounding
silence.

PIL can be found at:
    http://www.pythonware.com/products/pil/index.htm

The Image-SIG's t1python module is at:
    http://www.python.org/sigs/image-sig/t1lib/
(Incidentally this may have been updated since I first
looked at it, so my complaints here may no longer apply.)

Robert Kern's PyFT can be found at:
    http://starship.python.net/crew/kernr/Projects.html

PIL Plus from Pythonware includes TrueType support (I've read) 
but as a commercial product it was not an option for my 
project.

I looked for the gdmodule, and discovered Richard Jones 
wasn't supporting it anymore.  When I asked about taking it
over, he answered "Sure, go ahead."

So I did.  With help from others I have updated it to support
the 2.01 (beta) libgd.  There are weaknesses, in particular
the terrible JPEG output (rule 1: always use PNG with the
gdmodule), but at least all my TTF fonts work.

gdmodule also has a very simple interface; I don't have to 
know as much about an image file to work with it.  This 
may be related to the crummy JPEG output...  I dunno.

I'd love to see sample code to render either type of fonts
via PIL and one or more of the above libraries.  In fact, 
it seems like an ideal thing for a HOWTO or some such.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net





From cstrong at arielpartners.com  Wed Sep 25 23:02:40 2002
From: cstrong at arielpartners.com (Craeg K Strong)
Date: Wed, 25 Sep 2002 23:02:40 -0400
Subject: Is there a way to find the name of the method currently being executed?
Message-ID: <3D9278D0.9060307@arielpartners.com>

Hello:

class a:

    def func1(self):
        print "func1 being called"

    def func2(self):
        print "func2 being called"

In the code sample above, I would like to replace the "func1"
and "func2" strings with constants so the same code could
be copied verbatim into multiple functions and do the right
thing.

Is there a way to do this in Python?  For example, maybe you could
get your hands on the call stack and print out stack[0].__name__
or something....

Thanks in advance,

--Craeg

BTW, I could not find this in the python.org FAQ....





From keith_oconnell at blueyonder.co.uk  Thu Sep 19 13:45:08 2002
From: keith_oconnell at blueyonder.co.uk (Keith O'Connell)
Date: Thu, 19 Sep 2002 17:45:08 GMT
Subject: finding last match in a file
References: <slrnaojlin.8re.dag@omally.animagicnet.no>
Message-ID: <87r8fplvq4.fsf@smtp.blueyonder.co.uk>

Dag <dag at animagicnet.no> writes:

> I have a number of very large files where I have to find the last line
> which matches a certain experssion.  Currently I'm simply opening the
> file and looping through it noteing each time a line matches and at the
> end of the loop I see which was the last match.  However this is very
> slow and inefficient.  Is there some way to read through the file
> backwards and simply find first match starting from the end of the file.
> The files are really too big to read the entire thing into memory.


Would you not be able to use grep with the "-n" option. It will list
all your matches with the lines numbered so you can see the last match
having the highest line number. If you then use wc on the file you will
get the total number of lines. Take one from the other and you can see
how far from the end your last match is.

That seems to easy - Have I misunderstood you question?

Keith.

-- 
___________________________________   _
  Keith O'Connell.                   -o)
  Maidstone, Kent. (UK)              /\\
  keith_oconnell at blueyonder.co.uk   _\_v


From timr at probo.com  Fri Sep 20 00:19:29 2002
From: timr at probo.com (Tim Roberts)
Date: Thu, 19 Sep 2002 21:19:29 -0700
Subject: Destructor never called ???
References: <76710511.0209111450.40999dd8@posting.google.com> <tp70ou0r0j11as6enhpsaqul4oqa3mlu6t@4ax.com> <alp8n3$bsd$1@panix1.panix.com>
Message-ID: <4e8lou018f9uosl5p9mb0tt6691mfivamj@4ax.com>

aahz at pythoncraft.com (Aahz) wrote:

>In article <tp70ou0r0j11as6enhpsaqul4oqa3mlu6t at 4ax.com>,
>Tim Roberts  <timr at probo.com> wrote:
>>
>>However, Python is a garbage-collected system.  Unlike in COM, an object is
>>NOT destroyed as soon as its reference count goes to 0.  Your object will
>>only be destroyed if the garbage collector happens to run, and decides that
>>it might be a convenient time to actually free up the space your object
>>occupies.
>
>Wrong.  CPython uses reference counting backed by GC.

Well, then, what was all the fuss about deterministic desctructors in the
newsgroup last year or so?  Has this changed, or am I just remembering this
entirely wrong?
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From LogiplexSoftware at earthlink.net  Fri Sep  6 15:07:34 2002
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 06 Sep 2002 12:07:34 -0700
Subject: Soviet/Russian military
In-Reply-To: <3D66BF07.852B6858@worldnet.att.net>
References: <3D66A072.75F7E2B2@worldnet.att.net> 
	<3D66BF07.852B6858@worldnet.att.net>
Message-ID: <1031339254.19985.2630.camel@software1.logiplex.internal>

On Fri, 2002-08-23 at 16:03, Peter Koren wrote:
> Peter Koren wrote:
> > 
> > Dan,
> > 
> > I ran across a discussion of the Soviet/Russian military. It is
> > attached.
> > 
> > Love,
> > 
> > Dad
> 
> Gasp, apologies offered. I addressed this to my son, but this news group
> was selected in Mozilla so it got posted.

Even worse, you forgot to attach the article.

> Sorry,
> 
> Pete Koren
> -- 
> http://mail.python.org/mailman/listinfo/python-list

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From tim.one at comcast.net  Sat Sep  7 13:11:10 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 07 Sep 2002 13:11:10 -0400
Subject: None or 0
In-Reply-To: <k7ip4-0p3.ln1@junker.stroeder.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEMMBCAB.tim.one@comcast.net>

[Michael Str?der]
> I'm using the idiom
>
> a = a or b
>
> to get rid of statements like
>
> if a is None:
>    a = b

That's fine, although you may get burned someday if 'a' has an interesting
value that's not None but happens to evaluate as false.  Like, e.g.,

    def eat(number_of_pounds=None):
        # Default to 1 pound.
        to_eat = number_of_pounds or 1
        ...

If you're not hungry and try to do eat(0), you're in for a treat.  As a rule
of thumb, whenever you use None to mean "not there", you'll cut bugs by
testing for "is None" or "is not None" explicitly.  I've seen many subtle
bugs as a result of not doing this; sometimes code works fine for years, and
then someone happens to add a __len__ method to an object's class, and "all
of a sudden" evaluating to false no longer means a var is still bound to
None, but that it may also be bound to a non-None object that suddenly
decided to call itself empty.

> Now I wonder what happens if both a and b have zero-length values.

If A evaluates to false, 'A or B' returns B without testing B (so it doesn't
matter whether B does or doesn't evaluate to false -- B is simply evaluated
and returned then).

> Trying in Python 2.2.1 reveals:
>
>  >>> repr(0 or None)
> 'None'
>  >>> repr(None or 0)
> '0'
>  >>> repr('' or None)
> 'None'
>  >>> repr(None or '')
> "''"
>  >>>
>
> Is it guaranteed to work like this

Yes.  Similarly, 'A and B' returns B untested whenever A evaluates to true.

   A or  B  same-as  if A then A else B
   A and B  same-as  if A then B else A

B is never tested for true/false, and in the cases where A is the result the
B expression isn't even evaluated.

> or should that be avoided?

That's really a different question <wink>.





From itamar at itamarst.org  Fri Sep 27 21:16:35 2002
From: itamar at itamarst.org (Itamar Shtull-Trauring)
Date: Fri, 27 Sep 2002 21:16:35 -0400
Subject: ANN: Twisted 0.99.2 - now with stable internet
Message-ID: <20020927211635.327e6494.itamar@itamarst.org>

The Twisted team is proud to announce Twisted 0.99.2, which includes a API-stable version of twisted.internet, Twisted's event-driven networking core (like asyncore, but more so).

Twisted's twisted.internet pacakge provides:
 - event-driven TCP, SSL and unix socket networking
 - cross-platform (Unix and Win32) support
 - pluggable reactors, including select(), poll() and Kqueue
 - event loop GUI integration, for Win32, GTK+, Qt, wxPython and Tkinter
 - threading integration
 - scheduling
 - process running API

With the twisted.internet APIs now stable, we feel Twisted provides the best networking platform available today for Python. This also gives us a solid basis for working on stable APIs for the rest of Twisted's packages.

Download it now from http://www.twistedmatrix.com, or visit us on IRC on #twisted (irc.freenode.net servers, formerly known as openprojects).


What is Twisted?
================

Twisted is an event-based framework for internet applications.  It includes a
web server, a telnet server, a chat server, a news server, a generic client 
and server for remote object access, and APIs for creating new protocols and
services. Twisted supports integration of the Tk, GTK+, Qt or wxPython event
loop with its main event loop. The Win32 event loop is also supported, as is
basic support for running servers on top of Jython. Twisted works with
Python 2.1 and Python 2.2. Twisted even supports the CVS versions of Python, 
so it is ready for Python 2.3.

Twisted currently supports the following protocols, all implemented in pure
python, most of them as both servers and clients:
 
  - FTP
  - HTTP
  - NNTP
  - SOCKSv4 (server only)
  - SMTP
  - IRC
  - telnet
  - POP3
  - AOL's instant messaging TOC
  - OSCAR, used by AOL-IM as well as ICQ (client only)
  - DNS
  - MouseMan
  - finger
  - Echo, discard, chargen and friends
  - Twisted Perspective Broker, a remote object protocol
  - XML-RPC
  
For more information, visit http://www.twistedmatrix.com, or join the list at
http://twistedmatrix.com/cgi-bin/mailman/listinfo/twisted-python


What's new in 0.99.2
====================

- Improved HTTP support - bug fixes, caching APIs

- WOVEN, a web templating system (formerly known as DOMTemplate)

- microdom, a very small DOM implementation

- An inetd server written with Twisted (coming soon to a Debian mirror near you)

- Switched to epydoc for API docs (http://epydoc.sf.net - it rocks!)

- RDBMS storage for the NNTP server, and improved NNTP protocol support

- Finalized process-running API

- Huge numbers of bug fixes and feature improvements, and better docs



From vtail at yandex.ru  Mon Sep  9 09:57:29 2002
From: vtail at yandex.ru (vtail)
Date: 9 Sep 2002 06:57:29 -0700
Subject: Possible bug in WxPython
Message-ID: <9ff6b66.0209090557.3c0e3c94@posting.google.com>

Greetings.

When trying to run an example program from wxWiKi
http://wiki.wxpython.org/index.cgi/Getting_20Started, section 7.1.2, I
encounter following error. Could someone please explain it to me?

My python version is 
Python 2.2.1c2 (#33, Mar 26 2002, 13:04:18) [MSC 32 bit (Intel)] on
win32,
my wxpython version is 2.3.2.1 (from
c:/Python22/Lib/site-packages/wxPython/__version__.py)

I've attached the error message and the program that causes it.

Thanks for your help, Vtail.

====>

O:\_PUBLIC\Newdbase\Python\wxPython>python test4.pyw > aaa.txt
17:52:59: Debug: c:\Projects\wx\src\msw\app.cpp(542):
'UnregisterClass(no redraw
 canvas)' failed with error 0x00000584 (class still has open
windows.).
17:52:59: There were memory leaks.
17:52:59: ----- Memory dump -----
17:52:59: wxFrame at $FF8890, size 332
17:52:59: wxTextCtrl at $115EAA8, size 316
17:52:59: wxStatusBar at $778C38, size 260
17:52:59: wxMenu at $10D3AB0, size 112
17:52:59: wxMenuItem at $F9CE50, size 124
17:52:59: wxMenuItem at $F9CF08, size 124
17:52:59: wxMenuItem at $F9CFC0, size 124
17:52:59: wxMenuItem at $F9D078, size 124
17:52:59: wxMenuItem at $F9D130, size 124
17:52:59: wxMenuBar at $F8FD88, size 312
17:52:59: wxPyCallback at $F81C40, size 12
17:52:59: wxObject at $85DC50, size 28
17:52:59: wxPyCallback at $F81830, size 12
17:52:59: wxPyCallback at $F82638, size 12
17:52:59: wxBoxSizer at $F8FF98, size 96
17:52:59: wxButton at $F85128, size 264
17:52:59: wxSizerItem at $F7F998, size 60
17:52:59: wxButton at $10601A0, size 264
17:52:59: wxSizerItem at $F903A0, size 60
17:52:59: wxButton at $F90418, size 264
17:52:59: wxSizerItem at $F906D8, size 60
17:52:59: wxButton at $F90750, size 264
17:52:59: wxSizerItem at $F90A10, size 60
17:52:59: wxButton at $F82B08, size 264
17:52:59: wxSizerItem at $F82CE8, size 60
17:52:59: wxButton at $F82D60, size 264
17:52:59: wxSizerItem at $F82FF8, size 60
17:52:59: wxBoxSizer at $F83070, size 96
17:52:59: wxSizerItem at $F831A0, size 60
17:52:59: wxSizerItem at $F83218, size 60
17:52:59:
17:52:59:
17:52:59: ----- Memory statistics -----
17:52:59: 8 objects of class wxSizerItem, total size 480
17:52:59: 6 objects of class wxButton, total size 1584
17:52:59: 2 objects of class wxBoxSizer, total size 192
17:52:59: 1 objects of class wxObject, total size 28
17:52:59: 3 objects of class wxPyCallback, total size 36
17:52:59: 1 objects of class wxMenuBar, total size 312
17:52:59: 5 objects of class wxMenuItem, total size 620
17:52:59: 1 objects of class wxMenu, total size 112
17:52:59: 1 objects of class wxStatusBar, total size 260
17:52:59: 1 objects of class wxTextCtrl, total size 316
17:52:59: 1 objects of class wxFrame, total size 332
17:52:59:
17:52:59: Number of object items: 30
17:52:59: Number of non-object items: 0
17:52:59: Total allocated size: 4272
17:52:59:
17:52:59:

<====


====> test4.pyw
from wxPython.wx import *
import os
ID_ABOUT=101
ID_OPEN=102
ID_BUTTON1=110
ID_EXIT=200
class MainWindow(wxFrame):
    def __init__(self,parent,id,title):
        self.dirname=''
        wxFrame.__init__(self,parent,-4, title,
style=wxDEFAULT_FRAME_STYLE|
                                        wxNO_FULL_REPAINT_ON_RESIZE)
        self.control = wxTextCtrl(self, 1, style=wxTE_MULTILINE)
        self.CreateStatusBar() # A Statusbar in the bottom of the
window
        # Setting up the menu. 
        filemenu= wxMenu()
        filemenu.Append(ID_OPEN, "&Open"," Open a file to edit")
        filemenu.AppendSeparator()
        filemenu.Append(ID_ABOUT, "&About"," Information about this
program")
        filemenu.AppendSeparator()
        filemenu.Append(ID_EXIT,"E&xit"," Terminate the program")
        # Creating the menubar. 
        menuBar = wxMenuBar()
        menuBar.Append(filemenu,"&File") # Adding the "filemenu" to
the MenuBar
        self.SetMenuBar(menuBar)  # Adding the MenuBar to the Frame
content.
        EVT_MENU(self, ID_ABOUT, self.OnAbout)
        EVT_MENU(self, ID_EXIT, self.OnExit)
        EVT_MENU(self, ID_OPEN, self.OnOpen)

        self.sizer2 = wxBoxSizer(wxHORIZONTAL)
        self.buttons=[]
        for i in range(0,6):
            self.buttons.append(wxButton(self, ID_BUTTON1+i, "Button
&"+`i`))
            self.sizer2.Add(self.buttons[i],1,wxEXPAND)

        # Use some sizers to see layout options
        self.sizer=wxBoxSizer(wxVERTICAL)
        self.sizer.Add(self.control,1,wxEXPAND)
        self.sizer.Add(self.sizer2,0,wxEXPAND)

        #Layout sizers
        self.SetSizer(self.sizer)
        self.SetAutoLayout(1)
        self.sizer.Fit(self)

        self.Show(1)

    def OnAbout(self,e):
        d= wxMessageDialog( self, " A sample editor \n"
                            " in wxPython","About Sample Editor",
wxOK)
                            # Create a message dialog box 
        d.ShowModal() # Shows it 
        d.Destroy() # finally destroy it when finished. 

    def OnExit(self,e):
        self.Close(true)  # Close the frame. 

    def OnOpen(self,e):
        """ Open a file"""
        dlg = wxFileDialog(self, "Choose a file", self.dirname, "",
"*.*", wxOPEN)
        if dlg.ShowModal() == wxID_OK:
            self.filename=dlg.GetFilename()
            self.dirname=dlg.GetDirectory()
            f=open(os.path.join(self.dirname, self.filename),'r')
            self.control.SetValue(f.read())
            f.close()
        dlg.Destroy()


app = wxPySimpleApp()
frame = MainWindow(None, -1, "Sample editor")
frame.Show(1)

<====


From peter at engcorp.com  Mon Sep  2 11:57:26 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 02 Sep 2002 11:57:26 -0400
Subject: Must have Python modules?
References: <_THc9.42153$vY2.1039887@e3500-atl2.usenetserver.com> <slrnan6k4s.eg.gerhard.haering@haering.opus-gmbh.net> <FCIc9.30407$FJ1.5585669@e3500-atl1.usenetserver.com> <j4y9akjyoe.fsf@informatik.hu-berlin.de>
Message-ID: <3D738A66.9D78F20D@engcorp.com>

"Martin v. L?wis" wrote:
> 
> "Robert Oschler" <Oschler at earthlink.net> writes:
> 
> > In C++, which has thousands of libraries, snippets, etc. there were only a
> > few that I personally used religiously.
> [...]
> So if your question includes standard library modules as well: my
> absolutely must-have module is "os", followed by "sys".

I've written programs that need "sys" but not "os" more often
than the other way around(*).  "sys" ROOLZ!  :-)

* By "the other way around", I mean "more often than I write those with 
only 'os'", not "more often than those programs write me".

-too-early-on-a-long-weekend-to-be-serious-ly yr's,
 Peter


From marklists at mceahern.com  Tue Sep 10 16:11:25 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 10 Sep 2002 15:11:25 -0500
Subject: "str"(expr)"str" interpolation syntax
In-Reply-To: <Pine.GSO.4.33.0209102221070.10002-100000@techst02.technion.ac.il>
Message-ID: <JHEOKEOOLIGLDHCMAHMOGELPDNAA.marklists@mceahern.com>

> I was searching for a new interpolation syntax ...

<applying the brakes>

You might be interested to:

1.  Checkout the PEPs for string interpolation and simple string
substitution.
2.  Read this thread:


http://groups.google.com/groups?selm=ouit4e63fr.fsf%40andlx-anamika.cisco.co
m

3.  Read the archives from Python-Dev regarding the relevant PEPs.

After you do all that, you may reconsider or reframe your suggestion in
light of what has been discussed.

Or not.  <wink>

Cheers,

// m

-




From marco at reimeika.ca  Sat Sep 21 23:34:55 2002
From: marco at reimeika.ca (marco)
Date: 21 Sep 2002 23:34:55 -0400
Subject: "time" module under AIX?
Message-ID: <yibhegi7l40.fsf@coxeter.math.toronto.edu>

Hi,

I've just compiled python 2.2.1 under AIX and all seemed
to go well. However, I'm having some trouble importing
a few modules such as "time" and "math" e.g.

coxeter% python2                                            </home/marco> 23:11
Python 2.2.1 (#1, Sep 21 2002, 22:32:45) [C] on aix4
Type "help", "copyright", "credits" or "license" for more information.
>>> import time
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ImportError: No module named time
>>>

Other modules import fine. Compilation appeared to go well, for example:

building 'time' extension
cc_r -DNDEBUG -O -I. -I/usr/local/src2/marco/Python-2.2.1/./Include -I/usr/local
/include -IInclude/ -c /usr/local/src2/marco/Python-2.2.1/Modules/timemodule.c -
o build/temp.aix-4.3-2.2/timemodule.o
./Modules/ld_so_aix cc_r -bI:Modules/python.exp build/temp.aix-4.3-2.2/timemodul
e.o -L/usr/local/lib -lm -o build/lib.aix-4.3-2.2/time.so
building 'operator' extension
etc...

Any ideas? I'm using AIX 4.3.

Thanks!

--
marco at reimeika.ca
Gunnm: Broken Angel
http://reimeika.ca/


From gua81 at XXXyahoo.com  Tue Sep 24 20:46:15 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Wed, 25 Sep 2002 10:46:15 +1000
Subject: Strange python and cgi!
References: <amquos$rcu$1@lust.ihug.co.nz> <3D9101AC.842F0513@alcyone.com>
Message-ID: <amr0vu$sq5$1@lust.ihug.co.nz>

what kind of error am I expecting here?
.... errrrrrr I don't know how to access the web server error log.


> It's likely some other error occurred.  Check your Web server error log
> for the final word.
> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ As far as I'm concerned, being any gender is a drag.
> \__/ Patti Smith
>     Kepler's laws / http://www.alcyone.com/max/physics/kepler/
>  A proof of Kepler's laws.




From quiteblack at yahoo.com  Sun Sep 15 23:58:59 2002
From: quiteblack at yahoo.com (black)
Date: 15 Sep 2002 20:58:59 -0700
Subject: couldnt find __dic__
Message-ID: <7153651c.0209151958.4099ac3e@posting.google.com>

my book told me all dictionaries have __dic__ method but I searched
each dictionaries and find nothing, where is it ???


From sadams123 at optushome.com.au  Mon Sep 16 21:31:06 2002
From: sadams123 at optushome.com.au (Steven)
Date: Tue, 17 Sep 2002 11:31:06 +1000
Subject: wincerapi
Message-ID: <3d8685da$0$23169$afc38c87@news.optusnet.com.au>

For anyone who needs to get wincerapi up and running on Python 2.2 I've put
the necessary dlls and stuff up for download:

http://www.tetrica.com/quickdry/wincerapi.zip

I've only tried it out on my home machine Python 2.2, XP Pro, so your
mileage may vary, but I figured since not everyone has a compiler, and there
doesn't seem to be any other way to get wincerapi working...

Steven




From peter at engcorp.com  Mon Sep  2 12:05:08 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 02 Sep 2002 12:05:08 -0400
Subject: Keyerror problem
References: <_lyc9.7164$7N.29584@newsfep3-gui.server.ntli.net> <slrnan5f2o.4s9.gerhard.haering@lilith.my-fqdn.de> <0cFc9.1718$yB6.25319@newsfep1-gui.server.ntli.net>
Message-ID: <3D738C34.4E99461B@engcorp.com>

davbucko wrote:
> 
> Thanks!  I wasn't too sure about what the backticks meant, but I thought
> that they might be necessary.  Oh well

Sounds like a good time to use the interactive prompt to test out
the code before/as you write it.  Typing name and `name` would 
quickly show you what the difference was...

-Peter


From bdesth at nospam.free.fr  Sun Sep  8 21:56:07 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Sun, 08 Sep 2002 21:56:07 -0400
Subject: Larry Wall's comment on python...
References: <mailman.1031339711.632.python-list@python.org>
Message-ID: <3D7BFFB7.4050906@nospam.free.fr>

brueckd at tbye.com wrote:
> On 6 Sep 2002, Rod Stephenson wrote:
> 
> 
>>"Python is cool to look at small bits of, but I think the "outline"
>>syntax breaks down with larger chunks of code. I'm with Aristotle on
>>the structure of discourse--a story should have a beginning, and
>>middle, and an end. So should blocks"
>>
>>I'm not quite sure what he's trying to get at here - I guess that for
>>a long heavily indented chunk of code, you could lose track of the
>>overall structure, but I don't write code this way.
>>
>>Any comments?
> 
> 
> The very idea of Larry Wall lecturing on code readability is hilarious.
> 
> -Dave
> 
> 

C'mon, if he is able to read Perl code, he surely can say something 
about readability !-)

laotseu



From LogiplexSoftware at earthlink.net  Mon Sep 16 15:30:31 2002
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 16 Sep 2002 12:30:31 -0700
Subject: IDLE
In-Reply-To: <am557m$2nldt$2@ID-77524.news.dfncis.de>
References: <wpbf9.264320$aA.45691@sccrnsc02>
	<3D7D4F84.318CDB0F@bellatlantic.net> <v4Pg9.428631$me6.49396@sccrnsc01>
	<bs70f7ks.fsf@online.no> <am1efh$1ohti$1@ID-77524.news.dfncis.de>
	<EOjh9.2033$Cf.81@fe03>  <am557m$2nldt$2@ID-77524.news.dfncis.de>
Message-ID: <1032204632.17116.249.camel@software1.logiplex.internal>

On Mon, 2002-09-16 at 10:43, Thorsten Kampe wrote:
> 
> "Hardware: 386 enhanced, 75MHz, 32MB RAM."

This can't be correct.  The 386 never hit 75MHz.  I'm guessing you're
reading what Windows (3.x ?) reports in it's hardware settings which
only tells you whether it's running in "386 Enhanced" or "Standard"
mode.  

The actual CPU is either an AMD/Cyrix 486 clone or an early Pentium.


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From marklists at mceahern.com  Fri Sep  6 18:58:39 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 6 Sep 2002 17:58:39 -0500
Subject: Larry Wall's comment on python...
In-Reply-To: <200209062258.g86MwvVm096127@mail.hcsmail.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOMEAODNAA.marklists@mceahern.com>

[mmaddox at hcsmail.com [mailto:mmaddox at hcsmail.com]]
> Of course it does. I need a parser that can decipher the pure randomness
> of my own code. Basically, I'd like to run an SHA-512 algorithm against a
> poorly-typed text file describing my program. Then, I'd just compile the
> message digest.
>
> Of course, most of my current code has one big block, main(), and I just
> cut-and-paste the rest.

Your message brought me smiles, tears, and laughter--I haven't had that much
fun since watching Heehaw.

Cheers,

// m

-




From read-the-signature at send-spam-to-dev-null.com  Thu Sep  5 07:20:45 2002
From: read-the-signature at send-spam-to-dev-null.com (read-the-signature at send-spam-to-dev-null.com)
Date: Thu, 5 Sep 2002 11:20:45 +0000 (UTC)
Subject: newbie problem: use socke lib to retrieve one web page:
References: <czAd9.6803$yt3.3242044@newssrv26.news.prodigy.com>
Message-ID: <al7emd$m39$1@wnnews.sci.kun.nl>

koko <kokohh at hotmail.com> wrote:
> I write this to retrieve one web page using socket lib.
> [snip: code]
> s.connect((HOST,PORT))
> s.send(m)
> data=s.recv(4096)
> print data
> s.close()
> 
> -----------
> But I retrieved nothing and the program stuck.


Your code probably is _too quick_ for the server. That is, you tried to
read data while the server was still busy processing. You could "sleep" a
little, before reading. Oh yeah, are you sure that 

    data = s.recv(4096)

will _always_ give you all of the data? You might want to take a look at
asyncore [1] which provides a nicer way to handle these situations. I'm not
sure, but I expect that the Python cookbook [2] also has some interesting
things to say about this topic.

    Good luck!

        Bas

[1]   http://www.python.org/doc/current/lib/module-asyncore.html
[2]   http://aspn.activestate.com/ASPN/Python/Cookbook/

-- 
b a s {dot} v a n g i l s {at} c s {dot} k u n {dot} n l
BOFH excuse #63:
not properly grounded, please bury computer


From gerhard.haering at gmx.de  Sun Sep 22 06:46:41 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sun, 22 Sep 2002 12:46:41 +0200
Subject: Python HTTPSConnection/HTTPResponse Problem
In-Reply-To: <9abedeb.0209212200.5c669a50@posting.google.com>
References: <9abedeb.0209212200.5c669a50@posting.google.com>
Message-ID: <20020922104641.GB3780@lilith.ghaering.test>

* Edward Muller <edwardam at interlix.com> [2002-09-21 23:00 -0700]:
> I am trying to use HTTPSConnection and HTTPResponse objects [...]

Is your Python SSL-enabled?  Check with:

    import socket
    hasattr(socket, "ssl")

Which version of Python are you using? Be sure to use at least version
2.2, as it contains many SSL-related bugfixes.

-- Gerhard



From loewis at informatik.hu-berlin.de  Mon Sep 23 09:22:31 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 23 Sep 2002 15:22:31 +0200
Subject: Python Compiling
References: <amfveg$5g64k$1@ID-53055.news.dfncis.de> <j4lm5w9zft.fsf@informatik.hu-berlin.de> <tyf7khczziq.fsf@pcitapi22.cern.ch>
Message-ID: <j4r8fkg7s8.fsf@informatik.hu-berlin.de>

Jacek Generowicz <jacek.generowicz at cern.ch> writes:

> How would you reconcile the above statements in context of Common Lisp
> compilers ?  

It all depends on the programming language. I would, of course, expect
that a C compiler gives a performance improvement over a C interpreter
significant enough to make using C compilers worthwhile. 

Results from one language don't transfer to another. I don't know
(really) why Lisp compilers work that well, and what restrictions on
the Lisp language had to be made to make them work well. In general,
many of the type inference mechanisms done by lisp compilers don't
apply to Python.

Regards,
Martin


From steve at ferg.org  Wed Sep 11 10:16:20 2002
From: steve at ferg.org (Stephen Ferg)
Date: 11 Sep 2002 07:16:20 -0700
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid> <alln71$1s2s$1@nntp6.u.washington.edu>
Message-ID: <b16e4ef7.0209110616.48dfe04c@posting.google.com>

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> (what Joseph was doing was binding the button to the <Button-1> event,
> e.g. mybutton.bind("<Button-1>", afunction))
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
That example was mine.  It was based on the discussion in John Grayson's
book (p. 108), where Grayson says that (for mouse-button events) 
binding the action to <Button-1> has the same effect as
binding using "command".

But it seems that really is incorrect.  On 
http://www.tcl.tk/man/tcl8.4/TkCmd/button.htm#M6
it says that the "command is typically invoked when mouse 
button 1 is released over the button window."  So it looks like the
correct mouse-button binding to replace "command" would be to bind to 
<ButtonRelease-1>.

But it sounds like you are saying that even binding to 
<ButtonRelease-1> is not standard behavior ...

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Alas, too true. Button-1 fires when the mouse button is pressed (very 
nonstandard behavior) and ButtonRelease fires when the mouse button is 
released even if the mouse is no longer over the Button object. I don't 
know of any event binding that does a proper job of emulating the 
standard behavior of pressing a button. Button(command=...) is the way 
to go for this.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

What is the standard behavior?  Not <Button-Release-1>? 

Is there any place where definitions or explanations
or discussions of standard button behavior are written down
and available? A book? A Web page?

Do you know if "command" (at least for 
mouse-button behavior, ignoring
its response to keypresses) does anything other than 
simply respond to <Button-Release-1>?  


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Actually, I use a minor variant ...
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Thanks.  This is useful information.  

Also mucho thanks for something else.  I looked at your SUMMARY about 
a month ago and on your recommendation I picked up a copy of 
Brent Welch's "Practical Programming in Tcl and Tk".  It has been 
immensely helpful.  A good tip! :-)


From max at alcyone.com  Wed Sep 25 23:53:51 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 25 Sep 2002 20:53:51 -0700
Subject: a python puzzle
References: <mailman.1033002153.8360.python-list@python.org>
Message-ID: <3D9284CF.2B1161F2@alcyone.com>

Lulu of the Lotus-Eaters wrote:

> I think the various responders did this "by hand."  Probably having
> the
> Python source code made it even easier to just eyeball.  In a more
> general sense (just in case anyone doesn't know), the way to solve a
> Caesar cipher like this is to start counting letter frequencies.
> Comparing the frequencies to well-known English letter frequencies
> like:
> 
>     http://gnosis.cx/download/letterfrequency.gif
> 
> lets you start matching letters pretty quickly.  Of course, if you
> don't
> know it's English, the frequency table isn't as accurate.

This wouldn't have worked as well here, since the encrypted text was so
short; furthermore, the fact that it was mostly Python code would have
likely skewed the letter frequencies, since Python keywords, modules,
and builtin names appear more frequently in Python code than general
user-chosen identifiers; the letter frequency of the code would be
biased against the letter frequency of the common "words" in Python,
which is likely to be somewhat different from English as a whole.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Gods are born and die, but the atom endures.
\__/ Alexander Chase
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From amk at amk.ca  Tue Sep 10 16:36:16 2002
From: amk at amk.ca (A.M. Kuchling)
Date: Tue, 10 Sep 2002 20:36:16 -0000
Subject: Python iCalendar API?
References: <mailman.1031673134.15031.python-list@python.org> <slrnans8f6.jc.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <unslu0cjragpf0@corp.supernews.com>

In article <slrnans8f6.jc.gerhard.haering at lilith.my-fqdn.de>, 
   Gerhard H?ring wrote:
> Looks like there is:
> http://webcvs.kde.org/cgi-bin/cvsweb.cgi/kdepim/libical/src/python/

This seems to be a copy of the libical library; its official home page
is at http://softwarestudio.org/libical/ .

--amk


From Dick.Zantow at lexisnexis.com  Mon Sep  9 09:18:58 2002
From: Dick.Zantow at lexisnexis.com (rzed)
Date: Mon, 9 Sep 2002 09:18:58 -0400
Subject: Grumble about too strict an attitude about backward compatibility...
References: <mailman.1031563647.10076.python-list@python.org>
Message-ID: <ali764$rro$1@mailgate2.lexis-nexis.com>

"Pearu Peterson" <pearu at cens.ioc.ee> wrote in message
news:mailman.1031563647.10076.python-list at python.org...
>
> On 4 Sep 2002, Skip Montanaro wrote:
>
> >
> > I don't want to wake any sleeping dogs, however, as I'm sitting
here
> > watching a bunch of Fortran function names fly by as SciPy builds,
I'm
> > reminded of the fairly recent threads about backward
compatibility,
> > Python-in-a-tie, etc.   Here I am compiling with a Fortran 90/95
compiler
> > (Sun's Forte thing-a-ma-bob) and see function names like
> >
> >         lpni:
> >         klvna:
> >         chgubi:
> >         cyzo:
> >         klvnb:
> >         rmn2so:
> >         csphik:
> >
> > spew forth.  So, while it's great that this same large library
compiles and
> > runs on compilers back to at least Fortran 77 (and probably
earlier),
> > programmers are still stuck with the same cryptic function and
data names
> > they had to deal with 30+ years ago, all in the name of backward
> > compatibility.
>
> I don't think that its is due to the backward compatibility.
> There is huge amount of Fortran 77 code available and if one wants
to fix
> the cryptic names of functions to something more meaningful, then
one has
> to switch to newer Fortran standard (that allows longer names) and
edit
> huge amount of F77 code. I think nobody is willing to take this task
due
> to enormous amount of work (and it cannot be fully automated, some
> brain has to deside what are meaningful names), the backward
compatibility
> is a secondary issue (if an issue at all).
>
> > What's the Python connection?  Other than a reminder not to get to
slavish
> > about backward compatibility, I note that these same function
names will
> > then go on to pollute the Python namespace because all this
Fortran code is
> > automatically wrapped using f2py.
>
> The Fortran function names, that you refer to, are visible only in
the
> extension modules that are wrapping these functions (this is due to
how
> Python imports shared modules). All functions that f2py generates,
> have the f2py_ prefix to avoid name collision with names from Python
> or any other library.
> So, I don't understand your argument about name pollution.
>
> And btw, f2py supports mapping function names so that wrapping a
Fortran
> function with a name `tiarcf' can be accessed in Python side as
> `this_is_a_really_cool_function', for example.
>
> Pearu
>
>




From lpgray at UOP.com  Wed Sep 18 16:35:59 2002
From: lpgray at UOP.com (Gray, Lee P.)
Date: Wed, 18 Sep 2002 15:35:59 -0500
Subject: Python GUI app to impress the boss?
Message-ID: <1181DE2559EBD1118EF000805F57FC9C0426B22B@shvmail1>

	(Sorry for the screwed up quoting... I have to use Outlook at work.
:(  ) 

> From:	Christian Tismer [SMTP:tismer at tismer.com]
> Sent:	September 18, 2002 3:14 PM
> To:	Lee Gray
> Cc:	python-list at python.org
> Subject:	Re: Python GUI app to impress the boss?
> 
> Boa is in fact quite a bit alpha-ish, but after you're
> used to the quirks and limitations, it is really great.
> Also note that you should use the version from CVS
> which is *far* more developed than the downloadable.
> 
> I've written a quite big wizard for 3D image processing
> with wxPython, PIL, Stackless. If you like, I can prepare
> some screen shots.
> 
	[Gray, Lee P.]  ---------------
	Yeah, I tried to download from CVS, but being a complete noob
couldn't figure out how, short of tracking down and copying and pasting
every updated file.  Any tips would be appreciated.

	Yes, I'd definitely like to see your screenshots.

	Thanks,
	Lee



From wlfraed at ix.netcom.com  Thu Sep 26 03:48:45 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Thu, 26 Sep 2002 00:48:45 -0700
Subject: combining mailbox and email modules
References: <NPck9.581867$UU1.104088@sccrnsc03> <slrnap41re.1mh.matt@happy-hour.t-island.com> <LAxk9.79207$gA4.27843@sccrnsc02>
Message-ID: <u4euma.fc4.ln@ix.netcom.com>

Bob Horvath fed this fish to the penguins on Wednesday 25 September 
2002 11:14 pm:


> ... and here it the message...
> 
>  From mickey at verizon.net  Tue Aug 20 10:03:52 2002
> Return-Path: <mickey at verizon.net>
> Delivered-To: bob at horvath.com

        Is that indent on the "From ..." line really in the file, or just some 
artifact of pasting it into your post? I just checked my Eudora .mbx 
files and they don't show an indent on the "From ..." line

        If it /is/ in the file, that could be the bad line.

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From mgarcia at cole-switches.com  Wed Sep 18 20:37:18 2002
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: 18 Sep 2002 17:37:18 -0700
Subject: isFloat: Without Exception-Handling
References: <mailman.1032364350.15105.python-list@python.org>
Message-ID: <fec163ad.0209181637.3b523efc@posting.google.com>

sismex01 at hebmex.com wrote in message news:<mailman.1032364350.15105.python-list at python.org>...
> def isFloat(S):
>   R = 0
>   try:
>     float(S)
>     R = 1
>   finally:
>     return R

slightly shorter way:

def isFloat(x):
    try:
        float(x)
        return 1
    except:
        return 0

def test_print(a):
    print '%s -> %r' %(a,eval(a))

test_print('isFloat(1)')
test_print('isFloat(1.1)')
test_print('isFloat("1")')
test_print('isFloat("1.1")')
test_print('isFloat("a")')
test_print('isFloat( [] )')
test_print('isFloat( () )')
test_print('isFloat( {} )')


From thorsten at thorstenkampe.de  Sat Sep 14 04:28:30 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Sat, 14 Sep 2002 10:28:30 +0200
Subject: Coding Style: Quotes
References: <bt1g9.17429$Pf7.590136@news1.west.cox.net> <altgev$10u79$1@ID-77524.news.dfncis.de>
 <mailman.1031950037.30531.python-list@python.org>
Message-ID: <alurvl$17g58$1@ID-77524.news.dfncis.de>

* Michael Stenner
> On Fri, Sep 13, 2002 at 10:05:45PM +0200, Thorsten Kampe wrote:
>> * John Waycott
>>> I'm curious if others have adopted any standards for choice of single vs.
>>> double quotation marks to delimit strings. A look through the standard
>>> library reveals the choice of one over the other is rather arbitrary. I
>>> suspect it really makes no difference, but the question has come up during a
>>> code review.
>> 
>> Python itself seems to "prefer" single quotes slightly:
>> 1: >>> "String"
>> 1:     'String'
>> 2: >>>
>
> Just because it uses the single quote doesn't mean it prefers it.  It
> (or the authors, rather) had to choose.  They couldn't just not pick
> one.  Also, the method you use to input the string IS NOT part of the
> string, and so should be dropped after it's role (encoding the string)
> is complete.  Therefore, your example does not support your statement.
>
> Of course, I suspect it is for just this reason that you quoted
> "prefer". :)

The Python interpreter has it's own preferences of outputting/printing 
objects. So the question "'" versus """ is just a part of a wider 
discussion. I asked myself the same question when I started four weeks 
ago. How to put spaces, commas, string delimiter, indent etc.?

Instead of doing it my way, I decided to do it the Python way for sake 
of readability of my code.

1: >>> 1
1:     1
2: >>> { }
2:     {}
3: >>> [ 1 , 2 ]
3:     [1, 2]


Thorsten


From op73418 at mail.telepac.pt  Fri Sep  6 09:58:55 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Fri, 06 Sep 2002 14:58:55 +0100
Subject: Difference between static method and a function as a class attribute?
Message-ID: <pqchnu40hs1dgqc4bct2shi4lbpl1oe80i@4ax.com>

Hi,

What are the differences, if any, between 1 and 2?

1.
def funtest(*args):
    pass

class test(object):
    method = funtest

2.
class test(object):
    def funtest(*args):
        pass

    method = staticmethod(funtest)


And if there are any differences do they really matter, that is, does
anybody has any test cases where 1 and 2 are not interchangeable?

TIA and all the best,
Gon?alo Rodrigues


From pete at shinners.org  Sat Sep 28 15:17:48 2002
From: pete at shinners.org (Pete Shinners)
Date: Sat, 28 Sep 2002 19:17:48 GMT
Subject: pygame.event.Event?
References: <r_Lk9.174774$pX1.6183293@news2.tin.it> <up9gk8asa6g37c@corp.supernews.com>
Message-ID: <wfnl9.437869$kp.1232105@rwcrnsc52.ops.asp.att.net>

In article <r_Lk9.174774$pX1.6183293 at news2.tin.it>, Max wrote:
>I can't generate a QUIT event in pygame:
>pygame.event.Event(type, dict, [keyword_args]) -> Event
>
>What do I have to put in "dict"?

i should fix the docs up a little on that. the dictionary argument is
optional. it is a way to add any extra attributes to the event. the QUIT
event doesn't really have any attributes. you can ignore the argument.

also, an easier way to handle the extra attributes is with keyword argument
handling. "my_quit = pygame.event.Event(QUIT, fake=1)", and your program
could easily later determine if this was a "real" quit event or not.




From maxm at mxm.dk  Thu Sep 12 12:52:59 2002
From: maxm at mxm.dk (Max M)
Date: Thu, 12 Sep 2002 18:52:59 +0200
Subject: Coding Style: Quotes
References: <bt1g9.17429$Pf7.590136@news1.west.cox.net> <mailman.1031842455.20916.python-list@python.org>
Message-ID: <3D80C66B.5020206@mxm.dk>

Michael Stenner wrote:
> On Thu, Sep 12, 2002 at 02:25:11PM +0000, John Waycott wrote:
> 
>>I'm curious if others have adopted any standards for choice of single vs.
>>double quotation marks to delimit strings. A look through the standard
>>library reveals the choice of one over the other is rather arbitrary. I
>>suspect it really makes no difference, but the question has come up during a
>>code review.
> 
> 
> Since there is no "physical" difference between them, I can't imagine
> why it would matter.  The only reason there "needs" to be two is for
> nested quoting.  Sure, it would be annoying to alternate for each
> string, but beyond that...


It makes about as much sense to standardize the use of double and single 
quoting in Python as it does do try and genreate personal email adresses 
in a company from some kind of systematic approach.

There will allways be reasons to break the system.

Personally I use single quotes as much as possible because it is a 
direct key on may Danish keyboard. To double quote I have to press 
shift+2. :-/

regards Max M



From ralf_ahlbrink at web.de  Sat Sep 21 12:17:21 2002
From: ralf_ahlbrink at web.de (RA)
Date: Sat, 21 Sep 2002 18:17:21 +0200
Subject: PyQt setup on Linux Mandrake 8.2
References: <3D8B4898.20808@nospam.attbi.com>	<ddc19db7.0209210259.7beb52d4@posting.google.com>
Message-ID: <3d8c9b92$0$182$9b622d9e@news.freenet.de>

On Sat, 21 Sep 2002 03:59:41 +0000, Armin Steinhoff wrote:
> dwelch <dwelch91 at nospam.attbi.com> wrote in message news:<3D8B4898.20808 at nospam.attbi.com>...
[Snip]
>  
> I would suggest to dowload the latest version of Qt from the Trolltech
> homepage, just for a simple and clean recompilation.

Why not install a new compiler for a somewhat cleaner 
recompilation. And a vanilla kernel...

Bulls..t
At least if you find that e.g. KDE is working stable. And Mdk 8.2
is _not_ that bad.

Unfortunately, Mandrake is somewhat more perl than python biased.

> 
> And do the same with SIP and PyQt from the Riverbank homepage ...

Yes. You will need this, because the most recent RPMs are version
PyQt-3.1 and sip-3.3.2. 
Take a look at <ftp-mirror>/Mandrake-devel/cooker/contrib for SRPM
files or http://cvs.mandrakesoft.com/cgi-bin/cvsweb.cgi/contrib-SPECS/
just for RPM spec files.

If you know a bit about RPM, perhaps you, dwelch, can manage
to update these packages (rpm -Uhv xx.src.rpm, a few changes ..., 
rpm -ba SPECS/sip.spec).
Otherwise you could contact me. In about a week I've got time
to do this.

> 
> It seems to be better to forget that outdated stuff provided by the 
> Mandrake distribution ...

Mandrake's packages are predominantly rather new versions. But
see above.

> 
> Armin
> 
>> 
>> Installing software on Linux makes me feel like such an idiot!

The RPM  stuff (or DEB) has got real advantages compared to 
"setup.exe". 
And if the offered versions are outdated, fetch the
new source, learn from the (e.g.) sip.spec file somewhat about
the compilation, and make a new package!

>> 
>> Any help or pointers would be much appreciated.
>> 
>> Don



From jubafre at brturbo.com  Mon Sep 23 15:40:42 2002
From: jubafre at brturbo.com (jubafre at brturbo.com)
Date: Mon, 23 Sep 2002 16:40:42 -0300 (GMT-03:00)
Subject: writing in a file as binary mode????
Message-ID: <1401166435.1032810042920.JavaMail.nobody@webmail2.brturbo.com>

i open the a file for writing in a binary mode, but doesnt work

final3=['0020 0000x9', '0030 0000xa', '00B4 0000x8', '0030 0000xb', '00F0', '0001', '0003', '0005']
destino=raw_input("Digite o nome do arquivo .mem: ")
f = open(destino,'wb')
f.write('\x03AHM ')
for i in final3:
    f.write(i+' ')
f.close()


just the "f.write('\x03AHM ')" work, i open the file in the notepad and my file appears like a text file and not as a binary file, i dont know why??

>>file.mem
AHM 0020 0000x7 0030 0000x8 0010 0000x9 00F0 0003 0002 0000 

>>i have the file like this, how??
AHM  ? 0 ?  ? ? 




Juliano Freitas
www.gebrasil.hpg.com.br 

From tjlavoie at mb.sympatico.ca  Fri Sep 27 12:41:05 2002
From: tjlavoie at mb.sympatico.ca (Tim Lavoie)
Date: 27 Sep 2002 11:41:05 -0500
Subject: mod_python
References: <an1grv$moqid$1@hades.rz.uni-sb.de>
Message-ID: <87it0rqtb2.fsf@theasylum.dyndns.org>

>>>>> "Uwe" == Uwe Schmitt <uwe.schmitt at procoders.net> writes:

    Uwe> Hi, I tried mod_python these days and found it quite usefull.
    Uwe> Are there any modules for handling sessions, cookies, and
    Uwe> persistent data ?

Try the Webware toolkit, it has a ton of goodies there. It's at
http://www.webware.sourceforge.net. The main piece of it that you
would want for now is called WebKit, and it integrates nicely with
Apache via a couple of different methods, depending on what works best
for you. It feels a lot like Java servlet development to me, without
some of the nuisance factor of Java. So, you have persistent servlets,
which you can reload on the fly and so on, and which have access to a
whole mess of data.

As a side benefit, there are hooks to a couple of templating
mechanisms, so you don't have to write all the "print '<html>'" bits
manually.

  Cheers,
  Tim

-- 
Testing can show the presense of bugs, but not their absence.
                -- Dijkstra


From tjreedy at udel.edu  Fri Sep 20 09:40:22 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 20 Sep 2002 13:40:22 GMT
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com>
Message-ID: <azFi9.163782$5r1.6708151@bin5.nnrp.aus1.giganews.com>

"Benjamin" <phncontact.libero.it@> wrote in message
news:ifjloug0dkla5o3lvli464mmou5mbh47d5 at 4ax.com...
> i've been learning python not for becoming a programmer
> (i study modern languages at the university), but for having
> fun. my personal motto is "i want to know how it works".
['but I'm getting bored']

Perhaps you should look into some area of computer-aided language
analysis and processing.  Something too specialized (and perhaps too
difficult for most people) to be in any introductory text.  Example:
write a program to conjugate any Spanish verb given its infinitive
form.  If you succeed at that, write another one to de-conjugate any
conjugated form back to infinitive form (this is more difficult).  If
you still need a challenge 8<), write a third program to recognize
verbs in Spanish text.

Or pick a different language problem of more interest to you.  Perhaps
a cross-language
lexicon comparison.  For example, given the words for 'zero' to 'ten'
in several languages, devise an 'index of similarity' calculation.
(This is also non-trivial).  Calculate a similarity matrix and use it
to draw some conclusions.

I hope these ideas might serve to jog you out of the
standard-boring-problem rut you seem to be in.

Even if you do not become a 'programmer', you may easily find yourself
using programming skills and maybe even Python in particular in many
other careers.

Terry J. Reedy




From Padraig at Linux.ie  Fri Sep  6 08:36:03 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Fri, 06 Sep 2002 12:36:03 GMT
Subject: tree data structure
References: <FM0e9.2860$cP3.6165@news.iol.ie> <ala6c3$6i2$1@bob.news.rcn.net>
Message-ID: <Ti1e9.2864$cP3.6124@news.iol.ie>

Raymond Hettinger wrote:
> There are a number of ways to represent trees:
> -- instances of a treenode class
> -- a list of lists:  [l0, [s2, [s0, [q1]], [s1,[q2]], [s3]]]
> -- or, my favorite, a dictionary of lists:
> 
> 
> original = [('L0','S2'), ('S2','S0'), ('S2','S1'), ('S1','Q2'), ('S0','Q1'), ('S2','S3')]
> 
> tree = {}
> for key, value in original:
>     tree.setdefault(key,[]).append(value)
> tree
> {'S2': ['S0', 'S1', 'S3'], 'S1': ['Q2'], 'S0': ['Q1'], 'L0': ['S2']}

Interesting I didn't know about setdefault.
However how would you iterate over the above?
I would like to first get L0, then S2, ...
So I was thinking of something like:

('L0', ('S2', ('S0', ('Q1', None)), ('S3', None), ('S1', ('Q2', None))))

However I've only started programming python for the
first time today and am unsure of how to best acheive this.

thanks a lot!
P?draig.

> "Padraig Brady" <Padraig at Linux.ie> wrote in message
> news:FM0e9.2860$cP3.6165 at news.iol.ie...
> 
>>Hi,
>>
>>I've a graph like stucture represented
>>by a list of tuples:
>>
>>[(L0,S2), (S2,S0), (S2,S1), (S1,Q2), (S0,Q1), (S2,S3)]
>>
>>And would like to convert it to
>>a tree structure like:
>>
>>           S0 - Q1
>>         /
>>L0 - S2 - S3
>>         \
>>           S1 - Q2
>>
>>Any tips appreciated.
>>
>>Thanks,
>>P?draig.
>>
> 
> 
> 



From goodger at users.sourceforge.net  Sun Sep 15 12:59:48 2002
From: goodger at users.sourceforge.net (David Goodger)
Date: Sun, 15 Sep 2002 12:59:48 -0400
Subject: ANN: DocFactory wxPython app for Docutils
Message-ID: <B9AA34C3.28DFC%goodger@users.sourceforge.net>

Gunnar Schwant has contributed DocFactory, a wxPython GUI application for
Docutils.  It is in the preliminary stages.  For details, please see

    http://docutils.sf.net/sandbox/gschwant/docfactory/README.html

I have begun a "To Do" list:

    http://docutils.sf.net/sandbox/gschwant/docfactory/NOTES.html

The code is available via CVS or snapshot:

    http://docutils.sf.net/docutils-sandbox-snapshot.tgz

Please try it out.  Feedback is welcome: bug reports, patches, feature
ideas, etc.

-- 
David Goodger  <goodger at users.sourceforge.net>  Open-source projects:
  - Python Docutils: http://docutils.sourceforge.net/
    (includes reStructuredText: http://docutils.sf.net/rst.html)
  - The Go Tools Project: http://gotools.sourceforge.net/




From no at mail.com  Tue Sep 17 06:00:52 2002
From: no at mail.com (BT)
Date: Tue, 17 Sep 2002 12:00:52 +0200
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de><pUdh9.9377$yt3.4604908@newssrv26.news.prodigy.com><mailman.1032184821.5180.python-list@python.org> <c%qh9.102794$ub2.2120685@news1.tin.it> <mailman.1032212180.11973.python-list@python.org> <pKth9.118182$pX1.4250378@news2.tin.it> <3d867f4d$0$23173$afc38c87@news.optusnet.com.au>
Message-ID: <3D86FD54.92DD65F2@mail.com>

>> True.. that's why I'm planning to obfuscate the code then encrypt it.

>you don't want to go overboard with this, people still have to use the
>program - where are you going to keep the code that is 'now' (once
>decrypted) in plaintext?
>[...]
>if they're able to run the code, it'd be a fairly solid bet that they'd know
>about Python, if not, a quick search on Google would probaly net you all the
>info you need to find out just what that file is. After all, in order to run
>the program they'd need to at least install Python somehow - unless you're
>planning to embed Python in some other app, and run your code in that...

Yes, sorry... I forgot to mention I'm planning to embed the encrypted pyc in a
standard c application
who will load the file in memory and decrypt it.
I've read somethig about importing .pye files from a db, but now I am still
developing, never tried to do that:
I'll try do protect the app when it's completed.

For "obfuscate the code" I mean writing (or find) a small app that sobstitute
every variable and function  name
with a random code like 'd93jkm39'.
Yes, the software will be usable even in plain text, but I think everyone will
waste some time to figure out
what's happening.

I don't want to waste months to find a way to protect it, a week or two will be
sufficient.



From wpmelcher at snet.net  Mon Sep 30 17:21:41 2002
From: wpmelcher at snet.net (Bill Melcher)
Date: Mon, 30 Sep 2002 21:21:41 GMT
Subject: PySol: "global name 'win32_gethomedir' is not defined"
References: <all3u2$qgm$1@brokaw.wa.com>
Message-ID: <Ff3m9.1416$DB6.501602443@newssvr10.news.prodigy.com>

I think Alan's problem is that the Python distribution that he installed
does not include the Win32 Extensions.  The Python Labs install does not
include the extensions.  The ActiveState install does.


--
Cheers, Bill
TANSTAAFL!

"Alan Winston" <awinston at scn.org> wrote in message
news:all3u2$qgm$1 at brokaw.wa.com...
> I've been trying to run the PySol 4.81 pyc files under Python 2.2.1 and
> WinXP and Win98, with no success.
>
> The Python window closes much faster than I can read the error message (is
> there any way to force the window to stay open?), but I managed to catch
it
> with the print screen key:
>
> "global name 'win32_gethomedir' is not defined"
>
> PySol 4.60 runs fine, installed with the Windows install package, so this
is
> a low priority for me, but I am still curious.
>
>
>
>
> Alan Winston
> Seattle
>
>
>
>
>




From Padraig at Linux.ie  Mon Sep 30 09:12:39 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Mon, 30 Sep 2002 13:12:39 GMT
Subject: Find out link location
References: <7uf9na-vd9.ln@lbdb01.gsi.de> <3D984AF8.3070908@Linux.ie>
Message-ID: <3D984D80.9040509@Linux.ie>

Padraig Brady wrote:
> Klaus Hoeppner wrote:
> 
>> Hi,
>>
>> with the os.path module I can use the islink function to check, wether
>> a file is symbolic link. But how can I find out, where the link points
>> to. E.g. if I have readme.txt --> ../packages/mypackage/readme.txt how
>> do I get the "../packages/mypackage/readme.txt" with python?
> 
> import os
> filter (lambda x: x.find("link") != -1, dir(os))

To be a little more modern:

[x for x in dir(os) if x.find("link") != -1]

P?draig.



From dag at animagicnet.no  Wed Sep  4 08:11:35 2002
From: dag at animagicnet.no (Dag)
Date: 4 Sep 2002 12:11:35 GMT
Subject: Matching a string up to a word
Message-ID: <slrnanbkbc.huj.dag@omally.animagicnet.no>

I have a string that contains an ID number a file path and a date, kind
of like this:

23545 /this/is a/path/to/a file Sun Sep 7 2002.

I want to extract the file name.  If I could be sure that the file path
had no spaces or if it would be quited it would be easy.  Unfortunatly 
many files will have spaces and changing the format of the output I'm
dealing with isn't possible.  I do however know that the file name starts 
after the first whitspace and that it's always followed by a three letter 
day name (like Mon, Tue etc.)  Given this is there an easy way to extract 
the whole file path from the string.

Dag



From skip at pobox.com  Tue Sep 17 22:09:41 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 17 Sep 2002 21:09:41 -0500
Subject: Python GUI app to impress the boss?
In-Reply-To: <96904b50.0209171305.6d65f574@posting.google.com>
References: <96904b50.0209171305.6d65f574@posting.google.com>
Message-ID: <15751.57445.514521.644297@12-248-11-90.client.attbi.com>

    Lee> Can someone point me to a good Python GUI app to demonstrate (other
    Lee> than IDLE)?

Lee,

At IPC 10, Michel Sanner from Scripps took home the best paper award for his
ViPEr work.  References:

    http://www.python10.org/p10-awards.html
    http://www.python10.org/p10-papers/10/index.htm
    http://www.scripps.edu/~sanner/python/viper/index.html

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From erick_papadakis at yahoo.com  Sun Sep 15 21:02:17 2002
From: erick_papadakis at yahoo.com (Erick Papadakis)
Date: 15 Sep 2002 18:02:17 -0700
Subject: can wxpython do this?
Message-ID: <737aefee.0209151702.3f735252@posting.google.com>

hi, i am considering learning wxpython and have a few questions before
i
get into it to make sure it is worth it. 

1. what does it give me that java or c++ dont? is it cross-platform
too?

2. is it a good platform to develop desktop application in, given that
it
is a GUI app? how will i distribute it? is there any easy installer
that
will allow me to install python first on a layman user's computer
first,
then wxpython, then my own application? is this the concept of
wxpython
or am i missing something? 

3. does it allow me access to windows (or linux x-windows) native
APIs?
for instance, can i minimize my application to a user's desktop's
taskbar and have a right-click context menu on that taskbar entry?

4. do i have something like a browser control built in? i want my
desktop app to interact with my web server and display the response
just as if it was a browser (BUT with some more client-side richness
in features, hence the app to begin with). is this possible?

thanks very much for any ideas/resources that i can use to decide. 

cheers/.ep


From martin at v.loewis.de  Thu Sep 26 13:53:14 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 26 Sep 2002 19:53:14 +0200
Subject: fcntl module on Linux and nonblocking I/O
References: <7e9d8e38.0209260932.ed2c78@posting.google.com>
Message-ID: <m3lm5otz79.fsf@mira.informatik.hu-berlin.de>

skoranda at uwm.edu (Scott Koranda) writes:

> I just installed python 2.2.1 on my RH 7.3 Linux box and noticed
> that the fcntl module does not appear to have O_NDELAY, FNDELAY,
> or anything that appears similar.
[...]
> Without such a constant defined in the fcntl module how can I make a
> pipe non-blocking?

Use os.O_NDELAY.

Regards,
Martin


From ggittings at arraybiopharma.com  Fri Sep  6 18:46:54 2002
From: ggittings at arraybiopharma.com (Gabe Gittings)
Date: 6 Sep 2002 15:46:54 -0700
Subject: COM/Active X newbie question
Message-ID: <5523ec5.0209061446.1404cd10@posting.google.com>

I am a newbie to prgramming in Python and programming under windows. I
am trying to access an .ocx file. I can access the methods in python
after running makePy.py but when I run a test script and try to access
a method from the COM object I get this error.

Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
  File "C:\Python22\Lib\site-packages\win32com\gen_py\7ADCDC5B-D0B6-11D1-AA76-002078136998x0x1x0.py",
line 57, in GetRunManagerStatus
    return self._oleobj_.InvokeTypes(0xa, LCID, 1, (8, 0), (),)
com_error: (-2147418113, 'Catastrophic failure', None, None)


this is the code I am using

import win32com.client.gencache
import win32com.client

# Use these commands in Python code to auto generate .py support
# makes a .py for the Acquisition .ocx via UUID
from win32com.client import gencache
gencache.EnsureModule('{7ADCDC5B-D0B6-11D1-AA76-002078136998}', 0, 1,
0)

xdk = win32com.client.Dispatch('{7ADCDC5E-D0B6-11D1-AA76-002078136998}')
print xdk.GetSeqQueuePaused()


I am not really sure if .Dispatch is even appropriate in this
situation because I need to monitor an already running application
which I access through it's .ocx. Hope this makes sense.

So in case I wasn't clear there are two questions.
1. why am I getting the error message when I am accessing the COM
objects methods?
2. is this how you connect to a client side COM if you want to access
an already running app.

Thanks


From storedmails at gmx.net  Mon Sep 23 10:50:23 2002
From: storedmails at gmx.net (beachboy)
Date: 23 Sep 2002 07:50:23 -0700
Subject: Sessionhandling
Message-ID: <c5b3549a.0209230650.34a30fca@posting.google.com>

Hi!
Please can anyone help me.
My problem:
I'm trying to program a web application with python. This tool has to
provide diffrent languages. The user has to choose the language at the
beginning. So i have to save the information which language the user
chooses. I think this is a perfect example for using sessions.
The only problem is I've got no idea how to realise this in python.
Please help me as soon as you can because I have to solve this problem
very quickly.

hoping to hear from you soon
beachboy


From lists at gregfortune.com  Fri Sep 20 02:12:33 2002
From: lists at gregfortune.com (Greg Fortune)
Date: Thu, 19 Sep 2002 23:12:33 -0700
Subject: Catching results of an exec
References: <3d88f701$0$304$39cecf19@nnrp1.twtelecom.net> <3D88FC6E.59D0DD01@alcyone.com>
Message-ID: <3d8aba3c$0$308$39cecf19@nnrp1.twtelecom.net>

Erik Max Francis wrote:

> Greg Fortune wrote:
> 
>> Is there any way to "catch" the value produced by an exec?  I'm
>> considering
>> things like print statements that don't really assign a value, but do
>> send
>> a result to stdout.
> 
> The problem is "the value produced by an exec" is ambiguous, because
> execs (as don't return values).  If you want to capture some subpart of
> the execution of a statement, you'll have to handle it case by case.
> 
>> Right now I've got a nasty hack that redirects stdout, uses an
>> InteractiveInterpereter from the code module to execute the code,
>> writes to
>> an output file, reads the result from the file, restores stdout and
>> then
>> sends the result to the client.  Hopefully, that's a terrible way to
>> do it.
> 
> That's probably the best way to do it, actually -- or, at least, with an
> overridden sys.stdout in the globals dictionary that you're passing to
> eval/exec (which doesn't necessarily have to be the same dictionary as
> your main script is running).
> 

Thanks, I hadn't thought to use the namespace stuff on exec.  That'll clean 
up a couple more lines..

Greg


From jb at cascade-sys.com  Mon Sep 30 13:54:22 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 30 Sep 2002 10:54:22 -0700
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05>
Message-ID: <3D988FCE.6040502@cascade-sys.com>

Steve Holden wrote:

><sismex01 at hebmex.com> wrote...
>  
>
>[ ... ]
>OK, so don't use whole pennies (or smallest currency) as your
>base, use 10,000ths of a penny, but still use integer arithmetic
>for your calculations.
>
Normal Integers aren't big enough.  2^31 / 10K == 215K.

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From claird at starbase.neosoft.com  Fri Sep 13 16:33:02 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 13 Sep 2002 15:33:02 -0500
Subject: Tkinter listbox clicking problem
References: <sGog9.307183$aA.53945@sccrnsc02> <altfrk$1eva$1@nntp6.u.washington.edu>
Message-ID: <2D7E3697CF5D1521.5D9711540E657ED8.7EF0451C598438FC@lp.airnews.net>

In article <altfrk$1eva$1 at nntp6.u.washington.edu>,
Russell E. Owen <owen at nospam.invalid> wrote:
>In article <sGog9.307183$aA.53945 at sccrnsc02>,
> "Matthew Bonig" <mbonig at hotmail.com> wrote:
>
>>Right now I have it bound so when a Tkinter listbox gets a <Button-1>
>>(lb.bind("<Button-1>", someFunc)) it calls function someFunc. Now,
>>someFunc() simply prints the curselection(). The problem is it's lagged...
>>if I first select list item 1 (index 0), and then 2, and then 3 (array
>>indexes 0,1,2) it prints out themout incorrectly.
>>(I have someFunc actually do some try, except stuff so I don't get uncaught
>>exceptions) Instead of printing 0,1,2 it prints "nothing selected", 0, 1.
>>The function get's called before the listbox updates it's selection markers!
>>
>>Does anybody know how I could force listbox to update it's selection markers
>>so instead of printing "nothing selected, 0,1  it prints 0,1,2 like it
>>should?
>
>I'm hoping somebody has better ideas, but here are some to try:
>
>Call update_idletasks before you ask for the selection. I have no idea 
>if this will work, but it might.
>
>An ugly solution that would probably work:
>- extract the y coordinate from the event callback
>- call nearest(y_coord) to get the item closest to the mouse click
>- explicitly select that index item. You could try skipping this step, 
>but you risk a possible error between what you think they clicked on and 
>what is hightlighted. Unfortunately, this step will may be a real pain 
>for a list that supports multiple selected items.
>
>Poll for selection changes -- thus it may take a bit of time before your 
>code notices a change, but it will always be right once it gets there.
>
>-- Russell

Whoa!  There's a far easier solution.  Read up on 'bindtags'
in <URL: http://web.pydoc.org/2.2/Tkinter.html > and elsewhere.
<URL: http://www.cwi.nl/~jacob/textui.py.html > has a small
example that almost is pertinent.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From jn at who.net  Thu Sep  5 10:12:08 2002
From: jn at who.net (john nielsen)
Date: 5 Sep 2002 07:12:08 -0700
Subject: psyco is great!
Message-ID: <22cc1071.0209050612.7311a89c@posting.google.com>

Psyco is a great idea.

I just tried it out on _unmodified code_ that parses large logfiles
and it doubled the performance.

A nice convenince compared to things like swig, is that you don't even
need to have a C compiler around. Just point it to the functions you
want optimized.

import psyco

def fred(): print 'stuff here'

psyco.bind(fred)

fred()


john


From me at aaronsw.com  Wed Sep  4 23:06:53 2002
From: me at aaronsw.com (Aaron Swartz)
Date: 4 Sep 2002 20:06:53 -0700
Subject: Graham's spam filter
References: <mailman.1029992314.27908.python-list@python.org>
Message-ID: <c2ef5940.0209041906.4c3c5cba@posting.google.com>

I've thought about building something similar too. I don't need the
network code, I can just integrate it into my mail path. Here are some
useful resources:

python-cdb[1] is a Python interface to D. J. Bernstein's cdb
package[2], which supports fast lookups (normally just two disk
accesses), low overhead (24 bytes per record, no startup cost) and is
designed for mission-critical applications like email. In short, it's
absolutely perfect for this application.

[1] http://pilcrow.madison.wi.us/
[2] http://cr.yp.to/cdb.html

Bram Cohen has been thinking of modifications to the algorithm and
writing them up in Python[3].

[3] http://www.advogato.org/person/Bram/diary.html?start=17

Hope this helps,

Aaron Swartz [http://www.aaronsw.com/]


From timr at probo.com  Wed Sep 25 01:56:51 2002
From: timr at probo.com (Tim Roberts)
Date: Tue, 24 Sep 2002 22:56:51 -0700
Subject: Can't suspend a thread?
References: <QZ4j9.10630$Lg2.2393215@news2.news.adelphia.net> <amognf$r6g$1@panix1.panix.com>
Message-ID: <juj2puko63pmq5k4j4tm2eonoadvpup50d@4ax.com>

aahz at pythoncraft.com (Aahz) wrote:
>
>Robert Oschler <Oschler at earthlink.net> wrote:
>>
>>I was hoping to have a master thread that woke up and suspended several
>>other threads dynamically, as needed.  From the books I have, that seems to
>>be impossible in Python?  Is this true?  If so, is there a workaround?
>
>Threads suspend themselves; they can be woken up by other threads.

It is worth pointing out that this "restriction" is not a Python thing; it
is an operating system thing.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From kilaruba at copper.net  Wed Sep 18 16:59:58 2002
From: kilaruba at copper.net (bert)
Date: Wed, 18 Sep 2002 13:59:58 -0700
Subject: multiple index inconsistency
Message-ID: <TP5i9.9293$Rf3.1003108@newsfeed.slurp.net>

The following line of code
>>> a = 3*[range(3)]
produces
[[0, 1, 2], [0, 1, 2], [0, 1, 2]]

If I then write, say, 
>>> a[0][1] = 3.7
I get
[[0, 3.7, 2], [0, 3.7, 2], [0, 3.7, 2]]
and not
[[0, 3.7, 2], [0, 1, 2], [0, 1, 2]]

Now if I write
>>> a = [[0, 1, 2], [0, 1, 2], [0, 1, 2]]
instead, and then 
>>> a[0][1] = 3.7
I get the result I expected (and wanted):
[[0, 3.7, 2], [0, 1, 2], [0, 1, 2]]
Why the inconsistency? Is this a bug or am I missing something?
-- 
Bert


From mickey at tm.informatik.uni-frankfurt.de  Sun Sep  1 19:40:25 2002
From: mickey at tm.informatik.uni-frankfurt.de (Michael 'Mickey' Lauer)
Date: 2 Sep 2002 01:40:25 +0200
Subject: QPEApplication.setStylusOperation() woes.
Message-ID: <3d72a569@nntp.server.uni-frankfurt.de>

Hi,

I'm using PyQt on Qt/Embedded within a palmtop environment.
Although PyQt seems to wrap QPEApplication.setStylusOperation
(widget, QPEApplicaiton.RightOnHold), the effect on calling it is nada.

Did anyone got this to work?

(Already posted on the PyQt-Mailing list, but only got silence - I'm
hoping to get a larger audience here).

Yours,

:M:
--------------------------------------------------------------------------
Dipl.-Inf. Michael 'Mickey' Lauer   mickey at tm.informatik.uni-frankfurt.de 
  Raum 10b - ++49 69 798 28358       Fachbereich Informatik und Biologie  
--------------------------------------------------------------------------


From iqbal at colorado.edu  Wed Sep 11 13:44:51 2002
From: iqbal at colorado.edu (Naveed Iqbal)
Date: 11 Sep 2002 10:44:51 -0700
Subject: Python-FilemakerPro 5.5 Connectivity
Message-ID: <6afb9d2c.0209110944.74aeaca2@posting.google.com>

Is it possible to connect to a Filemaker pro database from python? If
so what are the modules required? and what are the connection
parameters?
thanks
naveed


From tim.one at comcast.net  Sun Sep 29 13:58:18 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 29 Sep 2002 13:58:18 -0400
Subject: Python's import: why doesn't it work?
In-Reply-To: <3d9739c7$0$19818$91cee783@newsreader01.highway.telekom.at>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEIMBHAB.tim.one@comcast.net>

]Jurie Horneman]
> ...
> However, I do stand by my criticisms of the Python docs.

Every page of the docs sez at the bottom:

    See About the Python Documentation for information on
    suggesting changes.

which link says, in part,

    General comments and questions regarding this document should be
    sent by email to python-docs at python.org.  If you find specific
    errors in this document, please report the bug at the Python Bug
    Tracker at SourceForge.

The docs are exactly as good as volunteers have made them (well, actually
better, thanks to Fred <wink>).  When you find an explanation inadequate to
your needs, suggesting specific improvements in the right place is the only
way it will improve.




From aleax at aleax.it  Fri Sep 20 03:32:47 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 07:32:47 GMT
Subject: sorry for this stupid question about struct module
References: <3d8a4ae7$0$280$7a628cd7@news.club-internet.fr>
Message-ID: <zaAi9.117236$ub2.2498874@news1.tin.it>

Alan Marchand wrote:

> Hi all,
> 
> I tryed this :
> 
> e:\temp>python.exe
> Python 2.2.1 ( #34, Apr 9 2002, ...................)
> ...........
>>>> import struct
>>>> x = struct.pack( '>I', 10 )
>>>> print repr( x )
> '\x00\x00\x00\n'                       # why not !
> 
> How can I obtain the result a expect : '\x00\x00\x00\0A' ?

If you need a specific repr that ignores the possibility of
representing some characters otherwise than as hex escapes,
I think you'll have to code it yourself -- most particularly
so if you want some hex escapes to have the x (as in the \x00
occurrences above) and others to lack it (as in the \0A above):
how would any standard repr ever be able to guess these reqs?!

I don't understand why you need such a weird repr.  What are
you going to do with it?  x is a 4-byte string made up of bytes
with numerical values 0, 0, 0, 10 -- repr shows the latter as
\n, of course, since 10 is the value of \n -- but why is repr's
behavior problematic to you?


> Remark :
> This, to send a message to a IBM Mainframe in order to start an IMS
> transaction ( via IMS CONNECT ).

You won't use repr for the purpose of sending that message,
thus I'm more and more perplexed about why you think there's
a problem here.  String x does seem to be what you want to send.


Alex



From martin at v.loewis.de  Sun Sep 22 16:23:13 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 22 Sep 2002 22:23:13 +0200
Subject: Wanted: Tix examples
References: <3d8dde58$0$30454$afc38c87@sisyphus.news.be.easynet.net>
Message-ID: <m3fzw16afi.fsf@mira.informatik.hu-berlin.de>

"Justin" <e_fax_t at hotmail._ZAPME_.com> writes:

> The module itself mentions the tixwidgets.py demo;
> where does one find it?

It's in the Demo/tix directory of the Python distribution.

Regards,
Martin


From eric.brunel at pragmadev.com  Wed Sep 11 08:17:03 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Wed, 11 Sep 2002 14:17:03 +0200
Subject: regular expression for space seperated quoted string
References: <PqFf9.3627$cP3.7118@news.iol.ie>
Message-ID: <alnbpb$ce5$1@wanadoo.fr>

Padraig Brady wrote:

> Hi, I'm trying to split a string that is seperated
> by spaces and also contains double quoted words which
> can contain spaces:
> 
> For e.g. using: re.split('[ ]*"?([^"]*)"?[ ]*', s1)
> on this string: s1='1 "2" "thre e"'
> gives:          ['', '1 ', '', '2', '', 'thre e', '']
> 
> Problem with this is the '' entries, but this isn't too bad.
> 
> However using the above re with: s2='1 2 "th ree"'
> I get:                           ['', '1 2 ', '', 'th ree', '']
> 
> any ideas?

What about:
>>> p = r'[^ \t\n\v\f"]+|"[^"]*"'
>>> re.findall(p, '1 2 3')
['1', '2', '3']
>>> re.findall(p, '1 2 "three"')
['1', '2', '"three"']
>>> re.findall(p, '1 "2" "thr ee"')
['1', '"2"', '"thr ee"']
>>> re.findall(p, '"yeah it seems to be working!" yeah it seems to...')
['"yeah it seems to be working!"', 'yeah', 'it', 'seems', 'to...']

It leaves the double-quotes around the values, but they can be removed 
quite simply...

I also tried to use look-ahead and look-behind features in re's, but it 
doesn't have the expected result:

>>> re.findall(r'(?<=")[^"]*(?=")|[^ \t\n\v\f"]+',
        '"yeah it seems to be working!" yeah it seems to...')
['yeah it seems to be working!', 'yeah', 'it', 'seems', 'to...']
>>> re.findall(r'(?<=")[^"]*(?=")|[^ \t\n\v\f"]+', '1 2 3')
['1', '2', '3']
>>> re.findall(r'(?<=")[^"]*(?=")|[^ \t\n\v\f"]+', '1 "2" "thr ee"')
['1', '2', ' ', 'thr ee']

The " " between 2 and three is also matched...

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From henrik.motakef at web.de  Thu Sep 19 11:27:26 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 19 Sep 2002 17:27:26 +0200
Subject: strong/weak - dynamic/static [Was: Getting started]
References: <mailman.1032445617.19510.python-list@python.org>
Message-ID: <87ofaut2xt.fsf@pokey.henrik-motakef.de>

"Mark McEahern" <marklists at mceahern.com> writes:

> $ python
> Python 2.2.1 (#1, Jun 25 2002, 10:55:46)
> [GCC 2.95.3-5 (cygwin special)] on cygwin
> Type "help", "copyright", "credits" or "license" for more information.
> >>> class foo:
> ...     def fooize(self):
> ...             print self.__class__.__name__
> ...
> 
> >>> class bar:
> ...     def barize(self):
> ...             print self.__class__.__name__
> ...

>>> f = foo()
>>> b = foo()
>>> b.__class__ = bar
>>> f.fooize()
foo
>>> b.barize()
bar
>>> type(f) == type(b)
1
>>> type(f)
<type 'instance'>
>>>

There's a difference between a type and a class, I guess...


From djc at object-craft.com.au  Fri Sep 20 22:35:49 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 21 Sep 2002 12:35:49 +1000
Subject: saving state in Python/CGI
References: <5.0.0.25.2.20020920134912.00ad3230@hms.harvard.edu> <mailman.1032559721.32440.python-list@python.org>
Message-ID: <m3lm5wnk6y.fsf@ferret.object-craft.com.au>

>>>>> "Tim" == Tim Churches <tchur at optushome.com.au> writes:

Tim> Frank Gibbons wrote:
>>  Hi,
>> 
>> I've been using Python (and Jython) for about 5 months now. So far,
>> so great. I'm building a CGI app that has about 5 different stages,
>> and I need to maintain state between them. Specifically, it's a
>> scientific tool that requires the user to upload a (possibly quite
>> large) datafile, processes it, does some computation, then hands it
>> back (formatted) with results.
>> 
>> I need access to the data on each page, yet don't want to pass it
>> as a hidden field, because of its potential size. In Perl, I could
>> just use the save() method from CGI.pm to save the form on the
>> server side, pass a session key in a hidden field, then retrieve
>> the data using the key. For the life of me, I can't find an
>> equivalent way in Python. cgiFormStorage looks like a dictionary,
>> but it's immutable, so you can't retrieve data from disk and add it
>> to the form. cgi.py appears to have no analog to CGI.pm's save()
>> method, at least not that I've been able to find. I've tried
>> pickling the form, but that doesn't work either (it saves something
>> to disk alright, but not the large data retrieved from a multi-part
>> form file upload).
>> 
>> Perhaps I've missed something. Maybe that's just not how you do
>> things in Python. Can anyone help?

Tim> You need to use one of the (many) Python Web app frameworks which
Tim> help (to greater or lesser degrees) with storing state and other
Tim> typical Web app tasks.  There is a list of such frameworks
Tim> available at
Tim> http://www.python.org/cgi-bin/moinmoin/WebProgramming

Tim> Of these, I have only used Albatross, but I can heartily
Tim> recommend it - it was specifically designed with highly stateful
Tim> applications in mind, is elegantly designed and well written and
Tim> has excellent documentation and the authors are very helpful. We
Tim> use it for a range of quite complex epidemiological applications.

Tim> However, many of the other frameworks list on the above site are
Tim> also reputed to be very good and you should consider all of them
Tim> before making a choice.

I am one of the authors Tim is referring to.  The application you have
described has very similar data capture and processing requirements to
Tim's application.  Albatross was developed specifically with that
type of problem in mind.

If you have the time you should try a few of the toolkits to see which
most suits the way you like to solve problems.  One of the big features
that Albatross has going for it is the ability to expose deeply nested
recursive data structures to HTML form input.

Read point 3 of the following page for a description of the alias feature:

  http://object-craft.com.au/projects/albatross/albatross/tag-input-alias.html

- Dave

-- 
http://www.object-craft.com.au


From gmuller at worldonline.nl  Sun Sep 22 14:56:51 2002
From: gmuller at worldonline.nl (GerritM)
Date: Sun, 22 Sep 2002 20:56:51 +0200
Subject: How to get memory usage on Windows?
References: <amilhk$bi8$1@reader1.tiscali.nl> <amkpo5$kur$1@reader1.tiscali.nl> <f51sou8o4j9pkilhc3g6jevkqmkk7jr3pl@4ax.com>
Message-ID: <aml3th$qu3$1@reader1.tiscali.nl>

"Wolfgang Strobl" <ws at mystrobl.de> schreef in bericht
news:f51sou8o4j9pkilhc3g6jevkqmkk7jr3pl at 4ax.com...
> "GerritM" <gmuller at worldonline.nl>:
>
> >Searching in books , documentation (a.o. win32api) and net did not yet
help
> >I found this article on the net:
> >http://www.developeriq.com/Magazinestories/02jul29memory.php3
> >but how to call this function from Python?
>
> I guess thanks to Mark Hammond you already have everything you need.
>
> Have a look at  \Python22\Lib\site-packages\win32\Lib\win32pdhutil.py,
> uncomment the test function, run it.
>
> I snarfed the  following snippet straight from win32pdhutil.py:
<...snip...>
> --
> Wir danken f?r die Beachtung aller Sicherheitsbestimmungen

Wolfgang,

thanks. I think this will work only on NT. Here at home I have "only"
windows98, which gives the following message:
"RuntimeError: The pdh.dll entry point functions could not be loaded."

I will have to try this at work, on a windows2000 machine.

regards Gerrit

--
www.extra.research.philips.com/natlab/sysarch/





From oren-py-l at hishome.net  Sun Sep  1 02:19:38 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Sun, 1 Sep 2002 02:19:38 -0400
Subject: python slowdown after a longish time of running (garbage collector)
In-Reply-To: <87adn2o7op.fsf@nospam.eml.cc>
References: <20020829152040.GA3061@foof.i3.cz> <20020830133244.V3893@prim.han.de> <mailman.1030710614.27326.python-list@python.org> <3D701C6F.6010402@ob_scure.dk> <87adn2o7op.fsf@nospam.eml.cc>
Message-ID: <20020901061938.GA9528@hishome.net>

On Sun, Sep 01, 2002 at 12:49:29AM +0000, Mahesh Padmanabhan wrote:
> I don't know how GC is implemented in Python, but GC by its very
> can only guarantee that it will close the socket eventually - it
> cannot guarantee *when*.

The GC is implementation-dependent. For Jython it uses the garbage collector 
of the underlying JVM.  CPython uses a combination of reference counting and 
cycle garbage collection.

In CPython the reference counting means that unless you create a cyclic
reference an object will be garbage-collected immediately when there are no 
more references to it.  It is unportable to rely on this behavior but 
apparently many programmers do.

	Oren



From aleax at aleax.it  Thu Sep 19 12:41:04 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 19 Sep 2002 18:41:04 +0200
Subject: strong/weak - dynamic/static [Was: Getting started]
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F06233024CCA79@mail.hebmex.com>
References: <F7DB8D13DB61D511B6FF00B0D0F06233024CCA79@mail.hebmex.com>
Message-ID: <E17s4Mz-0004Nz-00@mail.python.org>

On Thursday 19 September 2002 05:51 pm, sismex01 at hebmex.com wrote:

[snip: my code showing how an object can dynamically change type]

> > If you find a way to name this behavior "static", you'll be well
> > placed for the yearly Humpty Dumpty award.
	...
> Maybe you're being a bit rash, Alex?

Maybe, and maybe not.

[snip: sixmex01's code showing how not ALL dynamic type changes are allowed]

> The behaviour you expounded does not blindly work, it seems.

Of course, and, had I claimed Python is "BLINDLY dynamic", then I'd have
been "a bit rash".  As I cleverly omitted the damning adverb, I'm safe.

An object's type can be changed at runtime *as long as* this doesn't
alter the "layout" (__slots__ isn't the only example: you can't, for example,
dynamically change the type of a list to be a dictionary, etc, etc).

This still leaves Python vastly more dynamic than the great majority of
languages, where an object's type is determined, at the latest, at
object-creation time.  (SOME languages may indeed be even slightly
more dynamic: I think Ruby is, for example; but the fact that SOME
F1 drivers are even faster than Mr Barrichello -- one Mr Shumaker
comes to mind -- hardly means one is "being a bit rash" if one deems
Mr Barrichello a deucedly fast driver indeed).


Alex



From gmcm at hypernet.com  Fri Sep 13 08:00:55 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 13 Sep 2002 12:00:55 GMT
Subject: how can i make a executable program with a python2's script
References: <dd083388.0209060003.1188000c@posting.google.com> <3D7924BF.805@motorola.com> <mailman.1031351050.14744.python-list@python.org> <dd083388.0209111829.671b2a1b@posting.google.com>
Message-ID: <Xns9288519DACA30gmcmhypernetcom@199.171.54.215>

jgm2163 wrote:

[...]

> sorry,i want to make a release program on the computer that donot
> install python2.1 (because that python1.5 was installed on the redhat
> 7.2 defaultly).

Installer
 http://www.mcmillan-inc.com/installer_dnld.html
will do that, as long as the libc's are compatible.

-- Gordon
http://www.mcmillan-inc.com/


From danb_83 at yahoo.com  Tue Sep  3 03:03:14 2002
From: danb_83 at yahoo.com (Dan Bishop)
Date: 3 Sep 2002 00:03:14 -0700
Subject: Python to surpass C performance by 2030
References: <kuptw4o65q.fsf@lasipalatsi.fi>
Message-ID: <ad052e5c.0209022303.32319d0b@posting.google.com>

Erno Kuusela <erno-news at erno.iki.fi> wrote in message news:<kuptw4o65q.fsf at lasipalatsi.fi>...
> just for fun i ran pystone under python versions since 1.5.2 to
> current cvs. here are the results:
>
> [benchmark results snipped]
>
> since 1.5.2 is about 4 years old, the improvement per year is about
> 13.2% on average. if we assume that 1.5.2 is 50 times slower than
> C (probably conservative?), then at 13.2% per year since 1998 it
> will add up to 53x improvement in 32 years.
> 
> i wonder if this would be a good application for the time machine...
> 
> learn python today, and in 28 years you will have a job writing
> the speed-critical parts of c programs in python :)

Before that could ever happen, the Python interpreter would have to be
written in something other than C.


From eric.brunel at pragmadev.com  Thu Sep 26 08:45:37 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Thu, 26 Sep 2002 14:45:37 +0200
Subject: Multiple file open dialog
References: <3d92dce4$0$1286$cc9e4d1f@news.dial.pipex.com>
Message-ID: <amuv26$hdg$1@news-reader10.wanadoo.fr>

Nick B wrote:

> Hi,
> 
> Can anyone advise on how to get a list of multiple file names from a GUI
> dialog.  I am working exclusively in Win32 but so far only have GUI
> experience with Tkinter.  What I would like is access to the standard
> Win32 shift-click and control-click conventions.
> 
> Tkinter's askopenfilename can only return a single filename (?).

There should be a "multiple" option on askopenfilename that allows to 
select multiple files. However, it seems that's it's quite recent (didn't 
succeed in making it work with Tk 8.3.2). But maybe it works for you? Just 
do:
askopenfilename(..., multiple=1, ...)

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From max at alcyone.com  Wed Sep 25 02:34:28 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 24 Sep 2002 23:34:28 -0700
Subject: Strange python and cgi!
References: <amquos$rcu$1@lust.ihug.co.nz> <ajk2pukipmnmqtc662o36gnngiir552566@4ax.com>
Message-ID: <3D9158F4.F826F039@alcyone.com>

Tim Roberts wrote:

> If you are running Linux, check the log file at
> /var/log/httpd/error-log.

The problem is that the log location depends entirely on how Apache is
configured, which in turn depends upon the version, the distribution
(which often choses different locations for different log files), and
the local configuration.  It could be /var/log/error_log,
/var/log/errors, /var/log/apache/errors, etc.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Music washes away from the soul the dust of everyday life.
\__/ Berthold Auerbach
    7 Sisters Productions / http://www.7sisters.com/
 Web design for the future.


From pete at shinners.org  Thu Sep 12 11:19:30 2002
From: pete at shinners.org (Pete Shinners)
Date: Thu, 12 Sep 2002 15:19:30 GMT
Subject: "str.contains(part)" or alternatives?
References: <mailman.1031787694.27175.python-list@python.org>
Message-ID: <6g2g9.295995$aA.51455@sccrnsc02>

Delaney, Timothy wrote:
> In fact, this was the deciding factor in Guido changing his position on 'if
> s1 in s2'. A string has sequence-like behaviour (e.g. it is iterable,
> individual characters can be indexed), but also has its own behaviour (e.g.
> substring containment, all elements are single-character strings).

does this mean we can also now do this? if so, nifty

 >>> 1,2,3 in range(10)
1
 >>> 3,2,1 in range(10)
0




From raskpetersen at hotmail.com  Tue Sep 24 09:05:57 2002
From: raskpetersen at hotmail.com (Anders R)
Date: 24 Sep 2002 06:05:57 -0700
Subject: =?ISO-8859-1?Q?Decoding_url_with_country_specific_chars_(like_=E6=F8=E5)?=
Message-ID: <547475ee.0209240505.1c27bc9f@posting.google.com>

Hi,

i had a fustrating experience while doing some url decoding on a
unicode string that contains country specific chars:

The original string (unicode, so it might look strange here):
TestProduct?, opdateret 27/8 Pr?ver lige en umlaut: ?

The string in encoded form:
TestProduct%26%238482%3B%2C%20opdateret%2027/8%20Pr%C3%B8ver%20lige%20en%20umlaut%3A%20%26%23228%3B

the output i get when i decode it with urllib.unquote_plus(var):
TestProduct&#8482;, opdateret 27/8 Pr%C3%B8ver lige en umlaut: &#228;
in case special chars are escaped:
TestProduct&amp;#8482;, opdateret 27/8 Pr%C3%B8ver lige en umlaut:
&amp;#228;

So instead of '?' in the (danish) word 'Pr?ver' isn't decoded correct,
in fact it looks like Python (version 2.1) totally leaves the '%C3%B8'
alone...

Can anyone explain this behaviour?

It seems like unicode *is* supported fine, since the TM sign and the ?
(a with umlaut) is translated fine!

Maybe its only us danes getting discriminated ;-O

please help :-S

//Rask


From exarkun at meson.dyndns.org  Thu Sep 12 09:14:22 2002
From: exarkun at meson.dyndns.org (Jp Calderone)
Date: Thu, 12 Sep 2002 09:14:22 -0400
Subject: class variable won't icrement!
In-Reply-To: <7647A9F4B298E74DB6793865DA67285004ADF4@exchange.adrembi.com>
References: <7647A9F4B298E74DB6793865DA67285004ADF4@exchange.adrembi.com>
Message-ID: <20020912131422.GB29027@meson.dyndns.org>

On Thu, Sep 12, 2002 at 12:18:32PM +0200, Roman Yakovenko wrote:
> Hi it is simple just change 
> self.id = self.id + 1 
> 
> To 
> 
> test.id += 1
> 
> Roman
> 


  Actually, that turns out not to be the case.  += behaves differently
depending on its arguments - sometimes rebinding the name, sometimes
mutating in place.  For integers, it can not mutate in place (ints are
immutable, of course!) so "self.id = self.id + 1" is exactly the same as
"self.id += 1".  Were 'id' a mutable type, say a list, then the above
solution would work (assuming you changed 1 to something iterable).

 Jp

-- 
It is practically impossible to teach good programming style to
students that have had prior exposure to BASIC: as potential
programmers they are mentally mutilated beyond hope of
regeneration.        -- Dijkstra
--
 9:00am up 114 days, 9:53, 4 users, load average: 0.02, 0.01, 0.00
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 248 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020912/e970aba5/attachment.sig>

From charlotte at henkle.com  Tue Sep 17 18:23:09 2002
From: charlotte at henkle.com (Charlotte Henkle)
Date: 17 Sep 2002 15:23:09 -0700
Subject: boolean searching with keywords
Message-ID: <8f49f05a.0209171423.743c1383@posting.google.com>

Hello;

I'm trying to add some new functionality to a program I've written in
Python.  This is the first program I've ever written in Python, so
please forgive my newbieness.  I wouldn't be at all shocked if I
haven't picked the easiest solution to various problems.  ;)

This is how the program currently works.  The first thing the program
does when it starts is open a pop connection to see if there's any
mail in its mailbox.  If there's mail, it pulls down the first
message.  It parses the message looking for URL, pulls them out and
appends them to a URL list.  One it has a URL list, it begins to
follow the URLs so that it can pull down the HTML.

The program also has a list of key words.  The keywords text file is a
CR deliminated list of words to look for in the HTML.  If the program
finds ANY of the keywords, it sends off an email with a copy of the
HTML it's found.  Here are the important functions:

def generateKeywords():
        f = open('Keywords_Business.txt', 'r')
        keys = map(lambda s: s[:-1],f.readlines())
	return keys

def followURLs():
	for eachLine in urlList:
		print("Following this URL:" +eachLine)
		data=urllib.urlopen(eachLine)
		myData=data.read()
		text=stripper(myData)  # stripper strips out javascript, Etc
		if checkData(text):
			print('FOUND ONE!')
		        htmldata=urllib.urlopen(eachLine)
def checkData(d):
	regexp = r'\b(' + '|'.join(keys)+ r')\b'
	newsearch = re.compile(regexp)
	found = newsearch.search(d)
	if found:
		global subject 
		subject = found.group()
		return 1
	else: return 0


So far, so good (although if you can see improvements in this, please
let me know).

This works fine for single words on a line in a file.  However, now I
want to add a Boolean search capability.  IE, instead of searching for
"Frogs" or "Green" or "Lizards", I'd like to be able to search for
"frogs NOT green" or  "Lizards AND green" or "green AND scaly".  As
before, if any of the phrases are found, I want to break out of the
search and send mail.

I'm not sure how to do this.  I considered making a tree structure,
and then walking the tree, but I was unable to get it working
correctly.  Additionally, I was unsure if this was making the problem
too complex:  Is there an easier way to get the functionality I want? 
I would appreciate any help.

Thank you!


From peter at engcorp.com  Thu Sep  5 08:37:47 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 05 Sep 2002 08:37:47 -0400
Subject: Matching a string up to a word
References: <slrnanbkbc.huj.dag@omally.animagicnet.no> <3d75ffaa$1@news.sentex.net> <slrnanbti7.huj.dag@omally.animagicnet.no> <3d76c1f7$1@news.sentex.net> <al6nt0$vd$0@216.39.172.122>
Message-ID: <3d775000$1@news.sentex.net>

Bengt Richter wrote:
> On Wed, 04 Sep 2002 22:31:46 -0400, Peter Hansen <peter at engcorp.com> wrote:
> 
>>I was thinking of something like that.  This should do it then:
>>
>>  filename = ' '.join(field.split(' ')[1:-4])
>>
>>That is, provided you never expect to see more than one consecutive
>>space in your filenames...
> 
> Why should that hurt, if the head and tail items are consistently separated
> by single spaces?
> 
>  >>> line = '12345 /some  weird/path with multiple-    -spaces Wed Sep 4 2002.'
>  >>> line.split(' ')
>  ['12345', '/some', '', 'weird/path', 'with', 'multiple-', '', '', '', '-spaces', 'Wed', 'Sep', '4', '2002.']
>  >>> line.split(' ')[1:-4]
>  ['/some', '', 'weird/path', 'with', 'multiple-', '', '', '', '-spaces']
>  >>> ' '.join(line.split(' ')[1:-4])
>  '/some  weird/path with multiple-    -spaces'

Oops!  I was going by the behaviour with just .split(), which I tried 
originally (but which does other whitespace too), and didn't realize the 
explicit .split(' ') would do what it does.  Cool. :)

-Peter



From dsavitsk at e-coli.net  Fri Sep 13 10:36:24 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Fri, 13 Sep 2002 14:36:24 GMT
Subject: Newbie Looking for direction
References: <uo23o95eb8tg87@corp.supernews.com> <0Akg9.3474$WV.3470@fe03>
Message-ID: <IJmg9.8230$yt3.4316130@newssrv26.news.prodigy.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:0Akg9.3474$WV.3470 at fe03...
> "mowestusa" <justnotworking at yahoo.com> wrote ...
> > I have never programmed in a language before.
> > I have developed complex databases with Access (Office 95 version) just
> with
> > the visual tools.
>
>
> Languages are the only way. Graphical manipulations such as report
> generation in Access are all very well, but you hit the limits sooner or
> later.
>
> > I would like to learn a computer language or database as a hobby that I
> > would use to keep track of things in my real job.
> > I have worked through "Non-programmer's Tutorial for Python".
> > I'm reading "Programming Python (First Edition, got cheap)" right now.
> >
> "Programming Python" is a good book, but its focus is perhaps a little
> narrow. There are many online tutorials you can add to your book with, so
> you shouldn't *need* to buy anything else.
>
> > Most of the programs that I would like to develop would be relatively
> simple
> > databases.
>
> Nothing wrong with that. You can create them visually if that's easier.
You
> just need to learn how to use DB-API modules. For Access I'd recommend
> mxODBC, which doesn't have to be licensed for personal use.
>

* * *

> Steve Holden

also note that Access is easy to communicate with/control w/o installing 3rd
party tools.  You can easily use ADO, DAO, etc with python.  see
http://www.e-coli.net/pyado.html for help/examples. I will leave the debate
over the merits of one method as opposed to another for elsewhere.

-doug





From martin at v.loewis.de  Tue Sep 24 01:48:40 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 24 Sep 2002 07:48:40 +0200
Subject: PyXML Building on OS X?
References: <B9B3CED8.3B9%jwold@min.midco.net> <j4k7ldghtp.fsf@informatik.hu-berlin.de> <B9B55A89.6D8%jwold@min.midco.net>
Message-ID: <m3r8fk545j.fsf@mira.informatik.hu-berlin.de>

Jarrett Wold <jwold at min.midco.net> writes:

> extensions/expat/lib/xmlparse.c:3056: warning: `regparm' attribute directive
> ignored

I see. Please edit internal.h to replace

#if defined(__GNUC__)
#define FASTCALL __attribute__((stdcall, regparm(3)))
#elif defined(WIN32)
#define FASTCALL __fastcall
#else
#define FASTCALL
#endif

with

#define FASTCALL

> cc -L/sw/lib -bundle -flat_namespace -undefined suppress
> build/temp.darwin-5.5-Power Macintosh-2.2/pyexpat.o
> build/temp.darwin-5.5-Power Macintosh-2.2/xmlparse.o
> build/temp.darwin-5.5-Power Macintosh-2.2/xmlrole.o
> build/temp.darwin-5.5-Power Macintosh-2.2/xmltok.o -o
> build/lib.darwin-5.5-Power Macintosh-2.2/_xmlplus/parsers/pyexpat.so
> -flat_namespace
> /usr/bin/ld: -undefined: unknown argument: -lbundle1.o
> error: command 'cc' failed with exit status 1

Very strange. Can you please invoke

cc -L/sw/lib -bundle -flat_namespace -undefined suppress
build/temp.darwin-5.5-Power Macintosh-2.2/pyexpat.o
build/temp.darwin-5.5-Power Macintosh-2.2/xmlparse.o
build/temp.darwin-5.5-Power Macintosh-2.2/xmlrole.o
build/temp.darwin-5.5-Power Macintosh-2.2/xmltok.o -o
build/lib.darwin-5.5-Power Macintosh-2.2/_xmlplus/parsers/pyexpat.so -v

on the command line, and report the output?

Regards,
Martin



From dsavitsk at e-coli.net  Mon Sep 23 13:11:05 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Mon, 23 Sep 2002 17:11:05 GMT
Subject: Session handling
References: <c5b3549a.0209230511.5bbe6849@posting.google.com> <amn7dp$6cn$1@mamenchi.zrz.TU-Berlin.DE>
Message-ID: <IWHj9.264$841.144712@newssrv26.news.prodigy.com>

"Andreas Ulbrich" <andreas.ulbrich at gmx.net> wrote in message
news:amn7dp$6cn$1 at mamenchi.zrz.TU-Berlin.DE...
> beachboy wrote:
> > Hi!
> > Please can anyone help me.
> > My problem:
> > I'm trying to program a web application with python. This tool has to
> > provide diffrent languages. The user has to choose the language at the
> > beginning. So i have to save the information which language the user
> > chooses. I think this is a perfect example for using sessions.
> > The only problem is I've got no idea how to realise this in python.
> > Please help me as soon as you can because I have to solve this problem
> > very quickly.
> >
> > hoping to hear from you soon
> > beachboy
>
> Whenever I need to write a web-application I do use jython and write
> servlets. This combines the power of Servlet/JSP with the worlds best
> programming language.
>
> a jython servlet would look like this:
>
> from javax.servlet.http import HttpServlet
>
> class MyServlet(HttpServlet) :
>
> def doGet(self, request, response) :
>
> session = request.getSession(1)
> lang = session.getAttribute('language')
> if not lang :
> # initialise the session (i.e. choose language)
>
> # produce output suitable to lang
> if lang == 'DE' :
> print >> res.writer, "Hallo Welt!"
> else :
> print >> res.writer, "Hello World!"
>
> There are other solutions. I believe that Zope offers session-management
> too.

If the only state that needs to be maintained is a language variable, this
seems like a lot of work when a cookie will work fine.

-d





From aleax at aleax.it  Tue Sep 24 09:22:02 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 24 Sep 2002 13:22:02 GMT
Subject: list problem
References: <mailman.1032818129.8126.python-list@python.org> <Xns929364D72B214duncanrcpcouk@127.0.0.1> <tCYj9.157570$pX1.5675986@news2.tin.it> <Xns92938CC236BA2duncanrcpcouk@127.0.0.1>
Message-ID: <_FZj9.158106$pX1.5692316@news2.tin.it>

Duncan Booth wrote:

> Alex Martelli <aleax at aleax.it> wrote in
> news:tCYj9.157570$pX1.5675986 at news2.tin.it:
> 
>>> If you are going to use reduce, then you would be better off writing:
>>> 
>>>    import operator
>>>    reduce(operator.add, x)
>> 
>> Maybe (marginally).  But timing things is often a good idea:
> 
> Apologies for not being clearer. I actually felt that using operator.add
> read more cleanly than using a lambda. I should have made it more obvious
> that was what I meant. I wouldn't particularly worry which is faster
> unless one was actually proven to be a bottleneck, at which point the OP
> might well find the best way to optimise would be to not end up with
> nested lists in the first place.

My apologies in return -- you're perfectly right in your overall
attitude, and I even agree with you that lambda isn't clearest.
I don't know why I focused on performance in answering this
thread -- must have been on my mind for other reasons.

I _do_ think that reduce(operator.__add__, ...) is the only
really frequent use of reduce, with such dominance that it might
even be worth having as a built-in -- clearest and fastest.  It
might even specialcase _strings_ in particular, turning the op
into a ''.join in this case... just musing...

But I did like the nested list comprehension better in this
case as an issue of style, anyway.  You DID write "If you are
going to use reduce", being impeccable on this point too.  So,
sorry again for appearing to pick on your post for no good reason.


Alex



From carner at tscnet.com  Sat Sep 14 15:12:15 2002
From: carner at tscnet.com (Richard Kessler)
Date: Sat, 14 Sep 2002 12:12:15 -0700
Subject: Converting a string to time - simple task but don't know how to do it!
Message-ID: <7A6D2E53461F2BF2.AC83A3D5E1A01468.48CC780BD6530537@lp.airnews.net>

I have a string returned from a database expressing the time i.e. 2002-09-14
10:30:34.  I need to compare this value to the current time to determine the
amount of time elapsed between the two. I get the current time using
time.time() but I cannot find a way to convert the string from the database
into a time value (seconds from epoch) to do the simple subraction to find
seconds elapsed.

Help please...from anyone knowing how to do this simple task.

Thanks very much in advance.

Richard




From pknotz at sandia.gov  Thu Sep 26 13:56:45 2002
From: pknotz at sandia.gov (Pat Notz)
Date: 26 Sep 2002 10:56:45 -0700
Subject: How do you pronounce "__" (double underscore)?
Message-ID: <525eaba.0209260956.72bf6786@posting.google.com>

@ = "at"
! = "bang"
. = "dot"
* = "star"

Is there a short (one or two syllable) pronounciation for "__" (double
underscore).  "Underscore underscore" and "double underscore" are kind
of long.  Somthing like "splat"?  Anyone?

Pat


From mikemccand at bigfoot.com  Tue Sep 24 17:05:33 2002
From: mikemccand at bigfoot.com (Mike McCandless)
Date: 24 Sep 2002 14:05:33 -0700
Subject: how to check if a function signature "matches" without calling the function?
Message-ID: <a76c1561.0209241305.454f7ae6@posting.google.com>

Does Python 2.2.1 make it possible for me to see some details about
the expected arguments that a function takes, without actually calling
the function?

Specifically, if I have a function or method object, and I have a
tuple + dict of arguments I'd like to apply it on, is there a function
I can call that would check if the tuple + dict can be properly
"coerced" to the formal parameters?

Browsing through the Python source it looks like this capability is
not cleanly factored out of the "ext_do_call" function in
Python/ceval.c?

Thanks!!

Mike


From gerhard.haering at opus-gmbh.net  Thu Sep  5 03:17:41 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 5 Sep 2002 07:17:41 GMT
Subject: newbie-question: Nested functions and variable scope
References: <7a7157fd.0209041503.7b9a0f44@posting.google.com>
Message-ID: <slrnane15f.1v8.gerhard.haering@haering.opus-gmbh.net>

Leonardo Boiko wrote:
> Hi, I'm a Python newbie.

... and using something older than Python 2.2, it seems.

> I'm writing a program that uses expat. Some code looks like this:
> 
> def generate_main(filename="main.xml", (...)):
>     (...)
>     outfile = open (filename, "w")
>     in_reply = 0 # true if we are inside "reply" element
> 
>     def start_element(name, attrs):
>         if name == "reply":
>             in_reply = 1
> [...]

> UnboundLocalError: local variable 'in_reply' referenced before
> assignment

- Python < 2.1 doesn't have nested scopes; you'd have to play tricks with
  lambda IIRC; but then, I consider Python < 2.1 obsolete
- Python 2.1 has nested scopes, if the first import statement in a module
  is "from future import nested_scopes".
- Python 2.2 and later always use nested scopes

> If I put "in_reply" outside everything and use "global in_reply"
> inside each nested function, it works, but that's evil! What would be
> a better way of doing it?

Just put a "from __future__ import nested_scopes" at the top of your module
and it should just work.

-- Gerhard


From sross at connectmail.carleton.ca  Wed Sep 18 21:37:55 2002
From: sross at connectmail.carleton.ca (Sean Ross)
Date: 19 Sep 2002 01:37:55 GMT
Subject: Splitting a list of strings
References: <mailman.1032309381.3626.python-list@python.org> <w0Uh9.116763$5r1.4544705@bin5.nnrp.aus1.giganews.com> <am9tpv$9gp$1@driftwood.ccs.carleton.ca> <mailman.1032359962.30186.python-list@python.org> <o81i9.127490$pX1.4558733@news2.tin.it>
Message-ID: <amb9pj$hme$1@driftwood.ccs.carleton.ca>

Okay, then. For those still interested, here's several implementations of
sieving, based on peoples suggestions, plus a sample time trial. The trial
is on a file with over 8000 lines to process.

class ARFF:
    "Holds the lines of a *.arff file"
    def __init__(self, filename):
        file = open(filename)
        self.__lines = file.readlines()
        file.close()

    # used by sieve2()
    def __classify(self, char):
        "Returns classification index for attributes, comments, and data"
        if char == "@": return "attr"
        elif char == "%": return "comm"
        else: return "data"

    # based on suggestion by Emile van Sebille
    def sieve(self):
        "Sorts lines into data, attributes, or comments"
        results = [[],[],[]]  #[data, attributes, comments]
        for line in self.__lines:
            # classify line as attribute, comment, or data by testing first
char
            results["@%".find(line[0]) + 1].append(line)
        # discard first and last attribute
        results[1] = results[1][1:-1]
        return results[:-1]  # discard comments

    # based on suggestion by Alex Martelli
    def sieve2(self):
        "Sorts lines into data, attributes, or comments"
        results = {}
        for line in self.__lines:
            # classify line as attribute, comment,  or data by testing first
char
            results.setdefault(self.__classify(line[0]), []).append(line)
        # discard first and last attribute
        results["attr"] = results["attr"][1:-1]
        del results["comm"]  # discard comments
        return results

    # my original solution augmented with suggestion by Anthony Tuininga
    def sieve3(self):
        "Sorts lines into data, attributes, and ignores comments"
        attributes = [line for line in self.__lines if line[0] == "@"]
        data = [line for line in self.__lines if not line[0] in ("%", "@") ]
        return data, attributes[1:-1]

    # using filters and lambda
    def sieve4(self):
        "Sorts lines into data, attributes, and ignores comments"
        attributes = filter(lambda line: line[0]=="@", self.__lines)
        data = filter(lambda line:  not line[0] in ("%", "@"), self.__lines)
        return data, attributes[1:-1]

    # Emile's idea again, with a list comprehension
    def sieve5(self):
        "Sorts lines into data, attributes, or comments"
        results = [[],[],[]]  #[data, attributes, comments]
        [results["@%".find(line[0]) + 1].append(line) for line in
self.__lines]
        # discard first and last attribute
        results[1] = results[1][1:-1]
        return results[:-1]  # discard comments

    # based on Mark McEachern's suggestion
    def sieve6(self):
        attribute = "@"
        comment = "%"
        attributes = data = []
        for line in self.__lines:
            line = line.strip()
            if not line or line.startswith(comment):
                continue
            elif line.startswith(attribute):
                attributes.append(line)
            else:
                data.append(line)
        return data, attributes[1:-1]
if __name__ == "__main__":
    import time
    arff = ARFF('./mushroom.arff')

    t0=time.time()
    data, attr = arff.sieve()
    t = time.time()-t0
    print "sieve time: %r seconds" % (t)

    t0=time.time()
    result = arff.sieve2()
    t = time.time()-t0
    print "sieve2 time: %r seconds" % (t)

    t0=time.time()
    data2, attr2 = arff.sieve3()
    t = time.time()-t0
    print "sieve3 time: %r seconds" % (t)

    t0=time.time()
    data3, attr3 = arff.sieve4()
    t = time.time()-t0
    print "sieve4 time: %r seconds" % (t)

    t0=time.time()
    data4, attr4 = arff.sieve5()
    t = time.time()-t0
    print "sieve5 time: %r seconds" % (t)

    t0=time.time()
    data5, attr5 = arff.sieve6()
    t = time.time()-t0
    print "sieve6 time: %r seconds" % (t)

====================================================
Trial Time:
sieve time: 0.05000007152557373 seconds
sieve2 time: 0.12999999523162842 seconds
sieve3 time: 0.069999933242797852 seconds
sieve4 time: 0.081000089645385742 seconds
sieve5 time: 0.069999933242797852 seconds
sieve6 time: 0.20000004768371582 seconds


In case you're curious, I'm probably just going to use sieve3().
It's not much slower than sieve(), and it's very easy to understand what's
happening without a lot of commenting.

OK, so, there you are. Thanks to everyone for your time and suggestions,
Sean Ross




From gerhard.haering at gmx.de  Tue Sep 10 12:46:30 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 10 Sep 2002 16:46:30 GMT
Subject: Python iCalendar API?
References: <mailman.1031673134.15031.python-list@python.org>
Message-ID: <slrnans8f6.jc.gerhard.haering@lilith.my-fqdn.de>

Erik Price wrote in comp.lang.python:
> Is there a Python module that provides capability of processing files 
> in .ics (iCalendar) format?  I found nothing concrete on Google or at 
> python.org ...

Looks like there is:
http://webcvs.kde.org/cgi-bin/cvsweb.cgi/kdepim/libical/src/python/

I don't know how easy it is to rip that out of kdepim.

-- Gerhard


From skip at pobox.com  Mon Sep  9 17:29:28 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 9 Sep 2002 16:29:28 -0500
Subject: how to build wxPython w/ Sun's C++ compiler?
Message-ID: <15741.4792.752356.107952@12-248-11-90.client.attbi.com>

wxPython uses distutils to build.  I'm trying to build it on a Solaris
system using Sun's C and C++ compilers (cc & CC, respectively).  When I
execute

    python setup.py build

it tries to use "cc" instead of CC to compile C++ code, with predictable
results:

    running build_ext
    building 'wxc' extension
    cc -DNDEBUG -O -DSWIG_GLOBAL -DHAVE_CONFIG_H -DWXP_USE_THREAD=1 -Isrc -I/home/skip/local/SunOS/include/python2.2 -c src/helpers.cpp -o build/temp.solaris-2.8-sun4u-2.2/src/helpers.o -I/home/skip/local/SunOS/lib/wx/include/gtk-2.3 -I/home/skip/local/SunOS/include -DGTK_NO_CHECK_CASTS -D__WXGTK__ -D_REENTRANT -I/home/skip/local/SunOS/include/gtk-1.2 -I/home/skip/local/SunOS/include/glib-1.2 -I/home/skip/local/SunOS/lib/glib/include -I/usr/openwin/include
    cc: No input file specified, no output generated
    error: command 'cc' failed with exit status 1

How do I convince it to use CC for files with .cpp extensions.

Thx,

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From see_reply_address at something.invalid  Thu Sep 19 22:00:38 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Fri, 20 Sep 2002 14:00:38 +1200
Subject: Tkinter: scrolling the canvas items AND widgets
References: <3d06fae9.0209191255.63def2eb@posting.google.com>
Message-ID: <3D8A8146.3030506@something.invalid>

jmdeschamps wrote:

> This code lets me scroll the graphical items but the widgets stay put
> using PythonWin 2.2.1 on Windows 2000

 >

> for i in range(1):
>     f=Entry(canvas)
>     f.place(x=10,y=10)


I think you need to create a window item on the
canvas, with the window attached to it, something
like this :

       f = Entry(canvas)
       canvas.create_window(10, 10, window = f)

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From ws at mystrobl.de  Sat Sep 21 03:30:27 2002
From: ws at mystrobl.de (Wolfgang Strobl)
Date: Sat, 21 Sep 2002 09:30:27 +0200
Subject: Nokia 3310 SMS Application, done:connecting to serial port(linux/win), done:reading some hex like \x00 . . ., problem:write/send command! noresponse
References: <52facf1d.0209190533.3048e62c@posting.google.com>
Message-ID: <r77ooukb98cbfbpcqbitl9g53vuokkl79s@4ax.com>

ptbjrii at hotmail.com (Polerio Babao Jr.II):

>I tried almost three weeks writing or sending a command to the
>cell-phone but this gives me no response. I'm talking here about the
>nokia 3310 connected to my serial port. The command that I send is the
>AT command. From my pyserial I made an initialization like this
>ser=serial.Serial('/dev/ttyS1',19200) 
>ser.readline() 
>ser.write()
>ser.read()
>
>When I connect to the port for the first time I get the ser.read()
>output like this \x00\x0c I think. But that's two response is what I
>get on my first connect. After that when I envoke ser.read() again, I 
>get nothing.

Something like 

import serial
ser=serial.Serial(0,19200,timeout=3)
print ser.portstr
ser.write("ATZ\r")
print "got"
line=ser.read(10)
print line
ser.close()

worked on the first try on my w2k box, when connected to a Siemes S25
via COM1. I bet it will would with the Nokia, too, because "ATZ" is
almost universally understood. :-)

My wife had asked me to dump all her phonebook numbers from her handy
for backup purposes. Siemens only sold an somewhat expensive program
for the S25, which I didn't buy. A program that I've got with a newer
S45 unfortunately didn't even recognize the S25. 

The following qick&dirty snippet did the trick, and it works with both
the s25 and the s45.

# Strobl 19.9.2002  $Id: s25.py,v 1.3 2002/09/21 07:16:12 cvsroot Exp $
# read phonebooks from Siemens S25 or S45

import serial,re

def readuntilok(s):
    ol=[]
    while 1:
        c=s.read()
        if not c:
            break
        ol.append(c)
        ostring="".join(ol)
        if len(ol)>3 and ostring[-4:]=="OK\r\n":
            break
    return ostring

def cmd(s,cmd):
    s.write(cmd+"\r")
    r=readuntilok(s)
    r=r.split("\n")
    for i in range(len(r)):
        r[i]=r[i][:-1]
    return r

def showbook(ser,book):
    select= cmd(ser,'AT+CPBS='+book)
    if select[1]=="ERROR":
        print "Error in showbook"
        print
        return
    booksize=cmd(ser,"AT+CPBR=?")[1]
    print booksize
    siz=re.compile(r"\+CPBR: \((\d+)-(\d+)\)")
    von,bis= siz.match(booksize).groups()
    von=int(von)
    bis=int(bis)
    
    print "Phonebook",book
    for i in range(von,bis+1):
        buch=cmd(ser,"AT+CPBR=%d,%d"%(i,i))
        if buch[1]!="OK": print repr(buch[1])
    print

ser=serial.Serial(0,19200,timeout=3)
print ser.portstr
ser.write("ATZ\r")
print "sent"
line=ser.read(10)
print line

typ=cmd(ser,"AT+CGMI")[1]
print typ,
typ=cmd(ser,"AT+CGMM")[1]
print typ,
typ=cmd(ser,"AT+CGMR")[1]
print "Model",typ

print "IMSI",cmd(ser,"AT+CIMI")[1]
print "IMEI",cmd(ser,"AT+CGSN")[1]

uhrzeit=cmd(ser,"AT+CCLK?")
print "Uhrzeit",uhrzeit[1]

booksstring= cmd(ser,"AT+CPBS=?")[1]
print "Phonebooks",booksstring

for b in re.compile(r'"\w\w"').findall(booksstring):
    print "----Now reading book",b
    showbook(ser,b[1:-1])

print cmd(ser,"ATZ")
print  cmd(ser,'AT+CPBS=SM')
print "writable",cmd(ser,"AT+CPBW=?")

print "Done"
ser.close()


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

>pythonw -u s25.py
COM1
sent
ATZ

OK

SIEMENS S45 Model 23
IMSI xxxxxxxxxxxxxxxxxx
IMEI xxxxxxxxxxxxxxxxxx
Uhrzeit +CCLK: "02/09/21,09:11:57"
Phonebooks +CPBS: ("FD","SM","ON","ME","LD","MC","RC")
----Now reading book "FD"
+CPBR: (1-10),20,16
Phonebook FD

----Now reading book "SM"
+CPBR: (1-125),20,16
Phonebook SM
'+CPBR: 1,"xxxxxxxxxxxxxx",129,"xxxxx!"'
'+CPBR: 2,"xxxxxxxxxxxxxxxx",145,"xxxxx"'

etc.

['ATZ\r', 'OK', '']
['AT+CPBS=SM\r', 'OK', '']
writable ['AT+CPBW=?\r', '+CPBW: (1-125),20,(128-255),16', '', 'OK',
'']
Done

>Exit code: 0

-- 
Wir danken f?r die Beachtung aller Sicherheitsbestimmungen


From rjones at ekit-inc.com  Sun Sep  1 07:51:22 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Sun, 1 Sep 2002 21:51:22 +1000
Subject: ANN: New PEP Format: reStructuredText
In-Reply-To: <200209012145.51274.rjones@ekit-inc.com>
References: <mailman.1030681307.29889.python-list@python.org> <X5K05iAtPdc9Ewdt@jessikat.fsnet.co.uk> <200209012145.51274.rjones@ekit-inc.com>
Message-ID: <200209012151.22766.rjones@ekit-inc.com>

On Sun, 1 Sep 2002 9:45 pm, Richard Jones wrote:
> On Sun, 1 Sep 2002 6:46 pm, Robin Becker wrote:
> > In article <mailman.1030831309.22494.python-list at python.org>, Richard
> > Jones <rjones at ekit-inc.com> writes
> >
> > >This is done in the docutils sandbox under the richard/pythonpoint
> > > directory. I had this working ages ago (back in March, I think) but
> > > hadn't updated it when David merged the dps and restructuredtext
> > > projects to form the docutils project. That's done now.
> > >
> > >
> > >    Richard
> >
> > So does this appear as a CGI somewhere or should I go ahead and make a
> > working demo?
>
> No, there's no CGI. The pythonpoint script in the sandbox is a "working
> demo". Pass it a rest file as a command-line arg, and it'll spit out
> pythonpoint XML (read the example for limitations though).

ps. yes, it should be possible to merge the processing into one tool that 
spits out PDF, but I simply didn't get that far back in March, and "make" did 
a pretty good job... I had a presentation deadline to meet, and 
rest/pythonpoint were a huge help in getting there. No polish though :)

pps. not so sure about overcoming the limitations of the PDF layout - though 
again I really didn't have enough time when I wrote it to investigate


    Richard




From skip at pobox.com  Tue Sep  3 23:02:21 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 3 Sep 2002 22:02:21 -0500
Subject: Coredumps (Python 1.4)
In-Reply-To: <3D7579F1.1090500@nospam.yahoo.com>
References: <3D7579F1.1090500@nospam.yahoo.com>
Message-ID: <15733.30653.991781.186828@12-248-11-90.client.attbi.com>

    Anton> ... I coredump on this line ever since I linked with Electric
    Anton> Fence memory debugging lib. Any ideas why? 

Not off the top of my head...

    Anton> This is with Python 1.4 on SCO. Any know issues with memory
    Anton> allocation there? I'm building 2.2 now, but it will take a while.

Give up on 1.4.  It took me awhile (an hour or so) several months ago to get
Python 2.1 (I think) to build on SCO.  You'll be better off solving 2.2
build problems than struggling with what might be a bug in 1.4.

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From henrik.motakef at web.de  Mon Sep 23 15:33:21 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 23 Sep 2002 21:33:21 +0200
Subject: Dynamic loading of modules
References: <PrJj9.12293$Uu.3503@news1.bredband.com>
Message-ID: <87sn00zeke.fsf@pokey.henrik-motakef.de>

"Andreas Lydersen" <al at metroic.no> writes:

> Hi,
> I'd like to load a set of modules from a directory.  I have tried to use the
> __import__(...) statement to do so, but with no success.
> ?
> The static approch works (of course), but the dynamic does not!
> 
> static:
> import plugins.Password
> 
> dynamic atttempt:
> __import__("plugins.Password")

__import__ doesn't really import the module, it returns a module
object. So do something like

>>> pwd = __import__("plugins.Password")
>>> pwd.spam()
Spam!

hth
Henrik


From tdelaney at avaya.com  Wed Sep 25 20:09:52 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 26 Sep 2002 10:09:52 +1000
Subject: a python puzzle
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5DA@natasha.auslabs.avaya.com>

> From: Erik Max Francis [mailto:max at alcyone.com]
> 
> Michal Wallace wrote:
> 
> > Here's a python puzzle. Solve it and you win a prize. :)
> 
> tr hokdrafwfezxtculyjqmbipsvHIFg importsysngadfzjehulvcwbxICSk
> 
> I started by decoding the (obvious) Python source, which made the job
> considerably easier.

Trivially easy in fact - since it's just a substitution cypher, you've
already got a lot of information just from the first line of python code -
the first line obviously starts as

    import sys

and

    cdr kmxhe, irwka he uhk(farhez.myaayrf, ihkjyr):
    ^^^              ^^

is trivially

    for kmxhe, irwka in uhk(farhez.myaayrf, ihkjyr):
    ^^^              ^^

which gives you a few more letters. In fact, by that stage you already have
12 out of 26 characters.

I suspect it took Holger somewhat less than half an hour to solve ... 5
minutes at most, and most of that is writing the decryption code ;)

Tim Delaney



From tim.one at comcast.net  Mon Sep 16 18:29:32 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 16 Sep 2002 18:29:32 -0400
Subject: Difflib question
In-Reply-To: <886c5e4b.0209161208.25012958@posting.google.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEINBEAB.tim.one@comcast.net>

[Robin Siebler]
> Is there a way to get difflib to output *only* the lines that are
> different (instead of all of the lines)?

Read the docs for Sequence_Matcher.get_opcodes(), and ignore 'equal' tuples.



From aleax at aleax.it  Thu Sep 26 10:00:50 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 26 Sep 2002 14:00:50 GMT
Subject: List/Tuple bug or feature?
References: <mailman.1032905792.2164.python-list@python.org> <m15rma.pt4.ln@ix.netcom.com>
Message-ID: <mqEk9.170942$pX1.6084102@news2.tin.it>

Dennis Lee Bieber wrote:

> sismex01 at hebmex.com fed this fish to the penguins on Tuesday 24
> September 2002 03:12 pm:
> 
>> It's one of python's features :-)
>> 
>> For when you wanna write a long string, SQL for example:
>> 
>> Query = "SELECT what, when, where, why "
>>         "FROM t1, t2, t2 "
>>         "WHERE t1.id = t2.id_t1 and t3.id_t2 = t2.id and t1.id_t2 =
>>         t3.id " "AND t1.answer = 'OK'"

(Needs parenthese around it, or \ at line ends, of course).

>         What's wrong with...
> 
> query = """select what, when, where, why
>         from t1, t2, t3
>         where t1.id = t2.id_t1
>                 and t3_id2 = t2.id ...etc """

You can't intersperse comments in this way -- not a problem when
the string is simple enough, but some SQL queries can be pretty
complicated AND not all SQL dialects let you insert comments in
the query itself.


Alex



From lozinski at openstepnews.com  Mon Sep 16 18:41:14 2002
From: lozinski at openstepnews.com (lozinski at openstepnews.com)
Date: Mon, 16 Sep 2002 15:41:14 -0700 (PDT)
Subject: Python.jobmart.com Server Upgrade
In-Reply-To: <20020916221626.13483.93466.Mailman@mail.python.org>
Message-ID: <1020916154114.212AAC/w.lozinski@maya>

The python jobmart now allows for full text searches of the embedded resumes.
So hiring managers can find just the right person with whatever specialty skill
is required.

Your feedback is appreciated.

Regards
Chris

1-510-740-6486
lozinski at openstepnews.com
lozinski at jobmart.com



From kowk at earthlink.net  Wed Sep 25 13:47:03 2002
From: kowk at earthlink.net (Kow K)
Date: Wed, 25 Sep 2002 10:47:03 -0700
Subject: how to change font and its size in IDLE (in UNIX)?
In-Reply-To: <200209250852.27535.mfranklin1@gatwick.westerngeco.slb.com>
Message-ID: <CCF2BE1D-D0AE-11D6-862E-00050287C0A6@earthlink.net>

Hi Martin, your pointer was quite useful. Thanks -- Kow

On Wednesday, September 25, 2002, at 01:52 AM, Martin Franklin wrote:

> On Wednesday 25 Sep 2002 5:22 am, Kow K wrote:
>> Hi all,
>>
>> I'm trying to figure out how to change the font and its size in IDLE.
>> It's Courier in 10pt. I don't like Courier, and what's worse,
>> 10pt is too small for me. I wish I could change it to Fixed 7x13.
>>
>> Any input will be appreciated.
>>
>> Kow
>
> Take a look here (from http://groups.google.com/advanced_group_search)
>
>
> http://groups.google.com/groups?q=font+size+idle&hl=en&lr=&ie=UTF-8&scoring=
> r&selm=3a942c61_4%40corp.newsfeeds.com&rnum=1
>
>
> Cheers
> Martin
>
> --
> ### Python Powered Signature
> I started writting this email on
> Wed Sep 25 08:51:33 2002
>




From joec at mill.co.uk  Mon Sep  2 12:09:50 2002
From: joec at mill.co.uk (Joe Connellan)
Date: Mon, 02 Sep 2002 17:09:50 +0100
Subject: compiling extensions on linux
Message-ID: <3D738D4E.35B6C712@mill.co.uk>

I'm trying to port an extension of mine from windows to linux and am not
sure how to include the myExtension.def file in the compile - I'm using
g++

eg if I use

g++ -shared -o myextension.so

it compiles fine but I get the following error when importing it into
python

>>> import myextension
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ImportError: dynamic module does not define init function
(initmyextension)
>>>

Thanks

Joe





From cnetzer at mail.arc.nasa.gov  Wed Sep 25 19:58:01 2002
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Wed, 25 Sep 2002 16:58:01 -0700
Subject: a python puzzle
In-Reply-To: <amtgmd$c00$1@usenet.Stanford.EDU>
References: <Pine.LNX.4.44.0209251736470.16806-100000@hydrogen.sabren.com> <mailman.1032992492.8970.python-list@python.org> <amtgmd$c00$1@usenet.Stanford.EDU>
Message-ID: <200209252358.QAA25171@mail.arc.nasa.gov>

On Wednesday 25 September 2002 16:26, Andy Salnikov wrote:
> "Gary Herron" <gherron at islandtraining.com> wrote in message

> > The fact that this puzzle was easy to decrypt says something (good)
> > about Python: 
>
>   C'mon, "the fact that this puzzle was easy to decrypt says" that
> "encription" was really trivial. 

Yes, it was meant to be easy, because it is basically a slightly 
disguised SPAM message.  There was nothing "pythonic" about this puzzle 
at all, except a rather routine piece of code.  If I has bothered to 
'decrypt' it, I would be pissed.  (Reminds me of little Ralphie, in "A 
Christmas Story", waiting weeks to get his secret decoder wheel, only 
to find the secret message is 'Be sure to drink your Ovaltine')

-- 

Chad Netzer
cnetzer at mail.arc.nasa.gov



From martin at v.loewis.de  Thu Sep 19 02:45:26 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 19 Sep 2002 08:45:26 +0200
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <mailman.1032329366.6813.python-list@python.org> <j4ptvbaapm.fsf@informatik.hu-berlin.de> <3D890FB0.4030408@something.invalid>
Message-ID: <m3admeo4u1.fsf@mira.informatik.hu-berlin.de>

Greg Ewing <see_reply_address at something.invalid> writes:

> So, the only way to find out whether a particular
> extension is safe to use in Stackless is to try
> it. 

That is not true. It is more reliable to inspect the source code.

In general, it is very easy to do so: if there are no callbacks to
Python, the module is stackless-safe.

> That's why I say that I would be nervous about
> using Stackless. Not to say that would stop me
> from using it, but I'd feel nervous while I was
> doing it! :-)

That is perfectly fine with me. The question is whether incorporation
of Stackless Python in the core distribution would already make you
nervous. I can see no rational reason for that.

> Yes, I understand that, and I acknowledge that it
> won't cause any harm to have Stackless in the core
> if the Stackless stuff is switched off by default.

It is better than that: it does not need to be "switched off". Even if
switched on (at build time), it won't start copying slices of the
stack. An application actively has to import one of the modules, and
to call a function in those modules, before you need to get nervous.

Regards,
Martin



From martin at v.loewis.de  Wed Sep  4 03:09:50 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 04 Sep 2002 09:09:50 +0200
Subject: Memory fault - core dumped   while compiling 2.2.1 on SCO
References: <3D757EFA.9050401@nospam.yahoo.com> <yu99sn0qqw1r.fsf@europa.research.att.com>
Message-ID: <m3bs7etem9.fsf@mira.informatik.hu-berlin.de>

Andrew Koenig <ark at research.att.com> writes:

> Looks like the same symptom I had.
> 
> If you are using binutils 2.13, try going back to 2.12.1 and
> see if that fixes the problem.

Alternatively, use the system C compiler - Python does not support gcc
on SCO, apparently (or else it would not try to pass -belf).

It is probably easy to add gcc support - just adjust the OpenUnix
fragments in configure.in to SCO.

Regards,
Martin


From gmuller at worldonline.nl  Mon Sep  2 11:18:47 2002
From: gmuller at worldonline.nl (GerritM)
Date: Mon, 2 Sep 2002 17:18:47 +0200
Subject: Must have Python modules?
References: <_THc9.42153$vY2.1039887@e3500-atl2.usenetserver.com> <slrnan6k4s.eg.gerhard.haering@haering.opus-gmbh.net> <FCIc9.30407$FJ1.5585669@e3500-atl1.usenetserver.com>
Message-ID: <akvvkk$ab1$1@reader1.tiscali.nl>

"Robert Oschler" <Oschler at earthlink.net> schreef in bericht
news:FCIc9.30407$FJ1.5585669 at e3500-atl1.usenetserver.com...
<...snip...>

I want to know what the other favorites are "out there".
>
> thx
>
win32com, HTMLgen

I am still looking for a whois module, for ReverseNameLookup. Some snippets
exist, so far I did not yet find a reliable and
complete module.

regards Gerrit
--
www.extra.research.philips.com/natlab/sysarch/





From Dick.Zantow at lexisnexis.com  Wed Sep 18 12:31:59 2002
From: Dick.Zantow at lexisnexis.com (rzed)
Date: Wed, 18 Sep 2002 12:31:59 -0400
Subject: Python doesn't suck!
References: <e34e8ce6.0209180737.49868d9f@posting.google.com>
Message-ID: <ama9s2$e24$1@mailgate2.lexis-nexis.com>

"Domenico" <grayshade at libero.it> wrote in message
news:e34e8ce6.0209180737.49868d9f at posting.google.com...
> http://www.theperlreview.com/at_a_glance.html
>
> According to the most accurate algorithm, Python is the second
> less-sucking language.

Behind PHP (?!) if I read the charts right. Says a lot about the
languages.

--
rzed




From cotabas at earthlink.net  Sun Sep  1 15:46:05 2002
From: cotabas at earthlink.net (CotaBas)
Date: Sun, 01 Sep 2002 19:46:05 GMT
Subject: Executable Python Programs
References: <1rtc9.108$LI2.7238@newsread2.prod.itd.earthlink.net>
Message-ID: <18uc9.160$LI2.13183@newsread2.prod.itd.earthlink.net>

"CotaBas" <cotabas at earthlink.net> wrote in message
news:1rtc9.108$LI2.7238 at newsread2.prod.itd.earthlink.net...


Thank you,
I will check the FAQ before I post again ;-)


---------------------------
 -CotaBas





From loewis at informatik.hu-berlin.de  Sun Sep 15 14:10:08 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 15 Sep 2002 20:10:08 +0200
Subject: Will xml.dom work with HTML docs?
References: <k3Tg9.6083$Lg2.908752@news2.news.adelphia.net> <877khntzvt.fsf@pokey.henrik-motakef.de>
Message-ID: <j4y9a3qg3j.fsf@informatik.hu-berlin.de>

Henrik Motakef <henrik.motakef at web.de> writes:

> > if not is there a good parser module for HTML docs?
> 
> See the htmllib module in the standard library.

Alternatively, see PyXML. It parses an HTML document into a DOM
tree. It does so by knowing the HTML DTD, ie. it silently inserts
missing closing tags whereever needed.

Regards,
Martin



From peter at engcorp.com  Fri Sep  6 08:58:25 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 06 Sep 2002 08:58:25 -0400
Subject: Zope Status?
References: <Xns9281221A07167bosahvnetscapenet@66.185.95.104>
Message-ID: <3d78a657$1@news.sentex.net>

Bo wrote:
> I apologise if this isn't the proper forum to do this in.
> 
> 1) Can anyone tell me roughly if\when Zope is moving to python 2.2?
> 
> 2) Also any idea when they might release Zope 3?
> 
> 3) Anyone close to the Zope team who is willing to shed some light on the 
> near future for things Zope?

Very likely, going to http://dev.zope.org would answer these questions.



From ahmad at NOSPAMbitbuilder.com  Sat Sep  7 03:36:15 2002
From: ahmad at NOSPAMbitbuilder.com (Ahmad Baitalmal)
Date: Sat, 07 Sep 2002 07:36:15 GMT
Subject: Converting python object reference to int and back
Message-ID: <3D79AC95.1000000@NOSPAMbitbuilder.com>

Hi,
I need to keep references to python objects in a tree control's item.
The items can hold an integer as data.
How can I get an id (pointer/reference) from a python object and then 
get that object back using that integer?
-- 
-  - -- ----  ----------------------------------------- --- -- -   -
*Ahmad Baitalmal***
*BitBuilder***
web: http://www.bitbuilder.com
-  - -- ----
-------------------------------------------------------------- --- -- -   -



From aahz at pythoncraft.com  Sat Sep 21 12:32:37 2002
From: aahz at pythoncraft.com (Aahz)
Date: 21 Sep 2002 12:32:37 -0400
Subject: Destructor never called ???
References: <mailman.1032549652.29064.python-list@python.org>
Message-ID: <ami6v5$phc$1@panix1.panix.com>

In article <mailman.1032549652.29064.python-list at python.org>,
Tim Peters  <tim.one at comcast.net> wrote:
>
>Since you and I are approximately the only two people on the newsgroup
>whose heads haven't exploded from trying to understand exactly what the
>hell I'm talking about there <wink>, I think it's a Positive Good that
>Neil skipped all the SCC complications.  "Cycles and destructors: don't
>even think about mixing them" is much easier to remember, and has never
>done me any harm in real life.

As long as you're talking about approximations, make it three people;
I'm of the opinion that I understand roughly half of what you're talking
about and if I count as half a person, you can just round up.  I did try
to understand the rest and my brain didn't explode, but it's probably
leaking around the edges.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From prema at prema.co.nz  Sat Sep 28 14:07:37 2002
From: prema at prema.co.nz (Mike)
Date: Sat, 28 Sep 2002 18:07:37 GMT
Subject: py2exe and a single exe file?
References: <3D792351.3070008@motorola.com>
Message-ID: <Jdml9.252$zz3.2100@newsfeed.intelenet.net>

Hi Stephen !
I've been doing some builds quite successfully using this script from 
the ASPN Cookbook.  I name it Build.py and it sits in my tools folder.

Take a copy and put it in with your source folder then run
Build.py NewApp.py
Hope it helps
Best regards
Mike



Stephen Boulet wrote:
> I'm trying to get py2exe to work for me, and I must be doing something 
> wrong.
> 
> I have a script that uses easygui (which uses Tkinter) and Numeric. It 
> also imports a small module from the same directory that the script 
> resides in.
> 
> I try building it like this:
> 
>    python setup.py py2exe -d N:\ -i easygui
> 
> It builds and runs, but only with the other stuff in the build directory 
> with the generaged exe. The other stuff is:
> 
>   tcl directory containing tcl8.3 and tk8.3 directories
>   _numpy.pyd
>   _sre.pyd
>   _tkinter.pyd
>   multiarray.pyd
>   umath.pyd
> 
> What I would like to accomplish is just having a single exe file lumping 
> everything together. Is this possible?
> 
> -- Stephen
> 



From manuel at hendel.net  Tue Sep 17 04:39:58 2002
From: manuel at hendel.net (Manuel Hendel)
Date: Tue, 17 Sep 2002 10:39:58 +0200
Subject: two or more whitespace regular expression?
In-Reply-To: <yJWg9.96506$ub2.1934842@news1.tin.it>
References: <am12a3$qo0$1@newsreader.mailgate.org> <yJWg9.96506$ub2.1934842@news1.tin.it>
Message-ID: <20020917083958.GE23629@partagas.as.de.cw.net>

On Sun, Sep 15, 2002 at 07:33:50AM +0000, Alex Martelli wrote:
> import re
> twoormorewhites = re.compile(r'\s\s+')
                               ^
What is the "r" for? I tried to find something about this in the docs
(Regular Expression HowTo , etc.) but I couldn't find anything.

manuel

-- 
When you teach your son, you teach your son's son. 
-The Talmund  



From Padraig at Linux.ie  Mon Sep  9 06:13:05 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Mon, 09 Sep 2002 10:13:05 GMT
Subject: tree data structure
References: <FM0e9.2860$cP3.6165@news.iol.ie> <ala6c3$6i2$1@bob.news.rcn.net> <Ti1e9.2864$cP3.6124@news.iol.ie>
Message-ID: <Ru_e9.3322$cP3.6690@news.iol.ie>

Padraig Brady wrote:
> Raymond Hettinger wrote:
> 
>> There are a number of ways to represent trees:
>> -- instances of a treenode class
>> -- a list of lists:  [l0, [s2, [s0, [q1]], [s1,[q2]], [s3]]]
>> -- or, my favorite, a dictionary of lists:
>>
>> original = [('L0','S2'), ('S2','S0'), ('S2','S1'), ('S1','Q2'), 
>> ('S0','Q1'), ('S2','S3')]
>>
>> tree = {}
>> for key, value in original:
>>     tree.setdefault(key,[]).append(value)
>> tree
>> {'S2': ['S0', 'S1', 'S3'], 'S1': ['Q2'], 'S0': ['Q1'], 'L0': ['S2']}
> 
> How would you iterate over the above?

To complete this thread, Guido wrote an essay
on how to manipulate graphs like that generated above:
http://www.python.org/doc/essays/graphs.html

P?draig.



From whisper at oz.net  Tue Sep 17 06:22:20 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 17 Sep 2002 03:22:20 -0700
Subject: More on Protecting Source Code
In-Reply-To: <20020917062412.GA85457@hishome.net>
Message-ID: <GCEDKONBLEFPPADDJCOEOEEKFAAA.whisper@oz.net>

Oh yeah, and my views of intellectual property seems to be a lot closer to
the law of the land then yours.

Dave LeBlanc
Seattle, WA USA




From cliechti at gmx.net  Sun Sep  1 17:32:58 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 1 Sep 2002 23:32:58 +0200
Subject: del(instance.attr) -- no attribute
References: <un512ii7hjgq22@corp.supernews.com>
Message-ID: <Xns927CEF87372FAcliechtigmxnet@62.2.16.82>

Lee Harr <missive at frontiernet.net> wrote in 
news:un512ii7hjgq22 at corp.supernews.com:
> Is this a bug?

no

>>>> class C:
> ...  a = 1
> ... 
>>>> c = C()
>>>> c.a
> 1
>>>> del(c.a)
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> AttributeError: C instance has no attribute 'a'
>>>> hasattr(c, 'a')
> 1

yes, but

>>> del(C.a)

works

> I can see that maybe del(c.a) should not be allowed to delete
> the attribute, but I think the error message is misleading.

but the message is true. the instance has no attribute named "a", it's an 
attribute of the class "C". maybe "hasattr" is missleading to you, because 
it's showing class and instace attributes, including those of base classes.
but deleting has to occour in the correct spot.

chris

-- 
Chris <cliechti at gmx.net>



From SBrunning at trisystems.co.uk  Thu Sep 19 10:52:13 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 19 Sep 2002 15:52:13 +0100
Subject: strong/weak - dynamic/static [Was: Getting started]
Message-ID: <31575A892FF6D1118F5800600846864DCBD82B@intrepid>

> From:	Jacek Generowicz [SMTP:jacek.generowicz at cern.ch]
> Could someone provide a serious reference (i.e. something which would
> have gone through some sort of editorial/peer/technical review, rather
> than someone's random opinions on the web) explaining or defining the
> concepts of weak, strong, dynamic and static typing.
 
Words can mean anything that you want them to mean, and I don't know if
there is any such thing as definitive definitions of these words in this
context. Having said that, I know what *I* mean when I use them - see
<http://diveintopython.org/odbchelper_funcdef.html>. (Bottom of the page.)

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




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



From tdelaney at avaya.com  Mon Sep 23 20:05:09 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 24 Sep 2002 10:05:09 +1000
Subject: conditional expressions
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5CA@natasha.auslabs.avaya.com>

> From: Lulu of the Lotus-Eaters [mailto:mertz at gnosis.cx]
> 
> "Delaney, Timothy" <tdelaney at avaya.com> wrote previously:
> |This type of thinking is one very good reason why bool() is 
> a Good Thing
> |(TM). A large number (my guess is the majority) of Python 
> users do not know
> |about operator.truth().
> 
> Well...  Python users won't necessarily innately know about 
> bool() when
> it is available in 2.3 (or 2.2.x).  In either case, users 

No - however it will be documented as a core feature rather than a (to many)
obscure library. For current uers, it will appear as a new feature
(immediately bringing it to attention). For new users, it will be in an
easily accessible part of the documentation - the same place that integers
and other built-in types are discussed.

Tim Delaney



From mwh at python.net  Fri Sep 27 06:21:39 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 27 Sep 2002 10:21:39 GMT
Subject: if...else on the same line?
References: <mailman.1032622182.9002.python-list@python.org>
Message-ID: <lk4rcbwxe0.fsf@pc150.maths.bris.ac.uk>

David Garamond <davegaramond at icqmail.com> writes:

> i'm sorry for this terrorizingly basic question, but is it possible at
> all to write an if ... else statement on a single line in python?

No.

easy-ones-are-the-quickest-ly y'rs
M.

-- 
  > Emacs is a fashion statement.
  No, Gnus is a fashion statement.  Emacs is clothing.  Everyone 
  else is running around naked.
        -- Karl Kleinpaste & Jonadab the Unsightly One, gnu.emacs.gnus


From gua81 at XXXyahoo.com  Mon Sep 30 21:01:55 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Tue, 1 Oct 2002 11:01:55 +1000
Subject: PROBLEM with form in cgi
References: <anas2k$i2b$1@lust.ihug.co.nz>
Message-ID: <anas5g$i4j$1@lust.ihug.co.nz>

sorry it's "invalid syntax" not "syntax error"
Cheers.

"CheapSkate" <gua81 at XXXyahoo.com> wrote in message
news:anas2k$i2b$1 at lust.ihug.co.nz...
> HI,
> I'm having problem trying to figure out what is wrong with my cgi script.
> here's the code that's creating the problem in the apple.cgi
> print """<form action="delete.html" name="myform" method="post">"""
>                              ^
> it keeps on saying invalid syntax in the action(as indicated)
> and if I try to change it to this
> print """<form action="delete.html" name="myform" method="post">"""
>                              ^
> it gives the error at action.
> Please help.




From sismex01 at hebmex.com  Thu Sep 26 18:24:10 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Thu, 26 Sep 2002 17:24:10 -0500
Subject: Question about select()
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233026397F8@mail.hebmex.com>

Does select() work on bound sockets, to detect
incoming connections?

I'd like to be able to have a timeout when accepting
connections from a server socket.

Thanks!

-gus










Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From max at alcyone.com  Tue Sep 17 17:14:58 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 17 Sep 2002 14:14:58 -0700
Subject: CGI question
References: <slrnaoerk7.mgo.jboy@otaku.freeshell.org> <mailman.1032290361.21330.python-list@python.org>
Message-ID: <3D879B52.BD293951@alcyone.com>

Ian Bicking wrote:

> Look at cgi.FieldStorage.  Something like:
> 
> f = cgi.FieldStorage(keep_blank_values=1)
> 
> Will save these values.  (I don't know why it isn't set by default...
> by
> default you shouldn't throw information away, but FieldStorage does in
> this case)

It actually is not unreasonable.  If form elements exist but are left
blank, such as a text field, they will appear in the form submission
even though you may not care about them.  Since when you're receiving a
form submission, presumably you know what elements were in the form, you
always know what to look for and can simply default to something
reasonable if it isn't present.

Note that what he's trying for is something different than a mere blank
value; a blank value would look something like `?blank=' whereas what
he's talking about is `?blank'.  I don't know offhand if
keep_blank_values will accept this form; I'm hoping you tested it before
suggesting it.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ It's like being on a diving board that you know is too high.
\__/ Sade Adu
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.


From cliechti at gmx.net  Wed Sep 18 17:54:01 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 18 Sep 2002 23:54:01 +0200
Subject: Python GUI app to impress the boss?
References: <mailman.1032381445.13579.python-list@python.org>
Message-ID: <Xns928DF32683D0cliechtigmxnet@62.2.16.82>

"Gray, Lee P." <lpgray at UOP.com> wrote in 
news:mailman.1032381445.13579.python-list at python.org:
>> Boa is in fact quite a bit alpha-ish, but after you're
>> used to the quirks and limitations, it is really great.
>> Also note that you should use the version from CVS
>> which is *far* more developed than the downloadable.
>
> Yeah, I tried to download from CVS, but being a complete noob
> couldn't figure out how, short of tracking down and copying and pasting
> every updated file.  Any tips would be appreciated.

you're on windows right? then grab http://tortoisecvs.sf.net and install 
the latest version. after that right click on an empty folder in the 
windows explorer and select "cvs checkout", copy and paste the "CVSROOT" in 
the dialog. you can find the cvsroot, or at least how it looks like on the 
"browse cvs" page that you have apparently found.

the home of cvs is http://www.cvshome.org and these are interesting sites:
http://www.loria.fr/~molli/cvs-index.html, http://cvsbook.red-bean.com/

chris

-- 
Chris <cliechti at gmx.net>



From robin at execulink.com  Sat Sep 14 21:34:03 2002
From: robin at execulink.com (robin at execulink.com)
Date: Sat, 14 Sep 2002 21:34:03 -0400
Subject: html template system?
References: <btug9.688$Ee4.1090@news-server.bigpond.net.au>
Message-ID: <0ro7ouskb8ctv6f846gb8cfpqtl5vs1avc@4ax.com>

":B nerdy" <thoa0025 at mail.usyd.edu.au> wrote:

>what are good templating systems avaliable for python?

You have so many options, documented here:
http://www.python.org/cgi-bin/moinmoin/WebProgramming

These include my Wasp.

-- robin


From teaandbikkie at aol.com  Sat Sep 14 16:44:31 2002
From: teaandbikkie at aol.com (TeaAndBikkie)
Date: 14 Sep 2002 20:44:31 GMT
Subject: Skipping Parts of a For Loop
References: <mrMg9.95802$ub2.1909231@news1.tin.it>
Message-ID: <20020914164431.23208.00000400@mb-fp.aol.com>

>> Is there a way to skip some parts in a for loop?
>> 
>> 
>> for i in range(3):
>> print "i before", i
>> i=2
>> print "i after", i
>> 
An alternative to Alex's great suggestion, you could use an if statement:

for i in range(3):
    if 0<i<2: continue
    print i

Regards,
Misha


From bokr at oz.net  Sun Sep  8 02:53:24 2002
From: bokr at oz.net (Bengt Richter)
Date: 8 Sep 2002 06:53:24 GMT
Subject: python program to watch a file - example.
References: <mailman.1031464092.30848.python-list@python.org>
Message-ID: <ales54$2un$0@216.39.172.122>

On Sun, 8 Sep 2002 01:47:12 -0400, "Wayne R" <wayne at tbnets.com> wrote:

>This is a multi-part message in MIME format.
Why?
>
>------=_NextPart_000_0039_01C256D9.B8CB5320
>Content-Type: text/plain;
>	charset="us-ascii"
>Content-Transfer-Encoding: 7bit
>
> 
>I am trying to write a python program that will run as a deamon and
>watch a file for data to arrive and then wake up and perform duties on
>that data.   I found ways to make a python program run as a deamon but I
>am at a loss to find anything related to watching a file for data so
>that it can then process the data.  I didn't really want to do a sleep
>loop or something like, I am sure there is a better way.
> 
>Any ideas or links to more info?
>
Don't annoy people you want help from, for starters.
What system are you targeting? (Outlook in the headers says you're on windows
generating this abominable mix of actual message and useless cruft).

>------=_NextPart_000_0039_01C256D9.B8CB5320
>Content-Type: text/html;
WHY?!!

Regards,
Bengt Richter


From benji_75 at hotmail.com  Thu Sep 12 08:15:39 2002
From: benji_75 at hotmail.com (Ben Gerblich)
Date: Thu, 12 Sep 2002 21:45:39 +0930
Subject: Fatal Python error: PyThreadState_Get: no current thread
Message-ID: <F257a97F0epIncnz9h50000d230@hotmail.com>

Hi All,

My code has just started crashing. I get the message:

Fatal Python error: PyThreadState_Get: no current thread
Aborted (core dumped)

Is there a way to turn on debug mode or something that will allow me to see
where/why my app is crashing. "python2 -d main.py" does not shed any light.

It seems as though the app crashes because of threads. I have found parts of 
my code that when commented out, the app does not crash. BUT, if i leave 
this code in, and put some 'print' statements above it, then just (eg)3 out 
of the 10 prints will execute before the app crashes.

Thanks, Ben.
Im using:
Red Hat Linux 7.1
python2-2.1-5.i386.rpm
python2-devel-2.1-5.i386.rpm
python2-tkinter-2.1-5.i386.rpm
python2-tools-2.1-5.i386.rpm
wxPython-2.3.2.1-1-Py21.i386.rpm
wxGTK-2.3.2-1.i386.rpm
wxGTK-devel-2.3.2-1.i386.rpm



_________________________________________________________________
MSN Photos is the easiest way to share and print your photos: 
http://photos.msn.com/support/worldwide.aspx




From lee100 at sport.rr.com  Sat Sep 21 10:03:13 2002
From: lee100 at sport.rr.com (Lee Gray)
Date: Sat, 21 Sep 2002 14:03:13 GMT
Subject: pyrhon and oracle
References: <5b8834c2.0209201412.50e77861@posting.google.com> <pan.2002.09.20.23.29.54.67293.2172@verizon.net>
Message-ID: <B__i9.505088$q53.17034835@twister.austin.rr.com>

Satheesh Babu wrote:

> cx_Oracle might be the *easiest* to use. I use DCOracle while working
> with Zope, but for standalone development, cx_Oracle is tops.
> 
> Download the driver from
> http://www.computronix.com/utilities.shtml
> 
> go to Python DB-API specs
> http://www.python.org/topics/database/DatabaseAPI-2.0.html
> 
> Here's an example using DB-API
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/81189
> 
> vsb
> http://vsbabu.org/
> 

I'm just looking into Python, so these articles don't mean much to me yet... 
Does all this mean that Python doesn't talk to an Oracle DB "out of the 
box"?  For Windows, can Python use ADO?

Thanks,
Lee


From aleax at aleax.it  Sun Sep 15 12:59:12 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 15 Sep 2002 16:59:12 GMT
Subject: Tkinter in PyRun_SimpleString(char* s)
References: <am12be$4j0$1@schleim.qwe.de> <_OWg9.109539$pX1.3874894@news2.tin.it> <am2alp$3v3$1@schleim.qwe.de>
Message-ID: <A%2h9.98014$ub2.1982070@news1.tin.it>

Torsten Mohr wrote:

> Hi,
> 
> thanks for that hint.
> 
>> As you can see, the problem is with sys.argv, which you have
>> not set.  You can set it up from the Python code you execute,
>> before you import Tkinter, or with C-API call PySys_SetArgv.

...which needs to be called AFTER Py_Initialize -- I think that
may be your problem:

> /* C code */
>         Py_SetProgramName(argv[0]);
>         PySys_SetArgv(argc, argv);
> 
>         Py_Initialize();
>         PyRun_SimpleString(script);
> 
>         Py_Finalize();

Try switching the 2nd and 3rd ones of these five calls.


Alex



From loewis at informatik.hu-berlin.de  Thu Sep 26 05:27:42 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 26 Sep 2002 11:27:42 +0200
Subject: unicode woes
References: <amuhrk$4cn$05$1@news.t-online.com>
Message-ID: <j4r8fh85ip.fsf@informatik.hu-berlin.de>

Ulli Stein <mennosimons at gmx.net> writes:

> Some weeks ago the management decided to deploy unicode so that we can 
> handle every charset uniformly. The problem which showed at first: we 
> realized that we cannot change the encoding in site.py afterwards, i.e. we 
> have to specify the encoding all the way. 

You mean, every time? Not necessarily. If you want to write Unicode to
a file, I recommend to use codecs.open.

> It got worse and worse because 
> the unicode encoding spread like a virus through all the source code. 
> Simple Exceptions which we wanted to print to the console throw another 
> Exception:
> UnicodeError: ASCII decoding error: ordinal not in range(128)

I cannot understand this problem. I get

>>> def foo():
...   raise Exception(u"\u20ac")
... 
>>> foo()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 2, in foo
Exception>>> 

So even though it doesn't print the Unicode string, it does not cause
a UnicodeError, either.

> Furthermore, in Python you have no way to detect, which encoding a
> special unicode string atually is. We ended up having our own u()
> function.

What do you mean, 'actually'? A single Unicode string can be encoded
in many encodings, e.g. u"Hallo" can be encoded in ascii, latin-1,
koi-8r, or utf-8. Every Unicode string can be encoded in utf-8.

> The problem with that: we have to pass _every_ string through
> it. Ouch!

Again, I cannot understand the problem. Can you please show a bit of
code that demonstrates how you use that u() function, and what it
does?

> Now our source code is as ugly as our program is slow. Isn't there
> another way?

Not sure what problem you are trying to solve, so it is hard to tell
whether there is another way to solve it.

> When will the types string and unicode merged together like in Java? If 
> this is not changing and we do not find another way, we will have to port 
> _all_ source code to C++.

Don't despair; I'm certain that there is an easier way. Try to follow
the following principles:

- Never mix byte strings and Unicode strings (unless the byte strings
  are restricted to bytes <127, perhaps).
- In a Unicode application, convert all byte strings to Unicode as early
  as possible.
- Convert all Unicode data back to byte strings as late as possible.
- If you need to be sure that something can be printed for diagnostic
  output, use repr.
- For normal output, use an codecs.StreamWriter.

> def u(obj):

I'm not quite sure what this u function is supposed to do. If this is
meant to be an equivalent of str() which returns a Unicode object, I
recommend that you implement this as

def u(obj):
  try:
    return unicode(obj)
  except UnicodeError:
    return unicode(obj, getEncoding())

>         encoding = getEncoding()
>         if encoding == None:
>                 raise EncodingNotInitialized("There was no encoding defined previously.")

This is quite inefficient. Why is it that no encoding could be defined
previously? There is alway the system default encoding. Also, can't
you arrange encoding to be global, and change every time the encoding
changes? That would avoid the function call.

>         
>         if type(obj) == types.StringType:
>                 return unicode(obj, encoding)

I recommend to put the most frequent case first. I also recommend to
avoid type tests. Instead, just invoke 

   return unicode(obj, encoding)

and catch the exception you get if that fails (TypeError)

>         elif type(obj) == types.UnicodeType:
>                 # As we can't be sure what encoding the unicode object has we receive,
>                 # we will simply re-encode it.
>                 return unicode(obj.encode(encoding), encoding)

What good does that do? You perform a roundtrip conversion, so that
either fails (with a UnicodeError), or you get back the same object.
I recommend to write

                  return obj

instead. If you really think you need the UnicodeError if the Unicode
string cannot be encoded in encoding, write

                  obj.encode(encoding)
                  return obj

That saves atleast one of the conversions. However, it is not clear to
me why you need the UnicodeError at all.

>         elif type(obj) == types.IntType or type(obj) == types.LongType:
>                 return unicode("%d" % obj, encoding)
>         elif type(obj) == types.FloatType:
>                 return unicode("%f" % obj, encoding)
>         elif type(obj) in [ types.TupleType, types.ListType, types.Dic
> tType, types.DictionaryType ]:
>                 return unicode(str(obj), encoding)

These can be summarized as

                  return unicode(obj)

It is not clear why you special-case None; if this is a string
conversion function, returning None is just as surprising. Instead, I
would expect u(None) to work as

>>> unicode(None)
u'None'

> >>> "????%s"%unicode("???", "iso8859-15")
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> UnicodeError: ASCII decoding error: ordinal not in range(128)

What do you expect to happen? You have a byte string, and you want to
insert a unicode string. Those two cannot be mixed unless you convert
one form to the other. For that, you need to know what encoding to
use; Python uses the default encoding, which only converts ASCII in
the standard installation.

If you follow the suggestion to never mix byte strings and Unicode
strings, this would not happen.

Regards,
Martin


From max at alcyone.com  Fri Sep 20 23:30:08 2002
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 20 Sep 2002 20:30:08 -0700
Subject: Python about file I/O
References: <amgnnl$5qc9n$1@ID-49758.news.dfncis.de>
Message-ID: <3D8BE7C0.80853072@alcyone.com>

Ken wrote:

> 1.) How do you list the filenames within a given directory so user can
> choose which file to open?

Use os.listdir.

> 2.) How do you ensure the program doesn't crash if I specify a
> filename to
> open but the file doesn't exist? (I.E. how to check if file exist
> before
> opening the file?)

It doesn't "crash," it throws an error.  Simply catch the error and do
something meaningful:

	try:
	    inFile = file('soandso.txt')
	    ...
	except IOError:
	    print >> sys.stderr, "Couldn't open file"

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Woman was God's _second_ mistake.
\__/ Friedrich Nietzsche
    The laws list / http://www.alcyone.com/max/physics/laws/
 Laws, rules, principles, effects, paradoxes, etc. in physics.


From max at alcyone.com  Tue Sep 24 05:05:14 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 24 Sep 2002 02:05:14 -0700
Subject: instance creation
References: <265e380f.0209240052.4c2988f3@posting.google.com>
Message-ID: <3D902ACA.1188B5AE@alcyone.com>

Paul MG wrote:

> - Having two constructors seems to be illegal - or at least the second
> overwrites the first.

That's correct.  Python does not allow overloading functions with the
same name but different argument lists.

> - Assigning to 'self' in the load method seems to really not work!

Indeed, and you shouldn't have expected it to.  self is a variable like
anything else, and in Python, variables are merely bindings, not
reseatable references.  Assignment just rebinds the name, it doesn't
affect the bound object.

> - Putting the create() method outside of the class works fine - but I
> want it to be a class method on Appointment, goddammit!:) Is there no
> way to do this?

Later versions of Python have a staticmethod function for this.  You'd
use it like this:

	class C:
	    def f(x): # note, no self argument
	        ...
	    f = staticmethod(f)

This allows f to be called as C.f or even c.f, if c were an instance of
C.

> Any help would be appreciated, basically i am looking for the python
> idiom to do this, as my C++ and Smalltalk idiom's clearly aren't in
> the grain of the language!

Depending on your tastes, a module function might make more sense.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ I wish you well but / Right now I'm just bitter
\__/ Chante Moore
    Alcyone Systems' CatCam / http://www.catcam.com/
 What do your pets do all day while you're at work?  Find out.


From jylimd at yahoo.com  Wed Sep 18 17:30:54 2002
From: jylimd at yahoo.com (J. Li)
Date: Wed, 18 Sep 2002 17:30:54 -0400
Subject: [Help]: mailbox classes
Message-ID: <jylimd-DAD12F.17305418092002@skates.gsfc.nasa.gov>

Hi,

I am trying to extract a message from a unix mailbox file. Python does 
provide few classes for easy and uniform access. So it says! After 
having read the libiary references, I went nowhere. 

Can anyone give me an example as how to use these mailbox classes and 
what is it relationship with email.message classes.

thanks


From sholden at holdenweb.com  Thu Sep 12 00:04:39 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 12 Sep 2002 00:04:39 -0400
Subject: [FEEDBACK] Is this script efficient...is there a better way?
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <L9Rf9.168314$l_4.127784@atlpnn01.usenetserver.com> <4ZRf9.1782$Lg2.248733@news2.news.adelphia.net> <3hSf9.168947$l_4.98525@atlpnn01.usenetserver.com> <8PSf9.1796$Lg2.257378@news2.news.adelphia.net>
Message-ID: <_oUf9.170870$l_4.126599@atlpnn01.usenetserver.com>

"Bob X" <bobx at linuxmail.org> wrote ...
[ ... ]
> >
> Thank you very much...your shorter script did work. The keywords are all
> "normal" english words that would not have any special characters. I
> only added a re.IGNORECASE (since case doesn't matter to me) after a
> quick search of the python docs.
>
Was there any significant difference in the timings for the two different
methods, and how does the number of different keywords affect the timings?

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From peter at engcorp.com  Mon Sep  2 12:16:34 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 02 Sep 2002 12:16:34 -0400
Subject: Newbie's unsurmountable pb with IO stream trick
References: <aba30b75.0209020633.26ba4c70@posting.google.com>
Message-ID: <3D738EE2.3A8756D@engcorp.com>

Spendius wrote:
> 
> I'm about to give up learning Python because I've met
> insurmountable problems. I'm a newbie just having begun
> teaching myself this language a few days ago, and I
> already stall against unsolvable (at least to me) issues:
> so could you please help me and *show me* I'd be wrong
> giving up ?
> 
> I tried to rewrite something I wrote in Java, trying to
> communicate with an OS process (SQL*Plus in this occurence),
> but big time bucked against
> - the lack of methods of the so-called 'file objects' (if
>   there was an equivalent to java.io.InputStream.available() !)
> - the fact you don't declare variables: you just assign
>   values to them => I immediately hit against the 'global'
>   keyword issue (rather a global mess) and gave up...

Sounds like you're creating a few problems for yourself all
at once here.  For one thing, why are you trying to use 
global variables at all when the original code was in
Java, which doesn't have such things?

And what does not declaring variables have to do with
the "global keyword issue" (whatever that is)?

If you have an SQL database to access, why not use one of the
standard DBAPI modules to get to it?  You might be doing a lot
of extra work when you don't have to.

If you're just learning Python, you've taken on a fairly
complicated and large task right at the beginning.  Why not
spend some time learning to use it for simpler tasks before
you jump in head first?

Finally, from the description of what you're trying to do
with InputStream.available(), you might have a few options
including opening the file in non-blocking mode or using 
the select module, or something from the win32all package if 
you're running under Windows (you didn't say).

-Peter


From max at alcyone.com  Fri Sep 20 01:48:13 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 19 Sep 2002 22:48:13 -0700
Subject: newbie seeks advise.
References: <77c84db2.0209192137.51167ebb@posting.google.com>
Message-ID: <3D8AB69D.52A45E72@alcyone.com>

akirasugiura wrote:

> I've been learning python for about a month. I had no prior knowledge
> of programming.
> 
> My question is, I have a tuple which is, data = (69, 40, 74, 41, 56,
> 57, 38, 58, 26, 55). How I can make a function that picks up a largest
> number
> from this "data" tuple. If "data" was a  'List' I would be able to
> use "sort" but since sort doesn't  work with tuple. I am stuck here.

You could always convert the tuple to a list with the list function:

	data = (69, 40, 74, 41, 56, 57, 38, 58, 26, 55)
	dataAsList = list(data)

and then do what you intended.  Note that there is a max function which
already exhibits this behavior (which will work on any sequence):

	largest = max(data)

You could always do it manually; walk through the sequence, and keep a
tally of the largest one you've seen so far.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Man is preceded by forest, followed by desert.
\__/ (graffiti written during French student revolt)
    Bosskey.net / http://www.bosskey.net/
 A personal guide to online multiplayer first person shooters.


From nospam at horvath.com  Fri Sep 27 22:02:37 2002
From: nospam at horvath.com (Bob Horvath)
Date: Sat, 28 Sep 2002 02:02:37 GMT
Subject: combining mailbox and email modules
References: <NPck9.581867$UU1.104088@sccrnsc03> <slrnap41re.1mh.matt@happy-hour.t-island.com> <LAxk9.79207$gA4.27843@sccrnsc02> <1fj6jb2.62mx08f47amsN%fgranger@alussinan.org>
Message-ID: <158l9.428169$kp.1216679@rwcrnsc52.ops.asp.att.net>

Fran?ois Granger wrote:
> Bob Horvath <nospam at horvath.com> wrote:
> 
> 
>>The problem is that it looks good to me.  I've played around with 
>>it, and if I take out the MIME gunk (It actually is a Klez virus), 
>>it parses ok.
> 
> 
> I have had difficulties with the handling of multipart message with the
> email package. I did not found a solution yet.
> 

Actually I did get it working by using the HeaderParser class in 
Parser.py.  It doesn't parse the body.  All I needed was the 
headers, and be able to recreate the body.  This one worked great.



From syver-en+usenet at online.no  Sat Sep 14 07:17:05 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Sat, 14 Sep 2002 11:17:05 GMT
Subject: Calling Python from C# - please help
References: <6de16ac9.0209130508.5d1b56c6@posting.google.com> <j4elbyt3yw.fsf@informatik.hu-berlin.de> <ur8fyxau0.fsf@online.no> <0iCg9.2350$Ee4.6516@news-server.bigpond.net.au> <3D82FA5B.3050003@mxm.dk>
Message-ID: <uptvgvn1h.fsf@online.no>

Max M <maxm at mxm.dk> writes:

> Mark Hammond wrote:
> 
> > I have no examples of this, but I should.  I'm co-authoring a .NET
> > book which is nearly out :) -
> 
> 
> That sounds great. I enjoyed the win32 book, and still use it as a
> reference.
> 
> Are you by any chance still developing Pyhton .NET, or has it
> completely stopped?
> 
> regards Max M

Now that J# (Java .NET) has been released, I was wondering about the
possibility of compiling Jython with the MS J# compiler to create a
Jython .NET. Has anyone considered the possibility?

-- 

Vennlig hilsen 

Syver Enstad


From printers at sendme.cz  Wed Sep 11 09:14:30 2002
From: printers at sendme.cz (A)
Date: Wed, 11 Sep 2002 15:14:30 +0200
Subject: Paradox  table
Message-ID: <3D7F5DD6.8814.1363983@localhost>

Hi,
 Is there a module for reading Paradox database (.DB) files with Python?
I need to run the script under Linux, so I can not use ODBC.
Any suggestion would be appreciated
 Thanks for help
 Ladislav
 I look forward to hearing from you soon.

Best regards,
Ladislav Blazek( Mr.)

BMA TRADING Ltd.
email: export at sendme.cz
email2: export at bmatrading.com
Fax:/Tel +420 506 447921
Tel:+420 506 447920, +420 602 849309






From mertz at gnosis.cx  Thu Sep 26 21:11:05 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Thu, 26 Sep 2002 21:11:05 -0400
Subject: indexing and searching pdf files
References: <pan.2002.09.26.18.05.37.70342.4621@presidency.com>
Message-ID: <pA7k9kKkXIdL092yn@gnosis.cx>                

Rajarshi Guha <rajarshi at presidency.com> wrote previously:
|I have a load of pdf files and I would like to index them so that I can
|serach them for keywords. I was thinking of using pdftotext to generate
|the textfile and then create the index from that.
|My question - is there already something like this with python?
|Another question which is slightly off topic is, does anybody know of any
|articles/pages that talk about indexing text files efficiently - index
|generaion algorithms etc?

Carlo Bifulco created a program called PdfSearch do just what you want.
Its homepage is at:

    http://pdfsearch.sourceforge.net/

I have not tried the latest version, but I remember the project because
Carlo uses my indexer.py module as part of his project.

In terms of my indexer.py module, I wrote an article about full text
indexing at:

    http://gnosis.cx/publish/programming/charming_python_15.html

This was extended to cover XPATH indexing of XML documents in:

    http://gnosis.cx/publish/programming/xml_matters_10.html

The latter isn't your interest, but in concept it is somewhat similar to
Carlo's extension.

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------




From sabineellen at aol.com  Fri Sep 27 19:57:22 2002
From: sabineellen at aol.com (Sabineellen)
Date: 27 Sep 2002 23:57:22 GMT
Subject: windows xp command prompt
References: <mailman.1033165522.4243.python-list@python.org>
Message-ID: <20020927195722.29739.00005390@mb-mb.aol.com>

restarting the machine in windows xp includes all windows and programs shutting
down and restarting anew; he did that already


From pyth at devel.trillke.net  Wed Sep  4 20:22:33 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 5 Sep 2002 02:22:33 +0200
Subject: Python parser for cross references
In-Reply-To: <Xns927F9E6AC5904howardeegsoftwarecom@66.75.162.198>; from howard@eegsoftware.com on Wed, Sep 04, 2002 at 10:34:08PM +0000
References: <Xns927F9E6AC5904howardeegsoftwarecom@66.75.162.198>
Message-ID: <20020905022233.P3893@prim.han.de>

Howard Lightstone wrote:
> I would like to use Source Navigator to (re)examine a largish application.
> It parses the C, C++,assembly and Python jobs I am working right now and 
> has outputs that *statisfy* the naive QA people who think tables and 
> diagrams are *documentation*.
> 
> The Python parser for it was written by Thomas Heller (thank you very 
> much).
> 
> However, it does not emit cross reference information.  Of course, I first 
> thought "why not just add one?".  I can see why Mr. Heller did not.
> 
> A cross reference parser has to track scope....and determine function calls 
> within argument lists.....and track just 'name' references which are passed 
> around...and on and on.
> Is there a (C) parser for Python around which does a "just-good-enough" 
> parsing job to be adapted for this?

I don't think you could do this with a parser. 

It requires getting type information and that is only possible if
you run a program.  You probably would have to trace a program 
to get information which names refer to which object type(s).

regards,

    holger




From skip at pobox.com  Thu Sep 26 11:49:01 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 26 Sep 2002 10:49:01 -0500
Subject: off topic -- spam assassin
In-Reply-To: <200209261407.g8QE76Wo028975@ratthing-b246.strakt.com>
References: <200209261407.g8QE76Wo028975@ratthing-b246.strakt.com>
Message-ID: <15763.11373.573554.830138@12-248-11-90.client.attbi.com>

    Laura> On the advice of several of you here, I installed spam assassin,
    Laura> and had a pleasant six months of nearly spam free mail.  Thank
    Laura> you, each and every one of you who recommmended it.  Alas, the
    Laura> spammers have got smarter, and I am now getting 25+ pieces of
    Laura> spam that aren't caught by spam assassin every day again. :-(
    Laura> What's new for our side of the arms race?

Laura,

The tension between spammers and the people who develop spam-fighting
software can be likened to an arms race.  It's fairly clear the spammers use
tools like SA to make their mailings more likely to get through.  If you
installed SA six months ago, I'd suspect you're running 2.20 or thereabouts.
Since then 2.31 and 2.41 have been released.  You might want to upgrade.

Skip



From fredrik at pythonware.com  Thu Sep  5 13:27:16 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 05 Sep 2002 17:27:16 GMT
Subject: str() of a tuple
References: <H1xnHJ.1zE1@beaver.cs.washington.edu>
Message-ID: <UtMd9.9506$HY3.2176138@newsc.telia.net>

Douglas Zongker wrote:

> Why, oh why does the str() of a container use the repr()s of the
> objects inside?  This seems to violate what the docs claim str() does
> -- "return a nicely printable representation of an object."

"an object" refers to the container itself, not the objects
in the container.

</F>




From bokr at oz.net  Wed Sep  4 21:58:50 2002
From: bokr at oz.net (Bengt Richter)
Date: 5 Sep 2002 01:58:50 GMT
Subject: Memory fault - core dumped   while compiling python 2.2.1 on   SCO
References: <3D757EFA.9050401@nospam.yahoo.com> <yu99sn0qqw1r.fsf@europa.research.att.com> <m3bs7etem9.fsf@mira.informatik.hu-berlin.de> <3D765A05.6030004@nospam.yahoo.com> <j48z2h61j9.fsf@informatik.hu-berlin.de> <3D768753.8030302@nospam.yahoo.com> <m3n0qx8j3w.fsf@mira.informatik.hu-berlin.de> <3D76A139.1030304@nospam.yahoo.com> <al69pn$sg$0@216.39.172.122> <3D76BC56.7020503@nospam.yahoo.com>
Message-ID: <al6doq$au2$0@216.39.172.122>

On Wed, 04 Sep 2002 19:07:18 -0700, Anton Graph <aglyportREMOvethispart at nospam.yahoo.com> wrote:

>Bengt Richter wrote:
>> On Wed, 04 Sep 2002 17:11:37 -0700, Anton Graph <aglyportREMOvethispart at nospam.yahoo.com> wrote:
[...]
>> 
>> How about
>>     cc -v
>> ?
>> 
>> what does
>> 
>>     which cc
>> 
>> give you? And
>> 
>>     which cc | xargs -ld
>> ?
>
>
>
>$ which cc
>/bin/cc
>$ /bin/cc -v
>Usage: cc [ options ] files ...
>$ which cc | xargs -ld
>xargs: d is not a valid number of lines, must be a positive integer
>
Oops, sorry, that should have been

   which cc | xargs ls -ld

Regards,
Bengt Richter


From mhouck at eecs.tufts.edu  Thu Sep 12 12:48:53 2002
From: mhouck at eecs.tufts.edu (mike)
Date: Thu, 12 Sep 2002 16:48:53 GMT
Subject: dynamic includes from classes...
Message-ID: <Vz3g9.246505$kp.845705@rwcrnsc52.ops.asp.att.net>

Okay two questions...

First, is there a better way than exec to dynamically include a module?
i.e.
importvar = "foo"
exec("from %s import %s" % (importvar, importvar))

And, if there isn't, I tried the above code in a class that I wrote.  But,
it seems that the imported class foo is only visible inside the class, if I
send importvar to another function and do

a = eval(importvar)(args) #btw is there a better way to do this too?

it gives the error "Foo" undeclared.

Any suggestions on how to do this whole thing better, or how to make the
exec include global?

mike




From jdavis at empires.org  Tue Sep 24 21:37:09 2002
From: jdavis at empires.org (Jeff Davis)
Date: Wed, 25 Sep 2002 01:37:09 GMT
Subject: readlines() and "binary" files
References: <3d90cacf$0$30457$afc38c87@sisyphus.news.be.easynet.net>
Message-ID: <9r8k9.46422$V7.11170409@twister.socal.rr.com>

I think readlines() is just a shortcut for a very common task. Since your 
task isn't quite as common, I think it would be a better idea to use 
read() to read the whole thing, splitting the lines up by the 0x0d 0x0a 
pair (CR NL).

If it's a really large amount of data you can try to process it in chunks.

Regards,
        jeff davis

Justin wrote:

> Hi,
> 
> I have excel data with occasional multi-line fields,
> which when dumped to CSV translates to embedded CR's
> within a line, whereas the records/lines themselves
> are delimited by the CR+NL pair (this is MS-land).
> What I'd like to do is read those files and split every
> line apart on the semi-colon field separator. But it
> seems that whether the file is opened as text or not,
> (x)readlines() still considers the lone CR as a line
> delimiter and so not all my lines end up with the same
> number of fields as they should. Is there a way to handle
> this, or is readlines just not meant to work with anything
> but proper text files?
> 
>     f = open('testdata.csv','rb')
>     for line in f.xreadlines():
>         fields = line.split(';')
>         print len(fields) # should always be the same value



From whisper at oz.net  Tue Sep 17 03:31:49 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 17 Sep 2002 00:31:49 -0700
Subject: More on Protecting Source Code
In-Reply-To: <3d86b4a0$0$23173$afc38c87@news.optusnet.com.au>
Message-ID: <GCEDKONBLEFPPADDJCOEMEEBFAAA.whisper@oz.net>

> > Python may be easy to use, but if it's also easy to steal
> software written
> > in it, that's an impediment to it's being widely used in commercial
> > products, especially shrink-wrapped products. Perhaps that's a
> reason why
> > there seems to be so few commercial products in Python, beyond
> those that
> > are fairly closely licensed and/or have substantial parts of the app
> written
> > in C/C++.
>
> but if you're worried about protecting your IP, then wouldn't your package
> be under a closed license?
>
> > I suppose one solution is to modify the Python interpreter with
> different
> > op-codes and that ought to make it somewhat painful for the average
> hacker.
> > A better solution is to make a .pyc file approximately as hard
> as a binary
> > .exe file to decompile - however that could be done.
>
> From what I've read of that thread, the gist of it was
>
> "once its released, a determined person can, to some extent get your code"
>
> and
>
> "Security through obscurity is no security against a determined attacker"
>
> So, someone who is willing to hacka nd hack and hack away at your
> code, can
> get at it, and this is true for things coded in assembler, C, Python or
> Befunge. Most people aren't determined attackers, they're simply
> Joe Average
> who needs to get a  job done.
>
> An operating system is a complex beast, compiled into machine
> code its even
> more complex, how many people out there are using an unlicensed copy of
> Windows or Office? they didn't decompile, that would take skill, they just
> copied it bit for bit and posted the registration key.
>
> Do a very very simple risk analysis. Who is your target audience, what is
> the price of your software, will people bother going through your
> bytecode,
> or will they just post the registration key to a website? Will stolen
> software lead to a loss in revenue - or will those pirate users be people
> who would never have bought it in the first place? (yes, the pirate users
> are annoying, and you would feel ripped off, but OTOH you haven't lost
> money, and you've gained a bunch of users)
>
> Steven

Well, I think of it this way: machine coded binaries are more like a 128 bit
key and Python is more like a 40 bit key. I agree that nothing is safe from
reverse engineering; it's a matter of how much pain and money it takes to
unscrut the inscrutable. Python could do better - how I'm not totally sure.

Dave LeBlanc
Seattle, WA USA




From jdavis at empires.org  Tue Sep 24 05:31:37 2002
From: jdavis at empires.org (Jeff Davis)
Date: Tue, 24 Sep 2002 09:31:37 GMT
Subject: how to build up a guest book with python ?
References: <mailman.1032854050.26863.python-list@python.org>
Message-ID: <ZhWj9.39386$V7.10446716@twister.socal.rr.com>

http://python.org/doc/current/lib/module-cgi.html

might help you out. It's not the best tutorial out there I'm sure, but 
hopefully it can get you started.

Regards,
        Jeff

black wrote:

> 
> Hi,all~
> 
>     I'm learning python for almost 1 month and decide to build up a guest
>     book with it, but I really know nothing about web programming, any
>     tutorials for me please ?
> 
> 
> 
> ---------------------------------
> Do you Yahoo!?
> New DSL Internet Access from SBC & Yahoo!



From grabiller at 3dvf.net  Wed Sep 25 12:03:35 2002
From: grabiller at 3dvf.net (Guy Rabiller)
Date: Wed, 25 Sep 2002 18:03:35 +0200
Subject: Newbie References Question -> Exemple
References: <3d91c72f$0$8558$79c14f64@nan-newsreader-02.noos.net> <3d91d2cf$0$1484$79c14f64@nan-newsreader-01.noos.net> <3d91d7ab$0$13192$79c14f64@nan-newsreader-03.noos.net>
Message-ID: <3d91de4e$0$1485$79c14f64@nan-newsreader-01.noos.net>

Sorry I should have said:

Thanks to Gustavo, not sismex :-)

--
guy rabiller
3d animator / td
grabiller at 3dvf.net
http://grabiller.3dvf.net

Guy Rabiller wrote:
> Ok I solved it, thanks to sismex.
>
> If I want to change a coordinate, I have yo write:
>
> p[0][:] = [5,5]
>
> and not
>
> p[0] = [5,5]
>
> I understand that in the first case, the reference pointed ( via
> dictionary ) by p[0] is lost, a new value is created and rebinded to
> p[0], so quad and p[0] are not synchonised anymore.
>
> In the second case, the reference pointed by p[0] is updated with a
> new value. p[0] and quad are still synchronised.
>
> Is my understanding correct ?
>
> Thanks a lot to all.
>
>
> Guy Rabiller wrote:
>> Ok,
>>
>> thanks for your answers, I understand now that a variable name is
>> 'just' a name in a dictionary, and not a container.
>>
>> What I want to do is simple, and I suppose that this is my 'way of
>> thinking' regarding Python that is wrong.
>>
>> I have a list of points, and a list of polygons that should contain
>> 'references' to points.
>>
>> So for me:
>> -> points list:
>> p = []
>> p.append([0,0])
>> p.append([0,1])
>> p.append([1,1])
>> p.append([1,0])
>> -> polygon:
>> quad = []
>> quad.append(p[0])
>> quad.append(p[1])
>> quad.append(p[2])
>> quad.append(p[3])
>>
>> But now, if I change the coordinate of one point:
>> p[0] = [5,5]
>> I will have:
>>>>> p
>> [[5, 5], [0, 1], [1, 1], [1, 0]]
>> but:
>>>>> quad
>> [[0, 0], [0, 1], [1, 1], [1, 0]]
>>
>> For me the quad list of point is not 'updated' with the change of one
>> of its point.
>>
>> What I want to do is that, quad keeps a list of 'references' to
>> points coordinates.
>> Then if I change one point coordinates, the change is reflected in
>> the quad list of points.
>>
>> I have no clue how to do this.
>>
>>
>> Guy Rabiller wrote:
>>> Hi,
>>>
>>> let say I have:
>>> i1 = 1
>>> i2 = 2
>>> and
>>> p = [i1,i2]
>>>
>>> How can I have:
>>> p=[*i1,*i2]
>>> rather than
>>> p=[**i1,**i2] as it is curently ?
>>>
>>> ( Sorry for this nasty hybrid syntax )
>>>
>>> Textualy, how if I want that p[0] refere to the 'i1' container, and
>>> not to the refererence it contains ?
>>>
>>> What I want is that if now I set:
>>> i1 = 4
>>> that automaticaly:
>>> p -> [4,2]
>>> and not keeping [1,2]
>>>
>>> Thanks in advance.




From sebastian.wangnick at eurocontrol.int  Mon Sep  2 03:44:27 2002
From: sebastian.wangnick at eurocontrol.int (Sebastian Wangnick)
Date: 2 Sep 2002 00:44:27 -0700
Subject: stylistic question -- optional return value
References: <mailman.1030608587.4698.python-list@python.org>
Message-ID: <30292041.0209012344.ab2d4@posting.google.com>

> Andrew Koenig <ark at research.att.com> wrote:
> 
> > Hans> Still, I believe that rewriting the function (or splitting it up)
>  
> > Hans> might be a better idea. <0.3 wink>
> >
> > I've considered that, and I really don't think it works in this case.
> >
> > You might think of the optional result as being similar to debugging
> > information -- you really don't want to put the code that getnerates
> > debugging information into a separate function, because then it will
> > surely diverge from the code about which it is providing the informatio

If I understand you correctly then you'd like to be able to provide,
under circumstances determined by your function, additional
information back to your caller. Additionally, an "innocent" caller
should not be bothered with the additional information.

Why not create an additional optional *argument* (say a list, or a
function), to which you optionally provide your debug information?

def func(x,y,debug=None):
   if debug!=None: debug("func")
   return x+y

e = []
z = func(3,4,lambda info:e.append(info))
print z,e
 
Regards,
Sebastian

PS: I learned Python only last week, so take my posting with a grain
of salt.


From ahmad at NOSPAMbitbuilder.com  Mon Sep  9 07:06:01 2002
From: ahmad at NOSPAMbitbuilder.com (Ahmad Baitalmal)
Date: Mon, 09 Sep 2002 11:06:01 GMT
Subject: xml.minidom is stripping out my CRLF's in attrib values!!
Message-ID: <3D7C80C4.9070204@NOSPAMbitbuilder.com>

Hi,
I'm parsing an XML string which has nodes, the nodes has attributes, the 
attribute values have CRLFs. The string has the CRLFs before I parse,
After I create my DOM the values come back with no CRLFs at all!
What's the deal, is there a setting I can set?
Thanks
-- 
-  - -- ----  ----------------------------------------- --- -- -   -
*Ahmad Baitalmal***
*BitBuilder***
web: http://www.bitbuilder.com
-  - -- ----
-------------------------------------------------------------- --- -- -   -



From martin at v.loewis.de  Fri Sep  6 12:11:10 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 06 Sep 2002 18:11:10 +0200
Subject: help needed creating xml document with xml.dom
References: <m34rd33za5.fsf@localhost.localdomain>
Message-ID: <m31y87umht.fsf@mira.informatik.hu-berlin.de>

Brian Lalor <blalor at insight.com> writes:

> Good morning, all.  Several times, now, I've used xml.dom.minidom to create
> raw XML documents, but I don't think I've ever really done it the "right"
> way.  

I think that trying to create valid XML documents with DOM is
inherently wrong. Use print statements instead.

> I blame that on my seeming inability to find documentation on the
> "right" way to do it in Python, or in any language, for that matter.
> Pointers to said documentation are welcomed. :-)

Unfortunately, the DOM does not provide a way to create a Document
with an internal subset.

> Using minidom, I cannot seem to get either the namespace or doctype
> to print out.

Adding namespaces should be possible; adding a DOCTYPE declaration is
not.

>     # is this the right way, instead of directly importing xml.dom.minidom?
>     DOM = xml.dom.getDOMImplementation()

Yes, that's correct.

>     index_page.documentElement.setAttribute('xml:lang', 'en')
>     index_page.documentElement.setAttribute('lang', 'en')

If you want namespaces, you should use the namespace API:
setAttributeNS, createElementNS.

> Is this a bug in minidom, or (more likely) am I doing something wrong?  

Call it a bug. It's really a feature missing in DOM - while it *may*
work for an external DTD, there is no way to get it work with an
internal subset.

Again, using plain print statements will be easier.

Regards,
Martin


From duncan at NOSPAMrcp.co.uk  Mon Sep  2 04:47:05 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Mon, 2 Sep 2002 08:47:05 +0000 (UTC)
Subject: Executable Python Programs
References: <1rtc9.108$LI2.7238@newsread2.prod.itd.earthlink.net> <mailman.1030908003.28854.python-list@python.org>
Message-ID: <Xns927D61C17CF84duncanrcpcouk@127.0.0.1>

Gerhard H?ring <gerhard.haering at gmx.de> wrote in 
news:mailman.1030908003.28854.python-list at python.org:

> My personal opinion is that py2exe/Installer/etc. are for
> intermediate/advanced Python programmers. The reason is that it's
> sometimes difficult to fix the dependencies manually, if these tools are
> unable to hunt them down automatically for you.

I think you may be overstating the difficulty here. The vast majority of 
programs run through py2exe etc. with no problems at all. I can only think 
ot a couple of times that I have even had to think about the dependencies:

I had a case recently with py2exe was when it found some extra dependencies 
that I didn't want: I was using jabber.py which for not very good reasons 
imports site.py which indirectly imports a whole load of stuff such as 
tkinter. Explicitly excluding site.py fixed this. Admittedly it took me a 
little while to track down which import was the problem. An option to 
py2exe to dump out why it decided to pull in each module would have been 
helpful.

The only time I had trouble with an import that wasn't being found was when 
I used py2exe on a program using PIL. PIL demand loads a lot of converters 
for different graphics formats which was beyond py2exe's capabilities. 
Explicitly importing all the formats I wanted into the main program fixed 
that one.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From jb at cascade-sys.com  Sat Sep 14 04:09:34 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sat, 14 Sep 2002 01:09:34 -0700
Subject: threads or queue for this task
References: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com>
Message-ID: <3D82EEBE.5050405@cascade-sys.com>

robin at execulink.com wrote:

>I wish to have a server continually cycle (until interrupted by
>keyboard, signal, or whatever). It needs to perform several (say 6)
>different tasks, each of which may take significantly different
>lengths of time to perform.
>
I can think of two approaches that may solve your problem.

It sounds like you have a single stream of input requests which you wish 
to distribute among multiple service threads.

One way is to have your main thread read the input stream, parse the 
incoming commands and then launch a separate thread to handle each 
incoming request.  It would NOT wait for the threads to complete or 
otherwise try to regulate the number of threads.  Plus side is maximum 
parallelism.  Down-side is if commands come too fast you may exceed the 
maximum thread count and the app will fail.

Another alternative is to launch a single thread for each of your 
service routines.  Also create a Queue for each thread and have the 
threads each read commands from their respective queue.  Service threads 
would loop forever, reading and executing commands.  Main program reads 
input as before and feeds commands to the appropriate thread VIA each 
threads' respective queue.  Again, the main thread does NOT wait for any 
threads to complete.  Thus it will service requests as fast as possible 
using exactly N helper threads.  The degree of parallelism 0..N would 
depend on the input commands arrival rates.

If this is unacceptable -- if you need to enforce some ordering on the 
execution of commands, then things get a lot more complicated.  E.g., if 
you want to dispatch N commands, one to each helper thread, and then 
wait for them all to complete before doing anything else -- you need to 
consult the "threading" module (the one following "thread" in the 
library ref). If it's truly that simple (read N and then execute N in 
parallel) your main program could enforce this discipline, adding a 
threading.Semaphore() object to synchronize on the completion of all N 
threads.  Threading includes various synchronization functions that 
allow you to do this and more.  E.g., you also could do things like 
maintain a maximum of M[i] parallel instances of thread i (more than 1 
but a different limit for each command).

If the threads themselves access common data structures then you need to 
synchronize their access so they don't garble the results.  The thread 
module itself has a locking mechanism that's sufficient for implementing 
regions of mutual exclusion.  

If you're new to concurrent programming then you may need to consult 
some books on the subject, as it's a bit tricky.

Regards

--jb


-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From eugene1977 at hotmail.com  Sun Sep 15 21:31:01 2002
From: eugene1977 at hotmail.com (eugene kim)
Date: Mon, 16 Sep 2002 01:31:01 +0000
Subject: two or more whitespace regular expression?
References: <am12a3$qo0$1@newsreader.mailgate.org> <X5Ug9.40826$gf6.1295831@bin4.nnrp.aus1.giganews.com>
Message-ID: <am2isf$o1n$1@newsreader.mailgate.org>

thank you that worked..



From manuel at hendel.net  Wed Sep 25 03:54:16 2002
From: manuel at hendel.net (Manuel Hendel)
Date: Wed, 25 Sep 2002 09:54:16 +0200
Subject: problems defining an empty list
In-Reply-To: <200209250040.12209.shalehperry@attbi.com>
References: <20020925070654.GB57911@partagas.as.de.cw.net> <200209250040.12209.shalehperry@attbi.com>
Message-ID: <20020925075416.GC57911@partagas.as.de.cw.net>

On Wed, Sep 25, 2002 at 12:40:12AM -0700, Sean 'Shaleh' Perry wrote:
> it sounds like you have a hidden piece of whitespace.  Perhaps the code is 
> indented when it should not be?  If the code is not in a loop, class, if 
> structure, etc then it should not be indented.

Actually it is in a ... structure.
[....]
# empty list for lines per all domains
linesperdomains = []

for domain in domains:
    # empty list for lines
    domainlines = []
    for neededline in neededlines:
        if domain == neededline[0]:
            domainlines.append(neededline)
    print len(domainlines)

    # empty list for lines per domain
    linesperdomain = []
    linesperdomain = [domain, domainlines]
    print linesperdomain
[....]

This ist just a part of the script.

Manuel
-- 
I am two with nature. 
-Woody Allen 



From phr-n2002b at NOSPAMnightsong.com  Sun Sep 29 23:19:33 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Sep 2002 20:19:33 -0700
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05> <an7bvd$burft$1@ID-125932.news.dfncis.de> <an7p7t$cpg$0@216.39.172.122> <mailman.1033338759.6775.python-list@python.org> <7xd6qwqtgz.fsf@ruckus.brouhaha.com> <mailman.1033352259.30842.python-list@python.org>
Message-ID: <7xelbcgo56.fsf@ruckus.brouhaha.com>

"Chris Gonnerman" <chris.gonnerman at newcenturycomputers.net> writes:
> > >>> (.70 * .05) + .000000001
> > 
> > rounds to the right thing.
> 
> Cool.  Now PROVE that's right in all cases.

What do you mean by "right"?  What do you mean by "all cases"?

> Like I said, decimal arithmetic is STILL the only choice.

I'm not convinced decimal arithmetic is "right in all cases".


From mhammond at skippinet.com.au  Wed Sep 18 08:57:58 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 18 Sep 2002 12:57:58 GMT
Subject: PythonScript
References: <mailman.1032321023.29261.python-list@python.org>
Message-ID: <qL_h9.19471$Ee4.59690@news-server.bigpond.net.au>

Gavin Tomlins wrote:
> Greetings, 
> 
> Firstly apologies if this is slightly out of the scope of the mailing list
> but after searching the web, archives I've been unable to find references of
> PythonScript with its use with the activescript engine. I own March Hammond
> & Andy Robinson's Python Win32 Programming and have gone over there chapter
> of active scripting though there examples focus more on python and
> activescripting via the command shell. 
> 
> The problem I seem to be hitting is utilising python in applications that
> embed an ActiveScript Engine. To date I am unable to get one python script
> to execute. Other languages vbscript, jscript,perl etc, work in the same
> environment so I'm left wondering if its a configuration problem or
> something that I might be overlooking. If anyone could give me pointers,
> url's, snippets etc. it would be interested. The product I specifically
> having problems is FinalBuilder (http://www.atozedsoftware.com) though have
> seen this problem in other applications as well.
> 
It is likely to be some strange interaction between PythonScript and the 
host.  Python works fine in all MS hosts, and quite a number of 3rd 
party ones.  What application is it?

Mark.



From arosado at softhome.net  Sat Sep  7 20:59:15 2002
From: arosado at softhome.net (Andres Rosado)
Date: Sat, 07 Sep 2002 20:59:15 -0400
Subject: Why Python?
Message-ID: <5.1.0.14.0.20020907205827.00bbcc38@mail.softhome.net>

At 05:51 AM 9/4/2002 -0400, you wrote:
>I just begun to learn how to program. And I choosed Python for
>doing that. surfing around the web i found an article about this
>language, they said it's the ideal language for beginners to the
>world of programming. and i admit it, the syntax is very clear making
>the code very readable, i learn easily the concepts of programming
>and i soon wrote some nice programs. but since i finished my
>newbie-tutorial, i am wondering about one question: is it worth to go
>deeper into python? i talked to a friend of mine, he told me that i
>should try php..
>what should i do?? ^_^ in other words: what is python good for?
>why should i continue with python?

Just because you learn something new, it doesn't mean that you should 
forget the "old". Programming languages are like tools for a carpenter. 
Just because you got a new hammer, you don't have to throw the old 
screwdriver. Each programming language has it advantages and disadvantages. 
Learning to use them correctly is as important as learning one. Definetely 
I won't use Python for any low level programming (C is great here). But for 
other areas, Python can shine like (or better) than others.

Also, I started recently to learn PHP. I wouldn't recommend it for a newbie 
at programming, unless he/she needs it.


-----------------------------------
Andres Rosado
Email: andresr at despammed.com
ICQ: 66750646
Homepage: http://andres980.tripod.com/

Life is a yo-yo, and mankind ties knots in the string.




From jwbaxter at spamcop.net  Sat Sep  7 12:08:08 2002
From: jwbaxter at spamcop.net (John Baxter)
Date: Sat, 07 Sep 2002 09:08:08 -0700
Subject: None or 0
References: <k7ip4-0p3.ln1@junker.stroeder.com>
Message-ID: <jwbaxter-E9C054.09080807092002@corp.supernews.com>

In article <k7ip4-0p3.ln1 at junker.stroeder.com>,
 Michael Stroder <michael at stroeder.com> wrote:

>  >>> repr(0 or None)
> 'None'
>  >>> repr(None or 0)
> '0'
>  >>> repr('' or None)
> 'None'
>  >>> repr(None or '')
> "''"
>  >>>
> 
> Is it guaranteed to work like this or should that be avoided?

It seems very convenient to implement or (and and) by returning the 
first object which decides the answer.

My guess is that enough people are counting on the behavior that even if 
the underlying hardware or implementation language were to shift enough 
to make this implementation quite inconvenient, it would nonetheless 
survive.

Having said that, I tend not to like counting on it, because it 
stretches the extra-Python definition of the operators quite a bit.  
Particularly on the "and" side:

>>> print 5 and 1
1

Go into (English-speaking) classrooms, and you're going to hear chants 
of "5 and 1 is 6" not "5 and 1 is 1".

  --John


From TomH at optiscan.com  Tue Sep  3 20:31:48 2002
From: TomH at optiscan.com (Tom Harris)
Date: Wed, 4 Sep 2002 10:31:48 +1000 
Subject: Excluding specific module from  import
Message-ID: <7796875278c69d8306b528e1cf7d4f5d3d755479@optiscan.com>

Greetings

I have a library that tries to import mx.DateTime, and on failure uses its
own implementation. The problem is that I am building an executable, and I
would much rather not have mx imported, as it increases the size of the
executable by a factor of two. Could someone give me an idea how to frob
things so that import will fail to import mx.*. I cannot frob sys.path as it
only contains a path to site-packages. I would sooner not rewrite the
library.
Thanks

Tom Harris, Software Engineer
Optiscan Imaging, 15-17 Normanby Rd, Notting Hill, Melbourne, Vic 3168,
Australia
email tomh at optiscan.com     ph +61 3 9538 3333  fax +61 3 9562 7742

This email may contain confidential information. If you have received this
email in error, please delete it immediately,and inform us of the mistake by
return email. Any form of reproduction, or further dissemination of this
email is strictly prohibited. Also, please note that opinions expressed in
this email are those of the author, and are not necessarily those of
Optiscan Pty Ltd.






From marklists at mceahern.com  Fri Sep 13 22:45:14 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 13 Sep 2002 21:45:14 -0500
Subject: Restricting methods in derived classes
In-Reply-To: <slrnao4psk.7ej.huaiyu@gauss.almadan.ibm.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOIEKHDOAA.marklists@mceahern.com>

[Huaiyu Zhu]
> But seriously, your first solution actually contains all I needed.  Here's
> what I adopted in the end:

[snip]

> Maybe there is a simple way to wrap up restrict() in a metaclass, so that
> instead of restrict(foo) we can say class foo(dict, restricted): ...
> I'm still not so used to metaclasses to figure that out, however.

I like the solution you adopted and I used it to simplify (imo, anyway) my
metaclass approach (see below), which I still prefer, but I'm not quite sure
why.  One thing I don't quite understand is why you'd want to bother messing
with magic methods at all (which you're _common suggests you do).  That
seems like an invitation to disaster--but that's a statement that comes more
from ignorance than experience on my part.  In the code I attached, if you
wanted to suppress magic methods, you would simply remove the is_magic test.

Thanks for this opportunity to improve my understanding of metaclasses.

Cheers,

// mark

#!/usr/bin/env python

import inspect

def make_undefined(cls, attr):
    def undefined(self, *args, **kwargs):
        msg = "'%s' object has no attribute '%s'" % (cls.__name__, attr)
        raise AttributeError(msg)
    return undefined

def is_magic(attr):
    prefix = suffix = '__'
    return attr.startswith(prefix) and attr.endswith(suffix)

class Restricted(type):

    common = ['__class__', '__defined__', '__dict__', '__doc__',
              '__getattribute__', '__init__', '__new__']

    def __new__(cls, classname, bases, classdict):
        defined_key = 'defined'
        defined = classdict.get(defined_key, []) + Restricted.common

        # If a non-magic attribute from any base class is not in common or
        # defined, hide it with a method that raises a descriptive
        # AttributeError (mimicking Python).
        for base in bases:
            for k, v in inspect.getmembers(base):
                if not is_magic(k) and k not in defined:
                    classdict[k] = make_undefined(cls, k)
        return type.__new__(cls, classname, bases, classdict)

class RestrictedDict(dict):

    __metaclass__ = Restricted

    defined = ['put', 'get', 'keys', 'items', 'values']

f = RestrictedDict()
print f.keys
print f.keys()
f['a'] = 'b'
print f.keys()
try:
    f.clear()
except AttributeError:
    pass
print f.keys()

-




From aleax at aleax.it  Tue Sep 17 18:39:06 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 17 Sep 2002 22:39:06 GMT
Subject: Will "Python in a Nutshell" be too short?
References: <mailman.1032288261.15607.python-list@python.org> <LMNh9.15079$T4.1903@fe04>
Message-ID: <eaOh9.123493$pX1.4445629@news2.tin.it>

Steve Holden wrote:
        ...
> I wil be very surprised if Alex' book does not become the canonical
> reference for the language, since his treatment is exceptionally thorough
> and (even for him) more than usually precise.

Thanks -- I do appreciate this.

>> packages that had to be left out of the book's 380 pages (htmllib,
>> telnetlib, urllib2, CongigParser, unittest, tkinter, xml, etc.).
>>
> They clearly aren't "Essential" :-)

It's a hard judgment in each and every case.  I think unittest is (is it
ever!), for example -- somebody not quite as keen on testing might
not agree with me though.

> It's far too late into the production process for such pleas to be
> effective, no matter what the individuals concerned may think of their
> merits. It's probably at the printers already...

Not quite (and the delays of SOME technical reviewers in supplying
the technical reviews MIGHT be part of it... hint, hint...), but it IS in
the final-editing stage.


Alex



From imbosol at vt.edu  Mon Sep 30 14:03:43 2002
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 30 Sep 2002 14:03:43 -0400
Subject: rank-0 arrays?
References: <3d988160.79509684@news>
Message-ID: <ana3o6$1u8$1@solaris.cc.vt.edu>

brobbins333 at shaw.ca wrote:
> In Numeric Python, what are rank-0 arrays?
> Must not arrays have at least 1 dimension?

Well, in mathematics, a rank-0 array is a scalar, which is considered
a degenerate array.  The reason rank-0 arrays exist is because there
are some operations that produce arrays of various rank depending on
the rank of the operands, and in many cases these operations can
produce arrays of rank 0.

For example, consider cartesian tensors, which can be represented by
Numeric arrays.  Cartesian tensors have an operation called inner
product, which multiplies a rank-n tensor by a rank-m tensor to
produce a tensor of rank abs(n-m).  (I'm probably simplifying this too
much.)  Anyways, suppose you want the inner product of two rank-1
tensors (i.e., vectors).  The result is a rank-0 tensor, i.e., a
scalar.

Another example is a trace operation, which takes a tensor of rank n,
and produces a tensor of rank (n-2).  The trace of a rank-2 tensor is,
as with matrices, a scalar.

Evidently, Numeric supports rank-0 arrays to allow these operations to
be done consistently, without special exceptions to handle degenerate
cases.

However, IIRC, it is good advice not to play around too much with
rank-0 arrays in Python.  They should be converted to regular Python
floats if you're going to do something non-Numeric with them.



-- 
CARL BANKS
http://www.aerojockey.com


From jkraska at san.rr.com  Sat Sep 28 00:11:05 2002
From: jkraska at san.rr.com (Courageous)
Date: Sat, 28 Sep 2002 04:11:05 GMT
Subject: What does Python fix?
References: <e15b6fa3.0209271418.14cd5642@posting.google.com>
Message-ID: <flaapucbduj6g4ifrdf3oq99hdrepi1o3v@4ax.com>

>I'm curious to read a bit more about why Tim Peters (and presumably
>others) think that Lisp-inspired languages (even non-prefix ones) are
>doomed to eternal obscurity.  Would anyone care to comment and/or give
>me some pointers to commentary on the subject? 

Shrug. Lisp trades one form of complexity for another. By reducing
the solution to a universal grammar, they increase the symbolic load
on the user, who is awash in a page full of very similar symbols.
This is intimidating in particular to novices, for whatever reason,
who are instantly alienated. Without sweeping in novices, you lack
a grass roots movement, and without a grass roots, a language is dead.

The parts of lisp that really matter to the community have since
been adopted in part by many other languages. Lisp is dead. Long live
Lisp.

C//



From mcherm at destiny.com  Tue Sep 17 12:58:59 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Tue, 17 Sep 2002 12:58:59 -0400
Subject: Q: handling exceptions
References: <fa576652.0209170827.4ba29201@posting.google.com>
Message-ID: <am7mvu$23e$1@main.gmane.org>

Karsten Weinert wrote:
 > The problem is: how can I access the
 > message string of the exception?

Does this help?

Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help
 >>> def raiseit():
	raise Exception, 'This is the message string'

 >>> try:
	raiseit()
except Exception, s:
	print "Message is '%s'." % s

	
Message is 'This is the message string'.
 >>>

-- Michael Chermside






From duncan at NOSPAMrcp.co.uk  Tue Sep 24 05:07:31 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 24 Sep 2002 09:07:31 +0000 (UTC)
Subject: Three dumb questions (ordered by dumbness descending)
References: <amnvit$7kqdt$1@ID-77524.news.dfncis.de>
Message-ID: <Xns9293662F7B64Dduncanrcpcouk@127.0.0.1>

Thorsten Kampe <thorsten at thorstenkampe.de> wrote in news:amnvit$7kqdt$1 at ID-
77524.news.dfncis.de:

> 1. Why is 'zip(zip(x)) != x' (should be the same because it's the 
> transposed)

Because zip(x) doesn't transpose x. BUT, zip(*x) does transpose x.
So the question is what zip(*zip(*x)) does?

>>> x = [(1, 2), (3, 4), (5, 6)]
>>> x == zip(*zip(*x))
1

Provided x is a list of tuples, and you expand the list into separate 
arguments when calling zip, it is indeed its own inverse.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From bloke at ii.net  Sat Sep 28 02:42:42 2002
From: bloke at ii.net (Rob Hall)
Date: Sat, 28 Sep 2002 14:42:42 +0800
Subject: urllib2.urlopen(req).info() object
References: <an3fhi$3jr$1@news.iucc.ac.il>
Message-ID: <3d954f50$0$11909@echo-01.iinet.net.au>

I used ClientCookie.

Rob.


eli li <elimli at yahoo.com> wrote in message
news:an3fhi$3jr$1 at news.iucc.ac.il...
> The call "urllib2.urlopen(req).info()" returns an object similar to the
> dictionary, but it can have muliple values for some key. For example:
>
> Content-Type: text/html
> Set-Cookie: A=M
> Set-Cookie: WEBTRENDS_ID=192.168.120.21
> Set-Cookie: SINFO=ca9e
> Via: 1.1 colosus1 (NetCache NetApp/5.3R2D4)
>
> How can I retrive all values related to 'Set-Cookie' key?
>
>




From sholden at holdenweb.com  Thu Sep 26 08:40:19 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 26 Sep 2002 08:40:19 -0400
Subject: Getting started
References: <uojinmcfapjf45@corp.supernews.com> <mailman.1032442884.18747.python-list@python.org> <H3mi9.133635$pX1.4751676@news2.tin.it> <mailman.1032567390.29374.python-list@python.org> <AFik9.51323$b%3.25248@fe03> <mailman.1032979596.18965.python-list@python.org> <3D920950.323B25B3@alcyone.com> <mailman.1033012423.27861.python-list@python.org>
Message-ID: <GjDk9.5416$Lu6.2783@fe04>

"James J. Besemer" <jb at cascade-sys.com> wrote in message
news:mailman.1033012423.27861.python-list at python.org...
>
[attempts to be "WORTHY" enough ...]
> You can lead a pig to pearls....
>
> Regards
>

but-you-can't-stop-it-being-a-swine-ly y'rs  - steve
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From bdesth at nospam.free.fr  Sun Sep  8 23:09:34 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Sun, 08 Sep 2002 23:09:34 -0400
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com>        <4f49d232.0209050418.7668c21@posting.google.com>        <al7ub8$crj$1@panix2.panix.com>        <3D77FB95.3070300@nospam.free.fr>        <al8amr$d3q$1@panix2.panix.com> <mailman.1031257511.21417.python-list@python.org>
Message-ID: <3D7C10EE.3000509@nospam.free.fr>

Skip Montanaro wrote:

[snip]

> Alas, to even get to a phone screen these days, it appears you need "at
> least five years experience" with a buzzword language on your resume.  In
> most cases, neither Lisp nor Python qualify.
> 

Which would make me laugh it wasn't true :(

laotseu



From edwardam at interlix.com  Sun Sep 22 23:10:42 2002
From: edwardam at interlix.com (Edward Muller)
Date: 22 Sep 2002 20:10:42 -0700
Subject: Python HTTPSConnection/HTTPResponse Problem
References: <9abedeb.0209212200.5c669a50@posting.google.com> <mailman.1032691649.13632.python-list@python.org> <9abedeb.0209221343.74d27ad1@posting.google.com>
Message-ID: <9abedeb.0209221910.7fb75f6f@posting.google.com>

I *think* I found the problem. Using m2Crypto and urllib I was able do
process the transaction through Authorize.net's ADC Direct Response
mechanism and it *looks* like their server replied first with a ..

HTTP/1.1 100 Continue

and then with a ...

HTTP/1.1 200 OK

I don't know if this is valid or not (and I sent the same info to
Authorize.net).

Maybe someone else with more HTTP knowledge will know. I'd be willing
to test any patches or provide more data if anyone is interested in
looking at this.

Thanks,
Edward Muller

edwardam at interlix.com (Edward Muller) wrote in message news:<9abedeb.0209221343.74d27ad1 at posting.google.com>...
> I thought I had posted this already, sorry if it sorta doubles.
> 
> I can't use python 2.2, the latest I can use is python 2.1 because in
> the end this will be done with zope and zope currently only supports
> python 2.1.3 and earlier.
> 
> And yes, my python is SSL enabled. 
> 
> Gerhard H?ring <gerhard.haering at gmx.de> wrote in message news:<mailman.1032691649.13632.python-list at python.org>...
> > * Edward Muller <edwardam at interlix.com> [2002-09-21 23:00 -0700]:
> > > I am trying to use HTTPSConnection and HTTPResponse objects [...]
> > 
> > Is your Python SSL-enabled?  Check with:
> > 
> >     import socket
> >     hasattr(socket, "ssl")
> > 
> > Which version of Python are you using? Be sure to use at least version
> > 2.2, as it contains many SSL-related bugfixes.
> > 
> > -- Gerhard


From chris.walsh at real-sense.com  Wed Sep 11 12:28:59 2002
From: chris.walsh at real-sense.com (Chris Walsh)
Date: 11 Sep 2002 09:28:59 -0700
Subject: Getting __builtins__ into my local dictionary
Message-ID: <1691b372.0209110828.31501d82@posting.google.com>

Hi All,

I know this may sound like I am repeating a much discussed topic but I
want to create a multiple environments within the one Python
interpreter so that each envronment contains it's own local defs and
attributes but also understands functions such as str() (e.g. contains
__builtin__ module).

This is how I think I should do it:

================START=OF=CODE================
	PyObject *pMyMod  = PyModule_New("unique000");
	PyObject *pMyDict = PyModule_GetDict(pMyMod);

	PyDict_SetItemString(pMyDict, "__builtins__", PyEval_GetBuiltins());

	PyObject *pResultT;
	pResultT = PyRun_String("dir()\n",
		Py_eval_input, pMyDict, pMyDict);
	DumpAndRelease("dir()", pResultT);

	pResultT = PyRun_String("dir(__builtins__)\n",
		Py_eval_input, pMyDict, pMyDict);
	DumpAndRelease("dir(__builtins__)", pResultT);
=================END==OF=CODE================

The results I get from my Dump module (which just throws out the
contents of pResultT) is:

================START=OF=DUMP================
dir()
	List of 3 items: ['__builtins__', '__doc__', '__name__']

dir(__builtins__)
	List of 37 items: ['__class__', '__cmp__', '__contains__',
'__delattr__', '__delitem__', '__doc__', '__eq__', '__ge__',
'__getattribute__', '__getitem__', '__gt__', '__hash__', '__init__',
'__iter__', '__le__', '__len__', '__lt__', '__ne__', '__new__',
'__reduce__', '__repr__', '__setattr__', '__setitem__', '__str__',
'clear', 'copy', 'get', 'has_key', 'items', 'iteritems', 'iterkeys',
'itervalues', 'keys', 'popitem', 'setdefault', 'update', 'values']
=================END==OF=DUMP================

But... if I do a "dir(__builtins__)" from the Python interpreter there
are 118 items (not 37) including amongst other stuff, the str()
function.

How do I create my own module/dictionary which contains the
__builtin__ module which contains all the handy functions that
"dir(__builtins__)" shows up from the interpeter?

Kind Regards,

Chris.


From g_will at cyberus.ca  Fri Sep 20 13:50:01 2002
From: g_will at cyberus.ca (Gordon Williams)
Date: Fri, 20 Sep 2002 13:50:01 -0400
Subject: httplib 'getaddrinfo failed'
Message-ID: <026b01c260ce$24e23850$7507e640@wnt20337>

Hi All,

I recently got a DSL line to connect to the internet and my code is now
broken.  Could this be due to the DSL being different from a dialup
connection?

I'm using python 2.2 on W2000.

This is the traceback from the 2.2 httplib:


Traceback (most recent call last):
  File "G:\Python Files\Mutual Fund Updater\GetPrice.py", line 103, in ?
    f = u.open(url)
  File "G:\PROGRA~1\PYTHON22\lib\urllib.py", line 178, in open
    return getattr(self, name)(url)
  File "G:\PROGRA~1\PYTHON22\lib\urllib.py", line 288, in open_http
    h.putrequest('GET', selector)
  File "G:\PROGRA~1\PYTHON22\lib\httplib.py", line 452, in putrequest
    self.send(str)
  File "G:\PROGRA~1\PYTHON22\lib\httplib.py", line 394, in send
    self.connect()
  File "G:\PROGRA~1\PYTHON22\lib\httplib.py", line 362, in connect
    for res in socket.getaddrinfo(self.host, self.port, 0,
socket.SOCK_STREAM):
IOError: [Errno socket error] (2, 'getaddrinfo failed')


This is the traceback from the httplib in CVS that I got today.

Traceback (most recent call last):
  File "G:\Python Files\Mutual Fund Updater\GetPrice.py", line 103, in ?
    f = u.open(url)
  File "G:\PROGRA~1\PYTHON22\lib\urllib.py", line 178, in open
    return getattr(self, name)(url)
  File "G:\PROGRA~1\PYTHON22\lib\urllib.py", line 292, in open_http
    h.endheaders()
  File "G:\PROGRA~1\PYTHON22\lib\httplib.py", line 695, in endheaders
    self._send_output()
  File "G:\PROGRA~1\PYTHON22\lib\httplib.py", line 581, in _send_output
    self.send(msg)
  File "G:\PROGRA~1\PYTHON22\lib\httplib.py", line 548, in send
    self.connect()
  File "G:\PROGRA~1\PYTHON22\lib\httplib.py", line 516, in connect
    socket.SOCK_STREAM):
IOError: [Errno socket error] (2, 'getaddrinfo failed')


Same problem.  Any ideas?  I'm using Sympatico high speed edition which was
developed by Efficient Networks Inc. - if that is any help.

Regards,

Gordon Williams




From maxm at mxm.dk  Mon Sep  2 08:46:20 2002
From: maxm at mxm.dk (Max M)
Date: Mon, 02 Sep 2002 14:46:20 +0200
Subject: Must have Python modules?
References: <_THc9.42153$vY2.1039887@e3500-atl2.usenetserver.com> <slrnan6k4s.eg.gerhard.haering@haering.opus-gmbh.net> <FCIc9.30407$FJ1.5585669@e3500-atl1.usenetserver.com>
Message-ID: <3D735D9C.9060607@mxm.dk>

Robert Oschler wrote:

> I want to know what the other favorites are "out there".

win32all (if os in windows), mx*, PIL, reportlab, Nummeric

+ the standard library :-)


regards Max M



From martin at v.loewis.de  Sat Sep 21 06:56:57 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 21 Sep 2002 12:56:57 +0200
Subject: Embedding Python in C++ and get rid of it again
References: <amhhcq$ur4$05$1@news.t-online.com>
Message-ID: <m3ofark3uu.fsf@mira.informatik.hu-berlin.de>

Manuel Klimek <manuel.klimek at web.de> writes:

> I'm coding on xpermud, a scriptable mud client.
> This one features multiple scripting languages
> (perl and python at the very moment). 

For problems like this, it is not that helpful if you report what your
program does, but instead, how you have organized your program. Like:

- I have a Solaris 8 32-bit binary foo, which I generated with gcc
  2.95.2.
- I also have linked libpython22.a to that binary.
- I have compiled the bar.so extension module
- when I try to import bar, I get the error message "foobar not found"

> So why does python want to have "hard"-exported symbols in .so
> libraries?

I don't understand the question. What is a "hard"-exported symbol, and
why do think Python wants it?

Also, why do you think any of the Python symbols interfere with Perl?

> Special question: Is there a different way to
> get the python modules find the symbols they
> need, some magic internal undocumented python
> function we could call on the C++ side...?

What symbols are you referring to?

Regards,
Martin


From duncan at NOSPAMrcp.co.uk  Tue Sep  3 05:05:33 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 3 Sep 2002 09:05:33 +0000 (UTC)
Subject: string reading bug in ActiveState Python 2.2 ?
References: <3D74757D.1010405@molgen.mpg.de>
Message-ID: <Xns927E647FE7259duncanrcpcouk@127.0.0.1>

Axel Kowald <kowald at molgen.mpg.de> wrote in 
news:3D74757D.1010405 at molgen.mpg.de:

> I'm using ActiveStates Python 2.2.1 under win2000 to read in the 
> attached text file and it seems that python is only reading part of the 
> string. After
> 
> fp = open("bla.txt")
> z = fp.read()
> print len(z), z
> 
> I see that only 11 of the 15 characters have been read. I know they are 
> non-printable characters, but that shouldn't matter, should it ?  Using 
> another python on one of our unix machines everything works as expected 
> and all 15 chars are read in.  Is that there something special about 
> windows that I should know, or is this a bug in ActiveStates Python ?

Python is doing exactly what you asked it to do.

Please read the documentation for the 'file' builtin (of which 'open' is an 
alias):

'''Modes 'r+', 'w+' and 'a+' open the file for updating (note that 'w+' 
truncates the file). Append 'b' to the mode to open the file in binary 
mode, on systems that differentiate between binary and text files (else it 
is ignored). If the file cannot be opened, IOError is raised.

If mode is omitted, it defaults to 'r'. When opening a binary file, you 
should append 'b' to the mode value for improved portability. (It's useful 
even on systems which don't treat binary and text files differently, where 
it serves as documentation.)'''

Windows is a system that differentiates between binary and text files, so 
without the 'b' in the mode argument you are asking the system to read the 
file in text mode. In text mode Windows will drop carriage return 
characters when followed by linefeed, and will stop reading when it hits a 
control-Z character.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From martin at v.loewis.de  Mon Sep 16 11:58:51 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Sep 2002 17:58:51 +0200
Subject: Mutable Stings
References: <mailman.1032181643.21256.python-list@python.org>
Message-ID: <m3admidiys.fsf@mira.informatik.hu-berlin.de>

"Griebel, Peer" <Peer.Griebel at entire.de> writes:

> So my question: Is this correct - or did I overlook something?

You are not overlooking anything: Your analysis is correct, the code
is wrong. Feel free to submit a patch to SF.

Regards,
Martin



From rnd at onego.ru  Tue Sep  3 13:30:57 2002
From: rnd at onego.ru (Roman Suzi)
Date: Tue, 3 Sep 2002 21:30:57 +0400 (MSD)
Subject: more Re: Peculiar ImportError traceback
In-Reply-To: <Pine.SOL.4.44.0209031755160.7655-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <Pine.LNX.4.44.0209032124330.7110-100000@rnd.onego.ru>

On Tue, 3 Sep 2002, John J. Lee wrote:

BTW, I had something similar with different modules:

Python 2.2.1 (#1, May 21 2002, 09:12:04)
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import email
>>> email.message_from_file(open("000.msg"))
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python2.2/site-packages/PIL/__init__.py", line 35, in 
message_from_file
    # ;-)
  File "/var/tmp/python2-2.2.1-root/usr/lib/python2.2/email/Parser.py", line 
41, in parse
  File "/var/tmp/python2-2.2.1-root/usr/lib/python2.2/email/Parser.py", line 
134, in _parsebody
  File "/var/tmp/python2-2.2.1-root/usr/lib/python2.2/email/Parser.py", line 
45, in parsestr
  File "/var/tmp/python2-2.2.1-root/usr/lib/python2.2/email/Parser.py", line 
40, in parse
  File "/var/tmp/python2-2.2.1-root/usr/lib/python2.2/email/Parser.py", line 
81, in _parseheaders
email.Errors.HeaderParseError: Not a header, not a continuation

- I have no idea why PIL/__init__.py is even mentioned here!
I am pretty sure email package can't depend on it in any way.

My sys.path is like:

>>> sys.path
['', '/usr/lib/python2.2', '/usr/lib/python2.2/plat-linux2', 
'/usr/lib/python2.2/lib-tk', '/usr/lib/python2.2/lib-dynload', 
'/usr/lib/python2.2/site-packages',
'/usr/lib/python2.2/site-packages/Numeric', 
'/usr/lib/python2.2/site-packages/PIL']


Something is broken.
(the directory /var/tmp/python2-2.2.1-root doesn't exist)

>>>> from mx import DateTime
>*** You don't have the (right) mxDateTime binaries installed !
>Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
>  File "mx/DateTime/__init__.py", line 8, in ?
>    from DateTime import *
>  File "mx/DateTime/DateTime.py", line 9, in ?
>    from mxDateTime import *
>  File "/usr/local/lib/python2.2/site-packages/reportlab/__init__.py", line 13, in ?
>    myDir = os.path.split(__file__)[0]
>ImportError: No module named mxDateTime
>>>> from mx.DateTime import mxDateTime
>>>>
>
>
>Two obvious problems:
>
>1. Bizarre last line of traceback involving reportlab (whose
>__init__.py I list in full below).  Reportlab and mxDateTime, AFAIK,
>are entirely unrelated!
>
>2. mxDateTime's __init__.py (mx/DateTime/mxDateTime/__init__.py) does
>an import mxDateTime, which then fails, causing the error message
>about mxDateTime's binaries (ie. mxDateTime.so) to be printed.
>Obviously, Python isn't finding the library, but I can't remember how
>extension modules are located, nor could I find anything in the Python
>docs on this -- could anyone point it out for me?
>
>Even better, explain why on earth I get that line about reportlab!  Up
>until now reportlab has sat in /usr/local/lib/python2.2/site-packages
>quietly minding its own business.  The egenix-mx-base `mx' directory
>(2.0.3) is also in the same site-packages directory.  Grepping through
>reportlab, I see that one of its demos making use of mxTextTools, but
>I don't think that's relevant.
>
>
>John
>
>#copyright ReportLab Inc. 2000-2001
>#see license.txt for license details
>#history http://cvs.sourceforge.net/cgi-bin/cvsweb.cgi/reportlab/__init__.py?cvsroot=reportlab
>#$Header: /cvsroot/reportlab/reportlab/__init__.py,v 1.16 2001/08/13 12:17:12 rgbecker Exp $
>__version__=''' $Id: __init__.py,v 1.16 2001/08/13 12:17:12 rgbecker Exp $ '''
>__doc__="""The Reportlab PDF generation library."""
>Version = "1.09"
>
>def getStory(context):
>    if context.target == 'UserGuide':
>        # parse some local file
>        import os
>        myDir = os.path.split(__file__)[0]
>        import yaml
>        return yaml.parseFile(myDir + os.sep + 'mydocs.yaml')
>    else:
>        # this signals that it should revert to default processing
>        return None
>
>

Sincerely yours, Roman Suzi
-- 
rnd at onego.ru =\= My AI powered by Linux RedHat 7.2
aaro at onego.ru <- not to be mailed to




From eric.brunel at pragmadev.com  Mon Sep 23 04:05:15 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 23 Sep 2002 10:05:15 +0200
Subject: Why does Tkinter do this?
References: <f9ff3f62.0209201019.81fb220@posting.google.com>
Message-ID: <ammhgp$hj8$1@wanadoo.fr>

Mark Wright wrote:

> I'm trying to create a resizeable Label with an image.  bind()ing to
> the configure event allows me to resize the image whenever the label
> is resized, but if I set the size of the image to the new size of the
> label, it is too big (which produces cascading 'configure' events and
> a steadily expanding Label).  OK, so I try subtracting the borderwidth
> - which is no good either.  But if I subtract the borderwidth *times
> two*, it works.  Can anyone explain this to me?

Errrr, because the border is on both sides...?
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From claird at starbase.neosoft.com  Fri Sep  6 14:23:24 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 6 Sep 2002 13:23:24 -0500
Subject: Project idea
References: <3d77c0b4$1@news.nz.asiaonline.net>
Message-ID: <84246AF492892BD2.EEDDE31C3CDD0FDF.E45EA739B6484CC6@lp.airnews.net>

In article <3d77c0b4$1 at news.nz.asiaonline.net>,
Colin Brown <cbrown at metservice.com> wrote:
>Hi
>
>If you are interested in a Windows platform project for automated testing
>then read on.
>
>There is an excellent freeware package "AutoIt" available from
>http://www.hiddensoft.com/AutoIt/
>that provides a DLL module that could be readily interfaced to Python.
>
>"AutoIt" is a simple tool that can simulate key presses, mouse movements and
>window commands (maximize, minimize, wait for, etc.) in order to automate
>any windows based task (or even windowed DOS tasks).
			.
			.
			.
To encourage those who are considering this opportunity, I
observe that <URL: http://wiki.tcl.tk/autoit > illustrates
a simple Tcl binding.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From fb at frank-buss.de  Sun Sep  1 13:09:46 2002
From: fb at frank-buss.de (Frank Buss)
Date: Sun, 1 Sep 2002 17:09:46 +0000 (UTC)
Subject: A "for" with "list" question.
References: <3b55ea60.0208311330.4608ae71@posting.google.com> <mailman.1030830407.5536.python-list@python.org>
Message-ID: <akthkp$5nb$1@newsreader2.netcologne.de>

Sean 'Shaleh' Perry <shalehperry at attbi.com> wrote:

> oddballs = []
> for item in b:     # iterate over one list
>     if item in a:   # test if it is in the other list
>         print item
>     else:
>         oddballs.append(item)
> 
> for item in oddballs:
>     print item

That's wrong, because you won't get 0, which is in a, but not in b. 
Derived from the nice solution from Graham I would write:

>>> a = [0,1,2]
>>> b = [1,2,3]
>>> inBothLists = [item for item in a if item in b]
>>> inOneListOnly = [item for item in a+b if item not in inBothLists]
>>> print inBothLists
[1, 2]
>>> print inOneListOnly
[0, 3]

-- 
Frank Bu?, fb at frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de


From martin at v.loewis.de  Thu Sep  5 03:38:23 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 05 Sep 2002 09:38:23 +0200
Subject: Is it okay to compile 2.2.1 with gcc 2.96?
References: <3D769ACE.7000801@nospam.yahoo.com>
Message-ID: <m3fzwoubrk.fsf@mira.informatik.hu-berlin.de>

Anton Graph <aglyportREMOvethispart at nospam.yahoo.com> writes:

> My question is this. Is gcc 2.96:
> 
> Reading specs from /usr/lib/gcc-lib/i386-redhat-linux/2.96/specs
> gcc version 2.96 20000731 (Red Hat Linux 7.3 2.96-112)
> 
> considered to be a stable compiler for the purpose of compiling 2.2.1?

Certainly, yes.

Regards,
Martin


From DennisR at dair.com  Wed Sep 25 13:39:51 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Wed, 25 Sep 2002 17:39:51 GMT
Subject: ANN: comments on Python Windows EXE Factory 
Message-ID: <Hxmk9.3450$6x3.49072397@newssvr14.news.prodigy.com>

I have put up a page at http://www.autorunning.com/arpyexe.cgi which allows
building Python CGI interface programs over the web.  Drop your Python code
into a 50,000 character capacity text and press a "submit" type button.
Your Windows EXE is built and can be downloaded.

The default EXE comes prebuilt with 24 of the most common lib modules
required.  It also comes with python22.dll built in so that the EXE runs on
any Win 98 or later system and does not require that Python be installed.
Up to 3 additional lib modules can be installed.  The size of the sample
download is 432 Kb.

This web page is a fully functional demo for a possible commercial product.
The commercial product would run entirely on your local machine (no
submitting your code to my web page) and is likely to have greater capacity.

I have tested this and know about the following limitations.  I would
appreciate help in fleshing out any more:

  1) Will not run on Win 95 which has never had IE 4.0 or greater installed
(shell problem).  That is why I only claim Win 98.

  2) Running program times out in Opera if run in MIDI mode (one or more
simultaneous documents partially fills client area).  It works fine if the
browser window is first maximized within the client area (not same as
maximizing Opera to fill entire screen and which is not required).

Have I picked the right libraries to include?  What else would you recommend
to build on this approach?

Thanks to the great Python community.

--

Dennis Reinhardt

http://www.dair.com




From sam-nospam- at totallynerd.com  Mon Sep 23 11:58:37 2002
From: sam-nospam- at totallynerd.com (Sam Alexander)
Date: 23 Sep 2002 08:58:37 -0700
Subject: Zope Documentation and example sites
Message-ID: <503511c2.0209230758.74a64a3d@posting.google.com>

Hi all,

I've just installed Zope, and now I'm trying to find reesources to
learn it.  I have a few questions along these lines...

1) What online resources are out there to help a newbie learn Zope? 
I've downloaded the Zope Book, but it's a bit technical for my taste. 
I'm looking for something with examples or tutorials showing how to do
some basic stuff from scratch.  Any suggestions?  Either online or
books ???

2) What sites are using Zope?  I've noticed most if not all the Zope
sites I've seen don't mention they're using Zope.  Can someone give me
some examples of sites which use Zope so I can get a better idea of
the type of layouts I can do with Zope?  I know it's very
programmable, but other then the Plone site (which is great), I'm not
sure what other sites use Zope.

3) Finally, I've never used Python, but I'm seasoned in Linux and in
web/app programming (Perl, PHP, C++, Cold Fusion, etc).  How long
would it take to learn Zope/python well enough to create a production
site?

Thanks for your help and suggestions.


From kodokuchef at yahoo.com  Fri Sep 20 08:25:48 2002
From: kodokuchef at yahoo.com (akirasugiura)
Date: 20 Sep 2002 05:25:48 -0700
Subject: newbie seeks advise.
References: <77c84db2.0209192137.51167ebb@posting.google.com> <mailman.1032501210.16162.python-list@python.org>
Message-ID: <77c84db2.0209200425.4c723344@posting.google.com>

Thank you very much. I didn't know there is such a "max" function.
("max"
 is a function isn't it). I also tried "min" to find a smallest
number(lol) and it worked. "list(data)" also worked. Thanks again.


From hst at empolis.co.uk  Fri Sep 20 07:06:22 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Fri, 20 Sep 2002 12:06:22 +0100
Subject: finding last match in a file
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB04F41D@hendrix.empolisuk.com>

> In article <87r8fplvq4.fsf at smtp.blueyonder.co.uk>, Keith 
> O'Connell wrote:
> > Dag <dag at animagicnet.no> writes:
> > 
> >> I have a number of very large files where I have to find 
> the last line
> >> which matches a certain experssion.  Currently I'm simply 
> opening the
> >> file and looping through it noteing each time a line 
> matches and at the
> >> end of the loop I see which was the last match.  However 
> this is very
> >> slow and inefficient.  Is there some way to read through the file
> >> backwards and simply find first match starting from the 
> end of the file.
> >> The files are really too big to read the entire thing into memory.
> > 
> > 
> > Would you not be able to use grep with the "-n" option. It will list
> > all your matches with the lines numbered so you can see the 
> last match
> > having the highest line number. If you then use wc on the 
> file you will
> > get the total number of lines. Take one from the other and 
> you can see
> > how far from the end your last match is.
> > 
> > That seems to easy - Have I misunderstood you question?
> 
> The problem is that greping through large files with a fairly complex
> regexp is taking far too long, esp. when I know that the last 
> occurance is
> almost always within the last 100 lines.
> 
> Dag

It's not clear to me whether you need the number of the last line that matches, or just the contents of the last line. If you don't need the number then the following might be useful as  a basis.

def lastlinematching(fileobj, aregex):
# assume initially match will be in about the last 8K of the file
    fileobj.seek(-8192, 2)
    lines = fileobj.readlines()[1:]   #throw away possible incomplete line
    lastline = None
    for line in lines:
        if aregex.search(line):
            lastline = line
    if lastline is None:
        #not in last 8k, start again from the beginning
        fileobj.seek(0, 0)
        for line in fileobj.xreadlines():
            if aregex.search(line):
                lastline = line
    return lastline

Harvey

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.



From mal at lemburg.com  Wed Sep 25 12:08:50 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Wed, 25 Sep 2002 18:08:50 +0200
Subject: ANN: eGenix.com mx EXPERIMENTAL Package 0.7.0
Message-ID: <3D91DF92.4010603@lemburg.com>

________________________________________________________________________

ANNOUNCING:

          eGenix.com mx EXPERIMENTAL Extension Package for Python
                              Version 0.7.0

       Experimental Python extensions providing important and useful
                      services for Python programmers.


________________________________________________________________________

WHAT IS IT ?:

The eGenix.com mx Experimental Extensions for Python are a collection
of alpha and beta quality software tools for Python which will be
integrated into the other mx Extension Packages after they have
matured to professional quality tools.

Python is an object-oriented Open Source programming language which
runs on all modern platforms (http://www.python.org/). By integrating
ease-of-use, clarity in coding, enterprise application connectivity
and rapid application design, Python establishes an ideal programming
platform for todays IT challenges.

________________________________________________________________________

WHAT'S NEW ?

This release fixes a few minor bugs and solves the distutils
problem with mxNumber. It also comes with an updated distutils
setup which installs the RPMs into /usr/ rather than /usr/local/.

________________________________________________________________________

EGENIX.COM MX EXPERIMENTAL PACKAGE OVERVIEW:


mxNumber - Python Interface to GNU MP Number Types

       mxNumber provides direct access to the high performance numeric
       types available in the GNU Multi-Precision Lib (GMP).  This
       library is licensed under the LGPL and runs on practically all
       Unix platforms. eGenix.com has ported the GMP lib to Windows, to
       also provide our Windows users with the added benefit of being
       able to do arbitrary precision calculations.

       The package currently provide these numerical types:

       1. Integer(value)      -- arbitrary precision integers much like
                                 Python longs only faster
       2. Rational(nom,denom) -- rational numbers with Integers as
                                 numerator and denominator
       3. Float(value[,prec]) -- floating point number with at least
                                 prec bits precision
       4. FareyRational(value, maxden)
                              -- calculate the best rational represenation
                                 n/d of value such that d < maxden

mxTidy - Interface to HTML Tidy (HTML/XML cleanup tool)

       mxTidy provides a Python interface to a thread-safe, library
       version of the HTML Tidy. command line tool.

       HTML Tidy helps you to cleanup coding errors in HTML and XML
       files and produce well-formed HTML, XHTML or XML as output. This
       allows you to preprocess web-page for inclusion in XML
       repositories, prepare broken XML files for validation and also
       makes it possible to write converters from well-known word
       processing applications such as MS Word to other structured data
       representations by using XML as intermediate format.

mxURL - A URL Datatype

       mxURL provides a new datatype for storing and manipulating URL
       values as well as a few helpers related to URL building, encoding
       and decoding.

       The main intention of the package is to provide an easy to use,
       fast and lightwheight datatype for Universal Resource Locators
       (note the W3C now calls these URIs).

mxUID - A UID Datatype

       mxUID provides a fast mechanism for generating universal
       identification strings (UIDs). The intent is to make these UIDs
       unique with high probability in order to serve as object or data
       set identifiers.

       A typical use lies in generating session IDs. Other areas where
       unique IDs play an important role are RPC-implementations,
       ORBs, etc.

________________________________________________________________________

WHERE CAN I DOWNLOAD IT ?

The download archives and instructions for installing the packages can
be found at:

       http://www.egenix.com/

Note that in order to use the eGenix.com mx EXPERIMENTAL package you
will first need to install the eGenix.com mx BASE package which can
be downloaded from the same location.

________________________________________________________________________

WHAT DOES IT COST ?

The EXPERIMENTAL packages uses different licenses in its subpackages.
Please refer to the subpackage documentation for details. Some of them
may be integrated into the BASE package, others will be integrated
into the COMMERCIAL package.

The package comes with full source code

________________________________________________________________________

WHERE CAN I GET SUPPORT ?

Commercial quality support for these packages is available from
eGenix.com. Please see

http://www.egenix.com/files/python/eGenix-mx-Extensions.html#Support

for details about our support offerings.

________________________________________________________________________


Enjoy,
-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                        http://www.lemburg.com/python/






From tim.one at comcast.net  Mon Sep 16 01:07:39 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 16 Sep 2002 01:07:39 -0400
Subject: Source code
In-Reply-To: <am3ms1$2e49b$1@ID-53055.news.dfncis.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEENBEAB.tim.one@comcast.net>

[Terje Johan Abrahamsen]
> If I would like to write some code that would not be possible for
> others to read, after I distribute the program, would Python be
> suitable?

No, but I don't know of a language that would be.  Unless you're also going
to supply a restricted CPU that prevents stepping through the machine code,
you can't really hide anything from a determined reverse-engineerer of
ordinary skill.

> ...
> or should I look for a compiled language instead?

It's very easy to reverse-compile Python bytecode.  It's not so easy to
disassemble compiled machine code, but if you think there aren't lots people
who can read machine language almost as easily as high-level source, you're
in for a surprise.  Compilers follow predictable patterns too.  You could
try writing directly in highly obfuscated assembler; that would raise the
bar high enough to discourage almost everyone, although it would also
attract a class of cracker who would feel challenged to outwit you.

If you didn't know all that already, you're playing a game you can't win.  A
better idea is to question why you want to play it at all <wink>.




From bondpaper at earthlink.net  Mon Sep 30 21:11:45 2002
From: bondpaper at earthlink.net (Tom)
Date: Tue, 01 Oct 2002 01:11:45 GMT
Subject: Reloading a module
Message-ID: <bondpaper-648404.19153230092002@nnrp04.earthlink.net>

I'm developing some python mini applets on a Mac (using os 9 and the 
2.1.1 Python IDE), and I've found it impossible to delete and/or reload 
a module after making changes to the code during the 
debugging/verification process. Is there an easy and/or effective way to 
do this? If I execute:

del mymodule

and then execute a print dir(mymodule), it shows up as nonexistent. 
Then, when I execute:

import mymodule

followed by a print dir(mymodule), the changes do not show up. 


Thanks for the help.

Tom


From rajarshi at presidency.com  Thu Sep 26 18:05:37 2002
From: rajarshi at presidency.com (Rajarshi Guha)
Date: Thu, 26 Sep 2002 18:05:37 -0400
Subject: indexing and searching pdf files
Message-ID: <pan.2002.09.26.18.05.37.70342.4621@presidency.com>

Hi,
  I have a load of pdf files and I would like to index them so that I can
serach them for keywords. I was thinking of using pdftotext to generate
the textfile and then create the index from that.

My question - is there already something like this with python?
Another question which is slightly off topic is, does anybody know of any
articles/pages that talk about indexing text files efficiently - index
generaion algorithms etc?

Thanks,


From ianholmes01 at lycos.com  Thu Sep  5 07:54:18 2002
From: ianholmes01 at lycos.com (Ian Holmes)
Date: 5 Sep 2002 04:54:18 -0700
Subject: Psyco testing feed-back wanted
References: <mailman.1030139622.888.python-list@python.org>
Message-ID: <b31eb63a.0209050354.269c5f86@posting.google.com>

"Jimmy Retzlaff" <jimmy at retzlaff.com> wrote in message news:<mailman.1030139622.888.python-list at python.org>...
> Wow!!!
> 
> I've seen raves about Psyco, but I hadn't tried it myself until today.
> My test case was a function that tests if a point is in a polygon. I
> timed several runs with different approaches. All tests are whether 1
> point is in a polygon made up of 5,000,000 points. The timing was done
> on just the function call itself. Here are some results:
>

I had never heard of it before reading the "Python to exceed C in
2030" thread but as soon as it was mentioned I had to test it :)

I saw 3-5x speed increases on a big list / for loop (test data only)
that ran for 8s without psyco.

top product - code used as follows (fairly trivial I admit)
----
import string,time,random,psyco
from strings import * #imports a list of 4 strings

LOOPS = 100000

def method_one(body, x):
    for a in range(0, LOOPS):
        body.append(strings[x])
        x = random.randint(0,3)
    return body

print "Starting test one"
body = []
x = 0
START_TIME = time.time()
text = string.join(method_one(body,x))
print "append method (list) =>", time.time()-START_TIME,"s"
strlen = len(text)
string_size = strlen / 1024
print "Size is",string_size,"KB"

print 'starting psyco'
meth1 = psyco.proxy(method_one)

print "Starting test Two"
body = []
x = 0
START_TIME = time.time()
text = string.join(meth1(body,x))
print "append method (list) =>", time.time()-START_TIME,"s"
strlen = len(text)
string_size = strlen / 1024
print "Size is",string_size,"KB"


From gua81 at XXXyahoo.com  Mon Sep 23 20:07:16 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Tue, 24 Sep 2002 10:07:16 +1000
Subject: Python quote in cgi
Message-ID: <amoabi$sjh$1@lust.ihug.co.nz>

Hi, this might have been discussed b4,
can someone clarify for me, when writing cgi script in python,
does ' = """
eg
print """<a href="index.html">Main</a>"""




From lyle at users.sourceforge.net  Mon Sep  9 10:38:35 2002
From: lyle at users.sourceforge.net (Lyle Johnson)
Date: Mon, 09 Sep 2002 09:38:35 -0500
Subject: compile error on SWIG file
References: <3D7C55EE.D9278B1C@ipm.fhg.de>
Message-ID: <3D7CB26B.2040706@users.sourceforge.net>

Markus von Ehr wrote:

> compiling my c-file generated from SWIG
> (swig -c -python -o mvelib.c mvelib.i)
> with a setup-file, I receive the error:
> 
> c:\programme\vc98\include\eh.h(32) : fatal error C1189: #Error: "eh.h
> is only for C++!"
> 
> Anybody knows what I have to change?

Is it a C++ library that you're wrapping (sounds that way)? If so, you'd 
want to run SWIG with the -c++ option to tell it that.



From vze3myxh at verizon.net  Tue Sep 17 07:38:25 2002
From: vze3myxh at verizon.net (Satheesh Babu)
Date: Tue, 17 Sep 2002 11:38:25 GMT
Subject: More on Protecting Source Code
References: <mailman.1032232999.22661.python-list@python.org>
Message-ID: <pan.2002.09.17.11.38.36.695718.2096@verizon.net>

I love open source and generally prefer to use open source products,
mainly because I can't (or don't want to) keep telling my managers that
I'm waiting for the vendor to help me (usually, this involves several
hoops through some "customer service rep" before I get to talk to a real
engineer), while waiting to fix a problem. With open source, I have the
possibility to fix it myself. Strangely enough, I've seen that this is a
big reason why people don't use open source - because then they can no
longer use the excuse "waiting for the vendor".

Unless majority of the world thinks that having open source is
beneficial, there is a need for closed source software.

If I want to make money on the software that I write, there are
instances where I want to protect my intellectual property. What if:
  - I sell my open source product to a company
  - Some suit comes in and has a friendly relationship with another big
consulting firm and hands them contracts.
  - They take my code and if they find it worthy enough, they probably'll
repackage and sell it as their own. Little guys like me can't dedicate
that much legal resources to fight them.

Good thing is that thins don't happen often. If the software works,
companies are more than happy to pay you *reasonable money*. And they
don't really want to spend time to reengineer the code. Individual users
may be a different story - they might have a better marketing idea than
you do and will probably hijack your code.

For me, .py is enough. I agree with David on validity of his points
though.I've seen guys choosing Java because they can then distribute
.class files instead of source.

World is not ideal yet :-)

-vsb
http://vsbabu.org/


From niemeyer at conectiva.com  Thu Sep 12 18:17:01 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Thu, 12 Sep 2002 19:17:01 -0300
Subject: ANN: NINZ 1.2.1
In-Reply-To: <200209130756.46405.rjones@ekit-inc.com>
References: <20020912135926.A7584@ibook.distro.conectiva> <200209130756.46405.rjones@ekit-inc.com>
Message-ID: <20020912191701.A20774@ibook.distro.conectiva>

> And ZSI would be?

You're the second person to ask me about it, so here it goes:

---
ZSI, the Zolera SOAP Infrastructure, is a Python package that provides
an implementation of SOAP messaging, as described in The SOAP 1.1
Specification. In particular, ZSI parses and generates SOAP messages,
and converts between native Python datatypes and SOAP syntax. It can
also be used to build applications using SOAP Messages with Attachments.
ZSI is ``transport neutral'', and provides only a simple I/O and
dispatch framework; a more complete solution is the responsibility of
the application using ZSI. As usage patterns emerge, and common
application frameworks are more understood, this may change. 

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]



From wlfraed at ix.netcom.com  Fri Sep 20 18:43:52 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Fri, 20 Sep 2002 15:43:52 -0700
Subject: Newbia alert .. working my way through Easy Python Montana Edu page, need help..
References: <MPG.17f540bc75626c66989683@news.worldonline.nl>
Message-ID: <8b8gma.dj3.ln@ix.netcom.com>

Highwaykind fed this fish to the penguins on Friday 20 September 2002 
06:27 am:

> I'm asked ( on the website : www.honors.montana.edu\_jjc\easytut
> \node10.html ) to write a small programm to calculate rectangle,
> square and circle thingies( forgot the English word .. ) and I came up
> with this code :
> 
        <snip>

> But the thing keeps telling me width is not defined, also I can;t use
> the ELIF command .. THink I could do with some help :)
>

        Where to begin... I see way too many errors to intersperse corrections.

1)      Python indentation controls block structure!
2)      You never assign values to those names it is complaining about -- 
rather you are inputting to "a", "b", etc. but never passing /those/ to 
the functions.

        Compare:

import math     #get the long definition of PI

def area_rect(width, height):
        return width * height

def area_square(side):          # squares are special rectangles
        return area_rect(side, side)

def area_circle(radius)
        return math.pi * (radius * radius)      # r*r is likely faster than r**2


option = "prompt"       #filler for first pass
while option != "q":
        if option == "r":
                w = input("Width")
                h = input("Height")
                print "Area of Rectangle is ", area_rect(w, h)

        elif option == "s":
                s = input("Side")
                print "Area of Square is ", area_square(s)

        elif option == "c":
                r = input("Radius")
                print "Area of Circle is ", area_circle(r)

        else:
                print "Options:"
                print " 'r'     Calculate Area of Rectangle"
                print " 's'     Calculate Area of Square"
                print " 'c'     Calculate Area of Circle"
                print " 'q'     Quit program"
                print " <other> Display this help"

        option = raw_input("Option")


--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From huaiyu at gauss.almadan.ibm.com  Wed Sep 11 17:43:51 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Wed, 11 Sep 2002 21:43:51 +0000 (UTC)
Subject: Restricting methods in derived classes
Message-ID: <slrnanve8n.1to.huaiyu@gauss.almadan.ibm.com>

This should be something that the new style class solves but I can't figure
out exactly how:

I need a class that is similar to dict, but, besides providing several
additional methods, it only inherits a subset of dict methods.  What is the
best way to say that only a subset of methods are inherited?

Using the UserDict approach requires wrappers for all the inherited methods,
such as
    def keys(self):     return self.dict.keys()
    def items(self):    return self.dict.items()
This has a function call overhead.  I'd prefer something like
    items = self.dict.items
but of course self is not defined here.

Using the "class MyClass(dict)" approach, how do I say that methods such as
update, __setitem__ are not inherited?  I know the set/get attr trick, but
that adds a function call overhead on every method.

What I'm looking for is a way to say which methods from dict are inherited
in MyClass at the time the class is defined, without incuring a cost when
the methods are called.  Is this possible?

Huaiyu


From pan-newsreader at thomas-guettler.de  Tue Sep 24 15:46:48 2002
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Tue, 24 Sep 2002 21:46:48 +0200
Subject: ZCatalog without Zope
References: <amc9ke$e4p$02$1@news.t-online.com>
Message-ID: <pan.2002.09.24.21.46.47.802334.730@thomas-guettler.de>

On Thu, 19 Sep 2002 12:43:26 +0200, Thomas Guettler wrote:

> Hi!
> 
> Has anyone used ZCatalog from Zope
> with a standalone ZODB (without Zope)?
> 
> Background:
> I am thinking about moving away from Zope to quixote or just simple cgi.
> 
> If you are intereseted
> in the reason way I don't want to use zope:
> http://lists.zope.org/pipermail/zope/2002-August/119965.html
> 
>   thomas

For those who have the same interest. I just discoverd
IndexedCatalog:

http://www.async.com.br/projects/IndexedCatalog/

Did not try it yet.

 thomas


From sismex01 at hebmex.com  Fri Sep 27 18:25:43 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Fri, 27 Sep 2002 17:25:43 -0500
Subject: Python Process that can survive
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062330263A6DD@mail.hebmex.com>

> 
> Yeah I found this hack on the web and I can't believe that this is
> only way to do this in python. You're basically stripping the process
> of all it's stream and sending it on its way. This is really messy.
> 

It isn't messy, it's the correct way to do it. You're not
stripping the process of it's streams, your disconnecting
it from it's controlling TTY, which, if it closes before
the app, sends a DIE! signal to the app; that, or your
app freezes or dies when trying to use it's console.

>
> Plus in my lab there are a bunch of perl guys and a few python people.
> To date, what one can do in their script the python guys can do to
> (only nicer and much more readible). Perl has the ability to have a
> process continue after the underlying terminal is killed if the
> process is in the background. What is python doing that doesn't allow
> this?
> 
> Chris
>

I suspect someone's yanking someone else's chain; I used to
write perl programs also, and whenever I needed an app to
totally disconnect from it's parent, I *had* to close it's
link to the console, or it's parent's console, which basically
means that I had to redirect std{in,out,err} to /dev/null,
in whatever mode you want.

It's not that nasty, you know.

-gustavo










Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From mtremblay at golemlabs.com  Mon Sep 30 14:51:55 2002
From: mtremblay at golemlabs.com (news1.sympatico.ca)
Date: Mon, 30 Sep 2002 14:51:55 -0400
Subject: SWIG - C++ - Inheritance
Message-ID: <d31m9.3742$Qh1.491539@news20.bellglobal.com>

Hello,

    I'm having some problems with SWIG and Python.

    My base class is a CThread class which is defined in C++ and extended to
python using SWIG.  The class is an interface to easily create threads, the
process function must be defined in the child class and then when the run
function is called, the process function is executed in a separate thread.
Every function but the process function is already implemented.

class CThread
 {
 public:
      CThread();     //Constructor
      ~CThread();     //~Destructor
      virtual void process() = 0; //Pure virtual function to extend, thread
function
      bool run();
      bool suspend();
      bool resume();
      bool terminate();
 private:
      bool startThread();
      INT32 m_handle;?
};

The problem comes when I come to create a new class in python that inherits
from CThread.  Here is the python code sample
=============== test.py =================
import my_module

class MyThread(my_module.CThread):
    def __init__(self):
        pass
    def process(self):
        print "hey"

a = MyThread()
a.run()
============================================
my_module is exposed to Python using SWIG.  When I try to execute this, I
get the following error :

    #############################################
    Traceback (most recent call last):
      File "U:\Alexandra\test\pythontool\martin.py", line 14, in ?
        a.run()
      File "U:\Alexandra\test\pythontool\hector.py", line 649, in run
        def run(*args): return apply(_hector.CThread_run,args)
    TypeError: Type error. Expected _p_CThread
    #############################################

It appears that the swig module does not recognize the a variable as a
CThread.  I guess there is something I am doing wrong.  It does not even
call the process function so the problem does not appear to be with the pure
virtual function.  Anybody can help ? It would be really appreciated.
Thanks !

Mathieu Tremblay





From printers at sendme.cz  Sat Sep 28 11:56:58 2002
From: printers at sendme.cz (A)
Date: Sat, 28 Sep 2002 17:56:58 +0200
Subject: How to convert unicode
Message-ID: <3D95ED6A.17648.13FA357E@localhost>

Hi,
I have a script that downloads a webpage with  characters that are not plain ascii.From the 
web page header I can find out that there is UTF-8 coding( Unicode). I can display the page 
properly in a web browser if I send from my script
"Content-Type: text/html; charset=UTF-8 "
But I need to save the page( the text of the page) to a file and some time later import the text 
to a database (MS Access or similar).
How can I download the file or write to a  file so that the characters will be properly display in 
the file and in the database as well?
Thanks for help
Ladislav
  





From jdavis at empires.org  Wed Sep 18 23:57:44 2002
From: jdavis at empires.org (Jeff Davis)
Date: Thu, 19 Sep 2002 03:57:44 GMT
Subject: shared mem & semaphores module
Message-ID: <YWbi9.17569$V7.4455639@twister.socal.rr.com>

I found a module online by "INRIA" (I don't know who that is) that seems to 
wrap the C functions for shared memory management and semaphores.

http://gigue.peabody.jhu.edu/~mdboom/omi/source/shm_source/shm.html

My question is: are these functions included elsewhere? It seems logical 
that python would include them in the standard distribution, since those 
are a pretty important part of at least UNIX (although I know python isn't 
only for UNIX). If they don't, is there a reason other than cluttering the 
standard distribution?

Regards,
        Jeff


From s_gherman at yahoo.com  Tue Sep  3 15:34:04 2002
From: s_gherman at yahoo.com (Sorin Gherman)
Date: 3 Sep 2002 12:34:04 -0700
Subject: https in Python?
Message-ID: <ac259413.0209031134.3778c6ac@posting.google.com>

Hi,
Is it possible to handle https using Python's libraries? If so, what's
involved with using Python to read my hotmail messages, for instance?

Thanks in advance,
/sorin gherman


From wanner.dx at mellon.com  Thu Sep 12 06:05:56 2002
From: wanner.dx at mellon.com (dxterra)
Date: 12 Sep 2002 03:05:56 -0700
Subject: file upload using msvcrt
References: <20020911204906.00669.00000156@mb-fh.aol.com> <20020911210159.00669.00000162@mb-fh.aol.com>
Message-ID: <bb2665bf.0209120205.7d2224d8@posting.google.com>

I have tried both of the methods suggested. When I do a
len(fileinfo.value) before writing, the size is consistantly 8k
regardless of the actual file size. I am using Oracle 9iAS with
apache. I still think that the LimitResponseSize is being set
somewhere, but Oracle cannot tell me.

teaandbikkie at aol.com (TeaAndBikkie) wrote in message news:<20020911210159.00669.00000162 at mb-fh.aol.com>...
> >data = fileinfo.file.read( )
> >if not data: break
> 
> I just thought of another possible problem here, read() is supposed to read in
> the entire file, and maybe 8k is all data can take :)
> 
> So you could change it to read(1024) for 1k chunks.
> 
> Also, maybe read() recognises an EOF character in the word doc? Hopefully being
> binary read this should not happen...
> 
> Kind regards,


From tjreedy at udel.edu  Fri Sep 20 18:26:51 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 20 Sep 2002 22:26:51 GMT
Subject: Python Compiling
References: <amfveg$5g64k$1@ID-53055.news.dfncis.de>
Message-ID: <LgNi9.71150$216.2652928@bin4.nnrp.aus1.giganews.com>

"Terje Johan Abrahamsen" <spoermeg at voldelig.com> wrote in message
news:amfveg$5g64k$1 at ID-53055.news.dfncis.de...
> why is not Python a compiled  language?

This has been a subject of much discussion and work.

One answer: portability (see below) and ease of use for humans rather
than computers.

Another possible answer: it is, sort of.  CPython compiles Python code
to Python bytecode (stored in .pyc files).  Jython compiles Python
code to Java bytecode.  In either case, you can run code (raw ascii or
compiled bytecode) that does not use machine-specific features on most
any machine that has a decent C compiler or Java system -- without a
bunch of #ifdefs.

Psyco compiles Python functions (or parts thereof) to special-purpose
machine code.  But it is only available for machines whose machine
code it has been adapted to.

Pyrex adds statements needed to supply the information needed to
translate Python code to C and thence compile with a local C compiler.
There is at least one other system (name escaped me for the moment)
for converting numerical code to C and machine code.

Another answer is that the obvious, straigthforward method of
compiling Python would not gain as much as one would expect, because
of its dynamic name binding.  Much better to identify and rewrite in
C/C++ just the functions that are actual or potential bottlenecks.

Terry J. Reedy




From hst at empolis.co.uk  Tue Sep 17 07:35:55 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Tue, 17 Sep 2002 12:35:55 +0100
Subject: Regular expression help needed
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220D28@hendrix.empolisuk.com>

Torkil Grindstein wrote:

<snip> 
> Btw, I have no knowledge of SAX. Perhaps I should dig into that?
> 
> Torkil.
> -- 

If you are going to do much processing of XML, then certainly. With your current documents it would seem OK to use REs, but for more complicated processing that way madness lies.

Harvey

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.



From pinard at iro.umontreal.ca  Tue Sep  3 20:48:35 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Tue, 03 Sep 2002 20:48:35 -0400
Subject: efficient list merging
In-Reply-To: <20020903220944.475d7222.christophe.delord@free.fr> (Christophe
 Delord's message of "Tue, 3 Sep 2002 22:09:44 +0200")
References: <ced73313.0209031012.13624f3@posting.google.com>
	<mailman.1031080984.2275.python-list@python.org>
	<20020903220944.475d7222.christophe.delord@free.fr>
Message-ID: <oq8z2io9zw.fsf@titan.progiciels-bpi.ca>

[Christophe Delord]

>> I am not sure about the relative speed of everything, but I would be
>> tempted to try:
>> 
>>    dict(zip(list1 + list2, [None] * (len(list1) + len(list2)))).keys()
>
> This may be written using list comprehension:
>
>>>> dict([(k,1) for k in list1+list2]).keys()

The `for' loop, even within a list comprehension, will be compiled into an
explicit loop, likely to be slower than the implicit loops within `zip'.
Taking advantage of other clever suggestions, I would guess that:

    both = list1 + list2
    result = dict(zip(both, both)).keys()

would not be so bad, speed-wise.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



From bokr at oz.net  Sat Sep  7 17:11:10 2002
From: bokr at oz.net (Bengt Richter)
Date: 7 Sep 2002 21:11:10 GMT
Subject: returning values from FORTRAN with f2py
References: <c3835e5f.0209071122.338a787d@posting.google.com>
Message-ID: <aldq1e$qg9$0@216.39.172.122>

On 7 Sep 2002 12:22:16 -0700, spam at fisher.forestry.uga.edu (Chris Fonnesbeck) wrote:

>I am wondering how to persuade some fortran code that I have compiled
>with f2py to return its calculated values to me in python.  I have
>some code that does gaussian elimination (mostly from Numerical
>Reciopes), and is supposed to return an inverted matrix so that I can
                                         ^^^^^^^^-[1]
>solve a system of linear equations.  However, when run in python, the
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^-[2][3]
>function runs without error but returns no values:
>
>>>> a = [[2,1,3],[1,-1,1],[3,6,-1]]
>>>> b = [7,2,10]
>>>> import gauss
>>>> gauss.gauss(a,3,b,1)
>>>>
>
>Is there something that I have to change in my fortran code to make
>this work?
>
Have you looked at
http://www.pfdubois.com/numpy/html2/numpy.html
and
[1] http://www.pfdubois.com/numpy/html2/numpy-18.html#pgfId-303023
[2] http://www.pfdubois.com/numpy/html2/numpy-18.html#pgfId-303019
or maybe
[3] http://www.pfdubois.com/numpy/html2/numpy-18.html#pgfId-303085

Perhaps there is an easier route to your goal? (I haven't used the above,
but numpy seems to be recommended frequently).

Regards,
Bengt Richter


From skip at pobox.com  Tue Sep 10 08:46:43 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 10 Sep 2002 07:46:43 -0500
Subject: how to build wxPython w/ Sun's C++ compiler?
In-Reply-To: <23891c90.0209092359.11acb607@posting.google.com>
References: <mailman.1031607011.28525.python-list@python.org>
        <m3k7luojpq.fsf@mira.informatik.hu-berlin.de>
        <mailman.1031621231.16841.python-list@python.org>
        <23891c90.0209092359.11acb607@posting.google.com>
Message-ID: <15741.59827.147187.55289@12-248-11-90.client.attbi.com>

    Paul> What about configure programs, though? They tend to accept lots of
    Paul> command line arguments, and it's hard to get more "Unix-centered"
    Paul> than that these days.

Agreed.  However, disutils commands have this structure:

    python setup.py [ options ] command [ more options ] ...

which I think can get kind of confusing.

    Paul> Moreover, the config.status file nicely summarises what you
    Paul> presented to configure - much more convenient than environment
    Paul> variables, surely, because you get to remember how you built the
    Paul> package in question.

The problem with wxPython is that it's got stuff written in C++ while Python
is typically compiled with C.  I don't want what the system was built
with. ;-)

Skip




From bruce_deletethis_dawson at cygnus-software.com  Thu Sep 26 11:34:02 2002
From: bruce_deletethis_dawson at cygnus-software.com (Bruce Dawson)
Date: Thu, 26 Sep 2002 15:34:02 GMT
Subject: SET_LINENO
References: <3D92A46E.2030706@cygnus-software.com> <j47kh9b6kp.fsf@informatik.hu-berlin.de>
Message-ID: <3D9328C1.5030802@cygnus-software.com>

Thanks for all the replies - including the link to the unhandled 
exception handling option.

Martin v. L?wis wrote:

> Bruce Dawson <bruce_deletethis_dawson at cygnus-software.com> writes:
> 
> 
>>Executive summary: Has the SET_LINENO removal been tested with the HAP
>>debugger, and wouldn't a breakpoint instruction give much larger
>>benefits?

...


>>So, wouldn't a break-point opcode be a more substantial improvement
>>than removing SET_LINENO? Just bringing up that old topic again to see
>>if it gets any more traction :-)
>>
> 
> No. Removal of SET_LINENO does improve performance in a "free-running"
> application, which adding a breakpoint instruction would not.

A breakpoint instruction would be intended to improve performance in 
debuggers. My point was that the improvement there would be larger 
(probably 100% or more) than the improvement from SET_LINENO. But the 
improvement would not, as you say, affect "free-running" applications at 
all.


> That said: it might be still worthwhile to add a breakpoint
> instruction. There are a number of issues to be considered, such as
> getting the breakpoint instruction into the byte code, and getting the
> original instruction back in. So if you'ld like to see one, you
> probably have to write a PEP.


Some day I will do that. I have to dig deeper in the Python source 
before I can do that.


> 
>>P.P.S. A breakpoint instruction would also be a boon for asserts - I'd
>>love to be able to stop on the assert when running under a debugger,
>>to see what is wrong before the stack frame unwinds.
>>
> 
> I can't see how this helps. Wouldn't you have to set a breakpoint on
> all asserts for that? And wouldn't that require finding all asserts in
> the first place?

If assert was implemented with a hard-coded breakpoint then it would 
work very nicely. Under a debugger this would stop you on the relevant 
line. Not under the debugger it would throw an exception as before.

On x86 processors int 3 is the hard-coded breakpoint instruction. If you 
hit one it stops your program. If you're running under a debugger you 
can then step right over it. I was imagining something like that.

Asserts are great, but they are easier to interpret if when you trigger 
one you are dropped onto the assert statement. Just a mad idea...



From member at dbforums.com  Mon Sep 16 16:31:16 2002
From: member at dbforums.com (mcolli)
Date: Mon, 16 Sep 2002 20:31:16 +0000
Subject: looking for Python programmers
Message-ID: <1821724.1032208276@dbforums.com>

Hi,

I am looking for Python programmers to participate in a Zope/Pyhton
project in Buenos Aires, Argentina.

Those who may be interested please send me the CV.

Thanks
Mariela

--
Posted via http://dbforums.com



From martin at v.loewis.de  Mon Sep 30 01:45:45 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 30 Sep 2002 07:45:45 +0200
Subject: Catching user defined exceptions from foreign modules
References: <mailman.1033352618.4920.python-list@python.org>
Message-ID: <m3d6qwqbcm.fsf@mira.informatik.hu-berlin.de>

Derek Basch <dbasch at yahoo.com> writes:

> So, this code works fine for me. However, I have 2
> much larger modules that have the same code structure
> but in these larger modules the test1 refuses to catch
> the raised TestError unless I change the except clause
> to:
> 
> except test2.TestError:
>     print 'fudge'

Perhaps it isn't test2.TestError that is raised, but just TestError,
as the module is used as the application program?

Regards,
Martin



From op73418 at mail.telepac.pt  Mon Sep  9 16:27:40 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Mon, 09 Sep 2002 21:27:40 +0100
Subject: Python 2.3
References: <3d7cfc19_5@news.newsgroups.com>
Message-ID: <5t0qnu0dfbebm1uvk1v93msk4tespgp6bn@4ax.com>

On Mon, 09 Sep 2002 22:01:32 +0200, JB <jb at yahoo.de> wrote:

>Will rational numbers be implemented in Python 2.3? And when 
>is the first beta due?

There are no plans to implement rationals. There is however a Rat class
in the demo dir, implementing rationals in pure Python. For the 2.3
plans you can consult PEP 283 at

http://www.python.org/peps/pep-0283.html

>
>TIA,

Best regards,
G. Rodrigues


From anthony at computronix.com  Mon Sep 23 20:46:49 2002
From: anthony at computronix.com (Anthony Tuininga)
Date: Mon, 23 Sep 2002 18:46:49 -0600
Subject: list problem
In-Reply-To: <3D8F9F5E.DC48B13F@pop.ntlworld.com>
References: <mailman.1032818129.8126.python-list@python.org>
 <3D8F9F5E.DC48B13F@pop.ntlworld.com>
Message-ID: <1032828410.2133.1.camel@localhost>

If you have list comprehensions available, use

[b for a in x for b in a]

Anthony

On Mon, 2002-09-23 at 17:10, a.clarke11 wrote:
> How about:
> 
> for item in x:
>  while item:
>   y.append(item.pop(0))
> print'y=',y
> ?
> hope this helps
> Tony Clarke
> Glasgow
> 
> jubafre at brturbo.com wrote:
> 
> > i have:
> > x=[['0020', '0000x9'], ['0030', '0000xa'], ['00B4', '0000x8'], ['0030', '0000xb'], ['00F0'], ['0001'], ['0003'], ['0005']]
> >
> > and i want:
> > y=['0020', '0000x9', '0030', '0000xa', '00B4', '0000x8', '0030', '0000xb','00F0', '0001',
> > '0003', '0005']
> >
> > how to do this?
> >
> > Juliano Freitas
> > www.gebrasil.hpg.com.br
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list





From bokr at oz.net  Mon Sep  9 13:56:44 2002
From: bokr at oz.net (Bengt Richter)
Date: 9 Sep 2002 17:56:44 GMT
Subject: Python plugins
References: <alf7rc$1p06l1$1@ID-41653.news.dfncis.de>
Message-ID: <alincs$e43$0@216.39.172.122>

On 8 Sep 2002 10:13:00 GMT, Tobias Klausmann <klausman-un080902 at tuts.net> wrote:

>Hi there,
>
>For an open source project that needs to be extensible by its
>users, I would like to create a plugin architecture. Mind you,
>I'm not trying to use Python as a plugin language only like XChat
>or Gnumeric do. Let me illustrate:
>
>I have a directory somewhere in PYTHONPATH (let's just assume
>it's in the CWD) that contains lots and lots of *.py files with a
>defined set of functions (or, at least one: register()). What I'd
>like to do now is being able to load these plugins at program
>start and reload them if I'd like to later. Problem is that the
>program has to figure out by itself what *.py files are there -
>and I have to make sure somehow that I get a grip on the loaded
>files.
>
http://www.python.org/doc/current/lib/module-imp.html

Regards,
Bengt Richter


From b.maryniuk at forbis.lt  Tue Sep 10 06:33:05 2002
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Tue, 10 Sep 2002 12:33:05 +0200
Subject: UNIX processes
Message-ID: <200209101233.05796.b.maryniuk@forbis.lt>

Hi, all.
Somebody know how to list all the processes in RIGHT way on Linux/UNIX?
I need to find the zombies, stopped processes etc.

-- 
Regards, Bogdan

Linux: The Ultimate NT Service Pack




From whisper at oz.net  Fri Sep  6 23:56:46 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 6 Sep 2002 20:56:46 -0700
Subject: Larry Wall's comment on python...
In-Reply-To: <LNBBLJKPBEHFEDALKOLCOELCBCAB.tim.one@comcast.net>
Message-ID: <GCEDKONBLEFPPADDJCOEAEJDENAA.whisper@oz.net>

One could hardly expect Larry to say "I have seen Pearl 6 - and it is
Python!" :)

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Tim Peters
> Sent: Friday, September 06, 2002 20:30
> To: python-list at python.org
> Subject: RE: Larry Wall's comment on python...
>
>
> [Peter Hansen]
> > Anyone know how long is the largest chunk of code Larry Wall has
> > actually written in Python?
>
> It was 23 lines longer than the largest chunk of code Guido has actually
> written in Perl.  The difference is that Larry's code worked, and Guido's
> still doesn't.  I suspect Larry may have taken a wrong lesson from that
> <wink>.
>
> > The fact he has to say "I think" about this issue implies he is
> > guessing.  Maybe if he actually tried it he'd conclude differently.
>
> It was just a sound bite from an interview, and blocks aren't
> quite the same
> thing in Perl (for example, a Perl block can have its own variables, but a
> Python block can't).  I bet that if Guido had designed Perl, it
> would still
> have curlies, and I defy anyone to prove that wrong.  Heh.
>
> leave-rationality-to-the-rationals-this-is-language-design-ly y'rs  - tim
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list




From bokr at oz.net  Thu Sep 26 22:53:55 2002
From: bokr at oz.net (Bengt Richter)
Date: 27 Sep 2002 02:53:55 GMT
Subject: Help! This is driving me insane!
References: <FYMk9.12763$wH.1166@sccrnsc01>
Message-ID: <an0h83$vnh$0@216.39.172.122>

On Thu, 26 Sep 2002 23:43:33 GMT, "kara.klosterman" <kara.klosterman at attbi.com> wrote:

>Hello,
>I created a very simple class that totals some data
>but I am not getting the output from it that I would
>expect at all.
>Here is the class:
>http://www.geocities.com/dbasch/insane.txt
>
>I would expect to get:
>[0, 0]
>[0, 0]
>[[0, 0]]
>
Why?

>but I am getting:
>[0, 0]      <<--(a)
>[0, 0]      <<--(b)
>[[1, 1]]    <<--(c)
>
Why not? [I labeled the above (a) etc. for reference below].

>I dont understand how the 'grouptotals' list is
>getting the incremented values prior to the
>'grouptotalcounter' function being called.
What makes you think it is? The value you printed at (b) shows it's not.

>furthermore, I can place a 'print grouptotals' right
>after the errorlog append and it isnt the incremented
>values! Please tell me this isnt a normal thing!

Did you walk through the code and think through what
is happening?

class Test:
    def __init__(self):
        self.errorlog = []        
        pass

    def run(self):
        self.grouptotals = []
        for x in range(2):
            self.grouptotals.append(0)
            
        print self.grouptotals                    <<--(a), should be [0, 0]
        self.errorlog.append(self.grouptotals)    <<-- self.errorlog now [[0, 0]]
        print self.grouptotals                    <<--(b), no change, still [0, 0]
            
        self.grouptotalcounter()                  <<-- self.grouptotals changed to [1, 1]
                                                      (note that self.errorlog still refers to
                                                       the now changed self.grouptotals)
        
        return self.errorlog                      <<-- ref to changed grouptotals makes it [[1, 1]]
                                                  but there is no print statement. Reference to
                                                  self.errorlog is returned to caller (d) below
                                                  of the run method.

    def grouptotalcounter(self):
        for totalbyindex in range(2):
            self.grouptotals[totalbyindex] += 1   <<-- [0, 0] becomes [1, 1] here


If you put the above in insane.py and interactively made an instance thus:

 >>> import insane
 >>> t=insane.Test()
 >>> t.run()               # <<--(d)
 [0, 0]
 [0, 0]
 [[1, 1]]

that last line is the return value from t.run() being printed by the interactive
eval/print loop.

Everything as one would expect from the code.

Regards,
Bengt Richter


From duduca007 at yahoo.com  Tue Sep 24 13:46:41 2002
From: duduca007 at yahoo.com (Carlos Moreira)
Date: Tue, 24 Sep 2002 10:46:41 -0700 (PDT)
Subject: Pyro Error.
Message-ID: <20020924174641.40449.qmail@web11403.mail.yahoo.com>

Testing the simplest example of docs/8-example.html,
on client side, generate an error message:

Traceback (most recent call last):
  File "./client.py", line 10, in ?
    print jokes.joke("Irmen")
  File
"/home/pgs/carlos/python/Python-2.2/lib/python2.2/site-packages/Pyro/core.py",
line 315, in __invokePYRO__
    self.adapter.bindToURI(self.URI)
  File
"/home/pgs/carlos/python/Python-2.2/lib/python2.2/site-packages/Pyro/protocol.py",
line 224, in bindToURI
    raise ProtocolError('connection failed')
Pyro.errors.ProtocolError: connection failed

Where I must seek for inconsistences? Pyro.conf?

	[]'s

Cadu Moreira
KaduSoft President
--------------------------
[......]

__________________________________________________
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
http://sbc.yahoo.com



From gerhard.haering at gmx.de  Wed Sep 11 16:41:41 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 11 Sep 2002 20:41:41 GMT
Subject: Distutils/setup help
References: <mailman.1031769975.29462.python-list@python.org>
Message-ID: <slrnanvak5.3hq.gerhard.haering@lilith.my-fqdn.de>

Mahrt, Dallas wrote in comp.lang.python:
> I am in the process of making modifications to the Python build system to
> aid in the cross-compilation of Python. [...] Is this the right list to ask these questions?

I'd suggest python-dev, the list of the Python developers. You could
cc the distutils SIG's list.

-- Gerhard


From thomas at xs4all.net  Wed Sep 11 20:25:43 2002
From: thomas at xs4all.net (Thomas Wouters)
Date: Thu, 12 Sep 2002 02:25:43 +0200
Subject: Use glob not re for regexp?
In-Reply-To: <DQQf9.1761$Lg2.233877@news2.news.adelphia.net>
References: <DQQf9.1761$Lg2.233877@news2.news.adelphia.net>
Message-ID: <20020912002543.GA797@xs4all.nl>

On Thu, Sep 12, 2002 at 12:02:43AM +0000, Robert Oschler wrote:

> I saw a post where someone said to use glob not re for regular expressions?
> Is this true and if so why?  I use Secret Labs regexp package.

It depends on what you need. 'glob' doesn't do regular expressions, so if
you want regular expressions, don't bother with glob. Like its docstrings
say, glob is for matching filenames based on globular expressions, like most
shells (including DOS shells) do. The actual matching is done by fnmatch, so
if you want to use globular expressions (somewhat) like regular expressions,
you want fnmatch, not glob.

-- 
Thomas Wouters <thomas at xs4all.net>

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



From Peer.Griebel at entire.de  Wed Sep 18 07:35:38 2002
From: Peer.Griebel at entire.de (Griebel, Peer)
Date: Wed, 18 Sep 2002 13:35:38 +0200
Subject: Copy with __slots__
Message-ID: <673E8D215B3BD311A9C90090279BBB8402829E7A@mail.ebssoftware.de>

Hi,

I just discovered another problem. It seems that objects which are instances
of new style classes (derived from object) are not copied correctly. Here is
an example:


class C1(object):
    __slots__ = "s1";

class C2(C1):
    __slots__ = "s2";

class C3(C2):
    pass

o1 = C1()
o2 = C2()
o3 = C3()

print o1.__slots__	# prints s1
print o2.__slots__	# prints s2
print o3.__slots__	# prints s2

o1.s1 = 11
o2.s1 = 21
o2.s2 = 22
o3.a = 5

import copy
p2 = copy.copy(o2)
print p2.s2		# attribute s2 does not exist (has not been copied)


The last print statement fails with the an AttributeError. Since
o2.__slots__ does not know about the slots of the parent class (C1), it
would not be sufficient to simply copy all elements in o2.__slots__. All
parent classes had to be inspected also.

So how can o2 be copied efficiently?



connection reset by
  Peer

-- 
Dr. Peer Griebel                     Tel: +49(0) 731 / 9 74 95-0
Dipl. Inform.                       Fax: +49(0) 731 / 9 74 95-20
Entire Software AG                 mailto:Peer.Griebel at entire.de
Pfarrer-Wei?-Weg 10 - 12
D-89077 Ulm                                 http://www.entire.de




From lists at gregfortune.com  Thu Sep 12 15:38:14 2002
From: lists at gregfortune.com (Greg Fortune)
Date: Thu, 12 Sep 2002 12:38:14 -0700
Subject: problem with lists
References: <20020911215423.GA30455@habana.easygolucky.de> <mailman.1031782033.5370.python-list@python.org> <alq37d$lj726$3@hades.rz.uni-sb.de>
Message-ID: <3d80eb7f$0$311$39cecf19@nnrp1.twtelecom.net>

Uwe Schmitt wrote:

> Sean 'Shaleh' Perry <shalehperry at attbi.com> wrote:
>>>     Manuel
> 
>> You want to use :
> 
>> if fields[3:4] not in domains:
>>    domains.append(fields[3:4])
> 
>> .index() only works if the item is actually in the list.
> 
> At least you should wrap the line with .index in a try/except pair...
> But the solution with "in" is preferable.
> 
> Greetings, Uwe
> 
> 

Or simply use .count instead of a try/except block.

if(domains.count(fields[3:4]) == 0)

will work exactly like 'not in'.  In this case, I would prefer the 'not in' 
option, but wrapping a .index in a try/except seems like a huge overkill.  
IMHO, avoid exceptions if possible.

Greg Fortune
Fortune Solutions


From steve at ferg.org  Mon Sep  9 09:52:12 2002
From: steve at ferg.org (Stephen Ferg)
Date: 9 Sep 2002 06:52:12 -0700
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid>
Message-ID: <b16e4ef7.0209090552.3ad7691b@posting.google.com>

> you'd better include a clear warning... 
> that you'll explain how to do it correctly
> later...

Can you say a little more?  What is it that makes using "bind()"
incorrect?   And what is it about using "command" that avoids the
problems?

This is a genuine request for information, not a challenge to your
statement.  I am still learning about Tkinter features.


From BPettersen at NAREX.com  Mon Sep  9 17:52:08 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 9 Sep 2002 15:52:08 -0600
Subject: Windows/DOS: double clicking a .py file
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201253172@admin56.narex.com>

> From: Just [mailto:just at xs4all.nl] 
> 
> Sorry if this is a Windows FAQ, I'm a newbie to Windows, but not to 
> Python.
> 
> If I double-click a Python script (or use the start command 
> in the dos 
> prompt) a new dos prompt window is opened and my program is run. The 
> window goes away when the program is done. However, the window _also_ 
> goes away when the program raises an exception, making it virtually 
> impossible to actually _see_ the exception... Is there a way 
> to keep the 
> dos prompt window in this case?

Try something like the following... We've had much success with
automatically emailing the traceback back to the developer using the
cgitb module -- you can never trust users to give you good error reports
<wink> besides the cgitb module provides much more context than the
standard exceptions... Let me know if you'd like the code...

try:
    your code here
except:
    import msvcrt, traceback
    traceback.print_exc()
    print 'Hit a key to exit'
    while not msvcrt.kbhit():
        pass
    
-- bjorn



From cliechti at gmx.net  Mon Sep 23 17:00:02 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 23 Sep 2002 23:00:02 +0200
Subject: Python and cgi and <a> problem!
References: <ammn6e$lj7$1@lust.ihug.co.nz>
Message-ID: <Xns9292EA042C259cliechtigmxnet@62.2.16.82>

"CheapSkate" <gua81 at XXXyahoo.com> wrote in
news:ammn6e$lj7$1 at lust.ihug.co.nz: 

> Hi Im currently making a cgi script in python which list all the
> files&directory in a directory and list them into an html file with
> all the files linked to the actual file.
> ..... did what I just said make any sense?
> 
> Currently I can list all the files and directories and print it out in
> the html file.
> my problem now is how to add the anchor tag <a>???
> 
> here is the line which prints the files&directories in the html file:
> print crap
> 
> , where crap is a filename such as colour.txt

print '<A href="%s">%s</A>' % (filename, filename)

chris

-- 
Chris <cliechti at gmx.net>



From manuel at hendel.net  Wed Sep 11 07:51:18 2002
From: manuel at hendel.net (Manuel Hendel)
Date: Wed, 11 Sep 2002 13:51:18 +0200
Subject: deleting the first and the last character of a string
In-Reply-To: <3D7F2960.20507@Linux.ie>
References: <mailman.1031741971.20624.python-list@python.org> <3D7F2960.20507@Linux.ie>
Message-ID: <20020911115118.GE2054@partagas.as.de.cw.net>

On Wed, Sep 11, 2002 at 12:30:40PM +0100, Padraig Brady wrote:
> How about: line = line[1:-1]

This worked quite good, but I found another problem. There seams to be
trailing whitespaces. I tried to strip(s) to get rid of them, but this
doesn't work, I get the following error message:

Traceback (most recent call last):
  File "./pop3create.py", line 9, in ?
      print strip(line)
      NameError: name 'strip' is not defined

What's wrong here?

manuel

-- 
Technological progress has merely provided us with more efficient means for 
going backwards. 
-Aldous Huxley 



From keith_oconnell at blueyonder.co.uk  Fri Sep 20 06:20:48 2002
From: keith_oconnell at blueyonder.co.uk (Keith O'Connell)
Date: Fri, 20 Sep 2002 10:20:48 GMT
Subject: finding last match in a file
References: <slrnaojlin.8re.dag@omally.animagicnet.no> <87r8fplvq4.fsf@smtp.blueyonder.co.uk> <slrnaolc45.8re.dag@omally.animagicnet.no>
Message-ID: <877khh6jy7.fsf@smtp.blueyonder.co.uk>

>>>>> "Dag" == Dag  <dag at animagicnet.no> writes:

    Dag> The problem is that greping through large files with a fairly
    Dag> complex regexp is taking far too long, esp. when I know that
    Dag> the last occurance is almost always within the last 100
    Dag> lines.


        How about tail -n 100 <filename>?

Keith
-- 
___________________________________   _
  Keith O'Connell.                   -o)
  Maidstone, Kent. (UK)              /\\
  keith_oconnell at blueyonder.co.uk   _\_v


From dgallion1 at yahoo.com  Wed Sep 18 22:23:24 2002
From: dgallion1 at yahoo.com (darrell)
Date: Thu, 19 Sep 2002 02:23:24 GMT
Subject: Catching results of an exec
References: <3d88f701$0$304$39cecf19@nnrp1.twtelecom.net>
Message-ID: <wyai9.1626$k92.385398@twister.nyroc.rr.com>

exec("a=dir()")
print a

--Darrell


Greg Fortune wrote:

> Is there any way to "catch" the value produced by an exec?  I'm
> considering things like print statements that don't really assign a value,
> but do send a result to stdout.
> 
> For example, if I execute
> 
> exec('print dir()')
> ...



From spoermeg at voldelig.com  Mon Sep 16 00:31:38 2002
From: spoermeg at voldelig.com (Terje Johan Abrahamsen)
Date: Sun, 15 Sep 2002 23:31:38 -0500
Subject: Source code
Message-ID: <am3ms1$2e49b$1@ID-53055.news.dfncis.de>

If I would like to write some code that would not be possible for others to
read, after I distribute the program, would Python be suitable? I know that
there are a few programs like Py2Exe, that creates .exe files. But, I also
saw this:

--------------------
 A zip-compatible archive is built, containing all files from this
  directory as well as your script, and appended to a custom python
  interpreter supplied with py2exe.
--------------------

Doesn't sound like it is protected very well. Is there some alternative to
Py2Exe that can do the job, or should I look for a compiled language
instead?





From buzzard at urubu.freeserve.co.uk  Tue Sep 10 09:37:00 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Tue, 10 Sep 2002 14:37:00 +0100
Subject: Graph Layout Algorithms
Message-ID: <alksio$d2l$1@newsg3.svr.pol.co.uk>

Hello,
         Just wondering if anyone has implemented anything in Python.  I
need to lay out my graphs (directed acyclic / undirected / trees) in a
reasonably clear way.  So if anyone has already implemented anything, or if
anyone has advice on appropriate algorithms, I'd be glad to hear about it.
Cheers.

Duncan




From whisper at oz.net  Fri Sep 13 16:05:41 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 13 Sep 2002 13:05:41 -0700
Subject: Embed extended Python
In-Reply-To: <f760295a.0209130656.2096981f@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOECEHHEPAA.whisper@oz.net>

The boost libary is a great C++ embedding/extending library for Python.

www.boost.org

David LeBlanc
Seattle, WA USA 

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Mathieu Tremblay
> Sent: Friday, September 13, 2002 7:57
> To: python-list at python.org
> Subject: Embed extended Python
> 
> 
> Hello, 
>    I am quite new to Python and I'd like to embed extended python. 
> The problem is that the documentation I can find on the internet is
> always related to C standard types and not to C++ objects.
> 
>    To extend python I use the really good SWIG tool and everything is
> fine.  The problem comes when I try to call a python script that takes
> as an argument one of the class that I have defined in C++ through
> swig.
> 
> Exemple:
>  C++ class extended to python with SWIG, through the module wheel: 
>    CWheel
>    {
>      CWheel(){}
>      ~CWheel(){}
>      printWheel()
>      {
>        printf("I am a wheel");
>      }
>    };
>  Now, let's suppose that the module is successfully exported and can
> be used in python.  An example of simple python script would be :
>   import wheel
>  
>   def thefunction(awheel):
>      awheel.printWheel()
> 
>  Calling this function directly from python (supposing we provide the
> correct argument) works perfectly.  The problem comes when I want to
> call this script from C++.  Maybe It is a stupid question but here it
> is: How do you pass an existing CWheel as an argument to the function
> from C++??? How do you wrap a CWheel to a PyObject?
>  I guess the method would be to use the function
> PyObject_CallObject(pFunc, pArgs); the pArgs being a PyTuple where I
> can set a PyObject to a position X.  The problem is that all the
> examples that I find convert int, char*, long, double etc. to PyObject
> with built in python functions but no different/user defined class.
>  Anybody can help?
> 
>  Thanks a lot,
>   Mathieu
> -- 
> http://mail.python.org/mailman/listinfo/python-list



From rpm at wag.caltech.edu  Mon Sep  9 11:39:48 2002
From: rpm at wag.caltech.edu (Rick Muller)
Date: Mon, 09 Sep 2002 08:39:48 -0700
Subject: Can't exponentiate zero in NumPy
References: <al5gdq$18g$1@naig.caltech.edu>
Message-ID: <alifc4$gsi$1@naig.caltech.edu>

I found out more about this problem, and I'm still hoping that someone 
can suggest a good fix.

The problem comes when I take a power of a NumPy array that has very 
small values in it:

 >>> from Numeric import *
 >>> a = zeros(3,Float)
 >>> a**2
array([0., 0., 0.])
 >>> a[1] = 1.e-310
 >>> a**2
Traceback (most recent call last):
    File "<stdin>", line 1, in ?
OverflowError: math range error

This crashes on Linux, but not on Mac OS X.

Is this a bug or a feature?

If it's a feature (which it might be), can someone offer a one-line 
workaround? What I would like to do is set everything below a small 
value (say 1.e-16) to zero. Obviously,
 >>> for i in range(len(a)): if a[i] < 1.e-16: a[i] = 0
or something like it should work, but it will also be slow, and I was 
hoping that one of the Numeric UFuncs might do the same thing more 
effectively.

Thanks in advance for any help anyone can offer.

Rick



From andres.meza at nuevosmedios.ws  Thu Sep 12 11:37:04 2002
From: andres.meza at nuevosmedios.ws (Andres Meza)
Date: Thu, 12 Sep 2002 10:37:04 -0500
Subject: How to upload files using mod_python?
Message-ID: <1031845024.2957.154.camel@coraje>

Greetings. 

I have created a python CGI script that can upload files sucessfully.
Now, I want to create a mod_python script with the same feature using
the "publisher" handler and I get this error:   


[Thu Sep 12 08:53:41 2002] [error] PythonHandler mod_python.publisher:
Traceback (most recent call last):
[Thu Sep 12 08:53:41 2002] [error] PythonHandler mod_python.publisher:  
File
"/home/mrnet/rpm/tmp/mod_python-root/usr/lib/python2.2/site-packages/mod_python/apache.py", line 193, in Dispatch
[Thu Sep 12 08:53:41 2002] [error] PythonHandler mod_python.publisher:  
File
"/home/mrnet/rpm/tmp/mod_python-root/usr/lib/python2.2/site-packages/mod_python/publisher.py", line 104, in handler
[Thu Sep 12 08:53:41 2002] [error] PythonHandler mod_python.publisher:  
File
"/home/mrnet/rpm/tmp/mod_python-root/usr/lib/python2.2/site-packages/mod_python/publisher.py", line 308, in __init__
[Thu Sep 12 08:53:41 2002] [error] PythonHandler mod_python.publisher:
AttributeError: 'file' object has no attribute '__methods__'

Anyone have an idea about why I get this error?

Thanks for your time,
-- 
   Andr?s Meza
   Technical Manager
 __Nuevos Medios - NM S.A.____________________
   e-mail:         andres.meza at nuevosmedios.ws
   day time phone: (+572)6827794 ext.102
   url:            www.nuevosmedios.ws
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020912/13b0b78a/attachment.html>

From gerhard.haering at gmx.de  Tue Sep 10 12:39:19 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 10 Sep 2002 16:39:19 GMT
Subject: Trouble upgrading to 2.2.1
References: <aXof9.29813$w51.7932884@news2.news.adelphia.net>
Message-ID: <slrnans81k.jc.gerhard.haering@lilith.my-fqdn.de>

Robert Oschler wrote in comp.lang.python:
> I'm trying to upgrade from 2.1.1 to 2.2.1.  I'm using kpackage (KDE's rpm
> wrapper).

My recommendation is to learn to use rpm directly.

> When I try to install from the rpm file named
> "python2-2.2.1-1.i386.rpm" I get the following dependency errors:
> 
> libcrypto.so.2
> libsel.so.2

That should probably be libssl...

> How can I correct this, or, is there an easier way to get upgraded?

You should install the OpenSSL libraries in order to fulfill these
dependencies. If you do have the OpenSSL libraries installed, but the
libcrypto or libssl version numbers are different on your machine, you can
try to provide the necessary symlinks so rpm will shut up.

If you know what you're doing, you can always use the --force argument for
rpm, but unless you fix your dependencies, socket and all Python modules
that build on top of it (smtplib, urllib, ...) likely won't work.

I guess you have that Python RPM from the KRUD site, so perhaps you can
find upgrades for OpenSSL at the same place.

-- Gerhard


From bh at intevation.de  Tue Sep 24 08:36:50 2002
From: bh at intevation.de (Bernhard Herzog)
Date: 24 Sep 2002 14:36:50 +0200
Subject: Three dumb questions (ordered by dumbness descending)
References: <amnvit$7kqdt$1@ID-77524.news.dfncis.de> <82Zj9.136094$ub2.2966822@news1.tin.it>
Message-ID: <m3smzzo97h.fsf@intevation.de>

Alex Martelli <aleax at aleax.it> writes:

> Once you do have a random or pseudo-random generator that fully
> satisfies you, Ian Bicking's snippet reimplementing shuffle as a
> Decorate-Sort-Undecorate idiom is truly excellent -- a little jewel,
> IMHO, despite the little typo in it.  It does, alas, end up a
> little slower than Tim Peters' original -- but losing a speed
> competition to Tim is no shame, and Ian's version only loses very
> very little indeed.  To wit (having corrected Ian's typo):
[...]
> def shuffle_ian(x):
>     newList = [(random(), i) for i in x]
>     newList.sort()
>     x[:] = [i[1] for i in newList]

What's missing, though, is a proof that shuffle_ian produces each
possible permutation with the same probability (assuming random() has
suitable properties)

One problem I can see is that there's a non-zero chance that there will
be at least two equal random numbers in newList in which case the
corresponding values of x are used to sort newList. This skews the
probabilities and furthermore means that shuffle_ian can only shuffle
lists whose items can be ordered which among others exludes complex
numbers.

   Bernhard

-- 
Intevation GmbH                 http://intevation.de/ 
Sketch                          http://sketch.sourceforge.net/ 
MapIt!                          http://mapit.de/


From sismex01 at hebmex.com  Wed Sep  4 19:11:51 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Wed, 4 Sep 2002 18:11:51 -0500 
Subject: Bind threads to addresses -- Windows & urllib?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233022E50FC@mail.hebmex.com>

> 
> It's a machine with multiple IP addresses bound to one 
> adapter.  I want each spider thread to use a different
> address...
> 
> --
> Nick Arnett
> Phone/fax: (408) 904-7198
> narnett at mccmedia.com
> 

That's not gonna help you at all, because all threads will
be capped by the interface's top throughput, or by your
machine's processing power.

What you want, or rather, are trying, is, if I recall
correctly, is to bind a cpu (in case of a multi-cpu
machine) to a certain adapter. That's doable, but
you need a specific M$ utility to do that.

But in this case, since all threads are gonna be drinking
from the same fountain, is doesn't matter if one thread
likes orange, another strawberry, and another chocolate.

-gus
--

Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From aleax at aleax.it  Wed Sep 25 05:04:08 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 09:04:08 GMT
Subject: summary:  (was: callback + __call__ + TypeError: unbound method; on non method)
References: <lg8z1rtvzi.fsf@mu.biosolveit.local> <JiZj9.157856$pX1.5687433@news2.tin.it> <lgwupbs84w.fsf@mu.biosolveit.local> <lg1y7ize7d.fsf_-_@mu.biosolveit.local>
Message-ID: <c_ek9.140630$ub2.3071529@news1.tin.it>

<posted & mailed>

Volker Apelt wrote:

> ## Thank you, Alex Martelli for your help on this problem.

You're welcome, but:
        ...
>   class replacement for staticmethod and add a second
>   group of () to the call.

The "second group of ()" is needed only because:

> def call_me():
>     return time.time

CALLING call_me returns (doesn't CALL) the function object
time.time.  Therefore:

> class Callback_good:
>     _default_Func = staticmethod(call_me) #class wide default callback

now Callback_good._default_Func is a callable returning a
callable when called, therefore:

>     def __init__(self):
>         self._m_func = Callback_good._default_Func
>         
>     def __call__(self):
>         # uncomment one of these
>         #return self._m_func()() # for python < 2.2
>         #return self._m_func()   # for ptyton >= 2.2

Nope -- with your current definitions you need ()() in any
version of Python.


I have another preference, anyway:

class Callback_better:
    _m_func = staticmethod(call_me())  # record the final-callable
    def __call__(self): return self._m_func()

No need for __init__: self._m_func defaults to Callback_better._m_func
unless specifically set as an instance attribute of self.

This works either in 2.2, or 2.1 with the staticmethod class I
proposed (comes from the Python Cookbook, by the way -- both the
online version, and the enhanced one we edited into an O'Reilly
printed book:-0).


Alex



From loewis at informatik.hu-berlin.de  Mon Sep  9 10:13:59 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 09 Sep 2002 16:13:59 +0200
Subject: xml.minidom is stripping out my CRLF's in attrib values!!
References: <3D7C80C4.9070204@NOSPAMbitbuilder.com> <j4n0qr9yeo.fsf@informatik.hu-berlin.de> <3D7C8C76.8060700@NOSPAMbitbuilder.com>
Message-ID: <j41y839ro8.fsf@informatik.hu-berlin.de>

Ahmad Baitalmal <ahmad at NOSPAMbitbuilder.com> writes:

> That's not what my problem is, that I know about ( crlf's between nodes),,
> 
> Here is the deal:
> <cow>
> 	<hide value="spotted
> with black and white"></hide>
> </cow>
> 
> After the word "spotted" there is a crlf, -inside- the attribute
> value. Sholdn't it be treated as part of the value?
> 
> The value now comes stripped of that crlf.

What do you mean by "stripped of"? What is the exact value that you
get?

Regards,
Martin


From sholden at holdenweb.com  Sun Sep 29 06:18:34 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 29 Sep 2002 06:18:34 -0400
Subject: help
References: <amqclg$6uu$1@server.ltid.inpe.br>
Message-ID: <ktAl9.4777$pD4.140@fe05>

"Juliana Braga" <juliana at pdi.inpe.br> wrote in message
news:amqclg$6uu$1 at server.ltid.inpe.br...
> I need help.
> For example:
> x = ?name1?
> How can I get the value of variable x (in this case the value of variable
x
> is ?name1?). After , I?d like to do this:
> value of variable x = ?teste?.
>

Juliana:

Your question is a little difficult to understand, so you may get several
answers. I suspect that what you were looking for is:

>>> x = 'name1'
>>> exec x + "= 'teste'"
>>> name1
'teste'
>>>

While this works, however, please remember that almost all such uses of
"exec" are simply making trouble for you as a programmer. The code must be
compiled at run-time, and whenever you need to access the variables you
create in this way you have to resort to further use of the exec statement.

It will probably be more appropriate to use a dictionary to hold the values
you create. A dictionary can use many different datatypes as keys, and in
particular strings are fine. So you might want to try something like:

>>> adict = {}
>>> adict[x] = "teste"
>>> adict["name1"]
'teste'
>>>

There's a lot of similarity between dictionaries and Python namespaces (for
the altogether predictable reason that Python currently *uses* dictionaries
to implement namespaces). When you want to use program-determined "variable
names" then a dictionary is often the solution.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From ak at silmarill.org  Thu Sep 19 00:10:58 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Thu, 19 Sep 2002 04:10:58 GMT
Subject: Writing linux term color codes to a file
References: <e3gfouc157qqe6oj2i17g51edv8in2s26g@4ax.com> <am9s83$p6u$1@reader12.wxs.nl>
Message-ID: <h7jiou0081cotoeetkl6h28movlr7gvcb6@4ax.com>

On Wed, 18 Sep 2002 14:41:51 +0200, stasZ <stasZ at linuxisbeter.nl>
wrote:

>On Wed, 18 Sep 2002 03:09:37 +0200, Andrei Kulakov wrote:
>
>> Hello,
>> 
>> I ran into a bit of a problem here.. I'm making a small app that
>> visually creates bash prompts (just for fun of it), and at one point it
>> has to write the prompt to .profile, optionally with linux terminal
>> color codes, like \[044;\]m or something like that (you know what I'm
>> talking about). They don't come out right *in* the file. I tested the
>> simple case where I'd have a few such codes in a string in python prompt
>> and I'd look at it and it looks fine, but as soon as I write it out to a
>> file and look at the file, it's "cooked".
>> 
>> I hope you know what I mean, if anyone wants I can post direct pastes.
>> 
>> Looks like write() does some sort of processing on the text; is there
>> any way to turn that off?
>
>>>>prompt = r'\[044;\]'
>>>> prompt
>'\\[044;\\]'
>
>>>> f = open('/home/foo/prompt,'w')
>>>> f.write(prompt)
>>>> f.close()
>
>$ more prompt
>\[044;\]
>
>I think your problem is solved if you use a raw string (r'spam').
>With a raw string the backslash is just a backslash.

Thanks, I think that's it. I'll try it tonight. You've been doubly
helpful - that' sthe funniest sig I've seen this month. :P
>
>Stas Z



From rjones at ekit-inc.com  Wed Sep  4 20:13:48 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Thu, 5 Sep 2002 10:13:48 +1000
Subject: Daemonize?
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F06233022B66EC@mail.hebmex.com>
References: <F7DB8D13DB61D511B6FF00B0D0F06233022B66EC@mail.hebmex.com>
Message-ID: <200209051013.48248.rjones@ekit-inc.com>

On Wed, 4 Sep 2002 11:36 pm, sismex01 at hebmex.com wrote:
> Sorry to ask here, but I don't have any "unixy" python
> documentation at hand.
>
> Normally, I put an app into "daemon" mode, by first
> forking, then reopening stdin/out/err to /dev/null,
> and then forking again. Works good enough for my
> purposes.

Note you can also just assign an object with a write() method that does 
nothing to sys.std(out|err) and a similar object to sys.stdin. Why the second 
fork, btw?


> But, am I right? Is this the correct way to do this?
> Also, I can't find any "reopen" in python to apply
> to sys.stdin/out/err.

All that stuff is handled for you. Simply assigning to the std* attributes is 
enough. Note that in python 2.1+ the "old" sys.std(in|out|err) handles are 
still available as sys.__std(in|out|err)__ if you really need them. 


      Richard




From clspence at one.net  Wed Sep 11 17:17:10 2002
From: clspence at one.net (Chris Spencer)
Date: 11 Sep 2002 16:17:10 -0500
Subject: How do I capture just a single keystroke?
References: <pqavnu0vqp374e4155esnctk70on886ff7@4ax.com>
Message-ID: <7acvnu0p1maai2nnvle5ucjla1qk9v8k5u@4ax.com>

Nevermind...I found the answer.  For those who are interested, this is it.  It
should work on both Unix and Windows...
def getch():
	try:
		import msvcrt
		return msvcrt.getch()
	except:
		import tty,sys,termios
		fd = sys.stdin.fileno()
		old_settings = termios.tcgetattr(fd)
		try:
			tty.setraw(sys.stdin.fileno())
			ch = sys.stdin.read(1)
		finally:
			termios.tcsetattr(fd, termios.TCSADRAIN, old_settings)
		return ch

Chris.


From peter at engcorp.com  Sat Sep  7 00:14:00 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 07 Sep 2002 00:14:00 -0400
Subject: horizontal scrollitis? (was Re: Larry Wall's comment on python...)
References: <mailman.1031369109.16556.python-list@python.org>
Message-ID: <3d797cef$1@news.sentex.net>

mmaddox at hcsmail.com wrote:
> You're right. The PLACEMENT of blocks in my style is arbitrary. My
> intention was to indicate the STYLE of block delimiting was arbitrary

Probably also not true.  I believe many of the choices made for Python
in its early stages were made after both experience *and*
experimentation with alternatives.  Not sure about the block delimiters
(or invisibility thereof) themselves, but certainly things like the
use of the colon before a block were made because of studies suggesting
they were a Good Thing.

> Now, I LOVE Python. I also love C++. Smalltalk and Lisp are interesting.
> Perl is useful to someone somewhere. I said some things that are not as
> precise as I would have liked, and I don't want to explore the semantics
> of this mailing list and each individual post.	Python is cool. Let's talk
> about Python.

My post talked exclusively about Python, so I agree with that idea. :)
Maybe you would still respond to the other Python-related question
I asked:

 >>I definitely don't understand how Python, more or less than any other
 >>language, leads to more horizontal scrolling.  Pretty much all
 >>Python code I've ever seen (though I haven't seen yours yet :-) is
 >>roughly 80 characters max, and mine own has been about that since I
 >>started writing C code 16 years ago.
 >>
 >>Do you mean there is something special about Python not delimiting
 >>blocks in the traditional ways which somehow leads you to write
 >>much wider lines of code?

-Peter



From cmay4 at yahoo.com  Tue Sep 24 18:17:52 2002
From: cmay4 at yahoo.com (Chuck May)
Date: Tue, 24 Sep 2002 22:17:52 GMT
Subject: Limit number of concurrent threads
References: <Xns929393554CAB0cmay4yahoocom@165.112.130.8> <uwupbdwnj.fsf@online.no>
Message-ID: <Xns9293BA22C959Dcmay4yahoocom@165.112.130.8>

Ok...I think I've got it.  Thanks for everyone's help.

How does this look:

from threading import Thread
import Queue, time, random

maxthreads =  5  # maximum number of concurrent threads
totalops   = 80  # hold total number of operations

class Worker(Thread):
    def __init__(self, q):
        Thread.__init__(self)
        self.q  = q

    def run(self):
        all_done = 0
        while not all_done:
            try:
                table = self.q.get(0)
                time.sleep(random.randint(10, 1000) / 1000.0)
                print "id %d" % table
            except Queue.Empty:
                all_done = 1

if __name__ == "__main__":
    q = Queue.Queue()
    for i in range(totalops):
        q.put(i)

    threads = []
    for i in range(maxthreads):
        t = Worker(q)
        threads.append(t)
        t.start()

    # wait for all threads to complete
    for t in threads:
        t.join();

    print "Al done!"

Syver Enstad <syver-en+usenet at online.no> wrote in
news:uwupbdwnj.fsf at online.no: 

> Chuck May <cmay4 at yahoo.com> writes:
> 
>> I've searching for examples that fit my problem, and I haven't found 
>> any.  Most examples I've found look something like this:
>> 
>>     threads = []
>>     for i in range(number_of_tasks):
>>         thread.append(thread_for_task)
>>     
>>     # let the threads run - just wait for shutdown
>>     for t in threads:
>>         t.join() 
>> 
>> which whould work fine if I wanted to start all 80 threads at once.
>> 
>> I've read about the Queue, but I'm not sure if it's what I want, and
>> I 
>> 
>> can't seem to find any good examples on it.
> 
> I think Queue would be nice to use, Have a thread (the main thread
> perhaps) feed all the table names (or a table object if you please)
> into a Queue. 
> 
> Set up some worker threads with access to the Queue
> (global variable, member). The worker threads should loop something
> like this:
> 
> while 1:
>    queueItem = self._queue.get()
>    doSomethingWithQueueItemHere
> 
> Then you can start your workers.
> 
> (You'll have to find out some way to make the workers exit when the
> work is done, perhaps putting a number of None's equal to the number
> of workers last in the queue and testing for None in the working
> thread loop. 
> 
> 



-- 
Chuck May
IMS, Inc.



From marklists at mceahern.com  Wed Sep 18 17:17:38 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 18 Sep 2002 16:17:38 -0500
Subject: multiple index inconsistency
In-Reply-To: <TP5i9.9293$Rf3.1003108@newsfeed.slurp.net>
Message-ID: <JHEOKEOOLIGLDHCMAHMOOELCDPAA.marklists@mceahern.com>

[bert]
> The following line of code
> >>> a = 3*[range(3)]
> produces
> [[0, 1, 2], [0, 1, 2], [0, 1, 2]]
> 
> If I then write, say, 
> >>> a[0][1] = 3.7
> I get
> [[0, 3.7, 2], [0, 3.7, 2], [0, 3.7, 2]]
> and not
> [[0, 3.7, 2], [0, 1, 2], [0, 1, 2]]
> 
> Now if I write
> >>> a = [[0, 1, 2], [0, 1, 2], [0, 1, 2]]
> instead, and then 
> >>> a[0][1] = 3.7
> I get the result I expected (and wanted):
> [[0, 3.7, 2], [0, 1, 2], [0, 1, 2]]
> Why the inconsistency? Is this a bug or am I missing something?

Short answer:  lists are mutable.  ints aren't.

So when you do this:

  a = 3 * [range(3)]

you're basically doing this:

  l = range(3)
  a = 3 * [l]

The result is effectively:

  [[l], [l], [l]]

when you modify one of the lists, "all of them" are affected.

A better approach:

  x = 3
  y = 3
  a = [range(x) for x in range(y)]

(I assume x and y are different in theory.)

Now, try this:

  for x in a:
    print id(x)

and compare that to the list generated by 'a = 3 * [range(3)]'

Cheers,

// mark
-




From eworrall at netcomuk.co.uk  Sun Sep 29 13:15:19 2002
From: eworrall at netcomuk.co.uk (Eric Worrall)
Date: Sun, 29 Sep 2002 18:15:19 +0100
Subject: Linker can't find library (PyProlog setup)
References: <mdGl9.14197$Lg2.4064162@news2.news.adelphia.net> <0iGl9.14200$Lg2.4065036@news2.news.adelphia.net>
Message-ID: <3D973527.A9D39B4E@netcomuk.co.uk>

Robert Oschler wrote:
> 
> "Robert Oschler" <Oschler at earthlink.net> wrote in message
> news:mdGl9.14197$Lg2.4064162 at news2.news.adelphia.net...
> <snip>
> 
> Ugh, forgot to point out that I am running those commands from within a
> script file.
> 
> thx

Q&D solution - create symbolic links in the default library path and
point them to the libraries you want to include, for example:

ln -s /usr/lib/termcap/libtermcap.so.whatever /usr/lib/libtermcap.so

Eric Worrall

-- 
You have just received an Etech Solution
For all your Linux requirements contact
eworrall at netcomuk.co.uk


From aleax at aleax.it  Mon Sep 23 03:25:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 23 Sep 2002 07:25:24 GMT
Subject: mySQL and Python 2.2
References: <mailman.1032633449.19570.python-list@python.org>
Message-ID: <Elzj9.128797$ub2.2802895@news1.tin.it>

Greg & Janet LINDSTROM wrote:

> Hello-
> I am trying to convert from gadfly to mySQL for my database needs on
> my Windows 98 box. I have installed mySQL and it works (I have
> created my tables and can manipulate them manually), and have
> downloaded mySQL-python-0.9.1, but when I python setup.py build, I
> get an error about cl.exe not existing (and, sure enough, it
> doesn't :-).
> 
> I'm running Windows 98 with Activestate Python2.2. Any help you
> could supply (including code snippets of how to get connected to
> MySQL once I get it running) would be appreciated.

You seem to be trying to build from sources without having the
needed C compiler (that's what cl.exe is -- the executable that
calls the compiler and linker under Microsoft Visual C++).

If you don't have a C compiler, you can't build C-coded Python
extensions from sources: you need to find and install pre-built
binary versions of those extensions for the Python version you
are using.  You can find many such versions listed at (LONG
URL, mail sw might break it -- if so, then rejoin it...):

http://sourceforge.net/project/showfiles.php?group_id=22307&release_id=102893

Click on the link shown as (for example):
MySQL-python-0.9.1.win32-py2.2.exe
if you use Python 2.2 (there are other pre-built binaries
for older versions, should you need those).  Running that
EXE after downloading it should then install all you need.

I think you can also find pre-built binaries of MySQLdb 0.9.2,
for example at Gerhard H?ring's page, but I'm not sure about
the details -- Gerhard mentions said binaries at
http://www.cs.fhm.edu/~ifw00065/news.shtml but gives no
link and I can't find them.


Alex



From gerhard.haering at gmx.de  Fri Sep  6 19:50:26 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 6 Sep 2002 23:50:26 GMT
Subject: py2exe and a single exe file?
References: <3D792351.3070008@motorola.com>
Message-ID: <slrnanifq1.rh.gerhard.haering@lilith.my-fqdn.de>

Stephen Boulet wrote in comp.lang.python:
> I'm trying to get py2exe to work for me, and I must be doing something 
> wrong.
> 

> [...] It builds and runs, but only with the other stuff in the build
> directory with the generaged exe. The other stuff is:
> 
>    tcl directory containing tcl8.3 and tk8.3 directories
>    _numpy.pyd
>    _sre.pyd
>    _tkinter.pyd
>    multiarray.pyd
>    umath.pyd
> 
> What I would like to accomplish is just having a single exe file
> lumping everything together. Is this possible?

No. If you want to distribute the output of py2exe, use an installer
like NSIS or just zip it up.

Gordon McMillan's installer has means to do what you want, but it also
needs to unpack .pyd and .dll files at runtime, which it will delete
afterwards. So it basically cheats, which is logical, since you can't
statically link shared libraries, at least not on Windows.

-- Gerhard


From root at [127.0.0.1]  Mon Sep 16 22:37:43 2002
From: root at [127.0.0.1] (nobody)
Date: Tue, 17 Sep 2002 12:37:43 +1000
Subject: stdin on Win32
References: <3d865d34$0$22177$afc38c87@news.optusnet.com.au> <3d868e43$1@news.sentex.net>
Message-ID: <3d869577$0$22173$afc38c87@news.optusnet.com.au>

> Which flavour of Win32?

2K and XP

> Win98 needs you to explicitly invoke Python and pass the test.py name as a
command line

Yes that worked.

I guess I'll have to write a wrapper script around it..





From sholden at holdenweb.com  Tue Sep 17 12:23:47 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 17 Sep 2002 12:23:47 -0400
Subject: More on Protecting Source Code
References: <mailman.1032247760.982.python-list@python.org>
Message-ID: <UPIh9.11877$T4.5081@fe04>

"David LeBlanc" <whisper at oz.net> wrote in message
news:mailman.1032247760.982.python-list at python.org...
[ ... ]

> Well, I think of it this way: machine coded binaries are more like a 128
bit
> key and Python is more like a 40 bit key. I agree that nothing is safe
from
> reverse engineering; it's a matter of how much pain and money it takes to
> unscrut the inscrutable. Python could do better - how I'm not totally
sure.
>

Well, when you are, please submit a patch to SourceForge. Until then,
perhaps you could focus your efforts on solving the problem?

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From mhammond at skippinet.com.au  Sat Sep  7 20:04:08 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 08 Sep 2002 00:04:08 GMT
Subject: Using Python to call a DCOM server - "'Class not registered"
References: <3D787D95.9000907@ob_scure.dk> <5%ce9.1272$_l1.2960@news-server.bigpond.net.au> <3D7A1359.4010607@ob_scure.dk>
Message-ID: <Ytwe9.5037$_l1.12309@news-server.bigpond.net.au>

Thomas Jensen wrote:

> Is it safe to say that "ActiveX DLL" = InprocServer32 and "ActiveX EXE" 
> = LocalServer and that I need the later to use DCOM? (I have some 
> existing components that I'd rather keep as DLL - don't know why really? 
> :-)

Yep.  InProcServer32 means that the COM object is implemented in a DLL, 
and loaded into the calling process.  LocalServer means it is an 
executable, and loaded into a separate process and IPC used across 
boundaries.  For DCOM, there is no way the object can be loaded into the 
same process, so executable based servers must be used.

Mark.



From david.abrahams at rcn.com  Fri Sep 27 21:44:33 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Fri, 27 Sep 2002 21:44:33 -0400
Subject: Is Boost Python version 2 ready?
References: <MvZk9.12124$Lg2.3584877@news2.news.adelphia.net> <an2pt9$g5h$1@bob.news.rcn.net> <mailman.1033173996.13785.python-list@python.org>
Message-ID: <an339d$bqm$1@bob.news.rcn.net>

From: "Christian Tismer" <tismer at tismer.com>


> David Abrahams wrote:
> > From: "Christian Tismer" <tismer at tismer.com>
> >
> >>I have been asked whether Boost will be compatible with
> >>Stackless (or vice versa:).
> >>Who else could answer that better than you?
> >
> >
> > Well, I have never tested it with Stackless, but in the end it only does
> > "ordinary extension module stuff". What is it, precisely, you need to
watch
> > out for? Something about callbacks from C++ into Python with data on the
> > stack?
>
> Exactly :-)
> Probably nothing you would do unless enforced to?

AFAIK I'm never doing that myself. It's possible for someone to write a
construct in an extension module that does that, but they'd have to go out
of their way to shoot themselves in the foot.

HTH,
Dave

-----------------------------------------------------------
           David Abrahams * Boost Consulting
dave at boost-consulting.com * http://www.boost-consulting.com





From shalehperry at attbi.com  Wed Sep 11 17:53:23 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 11 Sep 2002 14:53:23 -0700
Subject: [FEEDBACK] Is this script efficient...is there a better way?
In-Reply-To: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net>
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net>
Message-ID: <200209111453.23751.shalehperry@attbi.com>

On Wednesday 11 September 2002 14:01, Bob X wrote:
>
> # build list of keywords
> kw = [ "some", "words" ]
>
> # loop through the list and print the lines to a file
> for line in inFile.readlines():
>      for badword in kw:
>          if line.find(badword) > -1:
>              result = '%s %s' % (badword, line)
>              print result            # Print the result
>              outFile.write(result)   # Write the result
>
> # close the files
> inFile.close()
> outFile.close()
>
> # let me know when it's done
> print "Finished processing file..."

1) readlines() loads the entire file into a list so if you have a 30+ mb file 
you just ate 30+mb of memory.  Try using xreadlines() instead, it reads the 
file line by line and is much more memory friendly.

2) do you expect to find more than one keyword in a particular line?  If not 
you could save some iterations by stopping the inner line.find() loop as soon 
as one item is found.

As a final comment you need to be aware that the more keywords you look for 
the slower this will be.  However there is not a way to get around that, it 
is just something to keep in mind.



From aglyportREMOvethispart at nospam.yahoo.com  Wed Sep  4 20:11:37 2002
From: aglyportREMOvethispart at nospam.yahoo.com (Anton Graph)
Date: Wed, 04 Sep 2002 17:11:37 -0700
Subject: Memory fault - core dumped   while compiling python 2.2.1 on
 SCO
References: <3D757EFA.9050401@nospam.yahoo.com> <yu99sn0qqw1r.fsf@europa.research.att.com> <m3bs7etem9.fsf@mira.informatik.hu-berlin.de> <3D765A05.6030004@nospam.yahoo.com> <j48z2h61j9.fsf@informatik.hu-berlin.de> <3D768753.8030302@nospam.yahoo.com> <m3n0qx8j3w.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3D76A139.1030304@nospam.yahoo.com>

Martin v. Loewis wrote:
> Anton Graph <aglyportREMOvethispart at nospam.yahoo.com> writes:
> 
> 
>>python -E setup.py build in gdb:
>>
>>(gdb) where
>>#0  0x80016d06 in do_reloc () from /usr/lib/libc.so.1
> 
> 
> That sounds like a question to bring to an SCO forum. Apparently, this
> crashes when doing relocations while loading the extension
> module. This raises a number of questions:
> 
> - Why does it do relocations at all? Could it be that you have object
>   files that a not position-independent? I.e. is distutils using the
>   compiler correctly both for compilation, and for linking?
> 
>   It might be correct that it tries to perform relocation, e.g. for
>   the global offset table, and the procedure linkage table - but it
>   *is* troubling that relocations is where it crashes.
> 
> - What is the precise version of the operating system you are using?
> 

SCO_SV bongo 3.2 5.0.5 i386

> - I'm troubled by the fact that in both configurations (with an
>   without gcc), the compiler is always referred-to as "cc". Are you
>   sure you are not using gcc?
> 
I'm positive it's SCO cc:

Wed Sep  4 16:17:04 PDT 2002 bongo:/u/src/agl/txb/txt $ gcc -v
Reading specs from /usr/local/lib/gcc-lib/i386-pc-sco3.2v5.0.5/2.95.2/specs
gcc version 2.95.2 19991024 (release)
Wed Sep  4 16:17:07 PDT 2002 bongo:/u/src/agl/txb/txt $ /usr/local/bin/cc
bash: /usr/local/bin/cc: No such file or directory
Wed Sep  4 16:17:14 PDT 2002 bongo:/u/src/agl/txb/txt $ /usr/bin/cc
Usage: cc [ options ] files ...
Wed Sep  4 16:17:18 PDT 2002 bongo:/u/src/agl/txb/txt $ /bin/cc
Usage: cc [ options ] files ...


>   Please have a look at
> 
>   http://gcc.gnu.org/ml/gcc-testresults/1999-04/msg00239.html
> 
>   Notice that the failure mode is quite similar... Likewise,
> 
>   http://www.bitmechanic.com/mail-archives/dbi-users/Jun1999/0485.html
> 
>   reports such a problem when using gcc on SCO.

I see. I doubt the problem is with the compiler though because I have a 
very similar problem with SCO cc. Maybe it's buggy dynamic linker, libs 
or 5.0.5 kernel?


This maybe irrelevant, but
the reason why I thought of migrating to python 2.2.1 was because my 
program linked against python 1.4 libs all of a sudden began to crash in 
memory allocation routines. I linked against Electric Fence and it shown 
a problem caused by this code:
....
PyObject *tuple = PyTuple_New(2);
....
  PyTuple_SET_ITEM(tuple, 1, layout[1]);
^^^^^^^^^^^^^^^^^^^^^^^^^^ bails out with SIGSEGV whenever linked 
against Electric Fence and python 1.4 libs.



Could someone out there tarball Python 2.2.X built on SCO and put it on 
a ftp site?



From aleax at aleax.it  Tue Sep 24 10:18:43 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 24 Sep 2002 14:18:43 GMT
Subject: Three dumb questions (ordered by dumbness descending)
References: <amnvit$7kqdt$1@ID-77524.news.dfncis.de> <82Zj9.136094$ub2.2966822@news1.tin.it> <2HZj9.260778$AR1.10903907@bin2.nnrp.aus1.giganews.com>
Message-ID: <7v_j9.137081$ub2.2977680@news1.tin.it>

Terry Reedy wrote:

> 
> "Alex Martelli" <aleax at aleax.it> wrote in message
> news:82Zj9.136094$ub2.2966822 at news1.tin.it...
>> Once you do have a random or pseudo-random generator that fully
>> satisfies you, Ian Bicking's snippet reimplementing shuffle as a
>> Decorate-Sort-Undecorate idiom is truly excellent -- a little jewel,
> 
> How is an O(n*log(n)) algorithm a 'jewel' compared to the simple and
> standard O(n) algorithm?

Not performance-wise, by any means -- it just LOOKS deucedly pretty!-)


>> It does, alas, end up a little slower than Tim Peters' original --
> [...]
>>and Ian's version only loses very very little indeed.
> 
> Increase n enough and it becomes arbitrarily slower and loses
> arbitrarily much.

You're right -- I should have mentioned that.  If you must shuffle
lists whose length N you cannot bound, AND you have no other
bottlenecks in your program that are worse than O(N log N), then
introducing an O(N log N) bottleneck is clearly out of the
question.


Alex



From thorsten at thorstenkampe.de  Tue Sep 10 17:54:31 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Tue, 10 Sep 2002 23:54:31 +0200
Subject: Python primer - comments appreciated!
References: <aljdhe$1q63hj$1@ID-77524.news.dfncis.de> <w3cf9.535714$m91.21695020@bin5.nnrp.aus1.giganews.com>
Message-ID: <allpms$1r005a$1@ID-77524.news.dfncis.de>

* Terry Reedy
> "Thorsten Kampe" <thorsten at thorstenkampe.de> wrote in message
> news:aljdhe$1q63hj$1 at ID-77524.news.dfncis.de...
>> def quotient_set(seq, func, partition='nonbool'):
>>     """ partition <seq> into equivalence classes
>
> To my mind, you should return a dictionary mapping each realized value
> of func (assuming hashable)  to the list of items for which func has
> that value.  In other words, result.values() would be the list of
> equivalence classes.  Let caller toss away info of value defining
> class.  In any case, using a dict to sort items into such classes
> should be much faster that what you wrote.
>
> def equiv_class(seq, func, part = 'nonbool'):
>  if part == 'bool':
>    result = {False:[], True:[]}
>    for item in seq:
>      result[bool(func(item))].append(item)

Using the 'bool' function and dictionaries makes my whole "if part == 
'bool'" branch superfluous - see below.

>      result.setdefault(func(item),[]).append(item)

To make every item hashable, I made this:
result.setdefault(repr(func(item)),[]).append(item)

>      # {}.setdefault(key,default) was added for just this sort of usage

Sorry, I just don't get it: your mixing (chaining?) a dict method and 
a list method as if it was a f(g(x)) and /it works/!?! Where is that 
documented? How is is that evaluated (left to right)?

> >>> f = lambda x: x % 3
> >>> equiv_class([0,1,2,3,4,5,6,7,8,9],f,part='bool')

Now this is: 
quotient_set([0, 1, 2, 3, 4, 5, 6, 7, 8, 9], lambda x: bool(x % 3))

(Mathematically spoken, the quotient set is the set of all equivalence 
classes induced by func, which is what we return)


Thorsten


From aahz at pythoncraft.com  Wed Sep 11 13:06:13 2002
From: aahz at pythoncraft.com (Aahz)
Date: 11 Sep 2002 13:06:13 -0400
Subject: Multi-threading python
References: <almvvt$34p$1@news.netpower.no> <mailman.1031751975.3372.python-list@python.org> <alnmcs$gdt$1@news.netpower.no>
Message-ID: <alnt65$qn8$1@panix1.panix.com>

In article <alnmcs$gdt$1 at news.netpower.no>,
David Brown <david at no.westcontrol.spam.com> wrote:
>
>If C code (or other extensions) that take significant time, or block for
>other reasons, release the GIL then there I suppose there is not going to be
>much of a problem.  The point is to maximize the throughput, and avoid
>threads blocking each other unnecessarily.

Yup.

>It sounds like that is only true for code that does a lot of work in pure
>python (rather than with time-consuming, GIL-released C code).

Yup.

>Thanks for the information - it was definitely helpful.  It looks like the
>system used by Python will work fine in all but a few situations (SMP system
>in which a lot of work is being done in Python itself, rather than in
>extension modules - or in badly written extensions which don't release the
>GIL before doing heavy work).  Such situations can be dealt with by running
>two python processes - if a program really needs to do hard SMP work in
>python, then that is a small price to pay since I image the GIL simplifies a
>lot of things.  I had been worried that the Python threading scheme would
>have caused more inefficiences (such as all threads stopping while a single
>thread was blocked for I/O).

Yup.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From see_reply_address at something.invalid  Wed Sep 11 21:49:13 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Thu, 12 Sep 2002 13:49:13 +1200
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid> <b16e4ef7.0209090552.3ad7691b@posting.google.com> <V57f9.77078$l_4.29066@atlpnn01.usenetserver.com> <b16e4ef7.0209101725.6f508364@posting.google.com>
Message-ID: <3D7FF299.2020509@something.invalid>

Stephen Ferg wrote:

> What I wanted to do, though, in "Thinking in Tkinter" was more of an
> explanation of the basic concepts that you need in programming with
> Tkinter.  From that standpoint, I think it is better to explain about
> binding -- that is, binding widgets, events, and callback handlers --
> by explaining "bind()".  Then, once the basic concepts are explained,
> I can note how "command" can be used as a labor-saving technique.


It might be better to start with some object
other than a button for the purpose of introducing the
concepts of event binding. Otherwise you're teaching
people an incorrect way of using buttons that they're
going to have to unlearn later.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From h.g.normann at gmx.de  Fri Sep 27 17:05:14 2002
From: h.g.normann at gmx.de (Hans - G. Normann)
Date: Fri, 27 Sep 2002 23:05:14 +0200
Subject: how to create QLineEdit.focusOutEvents?
Message-ID: <an2eos$9v7m1$1@ID-143433.news.dfncis.de>

Hello

I'm new in QT and Python. My first project: creating a GUI for Linux tar 
command. My prob:

A QLineEdit contains a directory name. I can create this name by clicking 
(Fileselection maodule) or by type in. On lost focus, I want check the 
Directoryname, but how can I create slots, signals and the event for Line 
Edit?

For creating gui, I'm using QT's designer. It's works fine, except 
focusOutEvent. Is there a (easy) sample?

Hans


From jubafre at brturbo.com  Tue Sep 17 15:57:16 2002
From: jubafre at brturbo.com (jubafre at brturbo.com)
Date: Tue, 17 Sep 2002 16:57:16 -0300 (GMT-03:00)
Subject: hexacodes in python?
Message-ID: <-1995099244.1032292636635.JavaMail.nobody@webmail1>

what is the hexacode for '?' in python?
where can i take the hexacodes?	
	
Juliano Freitas
www.gebrasil.hpg.com.br 

From LogiplexSoftware at earthlink.net  Thu Sep  5 14:28:14 2002
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 05 Sep 2002 11:28:14 -0700
Subject: Gratitude due to Red Hat? More work for the PBF?
In-Reply-To: <xuyptw3knfx.fsf@halden.devel.redhat.com>
References: <Iyn89.11545$aC5.4453@atlpnn01.usenetserver.com>
	<mailman.1030142061.12228.python-list@python.org> 
	<xuyptw3knfx.fsf@halden.devel.redhat.com>
Message-ID: <1031250495.19985.1180.camel@software1.logiplex.internal>

On Wed, 2002-08-28 at 08:33, Trond Eivind Glomsr?d wrote:
> Cliff Wells <LogiplexSoftware at earthlink.net> writes:
> 
> > On Tue, 2002-08-20 at 01:59, Steve Holden wrote:
> > > I'm a little hesitant to say this, but today (while, of all things, putting
> > > a sales proposal together) I realised that my attitude to the "Red Hat comes
> > > with 1.5.2" question has now altered. When considering a hosting environment
> > > I am happy to *exclude* any company who can only offer 1.5.2, as they are
> > > clearly not serious about supporting the language environment. This is a
> > > positive thing, as supporting Python isn't yet a common enough skill to be
> > > available from the Interlands and the like.
> > 
> > I submitted a bug report to RH regarding this issue and was basically
> > brushed off with "the RH tools that depend on Python should all work
> > with 2.x if you install all the required packages" [required packages
> > thoughtfully omitted].  Heaven forbid *they* should fix anything.
> 
> Fixing it would mean it was a problem.
> 
> A program originally using python 1.5 and newt would, in almost all cases, work
> just as well with python 2.1 if the newt module was available.

And what of 2.3?  2.4?  RedHat's response assumes
backwards-compatibility of future versions of Python with all of their
scripts and modules.  Unless Guido has loaned them his time-machine this
probably isn't a safe assumption.  Besides "in almost all cases" isn't
exactly reassuring.  

The reality is that most Python developers stay up-to-date with releases
of Python and the RedHat-sanctioned version lags behind by months (you
refer to 2.1 when I've been using 2.2 for quite some time now).

The aggravating thing is that this issue would be a snap for RedHat to
fix.  Instead of preempting /usr/bin/python for their own use, they
could maintain a /usr/bin/python-rh and leave python for user-installed
versions (instead of requiring the user version to be called "python2").

Problem solved, everyone happy.

> -- 
> Trond Eivind Glomsr?d
> Red Hat, Inc.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From shalehperry at attbi.com  Wed Sep  4 11:58:12 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 4 Sep 2002 08:58:12 -0700
Subject: Regexp Over Multiple Lines
In-Reply-To: <565e316f.0209040740.36573b56@posting.google.com>
References: <565e316f.0209040740.36573b56@posting.google.com>
Message-ID: <200209040858.12863.shalehperry@attbi.com>

On Wednesday 04 September 2002 08:40, Keith wrote:
> In Python can I have a regular expression that spans multiple lines...
> e.g. In Perl I may define a regular expression like:
>

yes you pass the re.compile() function an option of re.WHITESPACE as I recall 
(but check the docs to be sure of the exact option).



From awinston at scn.org  Tue Sep 10 11:42:57 2002
From: awinston at scn.org (Alan Winston)
Date: Tue, 10 Sep 2002 08:42:57 -0700
Subject: PySol: "global name 'win32_gethomedir' is not defined"
Message-ID: <all3u2$qgm$1@brokaw.wa.com>

I've been trying to run the PySol 4.81 pyc files under Python 2.2.1 and
WinXP and Win98, with no success.

The Python window closes much faster than I can read the error message (is
there any way to force the window to stay open?), but I managed to catch it
with the print screen key:

"global name 'win32_gethomedir' is not defined"

PySol 4.60 runs fine, installed with the Windows install package, so this is
a low priority for me, but I am still curious.




Alan Winston
Seattle






From emile at fenx.com  Thu Sep  5 00:59:04 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 4 Sep 2002 21:59:04 -0700
Subject: newbie-question: Nested functions and variable scope
References: <7a7157fd.0209041503.7b9a0f44@posting.google.com>
Message-ID: <al6l5i$sjq$1@main.gmane.org>

Leonardo:
> If I put "in_reply" outside everything and use "global in_reply"
> inside each nested function, it works, but that's evil! What would be
> a better way of doing it? Do I need to use a class?

Take a look at nested_scopes:

ActivePython 2.1, build 210 ActiveState)
based on Python 2.1 (#15, Apr 19 2001, 10:28:27) [MSC 32 bit (Intel)] on
win32
Type "help", "copyright", "credits" or "license" for more information.
>>> def level1():
...     ir = 1
...     def lvl2b():
...             print ir
...
<stdin>:1: SyntaxWarning: local name 'ir' in 'level1' shadows use of
'ir' as glo
bal in nested scope 'lvl2b'
>>> from __future__ import nested_scopes
>>> def level1():
...     ir = 1
...     def lvl2b():
...             print ir
...


HTH,

--

Emile van Sebille
emile at fenx.com

---------






From maxm at mxm.dk  Tue Sep 24 04:53:04 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 24 Sep 2002 10:53:04 +0200
Subject: file = open(path,'wb')?????
References: <mailman.1032828810.21572.python-list@python.org>
Message-ID: <3D9027F0.4080603@mxm.dk>

jubafre at brturbo.com wrote:


> i open the a file for writing in a binary mode, but doesnt work
> 
> final3=['0020 0000x9', '0030 0000xa', '00B4 0000x8', '0030 0000xb', '00F0', '0001', '0003', '0005']
> destino=raw_input("Digite o nome do arquivo .mem: ")
> f = open(destino,'wb')
> f.write('\x03AHM ')
> for i in final3:
>     f.write(i+' ')
> f.close()
> 
> 
>  i open the file in the notepad and my file appears like a text file and not as a binary file, i dont know why??
> 
> 
>>>file.mem
>>
> AHM 0020 0000x7 0030 0000x8 0010 0000x9 00F0 0003 0002 0000 
> 
> 
>>>i want the file like this, how??
>>
> AHM  &#8364; 0 ?  &#402; ? 


You are indeed writing strings to the file. Setting a file to binary 
mode only affects a few newline characters on Windows. Nothing else.

You should use the struct module for writing binary values.

4.3 struct -- Interpret strings as packed binary data

regards Max M



From LogiplexSoftware at earthlink.net  Fri Sep 20 09:36:53 2002
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 20 Sep 2002 06:36:53 -0700
Subject: More on Protecting Source Code
In-Reply-To: <3D86E466.6010208@cascade-sys.com>
References: <GCEDKONBLEFPPADDJCOEAEDDFAAA.whisper@oz.net>
	<20020917062412.GA85457@hishome.net>  <3D86E466.6010208@cascade-sys.com>
Message-ID: <1032529013.20661.194.camel@localhost.localdomain>

On Tue, 2002-09-17 at 01:14, James J. Besemer wrote:
> 
> Or if I reduce those thoughts to paper or to software or to wood and 
> metal that somehow everybody else is entitled to share the benefits of 
> my possibly substantial endeavor?  
> 
> What a load of bullshit.

Sort of.  The idea of "owning" ideas/information is indeed artificial. 
I don't see how you can argue otherwise.

However there are good reasons for having such a concept (as you outline
below).  However, also note that there are exceptions to intellectual
property laws (in the U.S. anyway) where the rights of the author/owner
are occasionally diminished in the interest of the common good.  It's a
balancing act between the two interests (creator and public).  The
problem that many (including myself) have with the current state of
copyright/patent law is that it is often enforced in such a way that the
interest of the public is disregarded.  Consider the fact that it is
still illegal to copy software contained in the ROMs from an Apple //e. 
Clearly Apple no longer has any monetary interest in maintaining the
copyright on that software (the hardware for which it was written is no
longer sold) and there is possibly some benefit to releasing it to the
public domain (for emulators and whatnot, or possibly as examples for
education), but the law continues to protect Apple's "interest" in this
piece of code for no apparent reason.  I'd suggest that 5 years is
adequate time to protect a piece of software, not 35 or 95 or 120 or 70
years after the death of the author (depending upon the particular work
being considered).

> An artist who creates an image or the specifications for a musical 
> performance has every right to own his creation and benefit from its 
> use.  This is no different for a programmer and code of his own creation.  

It is different in that software has a much shorter useful lifespan than
a work of art.  A book written in 1976 can still be sold and hence has
some value.  I'll sell you my copy of Wordstar for CP/M, if you're
interested.  I'm guessing you're not.

> Even if you go back to the early industrial age, key products of 
> industry was intellectual property.  E.g., engineers had to produce 
> drawings and written specifications before they could build a bridge or 
> a mousetrap.  That "information" was every bit their property as the 
> more tangible artifacts they later sold commercially.  An unscrupulous 
> competitor could use the IP to recreate the commercial artifacts at a 
> fraction of the original owner's cost.  Or he could reverse engineer the 
> IP and again enjoy an unfair  advantage.  That's blatantly unfair and 
> it's a good thing we have laws to protect against such theft.

All true.

> Now if somebody elects to give away their IP -- out of charity or in 
> hope of scoring points in some circles -- that's fine.  But if he 
> instead elects to profit from it, in a free world, that should be 
> perfectly reasonable and respectable also.
> 
> >but it
> >can be made more property-like by laws that restrict people's freedom to
> >use information. 
> >
> 
> Copyright laws have precedent dating back to pre colonial England. 
>  Interesting new books and music were intrinsically hard to create but, 
> (after Gutenberg) once created, they were easy for the unscrupulous to 
> copy and receive full benefit of their publication.  Robbing the 
> creators the fruits of of their labor is inherently unfair.  Not only 
> does the IP thief unfairly benefit from work he did not perform he also 
> robs the real owner of income he otherwise would have received.  
> 
> Thus IP laws naturally evolved to protect rightful property owners. 
>  It's meet and good.
> 
> People fought and died (and more importantly prevailed) in defense of 
> the rights to be free and to own property.

And others have fought and died to free property from the hands of a
privileged few.  What does this prove?  Nothing except that people fight
over property.

> Now I don't necessarily support some of the more recent changes and 
> extensions to copyright laws.  Some maybe go a step too far.  However, I 
> don't think they can or should be rejected altogether.  At their core 
> they're not arbitrary; rather in protecting the works' creators they're 
> the epitome of fairness.

The problem is that these laws are usually created as the result of
lobbying by groups that have a specific interest in protecting their
intellectual property and the interests of the public are just as often
neglected.  For instance, there is a "fair use" clause for copyright
that allows individuals to make archival copies of copyrighted works.  I
wonder what happens to that right when I buy a copy-protected CD of a
game?  Who is lobbying to protect that right?  The RCIAA?  Not likely.

 
> >The state has given you the 'right' to use the power of its court and 
> >police to forcibly restrict people that use information you have created 
> >in violation of copy-restriction laws.
> >
> 
> Yeah, sure, in your dreams we could instead have laws that say all 
> information was free.  I submit that such a situation would actually 
> stifle innovation.  Sure a few dreamers would still invent but the 
> volume and time to market would be greatly diminished.  
> 
> Lack of ownership is precisely why most socialist countries soon became 
> and forever remained vast economic wastelands. 

It may have been one reason, among many.  Gross over-simplification
weakens your argument.

 Denying individuals any 
> incentive for working harder than others, the economy soon deteriorates 
> to where it can hardly provide the necessities of life let alone invent 
> clever and innovative luxuries.  Worse, collectivist economics are 
> unnatural in that they fly in the face of most human instincts.

Probably true.

  Thus it 
> requires a large and oppressive government to enforce collectivism and, 
> in practice, those governments typically become corrupt.

The two or three examples we have available to examine would seem to
support this.  However, the common American view of equating an economic
system (socialism, communism) with a government system (totalitarianism,
dictatorship) is flawed.  It doesn't seem impossible that there could
exist a democratic socialist state any more than it would be impossible
to have a capitalist dictatorship.

  In any case, 
> economies invariably, immediately revert to capitalism soon as the 
> communist/socialist oppression ceases.

And Russia's economy has flourished ever since ;)  Even if we disregard
how U.S. trade with Russia has changed since the cold war (and its
effect on their economy) there isn't much of an argument to be made
here.

> It is precisely the promise of commercial reward that makes America the 
> incredible fount of invention and creativity that it is.  

And of course the natural resources to make those grand ideas a reality.

The simple 
> formula of rewarding people for their own hard work is what made America 
> the worlds only superpower.  Not only do we have the most powerful 
> military in the world our people also enjoy one of the world's highest 
> standards of living.  

And not at all because we have almost unlimited natural resources or
because we use our powerful military to make sure that we obtain other
countries' natural resources at a "reasonable" price.

Most socialist countries have had to choose one or 
> the other or don't have the choice of either.
> I just don't get you folks who push this Communistic view of economics. 
>  It's childish.  Marx sounded good on paper 100 years ago and yeah we 
> all believed it when we were idealistic teenagers. At the cost of 
> hundreds of millions of citizens (murdered by their own government) and 
> a significant fraction of the 20th century's world's human resources, we 
> gave Communism every chance to succeed.

Ah, by giving them a chance to succeed, you must mean by imposing
economic sanctions and encouraging them to bankrupt themselves on
military spending.  To pretend that the USSR collapsed without help from
outside forces is stretching it a bit.

  After almost a century of 
> experimentation with collectivist economics, the results are all in and 
> the bottom line is that it simply doesn't work.  The net cost to The 
> People in the long run is always greater than the net benefits.  If you 
> don't understand this then you really haven't been paying attention to 
> recent world history.

And most of these countries were in the same position prior to adopting
socialism.  And they'll be in the same position after they switch to
capitalism.  Name one wealthy country that switched to socialism and
went broke.  Name a poor socialist country that switched to capitalism
and became wealthy (not counting billion dollar handouts the U.S.
rewards these governments with - perhaps that's where some of the
motivation to switch to capitalism comes from?).

> The problem with Capitalism is that not everyone shares equally in the 
> wealth.  The problem with Communism is that everybody DOES share equally 
> in the misery.  I prefer Capitalism.

I don't disagree with your conclusion, just how you got there.  Your
arguments sound too much like cold war propaganda.  

> 
> >But that is not enough for you. You also want the 'right' to complain about
> >something that was given to you freely and expect that righteous rants 
> >about your view of intellectual property laws will somehow compel people 
> >to modify it to your specification without compensation.
> >
> 
> You're being unfair.
> 
> Mr. LeBlanc was not complaining about anythinthe fact that during the Cold War the USSR was g that I could see.  He was 
> merely describing a tangible shortcoming of Python (one which launched 
> this entire thread) that restricts its use for some purposes.  He 
> further, humbly suggests how it might be made a little more useful to 
> commercial users.

And at last we agree.  I didn't see a complaint in David's post or
indeed anything that required such a response.

> --jb
> 
> -- 
> James J. Besemer		503-280-0838 voice
> 2727 NE Skidmore St.		503-280-0375 fax
> Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
> 				http://cascade-sys.com	

You know, I work like 2.1 miles from your house (as maps.yahoo.com
flies).  We could add two to the millions who have fought and died over
this topic, if you like ;)

Anyway, taking this to alt.neverending.arguments...

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From rjones at ekit-inc.com  Sun Sep 29 08:21:36 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Sun, 29 Sep 2002 22:21:36 +1000
Subject: What does Python fix?
In-Reply-To: <87znu1rugt.fsf@pokey.henrik-motakef.de>
References: <e15b6fa3.0209271418.14cd5642@posting.google.com> <vcmcpusnjqvpj9lavtqlqa7cs01h4i1iij@4ax.com> <87znu1rugt.fsf@pokey.henrik-motakef.de>
Message-ID: <200209292221.36348.rjones@ekit-inc.com>

On Sun, 29 Sep 2002 7:55 pm, Henrik Motakef wrote:
> Courageous <jkraska at san.rr.com> writes:
> > >Let's not forget that Python's syntax (most of all the significance of
> > >whitespace) is considered strange at least by about all novices. In
> > >both cases - significant whitespace and lots of irritating superflous
> > >parentheses - people don't just get used to it, but tend to like it
> > >after a while.
> >
> > I'm fairly sure that you meant to say "Lisp" where you said Python,
> > right?
>
> No, I'm talking about how newbies tend to react when you tell them
> that block structure in Python is determined by indentation.

What do newbies care? I'm sure if they have a problem with indentation for 
blocks, they're going to have more trouble with remembering to put additional 
punctuation either side of the blocks, in the right places...

Unless, of course, your "newbie" is actually someone who has had the C-style 
brace crap engraved on their mind before encountering Python...

Of course, if you're talking from experience with having taught newbie 
programmers, and found that "about all" of them consider Python's blocking 
syntax "strange", then I'll happily take all this back :)

And if those same people don't find lisp's syntax (and hell, lisp in general) 
"strange", then I'll doubly apologise :)


     Richard




From david.abrahams at rcn.com  Sun Sep 15 13:22:31 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Sun, 15 Sep 2002 13:22:31 -0400
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org>
Message-ID: <am2g4a$o6h$1@bob.news.rcn.net>

"Christian Tismer" <tismer at tismer.com> wrote in message
news:mailman.1032107006.6411.python-list at python.org...

> The changes to Python are also so small now, that
> it doesn't matter if it will be merged or not.
> Porting Stackless to a new Python is a matter of a few
> hours. You may rely on getting a Stackless version
> of any new Python version in almost no time.

I think it still matters. You don't want people who need stackless'
capabilities to have to use a non-standard Python forever, do you?

--
-----------------------------------------------------------
           David Abrahams * Boost Consulting
dave at boost-consulting.com * http://www.boost-consulting.com





From skip at pobox.com  Wed Sep 18 19:56:49 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 18 Sep 2002 18:56:49 -0500
Subject: linking extension modules w/ "ld" on Solaris - to be expected?
Message-ID: <15753.4801.895937.802904@12-248-11-90.client.attbi.com>

On a Solaris machine I configured Python (2.1.3 at the moment, but 2.2.1 is
in the wings) using:

    ./configure --with-gcc=mycc --with-cxx=mycc

where "mycc" is a shell script which decides whether or not to invoke Sun's
C or C++ compiler.  Some (but not all) 3rd party extension modules need to
be compiled with C++, but I don't want to force all to be C++-compiled.
(Gcc is not an option here.)

I noticed in the configure output that LDSHARED got set to "ld -G".  Is this
typical for Solaris when building with a non-gcc compiler?

-- 
Skip Montanaro - skip at pobox.com
The need for gutters to be cleaned is directly proportional to how hard it
happens to be raining at the moment.



From sholden at holdenweb.com  Sat Sep  7 17:28:54 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 7 Sep 2002 17:28:54 -0400
Subject: web crawler help?
References: <JARd9.6927$yt3.3340577@newssrv26.news.prodigy.com>
Message-ID: <_due9.18451$l_4.1926@atlpnn01.usenetserver.com>

"koko" <kokohh at hotmail.com> wrote in message
news:JARd9.6927$yt3.3340577 at newssrv26.news.prodigy.com...
> is there any sample for basic web crawler, that ask for a starting url and
> log the url and extract the hyperlinks?
> thx


Take a look at tools/websucker in your Python distribution (you might need
to download it if you have used a limited binary installer).

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                        pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From rpm at wag.caltech.edu  Mon Sep  9 13:40:32 2002
From: rpm at wag.caltech.edu (Rick Muller)
Date: Mon, 09 Sep 2002 10:40:32 -0700
Subject: Can't exponentiate zero in NumPy
References: <mailman.1031591234.6624.python-list@python.org>
Message-ID: <alimef$lc9$1@naig.caltech.edu>

Tim Peters wrote:
> 
> Does this require wrapping in a Numeric array, or does it also happen in
> *any* of these straight Python expressions?:

Turns out it appears to be a more general problem than I thought.

> 
>     1e-310 ** 2

Works on RH linux/1.5.2, overflows on RH linux 2.2

>     pow(1e-310, 2)

Works on linux 1.5.2, overflows on linux 2.2

>     import math
>     math.pow(1e-310, 2)

Works on linux 1.5.2, works on linux 2.2 (??!)
> 
>     1e-310 ** 2.1

Works on linux 1.5.2, fails on linux 2.2

>     pow(1e-310, 2.1)

Works on linux 1.5.2, fails on linux 2.2

>     math.pow(1e-310, 2.1)
> 

Works on linux 1.5.2, works on linux 2.2
> In addition, exactly which version of Python are you using (2.0, 2.1, 2.1.1,
> 2.1.2, 2.1.3, 2.2, 2.2.1, ...)?

These are the standard python/python2 distributed with RedHat Linux 7.3.

"Works" means returns 0.0
"Fails" means raises an OverflowError

Can someone suggest a decent workaround?

Rick



From from_deja at alandaniels.com  Wed Sep 25 11:20:48 2002
From: from_deja at alandaniels.com (Alan Daniels)
Date: 25 Sep 2002 08:20:48 -0700
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <mailman.1031377930.4225.python-list@python.org> <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com> <3D8E6353.7030303@something.invalid> <dvbtouciidaufjneja3s4t89ds0r7jd76k@4ax.com> <3D911A1B.1000109@something.invalid>
Message-ID: <a6f9355c.0209250720.4512b3c3@posting.google.com>

Greg Ewing <see_reply_address at something.invalid> wrote in message news:<3D911A1B.1000109 at something.invalid>...

[snippage...]
> You use tabs to get up to the current statement indentation level, 
> and then spaces after that for the purpose of lining-up.

Unfortunately, that fails when one person writes their code with tabs
set to 8, and another person reads the code with tabs set to 4. Code
that was hand-aligned with a mix of tabs and spaces doesn't align
anymore.

Sometimes I wonder if Guido would have saved a lot of debate by declaring,
long ago, that Python code would never allow tabs in it, period. Which may
be a bit extreme, I'll admit. :-)


From aglyportREMOvethispart at nospam.yahoo.com  Wed Sep  4 15:07:49 2002
From: aglyportREMOvethispart at nospam.yahoo.com (Anton Graph)
Date: Wed, 04 Sep 2002 12:07:49 -0700
Subject: Memory fault - core dumped   while compiling 2.2.1 on SCO
References: <3D757EFA.9050401@nospam.yahoo.com> <yu99sn0qqw1r.fsf@europa.research.att.com> <m3bs7etem9.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3D765A05.6030004@nospam.yahoo.com>

Martin v. Loewis wrote:
> Andrew Koenig <ark at research.att.com> writes:
> 
> 
>>Looks like the same symptom I had.
>>
>>If you are using binutils 2.13, try going back to 2.12.1 and
>>see if that fixes the problem.
> 
> 
> Alternatively, use the system C compiler - Python does not support gcc
> on SCO, apparently (or else it would not try to pass -belf).
> 
> It is probably easy to add gcc support - just adjust the OpenUnix
> fragments in configure.in to SCO.
> 
> Regards,
> Martin

Ok. I run gmake distclean
and configure --without-gcc
(output from configure confirmed that it would not use gcc).

Compiling:

cc -G -KPIC -Ki486 -belf -Wl,-Bexport 
build/temp.sco_sv-3.2-i386-2.2/arraymodule.o -L/usr/local/lib -o 
build/lib.sco_sv-3.2-i386-2.2/array.so
/bin/sh: 1577 Memory fault - core dumped
gmake: *** [sharedmods] Error 139

Did not even take nearly as long for it to coredump. With optionless 
configure build was going at least for a half an hour and this one
bailed out pretty much right away...



From richie at entrian.com  Wed Sep 18 06:06:42 2002
From: richie at entrian.com (Richie Hindle)
Date: 18 Sep 2002 03:06:42 -0700
Subject: Earing money with Python software (was: More on Protecting Source Code)
References: <mailman.1032232999.22661.python-list@python.org> <m3elbtcg8l.fsf@mira.informatik.hu-berlin.de> <4jt7ma.fse.ln@lenashee.goebel-consult.de>
Message-ID: <8a6ba1da.0209180206.2d47f2bd@posting.google.com>

> How can I get some reward back when writing software like
> decompyle and bridgekeeper?
> [...]
> - Develop two different versions, a "free" one which is missing some
>   features and a commercial/registered/supported one offering more
>   sophisticated functionallity?

For library code (and tools like bridgekeeper which have a library
component), my approach is to release a single version under two
licenses, one free (as in both speech and beer) for use in Open Source
projects and one with a fee for use in closed source projects. 
Sleepycat seems like a good Open Source license to use - it's OSI
certified, GPL-compatible (but free from political problems) and
demands that projects using Sleepycat-licensed code must themselves be
open source (but not necessarily Sleepycat-licenced).  While there are
commercial users who release their source code, most don't, so this
pretty much catches all the people who I feel should pay.  There'll
always be exceptions but it's a simple way to get most of the way
there.

Making code free for Open Source projects effectively makes it free
for personal/evaluation use as well - if you're using it for a toy
project then you're probably happy to ship your source, and if you're
only using it for evaluation then you won't be redistributing your
project.  If your 'personal' use is in a project that you intend to
sell in a closed-source way, that makes you a commercial user.

Things are a bit different for tools like decompyle with no
redistributable components.  I don't have any better suggestions
there, other than a free (as in beer but not as in speech) license
that allows use only _with_ open source projects, and another paid-for
license allowing unrestricted use.  Whether licenses like that -
governing the class of software that a tool may be used with - are
feasible, I don't know.  I'm not a lawyer.

I've haven't properly started distributing code yet (it's all from
spare-time projects that my life and my day job leave me little time
for - what little there is is on my incomplete website at
www.entrian.com) so it's too early to say whether this approach is
working for me or not...

-- 
Richie Hindle
richie at entrian.com


From andrewa at calibredigital.com  Mon Sep 23 13:11:28 2002
From: andrewa at calibredigital.com (Andrew Alzner)
Date: Mon, 23 Sep 2002 13:11:28 -0400
Subject: parsing tabs in a string
Message-ID: <006201c26324$4148ae90$025e10ac@calibredd.com>

Hello,

I'm getting to know Python and have a quick question for something I don't understand.

I'm trying to parse information extracted from a spreadsheet which is tab separated. Here is a simplified version of what I have:

list=['"103-01a-17"\t2\t33\t256\t227\n']
y=list[0]
print y.find("\t")
print y.find("\t",1)

I thought the second .find should return the second occurence of "\t". To get the index of the second tab I have to type:

print y.find("\t",13)

This doesn't make sense to me. Can someone help me please.

Thanks,
Andrew
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020923/526ac8f6/attachment.html>

From nc-lessaudo at netcologne.de  Wed Sep 25 05:11:14 2002
From: nc-lessaudo at netcologne.de (Donald Lessau)
Date: Wed, 25 Sep 2002 11:11:14 +0200
Subject: [newbie] mod_python apache w32
References: <ami5rr$des$04$1@news.t-online.com>
Message-ID: <amrvgp$lfk$1@newsreader2.netcologne.de>

hi christopher,

i've got exactly the same problem... Did you find a solution in the
meantime?

thanks
Donald

"christopher knievel" <ch_knievel at gmx.de> schrieb im Newsbeitrag
news:ami5rr$des$04$1 at news.t-online.com...
> Hi all,
> i got some problems running mod_python with apache under windows.
> I am using python 2.2 and apache 1.3.26.
> As written in the manual i configured httpd.conf like this:
>
> LoadModule python_module modules/mod_python.dll
>
> AddModule mod_python.c
>
> <Directory "C:/Programme/Apache Group/Apache/python">
>   AddHandler python-program .py
>   PythonHandler mptest
>   PythonDebug on
> </Directory>
>
> I copied the mod_python.dll to C:/Python22/lib/mod_python.
>
> The Apache server is starting with "Apache/1.3.26 (Win32) mod_python/2.7.6
> Python/2.2.1 running..."
> So i guess it should be working....
>
> The test script looks like this:
>
> #! C:/Python22/python.exe
>
> from mod_python import apache
>
> def handler(req):
>     req.send_http_header()
>     req.write("<html><body>Hello World!</body></html>")
>     return apache.OK
>
> my error.log (last entry):
> [Sat Sep 21 18:00:13 2002] [error] [client 192.168.1.99] python_handler:
> make_obcallback returned no obCallBack!
>
> i've read several newsgroups and googled a lot but i didn't found a
working
> solution
> Thanks a lot in advance...
> christopher knievel
>
>




From BrenBarn at aol.com  Thu Sep 12 21:35:02 2002
From: BrenBarn at aol.com (OKB (not okblacke))
Date: Fri, 13 Sep 2002 01:35:02 GMT
Subject: Help with regular expression using findall and .*?
References: <mailman.1031862616.24171.python-list@python.org>
Message-ID: <Xns9287BBE855A1AOKB@199.45.49.11>

czrpb wrote:

> p=re.compile(r'macro.*?orcam',re.DOTALL)
> 
> for x in re.findall(p,s1):
>     print x
> 
> for x in re.findall(p,s2):
>     print x
> 

    	If you use re.findall directly you need to pass a pattern string 
in, not a compiled object, and I don't think there's a way to specify 
flags with that approach.  Instead, use the findall method on your 
compiled expression object:

for x in p.findall(s1):
    	print x

-- 
--OKB (not okblacke)
"Do not follow where the path may lead.  Go, instead, where there is
no path, and leave a trail."
	--author unknown


From e_fax_t at hotmail._ZAPME_.com  Tue Sep 24 16:31:20 2002
From: e_fax_t at hotmail._ZAPME_.com (Justin)
Date: Tue, 24 Sep 2002 22:31:20 +0200
Subject: readlines() and "binary" files
Message-ID: <3d90cacf$0$30457$afc38c87@sisyphus.news.be.easynet.net>

Hi,

I have excel data with occasional multi-line fields,
which when dumped to CSV translates to embedded CR's
within a line, whereas the records/lines themselves
are delimited by the CR+NL pair (this is MS-land).
What I'd like to do is read those files and split every
line apart on the semi-colon field separator. But it
seems that whether the file is opened as text or not,
(x)readlines() still considers the lone CR as a line
delimiter and so not all my lines end up with the same
number of fields as they should. Is there a way to handle
this, or is readlines just not meant to work with anything
but proper text files?

    f = open('testdata.csv','rb')
    for line in f.xreadlines():
        fields = line.split(';')
        print len(fields) # should always be the same value





From helm at dnv.com  Wed Sep 25 09:36:27 2002
From: helm at dnv.com (Jens Bloch Helmers)
Date: Wed, 25 Sep 2002 13:36:27 GMT
Subject: missing functions in os.path
Message-ID: <ud6r2jile.fsf@dnv.com>

I hope that a future version of Python will include the following
two functions:

     os.path.localpath( file )       # Returns the non-UNC name 
                                     # of file.

>>>file = r'\\computer\shared\myfile.txt'
>>>os.path.localpath( file )
'd:\somedir\shared\myfile.txt'



     os.path.relpath( file, dir )    # returns the relative path of
                                     # file relative to directory dir

>>>file = r'C:\test\mydir\hum.txt'
>>>dir  = r'C:\otherdir'
>>>os.path.relpath( file, dir )
'..\test\mydir\hum.txt'


From Oschler at earthlink.net  Fri Sep  6 13:20:50 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Fri, 06 Sep 2002 17:20:50 GMT
Subject: Interapp communication under Python[Linux]?
Message-ID: <St5e9.25080$w51.7041171@news2.news.adelphia.net>

My Python app will talk to 3 or 4 modules written in Java and C/C++.
Fortunately I have the source for the Java and C/C++ modules and I have
experience in both languages.  Currently the external modules have no
provision for being accessed outside of the process they are running as.  So
I'm wondering what the best protocol would be used to "glue it all
together".  There a dizzying array of possibilities:

1) Have everybody talk to each other via SOAP (web services).
2) Have everybody talk to each other via XML-RPC (Don't know much about this
method).
3) Use Boost's SWIG interface for C++ to link the C/C++ modules to Python.
How to link to Java in the same manner?
4) Use "sockets" and my own custom protocol for everybody.
5) Code in Jython instead and use various methods above to talk to C/C++
modules from Jython app.
6) CORBA? (yikes!)

So as I said, a lot of ways to get things done.  I'd like to hear from those
of you that have actually done a lot of cross-app communication with Python
on what your experiences have been, and the pros and cons you encountered.

thx





From kp at kyborg.dk  Tue Sep 24 08:11:29 2002
From: kp at kyborg.dk (Kim Petersen)
Date: Tue, 24 Sep 2002 14:11:29 +0200
Subject: __name__ in embedded python giving error
Message-ID: <3D905671.1070106@kyborg.dk>

Hi,

First: This is probably my fault in forgetting to consider some call 
when embedding python ....

When doing:

if __name__ == "__main__":

in modules i get the following error:

Traceback (innermost last):
   File "/var/www/cgi-bin/CGIX.py", line 641, in ?
     if __name__=="__main__":
TypeError: function requires at least 2 arguments; 1 given

Could some enlightened head give me a pointer as to what i've done wrong 
  (or am missing)?

Please query for more information - i haven't included much (because the 
program is a bit large)...

Error on Linux 2.4 - using python 1.5 (haven't tried yet with 2.2 - but 
i think the error is there as well)...



From loewis at informatik.hu-berlin.de  Mon Sep  2 06:57:13 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 02 Sep 2002 12:57:13 +0200
Subject: wie greife ich mit Python auf die c-bibliotheken zu ?
References: <3D7323BB.4070200@uni-koblenz.de>
Message-ID: <j4adn0lkvq.fsf@informatik.hu-berlin.de>

Oleg Seifert <misterxx at uni-koblenz.de> writes:

> k?nnte vielleicht jemand es mir an einem Beispiel zeigen. Ich habe als
> Attachments zwei kleine Anwendungen dazu geschickt.

Du musst einen Typ definieren; nimm als Beispiel Modules/xxmodule.c
aus der Python-Distribution. Dort sollte es dann ungef?hr lauten

typedef struct {
	PyObject_HEAD
	person 	*_person;
} PyPerson;

Die Methoden der C++-Klasse musst Du alle "wrappen", etwa

extern "C" PyObject*
PyPerson_setname(PyPerson* self, PyObject*args){
  char *n;
  if(!PyArg_ParseTuple(args, "s:setname", &n))
    return NULL;
  self->_person->set_name(n);
  Py_INCREF(Py_None);
  return Py_None;
}

Dann kann man das ganze vielleicht benutzen als

p = person.new() # Annahme: das ganze Modul heisst person, nicht xx
p.set_name("Peter")

Alternativ kannst Du auch set/get als Attribute wrappen, so dass man
dann schreiben kann

p.name = "Peter" # ruft implizit ->set_name()

Ciao,
Martin


From oren-py-l at hishome.net  Sat Sep  7 16:10:22 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Sat, 7 Sep 2002 16:10:22 -0400
Subject: Daemonize?
In-Reply-To: <20020907174957.GA25058@glacier.arctrix.com>
References: <F7DB8D13DB61D511B6FF00B0D0F06233022B66EC@mail.hebmex.com> <mailman.1031184848.32071.python-list@python.org> <unjt47pc5r8671@corp.supernews.com> <20020907161037.GA70503@hishome.net> <20020907174957.GA25058@glacier.arctrix.com>
Message-ID: <20020907201022.GA93398@hishome.net>

On Sat, Sep 07, 2002 at 10:49:57AM -0700, Neil Schemenauer wrote:
> Oren Tirosh wrote:
> >     devnull = os.open('/dev/null', 0)
> >     os.dup2(devnull, 0)
> >     os.dup2(devnull, 1)
> >     os.dup2(devnull, 2)
> > 
> > This is safer because it prevents these file descriptors from being reused.
> 
> Better:
> 
>     os.close(sys.__stdin__.fileno())
>     os.close(sys.__stdout__.fileno())
>     os.close(sys.__stderr__.fileno())
>     os.open('/dev/null', 0)
>     os.dup(0)
>     os.dup(0)

Any particular reason why you think this is better?

	Oren




From sholden at holdenweb.com  Tue Sep 17 18:10:43 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 17 Sep 2002 18:10:43 -0400
Subject: MarketProductLifecycle
References: <am818c$pam$1@reader1.tiscali.nl> <am81ff$pdh$1@reader1.tiscali.nl>
Message-ID: <cVNh9.15168$T4.10611@fe04>

"GerritM" <gmuller at worldonline.nl> wrote in message
news:am81ff$pdh$1 at reader1.tiscali.nl...
>
> Oops pressed the send key before verifying the destination...
>
> This does not belong here, but how to cancel from OE5?
>

Select your message (you have to be able to see it on your server...) and
select "cancel message" from the "Message" menu.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From jb at cascade-sys.com  Fri Sep 20 01:03:56 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Thu, 19 Sep 2002 22:03:56 -0700
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <bd751532.0209082142.54236307@posting.google.com> <allrrs$6hm$2@solaris.cc.vt.edu> <ambpgq$dre$1@nntp9.atl.mindspring.net> <006501c25fd5$ecc80740$0101010a@local>
Message-ID: <3D8AAC3C.9020300@cascade-sys.com>

Chris Gonnerman wrote:

>I just like that I don't have to type the #!$&@!! things 
>anymore.
>
>No smiley this time... I really hate typing curly after curly.
>It makes me wonder that anyone *likes* the freakin' things so
>much that they named a LANGUAGE after them.
>
They are a hell of an improvement over "begin" and "end".

But I agree, leaving them out altogether is even better.

--jb






From dag at animagicnet.no  Thu Sep 19 13:06:50 2002
From: dag at animagicnet.no (Dag)
Date: 19 Sep 2002 17:06:50 GMT
Subject: finding last match in a file
Message-ID: <slrnaojlin.8re.dag@omally.animagicnet.no>

I have a number of very large files where I have to find the last line
which matches a certain experssion.  Currently I'm simply opening the
file and looping through it noteing each time a line matches and at the
end of the loop I see which was the last match.  However this is very
slow and inefficient.  Is there some way to read through the file
backwards and simply find first match starting from the end of the file.
The files are really too big to read the entire thing into memory.

Dag


From timr at probo.com  Fri Sep 20 00:23:44 2002
From: timr at probo.com (Tim Roberts)
Date: Thu, 19 Sep 2002 21:23:44 -0700
Subject: Strange Telnet error
References: <4378fa6f.0209181245.7b06124d@posting.google.com>
Message-ID: <qj8louc5615jicn6covqe5b7df9sp8amdv@4ax.com>

mnations at airmail.net (Marc) wrote:
>
>I have been running some code that all the sudden stopped working. I
>can not figure out what's happened. I am doing a simple telnet
>connection that sends and receives info. Now I am getting this error:
>
>Traceback (most recent call last):
>  File "C:\Python22\lib\threading.py", line 408, in __bootstrap
>    self.run()
>  File "C:\Python22\lib\threading.py", line 396, in run
>    apply(self.__target, self.__args, self.__kwargs)
>  File "main.py", line 85, in run
>    command[0](*command[1:3])
>  File "key.py", line 92, in command
>    conn.write(com)
>  File "C:\python22\lib\telnetlib.py", line 272, in write
>    self.sock.sendall(buffer)
>AttributeError: 'int' object has no attribute 'sendall'
>
>Has anyone ever seen this before? If so, what could do this be related
>to?

I'm guessing you continued to try to do work after closing the connection.
Note this code at line 247:

    def close(self):
        """Close the connection."""
        if self.sock:
            self.sock.close()
        self.sock = 0    # <<<< this makes self.sock an integer
        self.eof = 1

Your code is multithreaded.  Could another thread have closed the socket on
you?
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From phr-n2002b at NOSPAMnightsong.com  Tue Sep 17 02:22:16 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 16 Sep 2002 23:22:16 -0700
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <pUdh9.9377$yt3.4604908@newssrv26.news.prodigy.com> <mailman.1032184821.5180.python-list@python.org> <c%qh9.102794$ub2.2120685@news1.tin.it> <mailman.1032212180.11973.python-list@python.org> <3d868ccc$1@news.sentex.net> <mailman.1032240440.21101.python-list@python.org>
Message-ID: <7xfzw9qgo7.fsf@ruckus.brouhaha.com>

"James J. Besemer" <jb at cascade-sys.com> writes:
> Furthermore, even if you formally assign the copyright to somebody
> else, I understand ownership eventually reverts to the original author
> after a lengthy period (45 years or something).  Evidently it
> preserves the copyright for the heirs of the author.

I heard the opposite.  In fact part of the debate before the 1976
copyright extension act was about who should get the additional years
of copyright of a work that had been assigned to a publisher, the
author or the assignee.  Of course the same corporate lobbyists who
got the extension passed in the first place made sure that publishers,
not authors, got 100% of the benefit of the extension.  See
http://eldred.cc for more info.


From tjreedy at udel.edu  Sat Sep 14 11:50:21 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 14 Sep 2002 15:50:21 GMT
Subject: How to represent sets
References: <alvd8p$1sq$1@news.hccnet.nl>
Message-ID: <1VIg9.16142$5r1.654088@bin5.nnrp.aus1.giganews.com>

"Anton Vredegoor" <anton at vredegoor.doge.nl> wrote in message
news:alvd8p$1sq$1 at news.hccnet.nl...
> In the past there have been some discussions on various platforms
about
> whether sets should be implemented using long integers (possibly
leading to
> longs becoming iterable arrays of booleans), or by using
dictionaries.

Framing this an an either/or question is simply wrong.  The 'best'
implementation of the abstract concept 'set' depends on the
application, the number of possible members, the density of actual
sets, and the operations needed.  Sets with up to 32 possible members
are well implemented by C longs (Python ints).  Somewhat dense sets of
ints in a much larger range can be longer bit maps.  A set of integers
in the range +- 10**20 has to be done with dicts.  Ditto with sets of
strings up to length, say, 32.

Similar false questions for other abstract data types: what is the
best way to implement numbers, sequences, or matrixes.  For matrixes,
linear arrays (with address calculation), arrays of arrays, and dicts
mapping pair->value all have their appropriate uses.

>  The matter seems to be mostly settled and there is a sets.py script
in the CVS
> directory at sourceforge which uses dictionaries.

Guido has agreed to support an experimental general solution as part
of the standard library, perhaps to eventually become a builtin module
or type.  This in no way precludes other implementations.  That is
part of way user classes with magic methods is for.

Choosing to support a general purpose set-of-anything-hashable rather
than a more limited set-of-integers-in-a-reasonable-range is in line
with Guido's other design decisions.

> I would like to reopen this discussion.

I don't remember any pronouncement from Guido or anyone else closing
it, except perhaps with regard to what he will personally support.

> Starting this discussion again when the CVS script is already very
mature
> seems a bit inappropriate, but I believe there is a way to make sets
faster. I
> only recently realized this so I could not act earlier.

Special-purpose implementations often have some advantage of the most
general implementation.  If not, they get forgotten about.
>
> To give some material for the discussion I am providing a beta
version of a
> possible implementation af a set class using long integers to
represent sets.
> It is basically the same script I presented a year ago but I have
made several
> improvements to this script so that it is now a bit faster, and I
hope the
> docstring is a bit better now. Please keep in mind this is just a
test script,
> for a safer and more complete implementation the script in CVS is
preferable.
>
> http://home.hccnet.nl/a.vredegoor/universe/universe.py
>
> or the html version:
>
> http://home.hccnet.nl/a.vredegoor/universe/universe.html

To be most useful as an alternative implementation, your
implementation should be (could be? is?) a plug-in replacement for
sets.py so one could gain its advantages by replacing 'import sets'
with 'import universe as sets' (or 'import longsets as sets').

Having one 'standard' implementation now defines a standard interface
that gives you a (pretty-well) fixed target to aim at for
compatibility.  Before this was hashed out among the prime developers,
and some arbitrary decisions made, there were multiple possible
targets.

I took a preliminary look at universe.py.  Your test() function would
be more useful as a regression test if you automated it with the
doctest module.  To do so, cut the contents of the function (perhaps
after dedenting) and paste it into the interactive interpreter.  (On
Windows, turn off 'fast pasting' first!!, and maybe paste just a few
lines at a time.)  Then cut and paste the result and put it in a doc
string and otherwise add the doctest boiler plate given in
doctest.__doc__.

Terry J. Reedy





From martin at v.loewis.de  Mon Sep  2 02:04:48 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 02 Sep 2002 08:04:48 +0200
Subject: MySQL and Python in Solaris 8
References: <mailman.1030570791.13266.python-list@python.org>
Message-ID: <m3sn0t6i67.fsf@mira.informatik.hu-berlin.de>

ganapathy murali krishnan <gmurali at cs.uchicago.edu> writes:

> cc -Xc -mt -O -I/opt/mysql/mysql-3.23/include/mysql
> -I/opt/python/default/include/python2.1 -c _mysql.c -o
> build/temp.solaris-2.8-sun4u-2.1/_mysql.o -fPIC
> cc: Warning: illegal option -fPIC
> "/opt/mysql/mysql-3.23/include/mysql/mysql.h", line 111: long long not
> allowed in Xc mode

Something is quite broken here. Can you tell where the Xc/mt options
come from? To my knowledge, Python/distutils never uses those on its
own - could it be that you set environment variables to that effect?
(perhaps mysqldb introduces them, but that would be surprising)

-fPIC usage indicates that you have been using gcc to build Python. Is
this the case? Do you have gcc on your system? GNU binutils?

> ImportError: ld.so.1: /usr/local/bin/python: fatal: relocation error: file
> /opt/python/default/lib/python2.1/site-packages/_mysql.so: symbol
> PyLong_FromUnsignedLongLong: referenced symbol not found
> -------
> 
> which suggests that the original build was not complete.

More precisely, it indicates that you have been linking with GNU
binutils, but Python failed to detect that it had to use
-Wl,-export-dynamic when linking python. What is your binutils
version?

Regards,
Martin



From cben at tx.technion.ac.il  Mon Sep 16 16:15:12 2002
From: cben at tx.technion.ac.il (Beni Cherniavsky)
Date: 16 Sep 2002 13:15:12 -0700
Subject: "str"(expr)"str" interpolation syntax
References: <mailman.1031688793.11971.python-list@python.org> <f7930f23.0209121050.5c1870e3@posting.google.com> <8a6ba1da.0209130002.37ede06a@posting.google.com>
Message-ID: <f7930f23.0209161215.5d4a5f03@posting.google.com>

richie at entrian.com (Richie Hindle) wrote:
> Guido's time machine strikes again:
> 
> >>> foo, bar, quux = 1, 2, '3'
> >>> print ("A template string - here: %s it goes\n" % foo +
> ...        "Also multiple values: %s, %s\n" % ( bar, quux ) +
> ...        "Quoting percent (%s) sings probably not needed...\n" % '%')
> A template string - here: 1 it goes
> Also multiple values: 2, 3
> Quoting percent (%) sings probably not needed...

Well, all right.  I'll find something more useful to do than repainting
the wheel ;-)


From tg5027 at citlink.net  Wed Sep 18 19:40:56 2002
From: tg5027 at citlink.net (terry)
Date: Wed, 18 Sep 2002 19:40:56 -0400
Subject: monetary applications (et al)
Message-ID: <E17rnyi-0000Mx-00@mail.python.org>

>>> Michael wrote:
for how to emulate numeric types.
>>>

Right, I have looked at that.  As near as I can tell there is a vast 
difference between 'emulation' and 'intrinsic' - and the burden of that 
difference falls on the programmer/customer in your assessment.  The 
difference involves the amount of effort required to anticipate and 
react to the use of emulation as opposed to the simplicity of using 
intrinsic data types.  Maybe in the C world, dealing with what ever 
comes your way in what ever way you deem appropriate for your current 
project/situation is a way of life, but for those of us who have to 
deal directly with a world in which money is a common data type, all of 
that extraneous OO stuff to represent what is really a common 
information form really is a crude approach to getting a job done.

just my $0.02  ;>)

terry



From cnetzer at mail.arc.nasa.gov  Sat Sep 28 22:23:04 2002
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Sat, 28 Sep 2002 19:23:04 -0700
Subject: Help please with menubars
In-Reply-To: <000f01c266f3$2903dbc0$2502a8c0@emily.ewndsr01.nj.comcast.net>
References: <000f01c266f3$2903dbc0$2502a8c0@emily.ewndsr01.nj.comcast.net>
Message-ID: <200209290223.TAA30009@mail.arc.nasa.gov>

On Saturday 28 September 2002 06:30, andy surany wrote:
> Thanks Chad.

No problem; I'm posting my reply to the newsgroup, because I think it 
is relevant.

> Yes, Tk.
>
> The problem with using add.menubar is that I end up with a cascade.

Huh?  Not in the example that I posted...

> But I hadn't thought of a bind. I know that I can bind to a button -
> but this is an item on a menubar. I'm not sure how to do that.

Me neither. :)  It really is more effort than it is worth; and I think 
you don't need it.

> BTW, what I am encoding is a menubar with 5 top level items. The
> first 4 are cacading/pulldown (no problems!), but the fifth is simply
> an "Exit" on the menubar. Click it, and the window should be history.

That was almost exactly what the code I posted did...  Hmmmph; Here is 
a revision that exits when you click on the "Exit" entry in the 
menubar, WITHOUT dropping down a menu.  I also added a separate menu 
entry that does cascade down, just to show you it can be done.

If this isn't what you need, then I have no clue what you mean; it 
would be best to post your code (or the relevant portions), to see how 
you are attempting to use menus.

I'm making it an attachment, so it should be easy for you to save and 
run it.  Please DO run it, if you haven't, before replying, and let me 
know if it doesn't work as I say it should.

Also, by using the "menu" option of a Toplevel, this code will 
integrate much better with the system you are on, than if you try to 
make your own custom "menubar", using a Frame and menubuttons, for 
example.  On the Mac, for example, this code follows the systems 
menubar conventions (per desktop, not per window).

-- 

Chad Netzer
cnetzer at mail.arc.nasa.gov
-------------- next part --------------
A non-text attachment was scrubbed...
Name: menubar_test2.py
Type: text/x-java
Size: 654 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020928/c0e9f2d0/attachment.java>

From bokr at oz.net  Tue Sep  3 10:23:47 2002
From: bokr at oz.net (Bengt Richter)
Date: 3 Sep 2002 14:23:47 GMT
Subject: syntax error in Pyton shell
References: <e895ad50.0209022304.44c32987@posting.google.com> <3D746BC1.2BFA05F8@alcyone.com>
Message-ID: <al2glj$r7k$0@216.39.172.122>

On Tue, 03 Sep 2002 00:58:57 -0700, Erik Max Francis <max at alcyone.com> wrote:

>lion wrote:
>
>> I'm learning exception handling, I type these code in Python shell:
>> 
>> >>> try: f=open("nofile")
>> ... except IOError: pass
>> ... print "The exception is just passed!"
>>   File "<stdin>", line 3
>>     print "The exception is just passed!"
>>         ^
>> SyntaxError: invalid syntax
>> 
>> But these code run without any error as a program(or script). Could
>> anyone tell me why?
>
>This is an artifact of entering in code manually in the shell, which
>normally would have more cues to go on.  In a script it would no that
>these two statements were unrelated; in the interactive interpreter it
>can't be sure so is waiting for a blank link to separate them.  You can
>tell this when the prompt is ... instead of >>>.
>
Why can't it be sure? IWT the dedent should tell it the previous chunk is complete.
And in any case, why not keep accepting lines until the blank line, once into accepting
more than one?

It is annoying to have to resort (interactively) to:

 >>> if 1:
 ...     for i in range(5):
 ...         print i,
 ...     print '[output desired on same line].'
 ...
 0 1 2 3 4 [output desired on same line].

because this screws up:

 >>> for i in range(5):
 ...     print i,
 ... print '[output desired on same line].'
   File "<string>", line 3
     print '[output desired on same line].'
         ^
 SyntaxError: invalid syntax

Regards,
Bengt Richter


From david.abrahams at rcn.com  Thu Sep  5 22:05:46 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Thu, 5 Sep 2002 22:05:46 -0400
Subject: try blocks in Python vs. C++
References: <mailman.1031270709.27012.python-list@python.org>
Message-ID: <al92np$o3n$1@bob.news.rcn.net>

"David LeBlanc" <whisper at oz.net> wrote in message
news:mailman.1031270709.27012.python-list at python.org...
> I think the biggest reason why try blocks aren't suggested for use in C++
is
> that they increase code size substantially (as do some other nice
features).

That's not a very good generalization. C++ implementations vary
substantially on the over (and under-) heads implied by exception-handling
constructs.

One reason you might want to avoid try blocks is that generally you want to
do cleanup unconditionally and rethrow the exception unmodified, and C++ has
no "finally" construct. Usually it works a lot better to use destructors of
local objects for this sort of thing, since you can re-use them and you
can't forget to "throw;" at the end.

BTW, saying "Stroustrup's C++ Book" isn't very specific (he has several),
and early editions of TC++PL contained some bad advice about C++ EH. See the
downloadable Appendix E at http://www.research.att.com/~bs/3rd_safe0.html
for some good advice.


> Try...except is good practice in Python IMO.

Yeah, it's a completely different story in Python:

    Python has 'finally'

    __del__ methods don't execute as the stack is unwound
    (whereas C++ destructors do)

    throwing an exception to deal with ordinary conditions
    (e.g. StopIteration) is common, since unlike most C++
    implementations throwing an exception is no more
    expensive in Python than returning from a function.

-----------------------------------------------------------
           David Abrahams * Boost Consulting
dave at boost-consulting.com * http://www.boost-consulting.com

>
> > -----Original Message-----
> > From: python-list-admin at python.org
> > [mailto:python-list-admin at python.org]On Behalf Of Roy Smith
> > Sent: Thursday, September 05, 2002 16:54
> > To: python-list at python.org
> > Subject: try blocks in Python vs. C++
> >
> >
> > I'm reading Stroustroup's C++ book, and came upon a piece of advice to
> > avoid using try blocks.  I use try blocks a lot in Python and gather
> > it's common practice among pythonistas.
> >
> > Is there something fundamentally different about try blocks in Python
> > and C++ which makes them more attractive to use in one vs. the other?
> > On the surface, they appear to be virtually identical in just about
> > every respect.  Is it simply a cultural difference?  Or just the
> > author's own personal style?
> > --
> > http://mail.python.org/mailman/listinfo/python-list





From snyderf at loki.stockton.edu  Tue Sep 24 21:14:38 2002
From: snyderf at loki.stockton.edu (Fred Snyder)
Date: Tue, 24 Sep 2002 21:14:38 -0400
Subject: IDLE Default Save Directory
Message-ID: <3D910DFE.F74644B4@loki.stockton.edu>

Hi -

I'm just getting up and running with Python(2.11)/IDLE(.8) on SuSE 7.3.
Hope this is not a too "duh" question.   I've looked but can not find
how to change the default "Save" and "Save As" directories in IDLE.
Have looked at the various configuration files but don't know the syntax
or the [Section] to make these changes.

Would someone please give me some examples of a customized config file
(maybe .idle) so I can make some changes.

Thanks,

Fred



From skpeternospam at ucdavis.edu  Mon Sep 30 17:25:40 2002
From: skpeternospam at ucdavis.edu (Sam Peterson)
Date: 30 Sep 2002 17:25:40 -0400
Subject: Python Process that can survive
References: <mailman.1033137937.24205.python-list@python.org> <571739c3.0209271111.6514c32f@posting.google.com>
Message-ID: <slrnaphhpt.esj.skpeternospam@hip.ucdavis.edu>

On Fri, 27 Sep 2002 at 19:11 GMT, Christopher Peery <peeryc at hotmail.com> spake thus:
> sismex01 at hebmex.com wrote in message news:<mailman.1033137937.24205.python-list at python.org>...
> Plus in my lab there are a bunch of perl guys and a few python
> people.  To date, what one can do in their script the python guys
> can do to (only nicer and much more readible). Perl has the ability
> to have a process continue after the underlying terminal is killed
> if the process is in the background. What is python doing that
> doesn't allow this?

My guess is that they're using something like

 $ nohup perl <perlscript>.

This catches the HANGUP(HUP) signal sent from the terminal and
effectively daemonizes a process.  All the perl I've scripted before
would die as soon as I'd logout, and I'd use the nohup program to keep
it running.

-- 
Sam
"Da Man"
Aren't I an egotist?
s/nospam/son/ -- to email me.


From tdelaney at avaya.com  Sun Sep 22 19:51:15 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 23 Sep 2002 09:51:15 +1000
Subject: Minimal debug/rep functionality
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5C0@natasha.auslabs.avaya.com>

> From: richie at entrian.com [mailto:richie at entrian.com]
> 
> The first version used the AST, but there are several holes in that -
> 'else' lines, for instance, are identified by the parser as executable
> lines, but the trace function doesn't always get called for them.

Something I would have had to find out myself ... thanks :)

> found out that SET_LINENO is going away in Python 2.3...

The reason I didn't do the same thing ...

> My current system (which I think works 8-) uses co_lnotab, which is a
> delightful data structure that describes the relationship between
> bytecodes and line numbers.  Here's the relevant piece of code, which
> builds in 'lineNumbers' a list of the execuable line number in the
> code object 'code':

Thank you. This should help immensely.

> Hope that helps.  The coverage tool that this belongs to is finished
> but unreleased (mostly because of lack of thorough testing, but I'm
> using it myself with no problems) - if you'd like a copy, drop me an
> email.

Please do - it will be interesting to see how our approaches differ. Don't
bother about tidying it up unless it's a real mess.

I'd better get onto asking my employer if I can pass this code onto other
people I guess :)

Tim Delaney



From bloke at ii.net  Sat Sep 14 14:38:11 2002
From: bloke at ii.net (Rob Hall)
Date: Sun, 15 Sep 2002 02:38:11 +0800
Subject: convert python package to rose uml
References: <3d838069$0$22311@echo-01.iinet.net.au>
Message-ID: <3d838203$0$22326@echo-01.iinet.net.au>

oh yeah, should have mentioned, it simply dumps the convertion into a file
called import.ptl in the CWD.

Rob




From alanmk at hotmail.com  Tue Sep 24 12:40:04 2002
From: alanmk at hotmail.com (Alan Kennedy)
Date: 24 Sep 2002 09:40:04 -0700
Subject: Strange values in Accept-Language header?
References: <f72dac0d.0209231303.12349f2e@posting.google.com> <8FWj9.39597$V7.10466682@twister.socal.rr.com>
Message-ID: <f72dac0d.0209240840.20fafc70@posting.google.com>

Alan Kennedy:

>> Accept-Language: en-us,x-ns1rDGeT4e2FpA,x-ns2f67971hgDw1
>> Accept-Language: x-ns13V8cN8S1Xz9,x-ns2T329gfxKa7d
>> 
>> I have no idea what these could mean (I've changed the actual
values
>> BTW, on the remote off-chance that somebody's highly personal
>> information was coded in there)

Jeff Davis:

> I really don't know what those extra characters mean. Do you think it might 
> have to do with some related preference like fonts or the screen or 
> something? 

I should have mentioned: although I changed the value, I didn't change
the first four characters, so the "x-ns1" and "x-ns2" at the beginning
of each string are original. To me, that indicates some form of
identifier structure, like a URN, for example......

However, if it is a crack attempt, perhaps the "x-ns1" and "x-ns2" are
keys in some dictionary/hashmap that are vulnerable in some server?

Jeff Davis:

> Perhaps if you posted the secret info someone else might notice something.

Yes, I think I'm being too squeamish by not posting the values. Here's
two of the headers, unmodified

Accept-Language: x-ns16W5tM7A2Nh6,x-ns2U210btwUq5f
Accept-Language: en-us,x-ns1jZKcD2t3NhQ,x-ns2r3509OnmPe2

If it was some form of request modifier for custom
clients/proxies/servers, then why not simply use a custom HTTP header,
like

X-Custom-Header: x-ns16W5tM7A2Nh6,x-ns2U210btwUq5f

Odd.

Alan.


From roy at panix.com  Thu Sep  5 19:53:48 2002
From: roy at panix.com (Roy Smith)
Date: Thu, 05 Sep 2002 19:53:48 -0400
Subject: try blocks in Python vs. C++
Message-ID: <roy-13BC67.19534805092002@reader1.panix.com>

I'm reading Stroustroup's C++ book, and came upon a piece of advice to 
avoid using try blocks.  I use try blocks a lot in Python and gather 
it's common practice among pythonistas.

Is there something fundamentally different about try blocks in Python 
and C++ which makes them more attractive to use in one vs. the other?  
On the surface, they appear to be virtually identical in just about 
every respect.  Is it simply a cultural difference?  Or just the 
author's own personal style?


From aleax at aleax.it  Thu Sep 19 11:29:43 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 19 Sep 2002 15:29:43 GMT
Subject: Getting started
References: <uojinmcfapjf45@corp.supernews.com> <mailman.1032442884.18747.python-list@python.org>
Message-ID: <H3mi9.133635$pX1.4751676@news2.tin.it>

holger krekel wrote:

> Lance wrote:
>> Hi All,
>> 
>> Last night I attended a Weak typing, C++ Templates, and Python talk by
>> Bruce Eckel. It was great, my introduction to Python.
>> 
>> I'm sold on Python and weak typing.
> 
> I am not. But i much prefer dynamic over static typing :-)
> What is 'weak typing' supposed to mean, anyway?  I have heard that a
> couple of times from java-people because they tend to think that python
> is not strongly typed.

$a = "2.3";
$b = 45;
print $a+$b;

emitting 47.3 is pretty close to what I consider "weak typing".  I
find it very hard to produce substantial, large programs using
weakly-typed languages.  Python is strongly, albeit dynamically,
typed, as you say.  Unfortunately, some seminal material of a few
decades ago confused terminology (I vaguely remember something
about it in some early article about Scheme, for example).


Alex



From Oschler at earthlink.net  Sun Sep  1 19:13:53 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sun, 1 Sep 2002 19:13:53 -0400
Subject: Add methods to a class at runtime?
References: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com> <slrnan4dl3.18h.gerhard.haering@lilith.my-fqdn.de> <wmvc9.21757$FJ1.3181934@e3500-atl1.usenetserver.com> <mailman.1030916762.5578.python-list@python.org>
Message-ID: <Eaxc9.33161$vY2.738637@e3500-atl2.usenetserver.com>

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote in message
news:mailman.1030916762.5578.python-list at python.org...
>
> I never saw such a note. Maybe you were reading a message from the
> (dormant) types-sig, where ways to introduce optional (!) stronger
> typing into Python are discussed. I never saw Guido make any comments
> about removing dynamic features from Python, though. With /one/
> exception, which you didn't mention, yet: assigning to the __class__
> attribute:
>

Yes it was in a thread where that was mentioned.  That's not a feature I
would miss, kind of scary (dynamically switching an instance's class).  The
add method stuff though is superb.  It's used to wonderful effect in XML
parsers I've seen for parsing.  If your're curious about the __dynamic__
keyword, here's the thread.  Look in the neighborhood of message #68 and
find the keyword __dynamic__ (wow what a great password this link would
make):

http://groups.google.com/groups?hl=en&lr=lang_en&ie=UTF-8&oe=UTF-8&safe=acti
ve&threadm=mailman.997727540.19628.python-list%40python.org&rnum=1&prev=/gro
ups%3Fhl%3Den%26lr%3Dlang_en%26ie%3DUTF-8%26oe%3DUTF-8%26safe%3Dactive%26q%3
DRe%253A%2BCopy%2Bconstructors%2Bpython

thx






From umop at psyon.org  Wed Sep 18 17:34:09 2002
From: umop at psyon.org (Eric Arnold)
Date: Wed, 18 Sep 2002 14:34:09 -0700
Subject: Regex question
References: <amapuf$45t8t$1@ID-144898.news.dfncis.de>
Message-ID: <amarmj$4cj7d$1@ID-144898.news.dfncis.de>

Okay, I have it down to:

(?<eventName>.*):(?<eventParam>[^,]*(,|$))+

which almost works ...

 <eventName>: "Name"
 <eventParam 1>: "AAA"
 <eventParam 2>: "BBB"
 <eventParam 3>: "CCC"
 <eventParam 4>: ""

it's picking up the <eventParam4> on the last character (right after "C")
and it's an empty string.  Shouldn't "CCC" (as captured with the $ end of
string as a delimiter) be the last thing found?

Eric

"Eric Arnold" <umop at psyon.org> wrote in message
news:amapuf$45t8t$1 at ID-144898.news.dfncis.de...
> Hi folks,
>
> I'm trying to figure out why this is not working, and wonder if there are
> any regexperts out there who can give me a pointer.
>
> The string I'm capturing is as follows:
>
> Name:AAA,BBB,CCC
>
> where I'd like to extract the following data:
>
> <eventName>: Name
> <eventParam 1>: AAA
> <eventParam 2>: BBB
> <eventParam 3>: CCC
>
> this is the regular expression I wrote to do so:
>
> (?<eventName>.*):(?<eventParam>([^,]*(?=,))|((?<=,)[^,]*))+
>
> I tried it with the (?<=,) as another (?=,) too, but no luck.  I imagine
> that the problem is that when the (?=) is used once, it can't be used with
> the second group, but I don't know a way around that.
>
> Thanks!
>
> Eric
>
>




From DGustafson at attbi.com  Fri Sep 20 06:48:57 2002
From: DGustafson at attbi.com (Dave Gustafson)
Date: Fri, 20 Sep 2002 10:48:57 GMT
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com>
Message-ID: <t2Di9.336220$kp.1062154@rwcrnsc52.ops.asp.att.net>

I've been programming for, hard to believe when I think of it, 30 years...
I just recently picked up a book that really brought back som fo the Joy of
programming and programming languages.  I wish I'd picked it up much
earlier!  It's not about Python, but rather about a dialect of Lisp called
Scheme, and is titled "The Little Schemer".  It's been really fun to work
through and really "grok" what's going on in programming, also fun to
translate the concepts there in to Python which to me is a more real-world
language.  I'd suggest it as a way to have some fun with this art...

Dave






From jb at cascade-sys.com  Thu Sep 26 02:39:15 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Wed, 25 Sep 2002 23:39:15 -0700
Subject: Larry Wall's comment on python...
References: <F7DB8D13DB61D511B6FF00B0D0F06233025D48F5@mail.hebmex.com> <mailman.1033004794.21467.python-list@python.org> <mq25pu0bs0159c70pp16rutfoabgh9f8hn@4ax.com>
Message-ID: <3D92AB93.8060102@cascade-sys.com>

Courageous wrote:

>>Once upon a time, a large class of computers, and their attendant 
>>peripherals, OSs and utility programs universally implemented tabs 
>>hard-wired at 8 stops.
>>    
>>
>
>Indeed; such is this issue that I leave hard tabstops as 8 in my
>editor, and set soft tab stops set to 4. 
>
I'm aware of that technique and I know of others who favor it.  And it 
does give you 4 column stops without having to dink with all your 
printers and other utilities.

Personally, though, I don't care for it.  VI is not my programming 
environment all the time (maybe 50% these days) and some other tools I 
use don't have this feature.  Then I'm stuck running into spaces where I 
don't expect them.

Another editing nit that I use is to cursor left one or some other 
specific number of "indents" and then cursor up or down to where the 
enclosing code lines up.  If you have spaces or a mix then half the time 
you have 3 extra character positions to move to get where you need to be.  

Incidentally, mixing spaces and tabs is officially less desirable in the 
official code style guides than tabs or spaces.

All my life I've relied on indent being 1 tab and I'm too old to change 
that, though I can live with 4 char tab stops.  When I make slight 
tweaks to other people's program, I always strive to honor the 
pre-existing standards.  If I'm doing a lot of work, then I might well 
convert to all tabs while I'm working on it and convert back, if 
necessary, to whatever standard the customer requires.  Code I write 
just for myself of course generally is all tabbed at 8 columns.

But I'm NOT telling you what you should do.  You do what you want.  

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From manuel at hendel.net  Fri Sep 27 05:13:25 2002
From: manuel at hendel.net (Manuel Hendel)
Date: Fri, 27 Sep 2002 11:13:25 +0200
Subject: comparing all values of a list to regex
In-Reply-To: <0lUk9.149240$ub2.3291527@news1.tin.it>
References: <mailman.1032947731.3297.python-list@python.org> <Nbgk9.140934$ub2.3077596@news1.tin.it> <mailman.1033039492.1772.python-list@python.org> <0lUk9.149240$ub2.3291527@news1.tin.it>
Message-ID: <20020927091325.GA375@partagas.as.de.cw.net>

Sorry for forgetting some details.

On Fri, Sep 27, 2002 at 08:07:24AM +0000, Alex Martelli wrote:
> 
> def classify(login_email):
>     l_e = login_email.split(',')

I think that I nearly understand what you are doing here. But I don't
understand the assert part. I already checked at the documentations
but this doesn't helped me at all.

>     assert 1 <= len(l_e) <= 2, "More than one comma in (%s)" % login_email
>     if len(l_e)==2: return 2         # both
>     elif l_e.find('@'): return 1     # email, I guess
>     else return 0                    # other case (local account, I guess)

Manuel

-- 
From '86 until the summer of last year, wherever I went, people would say, 
"You would have made a great James Bond! Weren't you going to be James Bond? 
You should have been, you could have been, you may have been." Yes, yes, yes, 
yes, yes. It was like unfinished business in my life. I couldn't say no to it 
this time around. 
-Pierce Brosnan 



From Henrik.Weber at sys.aok.de  Fri Sep 27 05:37:25 2002
From: Henrik.Weber at sys.aok.de (Henrik Weber)
Date: 27 Sep 2002 02:37:25 -0700
Subject: DBI cursor behaviour with multiple statements?
Message-ID: <7e964d1d.0209270137.568a3f3@posting.google.com>

Hello,

I'm trying to write a DBI2 compliant module for Microsoft Access
databases. Now I'm not quite sure how to interpret the DBI2
specification when it comes to the execution of multiple data
returning statements with the same cursor.

For example if c is my cursor and I do:
c.executemany("SELECT * FROM table WHERE column = '%s'", [["parm1"],
["parm2"], ["parm3"]])

should the result be the union of all three statements or just the
result of the first or the last statement? Or should the user switch
from one resultset to the next with nextset?

And what if I execute another SELECT-statement before all results of
the previous execute or executemany have been fetched? Should the
cursor first return the rest of the previous statements results or
should the results of the last statement replace whatever there was
before?

Maybe someone reading this group can clarify those points for me.

Thanks,
Henrik Weber


From jb at cascade-sys.com  Fri Sep 20 13:48:24 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Fri, 20 Sep 2002 10:48:24 -0700
Subject: [Help]: mailbox classes
References: <jylimd-DAD12F.17305418092002@skates.gsfc.nasa.gov> <Sfei9.18259$V7.4642808@twister.socal.rr.com>
Message-ID: <3D8B5F68.2070906@cascade-sys.com>

Jeff Davis wrote:

>Ok, I'll assume your using mbox format, but this is easily adapted for 
>maildir, etc.
>
>import mailbox
>mbox_file = '/path/to/mbox'
>
>mbox_fp = open(mbox_file)
>mbox = mailbox.UnixMailbox(mbox_fp)
>msg = mbox.next()
>
>there, now you have a message object (a rfc822.Message object, to be 
>exact). Now just import the rfc822 module and do whatever you need. You'll 
>probably want to use a loop to process all of the messages.
>
I've found these packages to be somewhat fragile.  In particular, real 
world email includes a lot of spam and a fair amount of spam does not 
conform 100% to all the mime encoding rules.  A frequent error is that a 
surprising number of mime-encoded messages for some reason do not 
include the ending boundary marker.  This often results in errors while 
parsing that message and the loss of all messages following that point. 
 IIRC the error becomes unexpected EOF after all the remaining messages 
have been mistakenly skipped looking for that missing boundary.  

I was looking to Python and the email classes to whip up a quick and 
dirty spam filter, so "filter out the spam first" is not a solution to 
my problem.  FWIW, Netscape does not have any problem with this same data.

I've had moderate success by reading all the messages into a string, 
splitting the string on "^From ", passing individual messages to the 
email parsing classes, and treating the the unexpected EOF exception 
like a real EOF.  But with hundreds of messages daily, that's a rather 
clumsy and expensive operation.  It also defeats the effort the email 
classes go to to avoid having to have everything in memory.  

I'm curious if I'm missing something obvious here and if there's some 
easy way to get mailbox to work with large amounts of real world email 
with imperfect encodings.

If possible, perhaps mailbox should have some absolute sense of the end 
or beginning of a message ("^From "?) that overrides mime boundaries.   
That MUST be how Netscape does it.

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From martin at v.loewis.de  Mon Sep 30 17:20:21 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 30 Sep 2002 23:20:21 +0200
Subject: xml.dom.minidom parser error
References: <fa06e058.0209301232.58fdc0fd@posting.google.com>
Message-ID: <m3elbb2mzu.fsf@mira.informatik.hu-berlin.de>

revyakin at yahoo.com (revyakin) writes:

> Any ideas what this could mean? Writing a parser by myself does not
> sound exciting.

No, it just means that you have to install one. You can chose either
pyexpat, or xmlproc; if pyexpat, you can either use the wrapper that
comes with Python, or PyXML.

If you use the wrapper, install Expat 1.95.x, and point Modules/Setup
to the location where you have installed Expat.

If you want to use xmlproc, or don't want a separate Expat
installation, install PyXML.

Good luck,
Martin


From pete at shinners.org  Wed Sep 25 12:22:57 2002
From: pete at shinners.org (Pete Shinners)
Date: Wed, 25 Sep 2002 16:22:57 GMT
Subject: Pygame event problem
References: <41c203e.0209250702.160102d@posting.google.com>
Message-ID: <Bplk9.73049$gA4.26401@sccrnsc02>

Anand wrote:
>   I wrote a simple image viewer in python using pygame image api. The
> GUI is in wxPython. There are 2 classes, one for the window frame and one
> for the pygame image processing.

>    The effect is stopped by clicking with the mouse on the pygame
> display window. My problem is that if I try to invoke any other menu methods
> in the main GUI frame *without doing this*, the application ie both the
> wxFrame and the pygame display window, hangs.
> 
> 
>   I have a gut-feeling that this is because I am calling a method on
> the MyPyGameImgViewer class without exiting the while loop.
> 
>  Is there a way around this, ie to call any method in the
> MyPyGameImgViewer class from the GUI class, while one of its methods
> is executing a while loop

i'd be a little afraid of mixing wxpython and pygame like this, but it
sounds like you are pretty close to running. to exit the pygame loop you
could probably just pass it a mouse button event

e = pygame.event.Event(MOUSEBUTTONDOWN)
pygame.event.post(e)


of course, that also means you'll probably want to clear the pygame event
queue before starting your while loop. also, to clear the event queue it
will probably be better to get _all_ the events from the queue. otherwise
it will fill up with mouse motion and other types of events. change the
event part of the loop to this

if pygame.event.peek(MOUSEBUTTONDOWN):
   pygame.event.get()
   break

instead, you could also filter the pygame event queue, so only mouse button
events were put onto the queue. the code for that would be like this

pygame.event.set_allowed(None) #allow no events
pygame.event.set_allowed(MOUSEBUTTONDOWN) #only allow buttondown



also, there are ways to get SDL to work within an embedded window.
therefore embed into a wxpython control. basically it involves setting an
environment variable to an existing window id/handle. wxpython allows you
to get the id/handle you need. the code to embed the SDL window looks kind
of like this:


handle = wxpython_widget.get_window_id_i_forget_how()
os.environ['SDL_WINDOWID'] = str(handle)
if sys.platform == 'win32':
    os.environ['SDL_VIDEODRIVER'] = 'windib' #no directx embedded
pygame.init()
size_is_ignored = 10, 10
window = pygame.display.set_mode(size_is_ignored)

also, since you are likely only using the display portion of pygame, no
need to initialize the whole thing (like audio/cd/joystick/etc). you could
change the "pygame.init()" call do just "pygame.display.init()"

that should be pretty slick when you get it working. although it's been a
long time since i did anything like this. this should be all the correct info.

oh, also remember, when you embed the SDL window like this, it will not
receive any input events. you'll need to use wxpython's event handling for
that.




From mcherm at destiny.com  Tue Sep 10 12:44:07 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Tue, 10 Sep 2002 12:44:07 -0400
Subject: embedded Python sample
References: <2a0f4cc4.0209100520.1d3a723b@posting.google.com>
Message-ID: <all7gh$mn$1@main.gmane.org>

bruno Monnet wrote:
> Hi,
> 
> I'm trying to call a python script from my c++ code. All I need is to
> invoke the script and check the return status.
> 
> does anyone has a simple example ? All I found until now is very
> complex and ugly code :-((
> 
> Regards,
> Bruno

Why not just execute as a sub-process... invoke it as a shell command 
just as you'd do with "ls".

-- Michael Chermside







From k at kk.k  Sat Sep 21 17:53:06 2002
From: k at kk.k (Zed)
Date: Sat, 21 Sep 2002 21:53:06 GMT
Subject: Math optimization
Message-ID: <6T5j9.145247$pX1.5195840@news2.tin.it>

Hi!
What's faster in python? Divide or multiply?
(a= b*c  is faster than   a=b/c)?

Math expressions with constants are calcolated during .pyc creation or real
time?
(a=b*5   is faster than    a=b*5*2/2 )?

Thanks




From python at rcn.com  Fri Sep  6 08:15:54 2002
From: python at rcn.com (Raymond Hettinger)
Date: Fri, 6 Sep 2002 08:15:54 -0400
Subject: tree data structure
References: <FM0e9.2860$cP3.6165@news.iol.ie>
Message-ID: <ala6c3$6i2$1@bob.news.rcn.net>

There are a number of ways to represent trees:
-- instances of a treenode class
-- a list of lists:  [l0, [s2, [s0, [q1]], [s1,[q2]], [s3]]]
-- or, my favorite, a dictionary of lists:

>>> original = [('L0','S2'), ('S2','S0'), ('S2','S1'),
     ('S1','Q2'), ('S0','Q1'), ('S2','S3')]
>>> tree = {}
>>> for key, value in original:
 tree.setdefault(key,[]).append(value)

>>> tree
{'S2': ['S0', 'S1', 'S3'], 'S1': ['Q2'], 'S0': ['Q1'], 'L0': ['S2']}



Raymond Hettinger



"Padraig Brady" <Padraig at Linux.ie> wrote in message
news:FM0e9.2860$cP3.6165 at news.iol.ie...
> Hi,
>
> I've a graph like stucture represented
> by a list of tuples:
>
> [(L0,S2), (S2,S0), (S2,S1), (S1,Q2), (S0,Q1), (S2,S3)]
>
> And would like to convert it to
> a tree structure like:
>
>            S0 - Q1
>          /
> L0 - S2 - S3
>          \
>            S1 - Q2
>
> Any tips appreciated.
>
> Thanks,
> P?draig.
>




From tjreedy at udel.edu  Mon Sep  9 21:43:29 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 10 Sep 2002 01:43:29 GMT
Subject: Diving into the Python pool....
References: <mailman.1031613328.24522.python-list@python.org>
Message-ID: <57cf9.424498$Aw4.17832215@bin2.nnrp.aus1.giganews.com>

"Jordan Faris" <jfaris at sbcglobal.net> wrote in message
news:mailman.1031613328.24522.python-list at python.org...
>My primary focus is to create streamlined, rapid-deployment web
applications
>(PHP and JSP portal sites, etc), but from what I've seen, Python can
be used
>or adapted to achieve comparable results;

Your main problem may be choosing from the 5+ Python-based web app
frameworks already available.
Welcome.

Terry J. Reedy







From mstenner at phy.duke.edu  Fri Sep  6 09:51:41 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Fri, 6 Sep 2002 09:51:41 -0400
Subject: pygtk
In-Reply-To: <pan.2002.09.06.13.30.48.241597.8751@ns.kleinerdrache.org>; from martin.klaffenboeck@gmx.at on Fri, Sep 06, 2002 at 03:30:48PM +0200
References: <pan.2002.09.06.13.30.48.241597.8751@ns.kleinerdrache.org>
Message-ID: <20020906095141.A14067@phy.duke.edu>

On Fri, Sep 06, 2002 at 03:30:48PM +0200, Martin Klaffenboeck wrote:
> Hello,
>   I'm new here in Python, because a few people told me that the
> implementation of gtk is better in python than in Perl.  So now I try do
> do object orientated programming here...
> 
> Is this a good group to discuss pygtk things or is this just for python?

There is a separate pygtk list that would be most appropriate for
pygtk questions.  Obviously, the people there are familiar with
python.  It is not necessarily true that people here use/know pygtk.

http://www.daa.com.au/mailman/listinfo/pygtk

> (I think I have questions in both.)

Perhaps you should subscribe to both lists then.

> So now this message is also to say 'Hello' to everybody in this room.

Hello.

> Maybe you'll here much from me. ;-)

Wonderful.  We hope you enjoy python.

					-Michael
-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From gustav at morpheus.demon.co.uk  Sat Sep  7 06:02:02 2002
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: 07 Sep 2002 11:02:02 +0100
Subject: Smoothing a discrete set of data
Message-ID: <3csmdso5.fsf@morpheus.demon.co.uk>

This isn't so much a Python question, as an algorithms question. I
couldn't find a newsgroup for discussing algorithms (is there a good
one?) so as I'm implementing the code in Python I thought I'd try here
- I hope that's OK.

I have a set of data, basically a histogram. The data is pretty
variable, and I'd like to "smooth" it to find trends. Actually, this
comes up a *lot* for me - some examples: sampled IO rates on a machine
- I want to look for trends (is IO higher overnight or during the day,
etc) or fuel consumption for my car (do I use less fuel since I had
the service).

Normally, what I do with things like this is draw a graph, and try to
spot the trends "by eyeball". But it feels to me like I should be able
to write something which smooths the data out. I just don't see how
:-)

Take an example - here's a rough ASCII-art diagram

  9     XX                                  XX
  8     XX XX                               XX
  7  XX XX XX                               XX XX
  6  XX XX XX                            XX XX XX
  5  XX XX XX                            XX XX XX
  4  XX XX XX                            XX XX XX
  3  XX XX XX XX          XX          XX XX XX XX
  2  XX XX XX XX XX    XX XX XX    XX XX XX XX XX
  1  XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX

This seems to me to have very clear peaks at the left and right, and a
shallow point in the middle. I'd say that you could "smooth" this to
something that averaged 8 for the first 3 bars, then 2 for the next 9,
then 7 for the last 3. But try to make that precise...

When I've tried to formalise this, I've always stumbled over the fact
that it's a trade-off exercise (there most be some sort of "tolerance"
parameter in the algorithm) - on the one hand, the original data is
"best" (in the sense that it matches exactly) whereas on the other
hand, a single horizontal line at the average is "best" (fewest number
of steps).

My instinct is that you're looking for something that balances number
of steps vs difference from the original data. But for the life of me
I can't see how to make this intuition precise.

Can anyone help me? I can't believe that this problem has never come
up before, but I can't find any literature on it.

Thanks in advance,
Paul Moore.


From sschwarzer at sschwarzer.net  Sun Sep  8 12:37:05 2002
From: sschwarzer at sschwarzer.net (Stefan Schwarzer)
Date: Sun, 08 Sep 2002 18:37:05 +0200
Subject: Strange threads/stack size interaction on FreeBSD
References: <3D765A95.6090308@tu-clausthal.de> <pan.2002.09.04.21.43.19.620760.2752@metaslash.com>
Message-ID: <3D7B7CB1.4070201@sschwarzer.net>

Hi Neal

Unless noted otherwise, the descriptions below refer to the problematic stack size
of 512 MB.

Neal Norwitz wrote:
>>- Should I enter a bug report on this topic?
> 
> Probably, although, there are few people who can test with your
> configuration.  So you may be stuck debugging this yourself.
 >
 > If you have stack traces, are they very long?  where does the crash
 > occur?

It's rather a block than a crash. I was able to reproduce the problem with this
small test script:

-----
import threading

def hello():
     print "Hello"

t = threading.Thread(target=hello)
t.start()
t.join()
-----

The program blocks in t.join(). A backtrace from pdb.py just before blocking is

-----
(Pdb) bt
   <string>(1)?()
   /srv/home/svss/thread_test.py(8)?()
-> t.join()
   /usr/local/lib/python2.2/threading.py(450)join()
-> self.__block.wait()
 > /usr/local/lib/python2.2/threading.py(190)wait()
-> waiter.acquire()
(Pdb) s
-----

So, to be more precise, the program stops on calling waiter.acquire() in
_Condition.wait(). Apart from that, I could stop the python process only with
the kill command from another terminal, but not with Ctrl-C in the xterm where
Python was running.

I also got kernel trace output on FreeBSD and noted that I see the program
blocking here, too. That is, the dump doesn't grow larger when I let hang the
program around longer.

The last entries before the program hangs are:

-----
  21517 python   CALL  mmap(0xbfaef000,0x10000,0x3,0x400,0xffffffff,0,0,0)
  21517 python   RET   mmap -1 errno 12 Cannot allocate memory
  21517 python   CALL  __sysctl(0xbfbff160,0x2,0x8810c440,0xbfbff15c,0,0)
  21517 python   RET   __sysctl 0
  21517 python   CALL  poll(0x8175000,0,0)
  21517 python   RET   poll 0
  21517 python   CALL  poll(0x8104000,0,0)
  21517 python   RET   poll 0
  21517 python   CALL  poll(0x8104000,0,0)
  21517 python   RET   poll 0
  21517 python   CALL  gettimeofday(0x881a97ec,0)
  21517 python   RET   gettimeofday 0
  21517 python   CALL  poll(0x8104000,0x1,0xffffffff)
-----

which is shortly after reading atexit.pyc.

> What compiler are you using?

gcc 2.95.3

> Can you report all the tests that don't pass?

Output from make test:

...
test test_asynchat crashed -- socket.error: 61
...
test test_fork1 failed -- test failed
...
test test_pwd produced unexpected output:
**********************************************************************
*** mismatch between line 7 of expected output and line 7 of actual output:
- caught expected exception
+ fakeuid 4127 did not except pwd.getpwuid()
**********************************************************************
...
test test_socket failed -- (48, 'Address already in use')
...
161 tests OK.
4 tests failed:
     test_asynchat test_fork1 test_pwd test_socket
...

However, in order to let the tests succeed, I disabled all tests beginning with
"test_thread". Suprisingly, after make test was through, it didn't return to the
prompt but I could stop it with Ctrl-C.

> Have you tried with the latest and greatest source?
>   2.2.1+: cvs update -r release22-maint

I got "cvs [update aborted]: no such tag release22-maint".

>   2.3:    cvs update -A

I get all the same symptoms as described above; I didn't look for the kernel
trace dump, but the test script still hangs.

After just reactivating the old kernel with a stack size of 256 MB I get the
expected output "Hello" and the script returns to the prompt. The last entries
in the kernel trace dump are (from the same position onwards as above):

-----
    217 python   CALL  mmap(0xbfaef000,0x10000,0x3,0x400,0xffffffff,0,0,0)
    217 python   RET   mmap -1079054336/0xbfaef000
    217 python   CALL  setitimer(0x2,0xbfbff554,0)
    217 python   RET   setitimer 0
    217 python   CALL  mmap(0xbfade000,0x10000,0x3,0x400,0xffffffff,0,0,0)
    217 python   RET   mmap -1079123968/0xbfade000
    217 python   CALL  __sysctl(0xbfbff524,0x2,0x4810c440,0xbfbff520,0,0)
    217 python   RET   __sysctl 0
    217 python   CALL  poll(0x8193800,0,0)
    217 python   RET   poll 0
    217 python   CALL  fstat(0x1,0xbfafe80c)
    217 python   RET   fstat 0
    217 python   CALL  ioctl(0x1,TIOCGETA,0xbfafe840)
    217 python   RET   ioctl 0
    217 python   CALL  write(0x1,0x8199000,0x6)
    217 python   GIO   fd 1 wrote 6 bytes
        "Hello
        "
    217 python   RET   write 6
    217 python   CALL  clock_gettime(0,0xbfaedfa4)
    217 python   RET   clock_gettime 0
    217 python   CALL  poll(0x8104000,0,0)
    217 python   RET   poll 0
    217 python   CALL  poll(0x8104000,0,0)
    217 python   RET   poll 0
    217 python   CALL  gettimeofday(0x481a97ec,0)
    217 python   RET   gettimeofday 0
    217 python   CALL  sigaction(0x2,0xbfbff8fc,0)
    217 python   RET   sigaction 0
    217 python   CALL  sigprocmask(0x1,0x480dee60,0xbfbff9e0)
    217 python   RET   sigprocmask 0
    217 python   CALL  sigprocmask(0x3,0x480dee70,0)
    217 python   RET   sigprocmask 0
    217 python   CALL  setitimer(0x2,0xbfbffa00,0)
    217 python   RET   setitimer 0
    217 python   CALL  close(0x3)
    217 python   RET   close 0
    217 python   CALL  close(0x4)
    217 python   RET   close 0
    217 python   CALL  fcntl(0,0x3,0)
    217 python   RET   fcntl 6
    217 python   CALL  fcntl(0,0x4,0x2)
    217 python   RET   fcntl 0
    217 python   CALL  fcntl(0x1,0x3,0)
    217 python   RET   fcntl 2
    217 python   CALL  fcntl(0x1,0x4,0x2)
    217 python   RET   fcntl 0
    217 python   CALL  fcntl(0x2,0x3,0)
    217 python   RET   fcntl 2
    217 python   CALL  fcntl(0x2,0x4,0x2)
    217 python   RET   fcntl 0
    217 python   CALL  exit(0)
-----

Probably, the best is to discuss this on one of the FreeBSD mailing lists.
Anyway, thank you very much for your help. :-)

Stefan



From sross at connectmail.carleton.ca  Wed Sep  4 20:52:45 2002
From: sross at connectmail.carleton.ca (Sean Ross)
Date: 5 Sep 2002 00:52:45 GMT
Subject: Python Source Code Documentation
Message-ID: <al69st$n1f$1@driftwood.ccs.carleton.ca>

Hi. I'm interested in trying to get an understanding of Python's source
code. I'd like to understand how it works. I was wondering if there exists a
set of documents or guidelines I could read, alongside the source, that
would outline what each file in the cvs is, what it does, how it fits into
the overall design: sort of like a thread through the maze. Where to begin,
what's important, how things tie together. That sort of thing.
Thanks for your time,
Sean Ross




From markus.vonehr at ipm.fhg.de  Thu Sep  5 11:56:25 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Thu, 05 Sep 2002 17:56:25 +0200
Subject: upgrading python
Message-ID: <3D777EA9.F9A96C83@ipm.fhg.de>

Hi,

first of all I'd like programming in python.
Sometimes you need to grade up to a newer version.
I have to download everything again and install every single package,
for example:

-win32all
-PyOpengl
-Pyserial
-Numeric
-unzip Pmw
-PIL
-py2exe
-PyXML
etc.

So I just mentioned "only" 8 packages, additionally I have to
install the newer python version and to uninstall the old packages...

Couldn't it be made easier? Maybe I do something wrong and
am misunderstanding how to deal with it.

Now I wanted to grade up to tk84. It was quite hard to find 
_tkinter.c, then I downloaded the tcl-sources, tk-sources, 
installed MSVC6, found out that some X11 stuff is missing...

I know that all mentioned packages are different projects but could
it be made that there's a python distribution with all packages
up2date and the installer is asking for the desired ones?

Do you have similar experiences?

Markus


From mhammond at skippinet.com.au  Sat Sep  7 20:11:26 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 08 Sep 2002 00:11:26 GMT
Subject: untrusted dynamic module: win32api
References: <unk63h7q26lefe@news.supernews.com>
Message-ID: <OAwe9.5047$_l1.12416@news-server.bigpond.net.au>

Enoch wrote:
> Hi,
> 
> I'm a Softimage|XSI user, a 3D graphic software that support ActiveX
> Scripting on win32 platforms.
> 
> This software integrate a 'NetView', an integrated internet explorer.
> 
> To be able to communicate between the NetView and XSI I have to create the
> XSI.Application object.
> 
> Works fine with VBScript and JScript, however I'm running into a problem
> trying to make it work with Python.
> 
> Basicaly I have to include in my html page the following code:
> 
> <SCRIPT Language="Python">
> import win32com.client
> w = win32com.client.Dispatch("XSI.Application")
> </SCRIPT>
> 
> But when I load this page into the NetView ( or even in an external
> brother ) I have this error message:
> 
> 'ImportError: untrusted dynamic module: win32api'
> 
> Is there any way to 'force' the import ?
> 
> Sorry if this has been discussed earlier, I'm relatively new to Python.

It appears that SoftImage is configuring ActiveScripting for "untrusted" 
scripts - ie, SoftImage is requesting the scripting engine switch into 
untrusted mode.  Thus, you really need to ask SoftImage is the 
application can be configured to not request restricted execution mode 
of the language.

Mark.



From gerhard.haering at opus-gmbh.net  Wed Sep 11 06:23:22 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 11 Sep 2002 10:23:22 GMT
Subject: function pointer
References: <ffcedbc9.0209110215.644c1788@posting.google.com>
Message-ID: <slrnanu667.22k.gerhard.haering@haering.opus-gmbh.net>

Christina wrote:
> Does Python have function pointers? [...]

Functions, as well as modules, classes and about everything else are
objects in Python. Python doesn't have pointers, but it does have
references, so you can make a reference to a function like this:

def foo():  pass

# Create a reference to the 'foo' function:
bar = foo

Python is highly dynamic, so this is nothing special. You can make
references to functions, classes and even modules. You can dynamically
create functions and classes at runtime, and a lot more.

Servus,

Gerhard


From steve at ferg.org  Thu Sep 12 08:36:40 2002
From: steve at ferg.org (Stephen Ferg)
Date: 12 Sep 2002 05:36:40 -0700
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid> <alln71$1s2s$1@nntp6.u.washington.edu> <b16e4ef7.0209110616.48dfe04c@posting.google.com> <20020911150511.GA28720@unpythonic.net> <mailman.1031796619.4985.python-list@python.org>
Message-ID: <b16e4ef7.0209120436.53b39cd6@posting.google.com>

Thanks.  This has got to be the authoritative answer on what "command" does!


From BPettersen at NAREX.com  Tue Sep 17 02:37:17 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Tue, 17 Sep 2002 00:37:17 -0600
Subject: threads or queue for this task
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E201920125335E@admin56.narex.com>

> From: James J. Besemer [mailto:jb at cascade-sys.com] 
> 
> >James J. Besemer wrote:
> >  
> >>Generally, Queues never get "full" and you generally don't care if 
> >>they're empty.
> >
> >You can define Queues that can hold a maximum number of items; [...]
> >
> >Setting size is just an optional embellishment, though sometimes it 
> >does pay dividends.
> >
> 
> Setting size has it's place but otherwise introduces unnecessary 
> complications.  Among other things, it forces the queue 
> writer to deal with the situation of what to do if the queue is full.
Signal error 
> back to its client?   Sleep and try again later?  Silently 
> discard the request?  Abort?
> 
> "Generally," NOT having a limit is by far the more elegant solution.

[snip]

You must be kidding (or you can't have written any significant amount of
multithreaded code). In the general producer/consumer paradigm you want
the producer to look very similar to:

  try:
      while 1:
          work = createWork()
          if not work: break
          queue.push(work) # block if queue full
      if I'm the last producer: # kill consumers
          for i in range(numberOfConsumerThreads):
              queue.push(None)
  except Timeout:
      abendGracefully()

and the consumer to look very similar to:

  try:
      while 1:
          work = queue.pop() # block until not empty
          if not work: break
          process(work)
  except Timeout:
      abendGracefully()

If you have any amount of data to process (why else would you be
multithreading?) I've never seen producers and consumers being close
enough in performance to not have the system end up swapping to death --
unless, that is, if the queue is bounded. The Python Queue module
supports most of this out-of-the-box, and adding timeouts is fairly
trivial...

processing-too-many-16Gb-files-with-tight-SLAs-leaves-scares<wink>'ly
y'rs
-- bjorn



From mal at egenix.com  Thu Sep  5 13:24:05 2002
From: mal at egenix.com (M.-A. Lemburg)
Date: Thu, 05 Sep 2002 19:24:05 +0200
Subject: [DB-SIG] RE: What database should I use
References: <F3EF14447A98D311B981009027B0FB380172469E@IMACS_EXCHANGE>
Message-ID: <3D779335.1040603@lemburg.com>

> Hello,
> I am going to program a small application (in Python and wxPython) which
> consists about 2000 - 3000 records . Can you please recommend which database
> is the best for that. I use Windows. Maybe MS Access  or FoxPro or MySQL or
> better? I want to keep the program as small as possible. Thanks Ladislav

If you only use it internally, then I'd suggest downloading
and installing the MS MDAC package and using it via ODBC. You
get the Access Jet-Engine for free that way.

Another approach would be SAP DB which is Open Source and
works nicely on Windows too.

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
_______________________________________________________________________
eGenix.com -- Makers of the Python mx Extensions: mxDateTime,mxODBC,...
Python Consulting:                               http://www.egenix.com/
Python Software:                    http://www.egenix.com/files/python/




From david at dataovation.com  Thu Sep 26 19:51:10 2002
From: david at dataovation.com (David McInnis)
Date: Thu, 26 Sep 2002 16:51:10 -0700
Subject: POP3 and Multipart Mime
Message-ID: <005c01c265b7$97954d40$6400a8c0@bigdaddy>

Does anyone know where I can find a tutorial on how to accomplish the
following?
 
I need to log into a POP3 account (which I have been able to do using
poplib).
 
I need to be able to read and decode the message.  The problem is that
most of our users send word docs as attachments (1 or more).  I need to
take these files and save them to disk.
 
 
 
David McInnis
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020926/9ff38f60/attachment.html>

From aleax at aleax.it  Thu Sep 26 13:22:11 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 26 Sep 2002 17:22:11 GMT
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <amltd3$3k$1@glue.ucr.edu> <3D911FE1.1050801@something.invalid> <amvcev$cf6$1@s1.read.news.oleane.net>
Message-ID: <7nHk9.172208$pX1.6117207@news2.tin.it>

Pedro RODRIGUEZ wrote:
        ...
>>> But why use grotesque hacks in the first place?  "a ? b : c" is a
        ...
> Ada has two short-circuits operators : "and then" and "or else" that
> do the trick :
>     if A and then B then
>         ...
>     end if;
> 
> Looks cleaner to me.

Ada's "and then" is like Python's "and"; Ada;s "or else" is like
Python's "or".  Neither is like C's "a?b:c" (ternary operator),
though with some tricks boolean shortcircuiting operators can be
used to simulate ternary operators (a bit cumbersome in the
general case).

The reason
        zz = a and b or c
is not the same thing as
        zz = a ? b : c
is that, when a is true and b is false, the former returns c,
while the latter returns b.

The tricks deal with ways to work around this problem, by
ensuring that and's RHO can never be false, e.g.
        zz = (a and [b] or [c])[0]


The other traditional emulation of ternary operator is by
indexing, e.g.:
        zz = (b, c)[not a]
and here the defect is that this isn't short-circuiting --
both the b and c sub-expressions are evaluated anyway, no
matter what's the value of a (this is a problem when,
e.g., b is actually a function call such as "fb()").

Here, the workarounds can use and/or if it's OK to
evaluate _a_ multiple times:
        zz = (a and b, not a and c)[not a]
or lambdas:
        zz = (lambda:b, lambda:c)[not a]()


The deeper problem with all of these "solutions" is that
they're unholy kludges, of course.


Alex



From kingprad at mail.com  Fri Sep 27 16:09:52 2002
From: kingprad at mail.com (Carl)
Date: 27 Sep 2002 13:09:52 -0700
Subject: Python Process that can survive
References: <mailman.1033137937.24205.python-list@python.org>
Message-ID: <668b76ce.0209271209.36412b6b@posting.google.com>

That's pretty cool! I can't wait to try it.


From aleax at aleax.it  Wed Sep 18 11:41:08 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 18 Sep 2002 15:41:08 GMT
Subject: Splitting a list of strings
References: <mailman.1032309381.3626.python-list@python.org> <w0Uh9.116763$5r1.4544705@bin5.nnrp.aus1.giganews.com> <am9tpv$9gp$1@driftwood.ccs.carleton.ca> <mailman.1032359962.30186.python-list@python.org>
Message-ID: <o81i9.127490$pX1.4558733@news2.tin.it>

Emile van Sebille wrote:
        ...
> for line in lines:
>     vals["@#".find(line[:1])+1].append(line)
        ...
> Or, if you want to use an lc:
> 
> [ vals["@#".find(line[:1])+1].append(line) for line in lines]

The lc variant wantonly produces and throws away a list of
copies of None, one per line in lines.  It's thus likely to
be quite a bit slower than the simpler loop.

IMHO, vals should be a dictionary indexed by the character
used to discriminate -- using vals as a list and doing a
find operation repeatedly seems a waste.  However, I admit
I haven't timed the two alternatives, so I could be wrong!


Alex



From ahmad at NOSPAMbitbuilder.com  Mon Sep  9 07:55:55 2002
From: ahmad at NOSPAMbitbuilder.com (Ahmad Baitalmal)
Date: Mon, 09 Sep 2002 11:55:55 GMT
Subject: xml.minidom is stripping out my CRLF's in attrib values!!
References: <3D7C80C4.9070204@NOSPAMbitbuilder.com> <j4n0qr9yeo.fsf@informatik.hu-berlin.de>
Message-ID: <3D7C8C76.8060700@NOSPAMbitbuilder.com>

That's not what my problem is, that I know about ( crlf's between nodes),,

Here is the deal:
<cow>
	<hide value="spotted
with black and white"></hide>
</cow>

After the word "spotted" there is a crlf, -inside- the attribute value. 
Sholdn't it be treated as part of the value?

The value now comes stripped of that crlf.


Martin v. L?wis wrote:
> Ahmad Baitalmal <ahmad at NOSPAMbitbuilder.com> writes:
> 
> 
>>What's the deal, is there a setting I can set?
> 
> 
> XML specifies that line endings must be normalized by a processor, see
> 
> http://www.w3.org/TR/REC-xml#sec-line-ends
> 
> There is no option to report anything but #xA to the application.
> 
> Regards,
> Martin
> 


-- 
-  - -- ----  ----------------------------------------- --- -- -   -
*Ahmad Baitalmal***
*BitBuilder***
web: http://www.bitbuilder.com
-  - -- ----
-------------------------------------------------------------- --- -- -   -



From mtremblay at golemlabs.com  Fri Sep 13 10:56:32 2002
From: mtremblay at golemlabs.com (Mathieu Tremblay)
Date: 13 Sep 2002 07:56:32 -0700
Subject: Embed extended Python
Message-ID: <f760295a.0209130656.2096981f@posting.google.com>

Hello, 
   I am quite new to Python and I'd like to embed extended python. 
The problem is that the documentation I can find on the internet is
always related to C standard types and not to C++ objects.

   To extend python I use the really good SWIG tool and everything is
fine.  The problem comes when I try to call a python script that takes
as an argument one of the class that I have defined in C++ through
swig.

Exemple:
 C++ class extended to python with SWIG, through the module wheel: 
   CWheel
   {
     CWheel(){}
     ~CWheel(){}
     printWheel()
     {
       printf("I am a wheel");
     }
   };
 Now, let's suppose that the module is successfully exported and can
be used in python.  An example of simple python script would be :
  import wheel
 
  def thefunction(awheel):
     awheel.printWheel()

 Calling this function directly from python (supposing we provide the
correct argument) works perfectly.  The problem comes when I want to
call this script from C++.  Maybe It is a stupid question but here it
is: How do you pass an existing CWheel as an argument to the function
from C++??? How do you wrap a CWheel to a PyObject?
 I guess the method would be to use the function
PyObject_CallObject(pFunc, pArgs); the pArgs being a PyTuple where I
can set a PyObject to a position X.  The problem is that all the
examples that I find convert int, char*, long, double etc. to PyObject
with built in python functions but no different/user defined class.
 Anybody can help?

 Thanks a lot,
  Mathieu


From shalehperry at attbi.com  Fri Sep  6 12:56:35 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 6 Sep 2002 09:56:35 -0700
Subject: Getting a proccesses ID
In-Reply-To: <I%4e9.190045$_91.214443@rwcrnsc51.ops.asp.att.net>
References: <Qu3e9.188987$_91.214523@rwcrnsc51.ops.asp.att.net> <mailman.1031326272.10078.python-list@python.org> <I%4e9.190045$_91.214443@rwcrnsc51.ops.asp.att.net>
Message-ID: <200209060956.35530.shalehperry@attbi.com>

On Friday 06 September 2002 09:48, Micah Mayo wrote:
>
> Ah, but os.getpid get's my pid, I need to get another programs, and in
> fact, several other program's pid. :)
>

hence the smiley.

If you wanted to be cool about it you could write a python module which parsed 
the various /proc interfaces of BSD, Linux, Solaris and created process 
trees/maps/dictionaries from this data.  Kind of like writing a generic db 
frontend to read the os's backend.



From solonas73 at hotmail.com  Fri Sep 20 07:08:43 2002
From: solonas73 at hotmail.com (Solon Kazantzidis)
Date: 20 Sep 2002 04:08:43 -0700
Subject: access to palm with pilot-link pisock library (python binding)
Message-ID: <24165ed4.0209200308.4d6b3b7a@posting.google.com>

Hello,

I want to achieve the following:

- I want to use the python language to access my palm pilot Vx
- I found the pilot-link package which provides a python binding of
the pisock library written in c
- Unfortunately the documentation I need as newbie to this is listed
in the TODO list so I am searching a documentation
- I want to be able to connect to the Palm (I have some example for a
connection from the pyrite-publisher .py file)
- I want to be able make what the pilot-xfer -l command does
- I want to be able to create/delete/access databases stored on my
Palm
- I need to have access on data which I want to gather with the Palm

What I achieved so far:
1. downloaded pilot-link package (version 0.11.3)
2. build it and created the python binding (python version 2.2.1)
3. from file plugin_instpisock.py (pyrite-publisher) I got some
starting information to establish connection to the palm:

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

#!/usr/bin/python2.2
import pisock

port = "/dev/pilot"

sd = pisock.pi_socket(pisock.PI_AF_SLP, pisock.PI_SOCK_STREAM,
pisock.PI_PF_PADP)
if not sd:
	print "failed to create socket"
	exit
if (pisock.pi_bind(sd, (pisock.PI_AF_SLP, port))) == -1:
	print "failed to bind socket"
	exit
if (pisock.pi_listen(sd, 1)) == -1:
	print "failed to listen on socket"
	exit
ret = pisock.pi_accept(sd)
if ret == -1:
	print "failed to accept connection"
	exit
socket = ret[0]
pisock.dlp_OpenConduit(socket)
pisock.dlp_ReadDBList(sd,0,pisock.dlpDBListRAM,0)

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

As there is no documentation out there and I am not the best
programmer I can only say that the whole procedure stops with a:
"Speicherzugriffsfehler"-->(translated) "Memory access failure"

All works fine until the "ret = pisock.pi_accept(sd)". At this point
the python interpreter stops and waits for me to press the hotsync
button. After I press it the above mentioned error arises.


I need urgently help with this as I cannot with the start form my
thesis until the mechanism for accessing works. (and then the real
work of synchronising gathered data starts....)


Thanks alot for your help/comments,

Solon.


From martin at v.loewis.de  Fri Sep  6 12:03:28 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 06 Sep 2002 18:03:28 +0200
Subject: Getting a proccesses ID
References: <Qu3e9.188987$_91.214523@rwcrnsc51.ops.asp.att.net>
Message-ID: <m365xjumun.fsf@mira.informatik.hu-berlin.de>

Micah Mayo <astrophels at yahoo.com> writes:

> Can anyone tell me a cross platform way of grabbing a particular
> processes ID

Depends on how you want to identify the process. If the process is a
child of yours, you best safe the process id at the time of process
creation.

If the process is sshd, you read at /var/run/sshd.pid. If you don't
trust that sshd saves its pid file in /var/run, you read
/etc/sshd/sshd_config - it might be best to find out the location of
sshd.pid at installation time.

If you *really* need to invoke ps, I recommend to
- perform the searching in Python, instead of invoking grep/awk.
- you plan for two families of ps binaries: SysV and BSD. On Solaris,
  ps x says
  
usage: ps [ -aAdeflcjLPy ] [ -o format ] [ -t termlist ]
        [ -u userlist ] [ -U userlist ] [ -G grouplist ]
        [ -p proclist ] [ -g pgrplist ] [ -s sidlist ]
  'format' is one or more of:
        user ruser group rgroup uid ruid gid rgid pid ppid pgid sid taskid
        pri opri pcpu pmem vsz rss osz nice class time etime stime
        f s c lwp nlwp psr tty addr wchan fname comm args projid project pset

  On Solaris, it so happens that you can use /usr/ucb/ps x, but that
  won't be available on other SysV systems. The SysV equivalent is "ps
  -e".

- you plan for multiple sshd processes being found. In theory, you
  need to find the one that has "1" as its parent process, but killing
  the one with the lowest pid might be good enough in real life.

Regards,
Martin


From peter at engcorp.com  Thu Sep  5 23:51:53 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 05 Sep 2002 23:51:53 -0400
Subject: A Free Idea: Search Engine for Webpages
References: <m27kig8qw1.fsf@boiko_linux.demogr.mpg.de> <pan.2002.08.24.16.57.44.643843.728@thomas-guettler.de> <m27kigcdh7.fsf@boiko.zv.uni-rostock.de> <3D685D54.249F3350@engcorp.com> <m23ct3kl3u.fsf@boiko_linux.demogr.mpg.de> <477762c2.0208260615.3790945@posting.google.com> <3D6AB55D.63D36207@engcorp.com> <m2k7mc96f1.fsf@boiko_linux.demogr.mpg.de>
Message-ID: <3d78263f@news.sentex.net>

Serge Boiko wrote:
> It would be very kind of you to ask the name of the software.
> Many thanks.
> -Serge

Sorry, I lost track of this thread after you asked and before I
managed to find out, until Amit's posting brought it to my
attention again (and basically answered the question, too).

More specifically, http://toolbar.google.com should be where you go
for this thing.

-Peter



From david.abrahams at rcn.com  Sun Sep 29 16:15:27 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Sun, 29 Sep 2002 16:15:27 -0400
Subject: Why does QT Windows need Visual Studio?
References: <3ljepusnje8iuca90im2r9vceja5tfc5l7@4ax.com>
Message-ID: <an7p0b$kdi$1@bob.news.rcn.net>


"John Hall" <wweexxsseessssaa at telusplanet.net> wrote in message
news:3ljepusnje8iuca90im2r9vceja5tfc5l7 at 4ax.com...
> I was about to download the Windows non-commercial version of QT, but
> noticed that it requires MS Visual Studio, so didn't.
> This requirement surprises me - is there a way around that?

http://kde-cygwin.sourceforge.net/qt3/

--
-----------------------------------------------------------
           David Abrahams * Boost Consulting
dave at boost-consulting.com * http://www.boost-consulting.com




From dbasch at yahoo.com  Mon Sep 30 12:11:06 2002
From: dbasch at yahoo.com (Derek Basch)
Date: Mon, 30 Sep 2002 09:11:06 -0700 (PDT)
Subject: Catching user defined exceptions from foreign modules
Message-ID: <20020930161106.27910.qmail@web20805.mail.yahoo.com>

Everyone,
Well, after pulling my hair out last night trying to
figure out why my user defined exceptions were not
working as I would expect, I fired up the modules this
morning and they work exactly as I would expect them
too. Its enough to drive a man (or woman) insane!
However, I still would appreciate someone telling me
if it is normal to have to import user defined
exceptions. Thanks for all the help everyone and sorry
to waste your time :(
Thanks,
Derek Basch

__________________________________________________
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
http://sbc.yahoo.com



From bdesth at nospam.free.fr  Tue Sep  3 23:37:12 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Tue, 03 Sep 2002 23:37:12 -0400
Subject: ascii value
References: <akv4fn$eo2$03$1@news.t-online.com> <akv4r1$qd7$04$1@news.t-online.com>
Message-ID: <3D757FE8.2020907@nospam.free.fr>

Axel Grune wrote:
> A sorry, I missed ord(), shoot me

Bang !

laotseu



From tim.schulz at sdm.de  Wed Sep  4 05:22:38 2002
From: tim.schulz at sdm.de (Tim Schulz)
Date: 4 Sep 2002 02:22:38 -0700
Subject: using python to call java code??
Message-ID: <3cc90cd2.0209040122.17242a04@posting.google.com>

I am considering to use zope in order to develop an content management
portal. My problem is, that I have to integrate some ready-built java
code. Is there any possibility to call java code from python? Is there
a API for this kind problem?

Thanks in advance.

Tim Schulz


From Padraig at Linux.ie  Mon Sep 30 09:01:50 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Mon, 30 Sep 2002 13:01:50 GMT
Subject: Find out link location
References: <7uf9na-vd9.ln@lbdb01.gsi.de>
Message-ID: <3D984AF8.3070908@Linux.ie>

Klaus Hoeppner wrote:
> Hi,
> 
> with the os.path module I can use the islink function to check, wether
> a file is symbolic link. But how can I find out, where the link points
> to. E.g. if I have readme.txt --> ../packages/mypackage/readme.txt how
> do I get the "../packages/mypackage/readme.txt" with python?

import os
filter (lambda x: x.find("link") != -1, dir(os))

#see the 2nd item

P?draig.



From Oschler at earthlink.net  Sat Sep 28 23:43:34 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sun, 29 Sep 2002 03:43:34 GMT
Subject: Fast-track to installing/implementing a SWIP module?
References: <Mktl9.13603$Lg2.3951792@news2.news.adelphia.net> <slrnapcp8g.617.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <GFul9.13719$Lg2.3965715@news2.news.adelphia.net>

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote in message
news:slrnapcp8g.617.gerhard.haering at lilith.my-fqdn.de...
> Robert Oschler wrote in comp.lang.python:

> If I understand correcly, you only have to have SWI-Prolog in your PATH,
then
> the following:
>
>     python setup.py build
>
> will build everything and
>
>     python setup.py install
>
> will install your modules.
>
>
> Explanation: This function:
>
>     def get_atom(atom):
>         query = "pl -f none -g 'current_prolog_flag(%s, V), write(V),
nl.' -t 'halt'"
>         f = os.popen(query % atom)
>         ...
>
> in setup.py should do the magic of finding the include and library paths
to
> link against SWI Prolog.
>
> -- Gerhard

Gerhard,

Thanks that's exactly what I needed.  I'm still having a hard time with the
'ease' that things happen when doing things in Python.

thx





From tismer at tismer.com  Fri Sep 20 09:19:19 2002
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 20 Sep 2002 15:19:19 +0200
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <mailman.1032329366.6813.python-list@python.org> <j4ptvbaapm.fsf@informatik.hu-berlin.de> <3D890FB0.4030408@something.invalid> <m3admeo4u1.fsf@mira.informatik.hu-berlin.de> <3D8A7735.3090403@something.invalid>
Message-ID: <3D8B2057.5090002@tismer.com>

Greg Ewing wrote:
> Martin v. Loewis wrote:
> 
>> That is not true. It is more reliable to inspect the source code.
>>
>> In general, it is very easy to do so: if there are no callbacks to
>> Python, the module is stackless-safe.
> 
> 
> 
> That's a fair point, and I guess it takes care of
> the majority of extensions.
> 
> But the cases where there is doubt are typically
> going to be big, complicated things like GUI
> libraries. Making a line-by-line inspection of
> one of those isn't exactly something you do over
> a cup of tea.
> 
> And you don't just have to look at the code, you
> have to understand what it's doing, because the
> things which cause problems are dynamic rather
> than static features of the code.

Fine. But still I'm trying to see your true message.
Why is it bad to just switch off switching when
an unknown extension is entered?
Making that xtension stackless-safe is in fact
a project, I can't see anything bad with it.

Compare it to olde Stackless: It refused to switch
anything that wasn't the toplevel interpreter.
New Stackless can be restricted in the same way.
The only difference is that the new technology
is *able* to switch any stacks. This should only
be enabled if one knows what she's doing.

But there is no reason to damn a new technology
just because it is very powerful. Run it tamed
by default, and everything is fine and by no means
worse than before.
On the other hand, if you want to take advantage
of it with extension modules, then analyse these
modules. This is a project, and I'm greatful for
any contribution.

but-there-is-no-free-lunch -- ciao - chris
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From ianholmes01 at lycos.com  Tue Sep  3 12:09:45 2002
From: ianholmes01 at lycos.com (Ian Holmes)
Date: 3 Sep 2002 09:09:45 -0700
Subject: Python to surpass C performance by 2030
References: <kuptw4o65q.fsf@lasipalatsi.fi> <3D6B6D4B.B8EE2F7D@engcorp.com> <mailman.1030464340.15191.python-list@python.org>
Message-ID: <b31eb63a.0209030809.6d901f3e@posting.google.com>

holger krekel <pyth at devel.trillke.net> wrote in message news:<mailman.1030464340.15191.python-list at python.org>...
> We might as well consider sponsoring Armin Rigo with his fantastic
> Psyco effort to achieve order-of-magnitude increases for python3.
> 
>     holger

Yup pystoned it on my win2k machine and got a 4 fold increase. - ran a
self made test script based on the "why it is so dramatical" post

http://groups.google.com/groups?dq=&hl=en&lr=&ie=UTF-8&threadm=mailman.1031047453.13193.python-list%40python.org&prev=/groups%3Fhl%3Den%26group%3Dcomp.lang.python

and for the list join solution post a nice 3x speed up (the string +=
string2 was about the same unsurprisingly) - what ever Armin is doing
its amazing and he has my vote.

Ian


From max at alcyone.com  Mon Sep 16 02:06:42 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 15 Sep 2002 23:06:42 -0700
Subject: couldnt find __dic__
References: <7153651c.0209151958.4099ac3e@posting.google.com>
Message-ID: <3D8574F2.3929F514@alcyone.com>

black wrote:

> my book told me all dictionaries have __dic__ method but I searched
> each dictionaries and find nothing, where is it ???

I think you misread it.  What it probably said is that all _instances_
have __dict__ attributes.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Success and failure are equally disastrous.
\__/ Tennessee Williams
    REALpolitik / http://www.realpolitik.com/
 Get your own customized newsfeed online in realtime ... for free!


From Padraig at Linux.ie  Mon Sep 16 06:01:01 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Mon, 16 Sep 2002 10:01:01 GMT
Subject: representing a literal single slash
References: <dd5454fb.0209141828.1ad24260@posting.google.com> <3D84E9F9.8090101@linux.ie> <3D84FEE6.90600@linux.ie> <sh8h9.52633$AR1.2065934@bin2.nnrp.aus1.giganews.com>
Message-ID: <xZhh9.4239$cP3.10322@news.iol.ie>

Terry Reedy wrote:
> "Padraig Brady" <padraig at linux.ie> wrote in message
> news:3D84FEE6.90600 at linux.ie...
> 
>>>mystr = mystr.replace(r'\\', r'\')
>>
>>Emm this actually doesn't work. Why can't you:
>>s=r'\'
>>Is it a bug?
> 
> 
> Ref Manual 4.1:
> When an `r' or `R' prefix is present, a character following a
> backslash is included in the string without change, and all
> backslashes are left in the string. For example, the string literal
> r"\n" consists of two characters: a backslash and a lowercase `n'.
> String quotes can be escaped with a backslash, but the backslash
> remains in the string; for example, r"\"" is a valid string literal
> consisting of two characters: a backslash and a double quote; r"\" is
> not a valid string literal (even a raw string cannot end in an odd
> number of backslashes).


> Specifically, a raw string cannot end in a single backslash (since 
 > the backslash would escape the following quote character)

But there is no escaping when the prefix r is used?
I guess there is a good reason for this inconsitency?

P?draig.



From jacob at boris.cd.chalmers.se.cd.chalmers.se  Wed Sep 25 13:37:22 2002
From: jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen)
Date: 25 Sep 2002 17:37:22 GMT
Subject: Will "Python in a Nutshell" be too short?
References: <mailman.1032288261.15607.python-list@python.org> <LMNh9.15079$T4.1903@fe04> <915a998f.0209230848.60cc1415@posting.google.com>
Message-ID: <amss8i$e0e$1@nyheter.chalmers.se>

In article <915a998f.0209230848.60cc1415 at posting.google.com>,
Hamish Lawson <hamish_lawson at yahoo.co.uk> wrote:
>Steve Holden wrote:
>
>> I wil be very surprised if Alex' book does not become the canonical
>> reference for the language, since his treatment is exceptionally thorough
>> and (even for him) more than usually precise.
>
>I had no doubts about this - I'm just a greedy so-and-so who wanted to
>have Alex's treatment over as wide a range of material as would fit
>into a book that could still qualify to be called a nutshell! I'm
>therefore delighted to learn from Alex that his book will in fact
>cover a good number of the most popular third-party libraries, and
>I'll therefore get to throw away a bunch of printed-out notes after
>all (which is what I meant when I hoped that "Python in a Nutshell"
>would be a one-stop reference). I know Alex wanted to keep the book to
>about 400 pages, but for me the 600 pages that the book has grown to
>is still nutshell enough and a small price to pay for the breadth of
>coverage. I'm looking forward to it.
>
>
>Hamish

I have had a small peek at 2 chapters in Alex Nutshell book. All I can say is
that I think it is going to be the best Python reference on the market.

Jacob Hall?n

-- 


From phr-n2002b at NOSPAMnightsong.com  Fri Sep 20 07:06:28 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 20 Sep 2002 04:06:28 -0700
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com> <t2Di9.336220$kp.1062154@rwcrnsc52.ops.asp.att.net>
Message-ID: <7xznuc2a4r.fsf@ruckus.brouhaha.com>

"Dave Gustafson" <DGustafson at attbi.com> writes:
> I've been programming for, hard to believe when I think of it, 30 years...
> I just recently picked up a book that really brought back som fo the Joy of
> programming and programming languages.  I wish I'd picked it up much
> earlier!  It's not about Python, but rather about a dialect of Lisp called
> Scheme, and is titled "The Little Schemer".  It's been really fun to work
> through and really "grok" what's going on in programming, also fun to
> translate the concepts there in to Python which to me is a more real-world
> language.  I'd suggest it as a way to have some fun with this art...

That sounds like either a "port" or a re-interpretation of the classic
"The Little Lisper".

You should also of course have read "Structure And Interpretation of
Computer Programs" by Hal Abelson and Gerry Sussman:

  http://www-mitpress.mit.edu/sicp/

The complete text is on that website now!


From GavinT at ResMed.com.au  Wed Sep 18 01:11:02 2002
From: GavinT at ResMed.com.au (Gavin Tomlins)
Date: Wed, 18 Sep 2002 15:11:02 +1000
Subject: PythonScript
Message-ID: <362133AB12B2CE4EA3C01E12CD6CCC6E0254E08E@osprey.corp.resmed.org>

Thanks for the prompt reply David, yes unfortunately the suggestions you
proposed have been implemented though to no avail. I guess its a matter of
attrition. Shall repost if I find a solution.

Regards

Gavin



-----Original Message-----
From: David LeBlanc [mailto:whisper at oz.net]
Sent: Wednesday, 18 September 2002 2:35 PM
To: Gavin Tomlins; python-list at python.org
Subject: RE: PythonScript


<snip>
> The problem I seem to be hitting is utilising python in applications that
> embed an ActiveScript Engine. To date I am unable to get one python script
> to execute. Other languages vbscript, jscript,perl etc, work in the same
> environment so I'm left wondering if its a configuration problem or
> something that I might be overlooking. If anyone could give me pointers,
> url's, snippets etc. it would be interested. The product I specifically
> having problems is FinalBuilder (http://www.atozedsoftware.com)
> though have
> seen this problem in other applications as well.
>
> I've tried both ActiveState and Python distribution 2.2.1 with Win32
> 1.48all.
>
> Regards
>
> Gavin
>

Did you register Python as an ActiveScript host? I believe this is done for
you automatically if you use the installer and not done for you if you
don't. It's possible that it gets done for you the first time you run the
PythonWin IDE too.

Have a look at this file: \Python22\Lib\site-packages\win32com\readme.htm

HTH,

Dave LeBlanc
Seattle, WA USA



Warning:  Copyright ResMed.  Where the contents of this email and/or attachment includes materials prepared by ResMed, the use of those materials is subject exclusively to the conditions of engagement between ResMed and the intended recipient.

This communication is confidential and may contain legally privileged information. By the use of email over the Internet or other communication systems, ResMed is not waiving either confidentiality of, or legal privilege in,the content of the email and of any attachments.
If the recipient of this message is not the intended addressee, please call ResMed immediately on  +61 2 9886 5000 Sydney, Australia.




From dwelch91 at nospam.attbi.com  Fri Sep 20 01:45:58 2002
From: dwelch91 at nospam.attbi.com (dwelch)
Date: Fri, 20 Sep 2002 05:45:58 GMT
Subject: newbie seeks advise.
References: <77c84db2.0209192137.51167ebb@posting.google.com>
Message-ID: <3D89F5E4.1000002@nospam.attbi.com>

akirasugiura wrote:
> Hi,
> 
> I've been learning python for about a month. I had no prior knowledge
> of programming.
> 
> My question is, I have a tuple which is, data = (69, 40, 74, 41, 56,
> 57, 38, 58, 26, 55). How I can make a function that picks up a largest
> number
>  from this "data" tuple. If "data" was a  'List' I would be able to
> use "sort" but since sort doesn't  work with tuple. I am stuck here.
> 
> Thanks in advance
> 
> Sincerely,
> 

Simple way, probably not optimal:

 >>> max_element = max(list(data))
 >>> max_element
74



From hancock at anansispaceworks.com  Sun Sep  8 14:08:17 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Sun, 08 Sep 2002 11:08:17 -0700
Subject: horizontal scrolling (was Re: Larry Wall's comment on python...)
Message-ID: <3D7B9211.86A1F406@anansispaceworks.com>

Terry Hancock wrote:
> that you don't need () on the if-clause (they may think they
> will cause a syntax error  -- which they *do* in csh/tcsh,
> which they may have previous experience in, BTW). 

Duh. I don't know what I was remembering here, but it's
not true that tcsh generates an error when the if expression
is in ()s. There's plenty of other stuff to complain
about, though. I think csh was invented to torture sysadmins.
;-D

Cheers,
Terry

-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
Anansi Spaceworks                 
http://www.anansispaceworks.com 
P.O. Box 60583                     
Pasadena, CA 91116-6583
------------------------------------------------------



From rjones at ekit-inc.com  Sun Sep  1 07:45:51 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Sun, 1 Sep 2002 21:45:51 +1000
Subject: ANN: New PEP Format: reStructuredText
In-Reply-To: <X5K05iAtPdc9Ewdt@jessikat.fsnet.co.uk>
References: <mailman.1030681307.29889.python-list@python.org> <mailman.1030831309.22494.python-list@python.org> <X5K05iAtPdc9Ewdt@jessikat.fsnet.co.uk>
Message-ID: <200209012145.51274.rjones@ekit-inc.com>

On Sun, 1 Sep 2002 6:46 pm, Robin Becker wrote:
> In article <mailman.1030831309.22494.python-list at python.org>, Richard
> Jones <rjones at ekit-inc.com> writes
> >This is done in the docutils sandbox under the richard/pythonpoint
> > directory. I had this working ages ago (back in March, I think) but
> > hadn't updated it when David merged the dps and restructuredtext projects
> > to form the docutils project. That's done now.
> >
> >
> >    Richard
>
> So does this appear as a CGI somewhere or should I go ahead and make a
> working demo?

No, there's no CGI. The pythonpoint script in the sandbox is a "working demo". 
Pass it a rest file as a command-line arg, and it'll spit out pythonpoint XML 
(read the example for limitations though).


    Richard




From aleax at aleax.it  Wed Sep 18 10:09:15 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 18 Sep 2002 14:09:15 GMT
Subject: Splitting a list of strings
References: <mailman.1032309381.3626.python-list@python.org> <am8pvo$dcn$1@driftwood.ccs.carleton.ca> <w0Uh9.116763$5r1.4544705@bin5.nnrp.aus1.giganews.com> <am9tpv$9gp$1@driftwood.ccs.carleton.ca>
Message-ID: <fO%h9.109881$ub2.2324280@news1.tin.it>

Sean Ross wrote:
        ...
> filter, where you drop handfuls of change into the top of the device and
> they come out the other side sorted into pennies, nickels, dimes,
> quarters. In this case I had comments, attributes, and data. The comments
        ...
> You can probably just chalk this all up to an infatuation with list
> comprehensions.

Yep.  A list comprehension is a device that Python got from functional
programming (Haskell, specifically): in functional programming, data
structures don't mutate.  Thus, a list comprehension doesn't mutate
existing structures -- it just produces a new one.

For the "change sorter" effect, use a normal loop and a dictionary
of lists.  Say you have a function "classifier" that accepts any
item in the list and returns how the item should be classified (as
a number, say, or a string -- any hashables that distinguish item
classifications will do).  "Sieving" a list that way is then trivial:

def sieve(alist, classifier):
    result = {}
    for item in alist:
        result.setdefault(classifier(item),[]).append(result)
    return result

If you don't want the result as a dictionary of lists, but e.g. a
list of lists in ascending order of classifier-returned keys,
change the last statement to:
    keys = result.keys()
    keys.sort()
    return [result[key] for key in keys]

See, you get to use a list comprehension again in this case:-).


Alex



From ark at research.att.com  Wed Sep 18 17:08:31 2002
From: ark at research.att.com (Andrew Koenig)
Date: Wed, 18 Sep 2002 21:08:31 GMT
Subject: A list comprehension to do this?...
References: <3D88E4F5.B443918@quartercase.com>
Message-ID: <yu994rcnuhtc.fsf@europa.research.att.com>

joe> Is there a list comprehension or other 'cleaner' or 'prettier' way 
joe> to do the following:

>>>> xx = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10)
>>>> ww = []
>>>> for i in range(0, len(xx), 2):
joe> 	ww.append((xx[i+1],xx[i]))

How about

        [(xx[i+1], xx[i]) for i in range(0, len(xx), 2)]

?
-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From aahz at pythoncraft.com  Wed Sep  4 17:44:55 2002
From: aahz at pythoncraft.com (Aahz)
Date: 4 Sep 2002 17:44:55 -0400
Subject: quick question about generator.next()
References: <mailman.1031153647.10654.python-list@python.org>
Message-ID: <al5usn$80r$1@panix1.panix.com>

In article <mailman.1031153647.10654.python-list at python.org>,
Erik Price  <erikprice at mac.com> wrote:
>
>If I have a generator object, the first value produced by the generator 
>object does not actually get produced until the first time I call 
>generator_object.next() , correct?
>
>In other words, simply instantiating the generator object doesn't 
>"activate" the generator function, rather it just creates it -- only 
>when the next() method is called does the generator's internal logic 
>"activate", right?

Correct.  I find that although technically inaccurate, it's a lot easier
to think of generators as functions that return iterator objects.

>(Where are generator objects getting their next() method from anyway... 
>is it implicitly inherited behind the scenes when I use the "yield" 
>keyword to turn a function into a generator?)

Generator objects conform to the iterator protocol.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From erick_papadakis at REMOVEFORSPAM.yahoo.com  Tue Sep 17 11:30:28 2002
From: erick_papadakis at REMOVEFORSPAM.yahoo.com (Erick)
Date: Wed, 18 Sep 2002 00:30:28 +0900
Subject: can wxpython do this?
References: <737aefee.0209151702.3f735252@posting.google.com> <spcaou804m5oo6gah31oqaeuqsuj60h0ip@4ax.com> <am3d0b$jj$1@news01bc.so-net.ne.jp>
Message-ID: <am7hrt$9b3$1@news01bb.so-net.ne.jp>

sorry for restating, but is there a way for me to connect to the Internet,
make a call to a webserver as a browser would do, retrieve data and display
it inside my application?

perhaps an example of using sockets in wxpython?

thanks much/erick




From cliechti at gmx.net  Sat Sep  7 16:42:28 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 7 Sep 2002 22:42:28 +0200
Subject: Killing/Terminating a running thread
References: <mailman.1031419029.11871.python-list@python.org>
Message-ID: <Xns9282E6FBFE1C2cliechtigmxnet@62.2.16.82>

Heiko Wundram <heikowu at ceosg.de> wrote in 
news:mailman.1031419029.11871.python-list at python.org:

> Hi to all!
> 
> I've looked over google groups, but I've found no real answer that
> satisfied the needs I have. Just a little explanation:
> 
> I have a Queue module, which stores method-calls and delegates them to
> threads, which take over running the call. Now it happens that a method
> takes much longer to complete than I would like it to, and in this case,
> the thread running the code should just be terminated, discarding all
> values computed so far.

sound like the news thread "Interrupting Python" could be interresting for 
you...
 
> The basic problem I have that it is pretty much impossible to get at the
> PID of the thread running the code in question; but I have a reference
> to the thread class in a central "garbage-collector" (the queue object).
> And anyway, this approach would only work on Unix (I guess), when using
> pthreads, which create a "real" subprocess with its own PID for a
> thread.
> 
> Has anyone had a similar problem, and know of a solution?

the clean solution is to test for a flag and terminate if the flag is set.

e.g. you could provide a function:

def interrupted():
    	return global_flag_if_user_wants_to_abort

and in you function in the queue:

def worker():
    	while not interrupted():
    	    	do_stuff()

you can do that of course with or without a function. a function has the 
benefit that you could do additional checks, such as stopping only a 
specific thread/worker. (e.g. threading.currentThread() could be used to 
test if it the thread to be killed)

chris
-- 
Chris <cliechti at gmx.net>



From imbosol at vt.edu  Sun Sep  1 19:45:55 2002
From: imbosol at vt.edu (Carl Banks)
Date: Sun, 1 Sep 2002 19:45:55 -0400
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <mailman.1030907703.26266.python-list@python.org>
Message-ID: <akuifb$h0e$2@solaris.cc.vt.edu>

Gerhard H?ring wrote:
>> which means the real time is not spent in the compile() function, but
>> in the match or find function. So basically, couldn't one come up with
>> a *human readable* syntax for re, and compile that instead?
> 
> That's equally powerful? Most probably not.

Why not?  It won't be as fast, but it should be able to do anything a
regexp can do, and would be much more versatile.


-- 
CARL BANKS
http://www.aerojockey.com


From fperez528 at yahoo.com  Fri Sep 27 17:35:56 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Fri, 27 Sep 2002 15:35:56 -0600
Subject: Better terminal for windows?
References: <3D94B037.2020202@motorola.com> <mailman.1033158756.25404.python-list@python.org>
Message-ID: <an2ivs$8fg$1@peabody.colorado.edu>

Gerhard H?ring wrote:

>> If I want to run python interactively (and perhaps using ipython), is
>> there something better than the "cmd" terminal?
> 
> Well, you could install MSYS from the mingw project, which is sort of a
> Cygwin-lite :-)
> 
>> I'm used to bash on linux at home, but I'd like to avoid installing
>> cigwin for now. Thanks.
> 
> cmd.exe isn't that bad once you activate tab completion in the
> registry. Have you done that?

That is, as long as you aren't interested in tab completion for python :) The 
problem is lack of support for readline, which is how you get tab completion 
under python.

The other problem is that the windows terminal doesn't properly honor ANSI 
color escapes. I tried a lot to get ipython's coloring to work under windows 
and finally gave up in disgust. Why is it that MS can't be bothered to _ever_ 
follow the simplest standards?

So some of ipython's nicest functionality (full support for tab completion for 
both objects and files,  and color-enhanced information, tracebacks and even 
syntax highlighted source info) are gone under Windows.

If you know of a way to fix this, patches are welcome :)

Cheers,

f.


From Oschler at earthlink.net  Tue Sep 10 12:52:10 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Tue, 10 Sep 2002 16:52:10 GMT
Subject: Trouble upgrading to 2.2.1
References: <aXof9.29813$w51.7932884@news2.news.adelphia.net> <slrnans81k.jc.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <_qpf9.29862$w51.7937790@news2.news.adelphia.net>

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote in message
news:slrnans81k.jc.gerhard.haering at lilith.my-fqdn.de...
> Robert Oschler wrote in comp.lang.python:
> > I'm trying to upgrade from 2.1.1 to 2.2.1.  I'm using kpackage (KDE's
rpm
> > wrapper).
>
> You should install the OpenSSL libraries in order to fulfill these
> dependencies. If you do have the OpenSSL libraries installed, but the
> libcrypto or libssl version numbers are different on your machine, you can
> try to provide the necessary symlinks so rpm will shut up.
>

I have 0.9.6b-34 of OpenSLL installed.  Is this the wrong version or do I
have a different problem?

thx





From steve at ferg.org  Tue Sep 10 21:25:58 2002
From: steve at ferg.org (Stephen Ferg)
Date: 10 Sep 2002 18:25:58 -0700
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid> <b16e4ef7.0209090552.3ad7691b@posting.google.com> <V57f9.77078$l_4.29066@atlpnn01.usenetserver.com>
Message-ID: <b16e4ef7.0209101725.6f508364@posting.google.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message news:<V57f9.77078$l_4.29066 at atlpnn01.usenetserver.com>...
> "Stephen Ferg" <steve at ferg.org> wrote in message
> news:b16e4ef7.0209090552.3ad7691b at posting.google.com...

> The problem, from a learner's point of view...
> Steve Holden  

Thanks!  An excellent answer.

And I think it puts a finger on what I'm trying to do in "Thinking in
Tkinter" in contrast to other explanations of Tkinter that I've
encountered.

Most other introductions to Tkinter programming would start by
explaining how to use "command", on the grounds that it is simpler to
use, and will get you up and running faster.  A quicker way to get a
learner productive.  A perfectly reasonable approach.

What I wanted to do, though, in "Thinking in Tkinter" was more of an
explanation of the basic concepts that you need in programming with
Tkinter.  From that standpoint, I think it is better to explain about
binding -- that is, binding widgets, events, and callback handlers --
by explaining "bind()".  Then, once the basic concepts are explained,
I can note how "command" can be used as a labor-saving technique.

As you say, "it's rather like teaching someone to drive by explaining
the mechanical details of the steering and engine."  Maybe a good
example would be, "it is sort of like teaching someone to drive by
explaining the details of the differential".  It is possible to
explain the techniques of say, driving on slippery road surfaces -- or
getting out of ditches when you're stuck -- without explaining how the
differential works.  But if you know how the differential works, you
will better understand why you should -- or shouldn't -- do certain
things in certain situations.

I guess what I'm saying is that this kind of explanation gives you
more than just techniques.  It gives you an organizing set of deeper
concepts -- a mental model -- within which the techniques make sense.

That's the kind of explanation of Tkinter that I'd like to read. 
"Thinking in Tkinter" will never be the book that I'd like to read.  I
don't have the time to do that.  But I'd like it to cover a few of the
most basic points, in the way that such a book would cover them.  I'm
not saying that that kind of approach is better.  Only that it might
be a better approach for certain folks with certain learning styles.

... My apologies for the verbosity!

Anyway, Steve, thanks for your very helpful answer.  You're a good
explainer.  I'm going to go to Amazon right now and order a copy of
"Python Web Programming"!


From Doc at goodweb.de  Thu Sep  5 18:03:57 2002
From: Doc at goodweb.de (Thomas Korb)
Date: 06 Sep 2002 00:03:57 +0200
Subject: 4Suite / McMillan Installer / py2exe
References: <doc.m3admxkclw.fsf@goodweb.de>
Message-ID: <doc.m3it1k9jqq.fsf@goodweb.de>

Thanks for the answers! (@Thomas Heller: I really like your 
py2exe; it does a great job when using the standard Python-
distribution only; and it is very easy to use!)


I think I found a solution for Gordon McMillan's Installer:

In the 4Suite distribution, there is a module named 'Math.py'
in (e.g. for Python 2.2): 

  C:\Python22\Lib\site-packages\Ft\Xml\Xslt\Exslt\
  
This causes problems with Python's built-in math-module (most
probably since Win32 systems are not case-sensitive) when trying
to use Gordon McMillan's Installer. (The executable is built; but
when you try to use e.g. the XSLT-Processor of 4Suite, you get a 
lot of error-messages.)

Solution - Rename the above module to something like 'myMath.py'
and change the import of 'Math' in __init.py__ (in the above dir)
to 'myMath'.

Then everythingworks just fine. (I am only using the XSLT-Processor; 
so maybe the above changes affect other parts of 4Suite. Did not 
check that up to now. But as far as I could see, the Math.py - module 
is not used very often in the 4Suite-distribution.)

I will report this to the people at 4Suite.org. Maybe they still have
the chance to change this modules name in the next distribution.

-- 
-----------------  Dr. Thomas Korb  -  G O O D W E B . D E s i g n  ---
-----------------  DOMAIN: www.goodweb.de / E-MAIL: Doc at goodweb.de  ---


From dman at dman.ddts.net  Thu Sep 12 10:35:32 2002
From: dman at dman.ddts.net (Derrick 'dman' Hudson)
Date: Thu, 12 Sep 2002 10:35:32 -0400
Subject: Paradox  table
In-Reply-To: <3D7F5DD6.8814.1363983@localhost>
References: <3D7F5DD6.8814.1363983@localhost>
Message-ID: <20020912143532.GA17566@dman.ddts.net>

On Wed, Sep 11, 2002 at 03:14:30PM +0200, A wrote:
| Hi,
|  Is there a module for reading Paradox database (.DB) files with Python?
| I need to run the script under Linux, so I can not use ODBC.
| Any suggestion would be appreciated

One way to access the data is to export it from paradox in CSV format.
CSV is an open standard and several (python) parsers exist for it.
However, using this method won't allow you to easily access the data
from both paradox and from your script, instead you would need to
re-export it every time you modified it in paradox and you would not
be able to modify it using your script.  If you can, it would probably
be easiest to migrate to a different db such as postgresql.  Postgres
databases are much more accessible than paradox is (at least with the
paradox version I've seen, 3.5).

-D

-- 
Dishonest money dwindles away,
but he who gathers money little by little makes it grow.
        Proverbs 13:11
 
http://dman.ddts.net/~dman/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 248 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020912/c86d3109/attachment.sig>

From sholden at holdenweb.com  Fri Sep  6 18:11:00 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 6 Sep 2002 18:11:00 -0400
Subject: Dead Man Timer?
References: <fE6c9.145308$aA.31724@sccrnsc02> <pan.2002.08.26.17.48.00.759341@effectif.com> <cb035744.0209011247.81b42dc@posting.google.com> <3D728103.5060308@itamarst.org_NOSPAM> <FrOc9.110025$_91.144030@rwcrnsc51.ops.asp.att.net>
Message-ID: <eL9e9.464$l_4.49@atlpnn01.usenetserver.com>

"Neil Everhart" <neilhart at attbi.nospam.com> wrote in message
news:FrOc9.110025$_91.144030 at rwcrnsc51.ops.asp.att.net...
> Itamar,
>
> Thank you for the suggestion.  I will test the timeoutsocket.py code
> tomorrow
> on my target setup.  I have incorporated it into a version of my script
but
> can not really evaluated it until I am running in the terminal server
> environment.
>
> Neil.
>
> --
>
>
> "Itamar Shtull-Trauring" <itamar at itamarst.org_NOSPAM> wrote in message
> news:3D728103.5060308 at itamarst.org_NOSPAM...
> > 1. Use timeoutsocket.py - this lets you set timeouts on socket
operations
> > (find it on http://py.vaults.ca).
> >
> > 2. Use an event driven networking architecture - don't know of any that
> have
> > telnet client support though.
> >
> > Option 1 is your easiest solution.
> >

Note that similar functionality is currently planned for 2.3. Changes have
been checked in.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                        pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From eddie at holyrood.ed.ac.uk  Thu Sep 19 06:47:07 2002
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Thu, 19 Sep 2002 10:47:07 +0000 (UTC)
Subject: Strange Telnet error
References: <4378fa6f.0209181245.7b06124d@posting.google.com>
Message-ID: <amc9vb$8li$1@scotsman.ed.ac.uk>

mnations at airmail.net (Marc) writes:

>Hi,

>I have been running some code that all the sudden stopped working. I
>can not figure out what's happened. I am doing a simple telnet
>connection that sends and receives info. Now I am getting this error:


>Traceback (most recent call last):
>  File "C:\Python22\lib\threading.py", line 408, in __bootstrap
>    self.run()
>  File "C:\Python22\lib\threading.py", line 396, in run
>    apply(self.__target, self.__args, self.__kwargs)
>  File "main.py", line 85, in run
>    command[0](*command[1:3])
>  File "key.py", line 92, in command
>    conn.write(com)
>  File "C:\python22\lib\telnetlib.py", line 272, in write
>    self.sock.sendall(buffer)
>AttributeError: 'int' object has no attribute 'sendall'


Have you just loaded a new version of Python by any chance.  Line 272 in my
2.2 copy of telnetlib has 'send' instead of 'sendall', the latter seems to be
the better of the two so I guess it's a newer version.  Which makes me suspect
they've broken something else while working on it, in which case I'd report it
as a bug.  Unless of course, your code fiddles about with the underlying
sockets and things of the telnet object.

Eddie


From roy at panix.com  Thu Sep  5 15:18:51 2002
From: roy at panix.com (Roy Smith)
Date: 5 Sep 2002 15:18:51 -0400
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com> <4f49d232.0209050418.7668c21@posting.google.com> <al7ub8$crj$1@panix2.panix.com> <3D77FB95.3070300@nospam.free.fr>
Message-ID: <al8amr$d3q$1@panix2.panix.com>

I wrote:
>> If you're learning languages to study computer
>> science, lisp is looking a lot better.

laotseu  <bdesth at nospam.free.fr> wrote:
> I think once you know them both, you know *how* to program, whatever
> the language you have to use.

I believe we actually both said the same thing, just worded
differently :-)


From keilj_33 at yahoo.com  Tue Sep  3 16:25:20 2002
From: keilj_33 at yahoo.com (Jeffrey Keil)
Date: 3 Sep 2002 13:25:20 -0700
Subject: basic cgi question
Message-ID: <67f6bbe5.0209031225.56b1b7bd@posting.google.com>

Greetings:

I've used Perl in the past for cgi programs. I thought I'd play around
a little with Python. I've tried a few experiments.

I have programs that prints out to the browser. Unfortunately, it
prints out everything, including the "import" statements, "print",
etc. I'm not quite sure what I'm doing wrong.

Thanks,
Jeff Keil


From ark at research.att.com  Fri Sep 27 19:25:28 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 27 Sep 2002 23:25:28 GMT
Subject: Better terminal for windows?
References: <3D94B037.2020202@motorola.com>
Message-ID: <yu99u1kbuiaf.fsf@europa.research.att.com>

Stephen> I'm using python 2.2.1 on winnt.
Stephen> If I want to run python interactively (and perhaps using ipython), is
Stephen> there something better than the "cmd" terminal?

How about idle?

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From martin at v.loewis.de  Mon Sep 16 17:57:36 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Sep 2002 23:57:36 +0200
Subject: Mutable Stings
References: <mailman.1032181643.21256.python-list@python.org> <m3admidiys.fsf@mira.informatik.hu-berlin.de> <6qznuhx1je.fsf@thetis.intevation.de>
Message-ID: <m3u1kpegxb.fsf@mira.informatik.hu-berlin.de>

Bernhard Herzog <bh at intevation.de> writes:

> Not necessary. I submitted a bugreport when this came up the last time
> and it has been fixed in CVS since then.

You are right; I was using an old copy.

Regards,
Martin


From rsrchstr at hotmail.com  Sun Sep 29 17:26:05 2002
From: rsrchstr at hotmail.com (mike henley)
Date: 29 Sep 2002 14:26:05 -0700
Subject: idea for a much needed application
Message-ID: <576c1752.0209291326.56dcdf31@posting.google.com>

I remember a while ago reading about tim burners-lee and how his
vision of the WorldWideWeb as a medium where all can freely and easily
publish information has not materialized properly, and i felt it so
strongly over the past week when trying to look at options for hosting
a website. Hosts are either expensive or limit you in terms of what
languages, modules, applications, and OSs you can use. They also limit
you in terms of space, bandwidth used per period, number of executions
of CGIs... etc. Some of those who give good deals, seem difficult to
trust with either your credit card number or your content.

I wish i can just use my own computer, with whatever technologies I
choose, keeping my content that i build up over time safe here on my
own machine, but unless you have a broadband connection this is not
possible. It may be possible using a dynamic DNS provider, but then
again either they add cost, end up with an ugly URL, and, most
importantly, there's no way for others to know when you get online. So
here comes the idea that occured to me, and i post it here to ask if
you know of an application that already provides this functionality,
then please tell me what it is, or if you think it is worth doing on
an open source basis, or even proprietary, in which case i invite
anyone to do so, or if you think it's a completely stupid and
impossible idea, then perhaps you can please tell me why.

You probably guessed if you read as far as the second paragraph that
what i'm thinking about is an application functionality that enables
instant messaging to provide the dynamic DNS functionality. So instead
of giving others a URL I can just give them an IM username they can
add to their client, and when i go online, through my humble dial-up
connection that changes IP whenever i log in, they'll be aware of my
presence, and be able to connect to a server i host on my machine.

It sounds like a simple thing; IM presence, automatic IP query of my
username, click on my username and or use a contextual menu option or
whatever and a browser launches to go to the site i'm hosting on my
machine. Perhaps it can use existing open-source platforms, such as
jabber or gnutella, or as an add-on to existing open-source
applications/clients. Basically what i'm suggesting is a little tool
to enable a server such as apache to join the whole IM/p2p thing and
have "presence". The more automated it is, and the easier it is, the
better. I think such a functionality will be especially useful for
certain uses, in my case it will be hosting a wiki, as i leave my
machine online a lot through my dial-up, can't predict the number of
executions of CGI or the amount of space needed, wish to be free to
use whatever wiki application is best for my needs and in whatever
language regardless of what a host will support, and do not wish to
pay a monthly fee, extra charges, setup fee or a domain registration.
Perhaps there can be other areas where it is useful, and i'm sure
there are other areas where it will not be suitable. Perhaps there can
also be a simple way to enable a page hosted in such a way to be
submitted to search engines, i haven't thought about that yet, but
tell me if u see any.

What do you think? already done, worth doing, stupid, or impossible?
(p.s. please also let me know if there are other newsgroup to which
posting this will be a good idea)


From aleax at aleax.it  Wed Sep 25 05:21:58 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 09:21:58 GMT
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com> <3D8E6353.7030303@something.invalid> <96zj9.150553$pX1.5399868@news2.tin.it> <3D911CFA.4060405@something.invalid>
Message-ID: <Wefk9.163287$pX1.5841874@news2.tin.it>

Greg Ewing wrote:
        ...
>>>Using *tabs* only is the only way to get truly
>>>preference-independent indentation.
>>
>> ...unless other people use tools that do NOT let them change the
>> ways tabs are displayed, and that way is intolerably incompatible
>> with your use of tabs.
> 
> I agree there are good reasons to prefer spaces-only
> for code that will be transmitted or displayed by
> dumb software.

To put it differently: use tabs only if you ARE willing to
bet a very substantial amount that the source file in question
will NEVER be transmitted, displayed, nor otherwise processed
in any way, shape, or form, by software you consider "dumb"
(take it up with KDE KNode's maintainers whether their news
reader is "dumb"... I like it, though it doesn't show tabs:-).


> I was only pointing out that, whatever arguments there
> may be in favour of spaces-only, "it doesn't force the
> reader to view it the way you do" is not one of them!

I agree with you that this is a fact.  Only very smart
software would be able to display varying amount of
leading spaces in programmable ways, and software that
smart would no doubt be able to do anything with tabs too.


Alex



From prema at prema.co.nz  Sat Sep 28 14:08:18 2002
From: prema at prema.co.nz (Mike)
Date: Sat, 28 Sep 2002 18:08:18 GMT
Subject: py2exe and a single exe file? -- Oops
References: <3D792351.3070008@motorola.com>
Message-ID: <leml9.253$zz3.1979@newsfeed.intelenet.net>

This time !!!
:-)

from distutils.core import setup
import sys, os, py2exe

name = sys.argv[1]
sys.argv[1] = 'py2exe'
sys.path.append(os.path.dirname(os.path.abspath(name)))

setup(name=name[:-3], scripts=[name])



Stephen Boulet wrote:
> I'm trying to get py2exe to work for me, and I must be doing something 
> wrong.
> 
> I have a script that uses easygui (which uses Tkinter) and Numeric. It 
> also imports a small module from the same directory that the script 
> resides in.
> 
> I try building it like this:
> 
>    python setup.py py2exe -d N:\ -i easygui
> 
> It builds and runs, but only with the other stuff in the build directory 
> with the generaged exe. The other stuff is:
> 
>   tcl directory containing tcl8.3 and tk8.3 directories
>   _numpy.pyd
>   _sre.pyd
>   _tkinter.pyd
>   multiarray.pyd
>   umath.pyd
> 
> What I would like to accomplish is just having a single exe file lumping 
> everything together. Is this possible?
> 
> -- Stephen
> 



From jb at cascade-sys.com  Fri Sep 27 05:31:41 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Fri, 27 Sep 2002 02:31:41 -0700
Subject: Getting started
References: <3D911306.7080609@ucsd.edu> <mailman.1032982114.4627.python-list@python.org> <e15b6fa3.0209261128.4edcdfe8@posting.google.com>
Message-ID: <3D94257D.2020501@cascade-sys.com>

Adam Taylor wrote:

>This sounds like the same thing as I said, or damn close.  Am I
>missing something?
>

Not really.  Just different emphasis.

>Yes.  It seems to me that Table 1 in Cardelli:
>
>          Typed   Untyped
>Safe      ML      LISP
>Unsafe    C       Assembler
>
>Might be better like this:
>
>          Statically typed   Dynamically typed    Untyped
>Safe      ML                 LISP
>Unsafe    C                                       Assembler
>
>(Python would also be in the (Safe,Dynamically typed) position.)
>
>Although safety and strongly-checked are not the same thing, this
>table is also correct:
>
>                      Statically typed   Dynamically typed   Untyped
>Strongly checked      ML                 LISP
>Weakly checked        C                                      Assembler
>

I agree.  So where in this spacewould YOU put Python and Perl?

>>A literal reading furthermore defines "weakly typed" only in 
>>terms of static checking.
>>    
>>
>
>I don't understand what you mean by this.  Cardelli never uses the
>term "weakly typed" himself, he only uses it that one time, to say
>that it's the term most people use for the concept that he prefers to
>call "weakly checked".
>

I was translating his internal vocabulary to use the term we started 
with.  He defines "weakly checked" only in terms of "static checking". 
 Since they're synonyms, this reference only supports a def of "weakly 
typed" in terms of static checked languages.  So I'm saying the 
reference does not help us nearly as much as I would like.

>Or at
>least when people start talking about "strongly typed" vs "weakly
>typed" languages, that first thing to get straight is which axis
>they're talking about.
>
Fair enough.  

--jb

-- 

James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From senux at senux.com  Thu Sep 26 06:28:23 2002
From: senux at senux.com (Brian Lee)
Date: Thu, 26 Sep 2002 19:28:23 +0900
Subject: CGI server by Python
Message-ID: <20020926102823.GA17893@mercury.senux.com>

Hi,

I tested CGIHTTPServer.py in Python 2.2 and found that cgi files in 
sub directory does not run. Does anybody know about this?

/cgi-bin/test.cgi works.
/cgi-bin/a/test.cgi does NOT works.

$ python CGIHTTPServer.py
Serving HTTP on 0.0.0.0 port 9002 ...
avante.xxx.co.kr - - [26/Sep/2002 19:20:24] code 403, message CGI script is not a plain file ('/cgi-bin/a')
avante.xxx.co.kr - - [26/Sep/2002 19:20:24] "GET /cgi-bin/a/test.cgi HTTP/1.1" 403 -
avante.xxx.co.kr - - [26/Sep/2002 19:20:33] "GET /cgi-bin/test.cgi HTTP/1.1" 200 -

/cgi-bin/test.cgi and /cgi-bin/a/test.cgi is same file.

-- 
Brian Lee <senux at senux.com>



From marklists at mceahern.com  Fri Sep 27 18:02:06 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 27 Sep 2002 17:02:06 -0500
Subject: watching mutables?
In-Reply-To: <an2j43$8fg$2@peabody.colorado.edu>
Message-ID: <JHEOKEOOLIGLDHCMAHMOEECOECAA.marklists@mceahern.com>

Someone (I think Skip?) mentioned the possibility of using metaclasses.  The
OP has asked for a method that doesn't require modifying the original code.
This doesn't satisfy that request, but is interesting nonetheless (imho):

#!/usr/bin/env python

class Watched(type):

    def __new__(cls, cls_name, bases, classdict):
        # print "__new__(%s, %s, %s, %s)" % (cls, cls_name, bases,
classdict)
        def w_setattr(self, name, value):
            print "%s.__setattr__(%s, %s)" % (cls_name, name, value)
            bases[0].__setattr__(self, name, value)

        def w_setitem(self, i, el):
            print "%s.__setitem__(%s, %s)" % (cls_name, i, el)
            bases[0].__setitem__(self, i, el)

        def w_getitem(self, i):
            print "%s.__getitem__(%s)" % (cls_name, i)
            return bases[0].__getitem__(self, i)

        def w_getattribute(self, name):
            print "%s.__getattribute__(%s)" % (cls_name, name)
            return bases[0].__getattribute__(self, name)

        classdict['__setattr__'] = w_setattr
        classdict['__setitem__'] = w_setitem
        classdict['__getitem__'] = w_getitem
        classdict['__getattribute__'] = w_getattribute

        return type.__new__(cls, cls_name, bases, classdict)

class WatchedDict(dict):

    __metaclass__ = Watched

class WatchedList(list):

    __metaclass__ = Watched

l = []
l.append(1)

l = WatchedList()
l.append(1)

d = WatchedDict()
d['a'] = 1

print d['a']

**
Output:

WatchedList.__getattribute__(append)
WatchedDict.__setitem__(a, 1)
WatchedDict.__getitem__(a)
1




From pierpaolo.glave at ciaolab.com  Wed Sep  4 08:55:22 2002
From: pierpaolo.glave at ciaolab.com (Pier Paolo Glave)
Date: 4 Sep 2002 05:55:22 -0700
Subject: Tkinter, updateStringProc for type font
Message-ID: <1907a5e4.0209040455.64ae73b1@posting.google.com>

Hi all,

I'm facing a problem with my Python program, which uses Tkinter to
interface Tcl/Tk.
I sometimes get this error message from Tcl: 
"UpdateStringProc should not be invoked for type font".
As a consequence, the application crashes with SIGABRT (6).

I'm using: Python 2.2, Tcl/Tk 8.3.3-69.

I've seen in the tcl sources that this happens when a function named 
Tcl_GetString() is called for a "type" which has the "byte" field set
to NULL, and no "updateStringProc" function pointer: this seems to
happen for Tk type "font", sometimes in my application.
Anyway, I'm not able to understand more.

Has anyone encountered this problem before?

Thank you
--
Pier Paolo Glave


From padraig at linux.ie  Sun Sep 15 16:12:33 2002
From: padraig at linux.ie (Padraig Brady)
Date: Sun, 15 Sep 2002 20:12:33 GMT
Subject: representing a literal single slash
References: <dd5454fb.0209141828.1ad24260@posting.google.com>
Message-ID: <3D84E9F9.8090101@linux.ie>

Jonathan Claggett wrote:
> Here is an easy question (I'm hoping).
> 
> I'm trying to process a string so that all doubled up backslashes (\\)
> are replaced with a single backslash (\).
> 
> The obvious approach doesn't work since a single backslash can't end a
> string:
> 
> str = str.replace('\\', '\') # broken
> 
> So, how does one represent a literal single slash? unicode or hex
> values perhaps?

mystr = mystr.replace(r'\\', r'\')

P?draig.



From joe.sixpack at quartercase.com  Wed Sep 18 16:41:25 2002
From: joe.sixpack at quartercase.com (joe sixpack)
Date: Wed, 18 Sep 2002 20:41:25 GMT
Subject: A list comprehension to do this?...
Message-ID: <3D88E4F5.B443918@quartercase.com>

Is there a list comprehension or other 'cleaner' or 'prettier' way 
to do the following:

>>> xx = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10)
>>> ww = []
>>> for i in range(0, len(xx), 2):
	ww.append((xx[i+1],xx[i]))
	
>>> ww
[(2, 1), (4, 3), (6, 5), (8, 7), (10, 9)]
>>>

Thanks in advance...


From tundra at tundraware.com  Sun Sep  1 16:40:03 2002
From: tundra at tundraware.com (Tim Daneliuk)
Date: Sun, 01 Sep 2002 20:40:03 GMT
Subject: [ANN]: 'waccess' 1.61 Released
Message-ID: <nqttka.sf92.ln@boundary.tundraware.com>

I have just released version 1.61 of 'waccess':

      http://www.tundraware.com/Software/waccess/

'waccess' is a Quick And Dirty web access log analysis tool.  It's
useful if you want to see how many hits a particular file has seen on
your web site.  It will also attempt to get the names of the accessing
hosts by doing reverse lookups on their IP addresses.

'waccess' is not a replacement for a serious web log analysis tool,
but it is quite handy nonetheless.

'waccess' is written in Python and needs a reasonably current version
of the langugage on your machine.

'waccess' was written for and tested on FreeBSD 4.6R, but should port
to other systems pretty trivially.

This version adds two features:

          - User can now name the log to examine on the command line
            instead of having to edit the string constant in the source.

          - The program now supports 'ignoring' IP addresses.  You can
            designate which IP address spaces you do not care about and
            'waccess' will ignore all records from such addresses.

Changes to the FreeBSD port have been committed as well.
-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com



From 102030405 at gmx.net  Sat Sep 28 01:54:05 2002
From: 102030405 at gmx.net (Ralf Claus)
Date: Sat, 28 Sep 2002 07:54:05 +0200
Subject: zlib needs the python15.dll on python22 ???
Message-ID: <an3g6c$j5t$00$1@news.t-online.com>

Hello NG,
by importing the zlib module on python 2.2 (windows)
i get a popup that means "python15.dll not found in path 
//winnt/system32/......and so on...

On that computer, there was never installed python 1.5 !?

What is wrong ?

Thanks a lot
Ralf



From jeff at sasmor.com  Mon Sep  9 10:11:20 2002
From: jeff at sasmor.com (Jeff Sasmor)
Date: Mon, 09 Sep 2002 14:11:20 GMT
Subject: Zope Status?
References: <Xns9281221A07167bosahvnetscapenet@66.185.95.104> <cSOe9.3024$io.1100@nwrddc03.gnilink.net> <3D7C4833.2080308@mxm.dk>
Message-ID: <c_1f9.2$on1.1@nwrddc01.gnilink.net>

Hi Max,
What I meant was that since the initial implementation won't
support the old (Zope 2) APIs, you'd have to have the mindset
to dig in and see what's going on.  The learning curve for
Zope 2 is made easier due to the many existing apps and
examples that exist.  Even so, it's a complex system and
is daunting for a lot of folks.

I expect that the ratio of folks who use Zope to those who
create apps in Python for Zope is pretty high.

Right now, almost anyone who wants to play around can install
Zope 2 and get it to do things by installing some popular
apps like Squishdot (etc,etc).  You only have to know
how to install software to be able to be productive.

Until this same state of affairs exists for Zope 3 (and there's
nothing to say that it will take that long, one hopes) it will take
a more advanced (tyro < 'advanced developer' < Einstein)
developer who has the time and inclination to
poke into the innards and see how it works.  Especially as the
docs will likely  trail the sw for a while.

>From what I've seen, Zope 3 is more regular and better organized
than Zope 2, but it _is_ different.

Anyway, that's all I meant. Nothing bad, just inevitable. Anyway,
it's just my opinion, so it doesn't matter all that much!

jms
--
#--------------------------------
Jeff Sasmor
jeff at sasmor.com
"Max M" <maxm at mxm.dk> wrote in message news:3D7C4833.2080308 at mxm.dk...
> Jeff Sasmor wrote:
>
>
> > That's just one thing. Also, Zope 3 is a clean-slate redesign; that is,
> > it is being rearchitected with no initial regard for existing Zope
> > applications.  I suppose opinions go either way on this, but overall
> > it is probably a Good Thing.  I recall reading that there will
eventually
> > be backwards compatibility, perhaps with some interface layer, but
> > I suspect that the initial releases will be for very advanced
developers.
>
>
> As far as I can see from the examples you actually don't need to be an
> advanced developer. It seems that it will indeed be much more simple to
> develop for Zope3 than it is for the current version.
>
> Also they are taking care of some of the things that could make it
> difficult to develop in Zope. Like relations.
>
>
> regards Max M
>




From peter at engcorp.com  Mon Sep 16 09:20:04 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 16 Sep 2002 09:20:04 -0400
Subject: "if x == None" raises "'NoneType' object is not callable"
References: <af84f376.0209160501.4d21500@posting.google.com>
Message-ID: <3d85da78$1@news.sentex.net>

Luc Saffre wrote:
> """
> If I define __getattr__() and __setattr__() for a class, then I cannot
> test instances of this class for equality with None.
> 
> the code example below raises
> TypeError: 'NoneType' object is not callable
> 
> class Row:
> 
>    def __init__(self):
>       self.__dict__["_values"] = {}
> 
>    def __getattr__(self,name):
>       try:
>          return self._values[name]
>       except KeyError,e:
>          AttributeError,str(e)
>    
>    def __setattr__(self,name,value):
>       self.__dict__["_values"][name] = value
> 
> 
> row = Row()
> if row == None: # here it happens.
>    print "row instance is None!"

When when you call "row == None" it calls __getattr__ with '__eq__'
to see if you want to provide a equality comparator function.
Inside __getattr__ you are not going to find such a beast,
so you go to the exception handler.  Unfortunately, you left
out the "raise" keyword, so you are in the end returning None
as all methods do in Python which don't explicitly return something
else.

-Peter



From sholden at holdenweb.com  Sun Sep 29 07:00:34 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 29 Sep 2002 07:00:34 -0400
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org>
Message-ID: <E4Bl9.5382$pD4.397@fe05>

<sismex01 at hebmex.com> wrote...
> >
[ ... ]
OK, so don't use whole pennies (or smallest currency) as your
base, use 10,000ths of a penny, but still use integer arithmetic
for your calculations.

Floating point has too many rounding errors to be trustworthy;
or, looking at it from another angle, floating-point calculations
cannot be 100% accurate because of the way it's implemented.

"...has too many errors to be trustworthy..." is clearly an overstatement
given the huge number of computations each day which use floating-point
without any apparent problem. Floating-point got us to the moon and back,
surely it can handle a few money computations.

The real problem is that you have to grok the nature of floating-point to
use it well and appropriately. This is not a simple task (and you only have
to look at the timbot to realise that it can have disastrous consequences
;-). A typical modern double (64-bit floating-point number) will give you an
effective 53 bits of fraction, which equates to over 14 significant figures.
Since you always need two digits after the decimal point you can work with
amounts up to 999,999,999,999.99 without loss of accuracy as long as you
know what you are doing. Appropriate rounding before storage will avoid
accumulation of rounding errors.

Of course, none of this is necessarily an argument that f-p is the *best*
representation for money, but I'm always amused to see people arguing
against it on general grounds when it's perfectly acceptable for many such
purposes *when correctly used*.


regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From peter at engcorp.com  Wed Sep 18 09:52:33 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 18 Sep 2002 09:52:33 -0400
Subject: try except ImportError failing to catch error!
References: <mailman.1032341542.17022.python-list@python.org>
Message-ID: <3d888514$1@news.sentex.net>

Ben Gerblich wrote:
> I am importing all module files in the CWD. This part works fine.
> 
> I want to add the functionality to catch the error if there is a
> typo in the imported module file. The following does not catch
> the error!
> 
> try:
>    self.userDefs[file[:-3]]['class'] = __import__(file[:-3])
> except ImportError, e:
>    print 'cant do it', e
> 
> I still get the runtime error:
>[...] 
> TypeError: line() takes exactly 5 arguments (6 given)

That's not an import error, so your exception handler won't
catch it.  If you want to catch *any* error in the imported
module, you'll have to do "except Exception, e" instead.
Unfortunately, then you'll also have ImportErrors in the mix,
so you won't know if the module could be found but had its
own error, or if the call to __import__ failed...

Also, you won't know whether perhaps the file[:-3] expression
failed, or any of the other stuff that might fail in that
import line.

> And is there a way for my handler routine to return the line-number of 
> the imported module file that caused the error?

Look into the traceback standard module.

I don't understand why you want to intercept these exceptions
anyway.  If you want to catch the error, *and* the line number,
and print it, why don't you want Python to show you where
the problem came from?  Maybe there's a better approach than
what you've selected so far...

-Peter



From Hobbes2176 at yahoo.com  Thu Sep 19 15:46:55 2002
From: Hobbes2176 at yahoo.com (Mark)
Date: Thu, 19 Sep 2002 19:46:55 GMT
Subject: Minimal debug/rep functionality
References: <J%ai9.893$1D.574@nwrddc01.gnilink.net> <3d894106@news.sentex.net>
Message-ID: <PQpi9.16355$t%6.7733@nwrddc02.gnilink.net>

Peter Hansen wrote:
> 
> Sorry, I couldn't find a solution that was larger than six lines:
> 
>  >>> def trace(frame, event, arg):
> ...   if event == 'line':
> ...      print '%s: %s' % (frame.f_code.co_filename, frame.f_lineno)
> ...   return trace
> ...
>  >>> import sys
>  >>> sys.settrace(trace)
> 
> -Peter

Hey, 

This is exactly what I needed.  Now, can you explain to me how I hook in a 
"local" trace function to take over once I execute the function I'm 
interested in tracing?  I'd like to use the "event" argument to take note 
of calls that are 1 level deep (ie in my foo function, any calls made) and 
turn off tracing of these until they return back to the level of foo.

I presume I could implement this with a "if event == 'call':" condition, but 
I can't seem to get a global trace function to hook up to a local trace 
function ... it does one or two things right, then goes crazy.

Regards,
Mark


From mmoales at fluent.com  Thu Sep  5 10:40:02 2002
From: mmoales at fluent.com (Mark Moales)
Date: Thu, 05 Sep 2002 10:40:02 -0400
Subject: Privacy and Inheritance
References: <mailman.1031112136.4594.python-list@python.org>
Message-ID: <3D776CC2.CA928E38@fluent.com>

Dennis,

I think the point of the double underscores is to create "private"
attributes similar to C++'s or Java's private keyword.  In other words,
only the class that defines the attribute can access it.  Subclasses and
others can't.  We use a single underscore here to denote "protected"
attributes, or attributes that are accessible by subclasses.  However,
the only way to enforce this in Python is by having well behaved
programmers ;-)  Here's an example:

class BaseClass:
    def __init__(self):
        self.publicVar = 'Foo'
        self._protectedVar = 'Bar'
        self.__privateVar = 'Spam'

class SubClass(BaseClass):
    def aMethod(self):
        print self.publicVar
        print self._protectedVar
        # Can't do this because it's private
        # print self.__privateVar

Mark

"Dennis B." wrote:
> 
> Greetings;
> 
> Have enough confidence in Python to realize this is probably just my
> misunderstanding, but wondering why Python mangles private, double
> underscore attributes with the name of the base class even if they're
> inherited and the instance is that of the sub class.
> 
> Like the following doesn't even seem to work:
> 
> class StandardNormal:
> 
>     def letEventZScore(self, eventZScore):
>         self.__eventZScore = float(eventZScore)
> 
> class DerivativeNormal(StandardNormal):
> 
>     def getCentral(self):
>         if self.__eventZScore < 0:
> 
> Unless I change that last line to something like:
> 
>         if self._StandardNormal__eventZScore < 0:
> 
> Maybe just missing the point all together, but wondering if anyone'll
> point out anything a little more elegant and/or intuitive.
> 
> Thank you and have a great day,
-------------- next part --------------
A non-text attachment was scrubbed...
Name: mmoales.vcf
Type: text/x-vcard
Size: 281 bytes
Desc: Card for Mark Moales
URL: <http://mail.python.org/pipermail/python-list/attachments/20020905/5cd59866/attachment.vcf>

From marklists at mceahern.com  Tue Sep 10 15:42:48 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 10 Sep 2002 14:42:48 -0500
Subject: Reset a generator function?
In-Reply-To: <%Lrf9.30107$w51.7961593@news2.news.adelphia.net>
Message-ID: <JHEOKEOOLIGLDHCMAHMOOELNDNAA.marklists@mceahern.com>

[Robert Oschler]
> How can I reset a generator function so that it begins yielding
> values as it did when first called?

My braindead way of doing it is just to create another instance of the
generator:

$ python
Python 2.2.1 (#1, Jun 25 2002, 10:55:46)
[GCC 2.95.3-5 (cygwin special)] on cygwin
Type "help", "copyright", "credits" or "license" for more information.
>>> from __future__ import generators
>>> def mygen():
...     i = 0
...     while True:
...             i += 1
...             yield i
...
>>> x = mygen()
>>> x.next()
1
>>> x.next()
2
>>> x.next()
3
>>> x = mygen()
>>> x.next()
1
>>>

// m

-




From pyth at devel.trillke.net  Mon Sep 30 15:53:27 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 30 Sep 2002 21:53:27 +0200
Subject: Question about __getattribute__/__getattr__
In-Reply-To: <Pine.LNX.4.44.0209301503100.12980-100000@mercury.sabren.com>; from jpm@papercut.org on Mon, Sep 30, 2002 at 03:04:24PM -0400
References: <JHEOKEOOLIGLDHCMAHMOIEJDECAA.marklists@mceahern.com> <Pine.LNX.4.44.0209301503100.12980-100000@mercury.sabren.com>
Message-ID: <20020930215327.X30315@prim.han.de>

Joao Prado Maia wrote:
> On Mon, 30 Sep 2002, Mark McEahern wrote:
> 
> > >         if cached_value == None:
> > 
> > This is an aside, but comparison to None should be via identity rather than
> > equality; e.g.,
> > 
> >   if x is None:
> > 
> > or:
> > 
> >   if x is not None:
> > 
> 
> Yeah, I always forget that when switching from PHP to Python. Anyway, my 
> main question was how to get the list of arguments passed to a 
> method/function. Any ideas ? :)

Probably you are interested in the Memoize recipe at 

    http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52201

regards,

    holger



From Padraig at Linux.ie  Thu Sep 12 05:42:14 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Thu, 12 Sep 2002 09:42:14 GMT
Subject: Anyone got a logfile script? - Help a n00b
References: <mailman.1031821818.15125.python-list@python.org>
Message-ID: <3D806151.9020707@Linux.ie>

Daniel Stuart Kelly wrote:
> What I'm looking for is a little script when run will dig into a log file and 
> look for a variable.
> 
> When it finds the variable I want it to print that line and the 20 or so right 
> above it to either screen or a plaintext file. 
> 
> I know it sounds simple, but I've only started learning this and I've no idea 
> where to begin.. Any tips or pointers would also be helpful...
> 
> Thanks
> /Daniel
> 
would a "context grep" not be simpler?

grep --before-context=20 log.file

P?draig.



From hst at empolis.co.uk  Tue Sep 10 08:44:57 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Tue, 10 Sep 2002 13:44:57 +0100
Subject: Pickle problem
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220D10@hendrix.empolisuk.com>

Erik Price wrote:
> 
> On Tuesday, September 10, 2002, at 06:13  AM, Harvey Thomas wrote:
> 
> > I'm trying to pickle a complicated data structure which includes a 
> > dictionary whose entries are created by (simplified):
> >
> > a[x] = re.compile(y).match
> >
> > The pickle fails, however, with the error:
> >
> > PicklingError: Can't pickle <built-in method match of 
> _sre.SRE_Pattern 
> > object at
> >  0x00838700>: it's not found as __main__.match
> >
> > I'd be grateful if someone could tell me if it is possible 
> to pickle 
> > the data structure and what I'm doing wrongly.
> 
> Are you trying to store the MatchObject returned by a call to 
> a regular 
> expression object's match() method?  If so, it looks like you're 
> forgetting the parens at the end of "re.compile(y).match".
> 
> 
> 
> Erik
> 
No, I haven't forgotten the parentheses. I'd like to pickle the match method of the compiled regular expression.

Harvey

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.



From dougz at cs.washington.edu  Wed Sep  4 17:24:54 2002
From: dougz at cs.washington.edu (Douglas Zongker)
Date: Wed, 4 Sep 2002 21:24:54 GMT
Subject: str() of a tuple
Message-ID: <H1xnHJ.1zE1@beaver.cs.washington.edu>

I'm sure this is a FAQ, but I haven't found a good explanation for one
of Python's most irritating behaviors:

   >>> a = -0.4
   >>> repr(a)
   '-0.40000000000000002'
   >>> str(a)
   '-0.4'

So far, so good.
   
   >>> b = (a,)
   >>> repr(b)
   '(-0.40000000000000002,)'
   >>> str(b)                      # argggh.
   '(-0.40000000000000002,)'     

Why, oh why does the str() of a container use the repr()s of the
objects inside?  This seems to violate what the docs claim str() does
-- "return a nicely printable representation of an object."  Would a
sane implementation break something?

Is there a convenient way to print out a tuple of floats so that the
output is actually readable?

thanks,
dz


From steve at ferg.org  Wed Sep 11 10:41:29 2002
From: steve at ferg.org (Stephen Ferg)
Date: 11 Sep 2002 07:41:29 -0700
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid> <alln71$1s2s$1@nntp6.u.washington.edu>
Message-ID: <b16e4ef7.0209110641.256799a0@posting.google.com>

> I don't 
> know of any event binding that does a proper job of emulating the 
> standard behavior of pressing a button. Button(command=...) 
> is the way to go for this.

I think I found out why.  At 
http://www.tcl.tk/man/tcl8.4/TkCmd/button.htm#M6
 -- under "Button>Default Bindings" -- I found this :

   "[3] If mouse button 1 is pressed over a button and later 
   released over the button, the button is invoked. However, 
   if the mouse is not over the button when button 1 is 
   released, then no invocation occurs."

I take it that this default binding is what "command" is 
attached to.

So no combination of "bind()" statements (well, no *easily coded*
combination of bind() and callback logic) can emulate this behavior.


From cnetzer at mail.arc.nasa.gov  Thu Sep 26 19:13:08 2002
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Thu, 26 Sep 2002 16:13:08 -0700
Subject: Who knows somefunction?
In-Reply-To: <00d701c265a5$053d0710$df7e4e18@brianspiv1700>
References: <00d701c265a5$053d0710$df7e4e18@brianspiv1700>
Message-ID: <200209262313.QAA18247@mail.arc.nasa.gov>

On Thursday 26 September 2002 14:38, Brian Quinlan wrote:
> Nope. You might want to consider creating your own dictionary to map
> ids to the objects that you are interested in.

That got me thinking about whether there is an easy (and robust) way to 
built a dictionary of all the object-id:object pairs in the current 
environment.

The following is some dumb code to start with.  Does anyone know of a 
(non-ridiculous) way to improve it, out of curiosity?

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
def add_ids_to_dict( l, d ):
    all_previously_added = 1

    # Add all objects to dict, unless already there
    for o in l:
        if not d.has_key( id( o ) ):
            d[ id( o ) ] = o

            # Adding a new object means they weren't already there
            all_previously_added = 0

    # Give up if all objects in list were already in dict, otherwise
    # recurse on all objects in the list
    if not all_previously_added:
        for o in l:
            if hasattr( o, '__dict__' ):
                add_ids_to_dict( o.__dict__.keys(), d )
                add_ids_to_dict( o.__dict__.values(), d )
            else:
                add_ids_to_dict( dir( o ), d )

    return

if __name__ == '__main__':
    objects_dict = {}
    add_ids_to_dict( globals(), objects_dict )

    print objects_dict
    print len( objects_dict.keys() )
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Chad Netzer
cnetzer at mail.arc.nasa.gov



From peter at engcorp.com  Tue Sep  3 08:11:49 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 03 Sep 2002 08:11:49 -0400
Subject: RedHat 7.3 and Python2.2.1
References: <mailman.1031020923.22683.python-list@python.org>
Message-ID: <3d74a6e8@news.sentex.net>

Kow Kuroda wrote:
> 
> As a last thought, doesn't Red Hat have different versions of the 7.3 
> distribution?

Maybe you installed a beta?  I'm pretty sure there's only one release
version to date.  I could be wrong.  The one I installed I took from
their web site just two weeks ago.  uname says it's version:

   2.4.18-3 #1 Thu Apr 18 07:37:53 EDT 2002 i686

and /etc/redhat-release says "release 7.3 (Valhalla)"

Oh well, there's no telling with software. :)

-Peter



From kennethg at pd.jaring.my  Fri Sep 20 12:10:28 2002
From: kennethg at pd.jaring.my (Kenneth Gomez)
Date: Fri, 20 Sep 2002 16:10:28 GMT
Subject: Local Python Server Installation
Message-ID: <3d8b47c6.277669@news.jaring.my>

Dear Group,

I would like to know wich software to install if I would like to learn
to use Python as a web development tool? 

I have installed Python 2.2.1 and Apache but I belive I would need a
Python Server Program to enable Apache to render .py files, am I
right?

Could you kindly advise me? 

I am running Windows Me.
Thank you.

Regards,
Kenneth.
Cheers,
Kenneth Gomez.


From loewis at informatik.hu-berlin.de  Sun Sep 15 14:00:27 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 15 Sep 2002 20:00:27 +0200
Subject: file mode attributes...
References: <mailman.1031961917.2057.python-list@python.org>
Message-ID: <j4fzwbrv44.fsf@informatik.hu-berlin.de>

"Bjorn Pettersen" <BPettersen at NAREX.com> writes:

> I was wondering if this was the best way of implmenting these functions
> (windows)?

Yes.

> It seems rather round-about, especially since the stat attributes I'm
> using aren't even mentioned in the manual...

On Unix, you'd normally add S_IRGRP|S_IROTH as well. Permissions, on
Unix, are really a bitmap thing, so manipulating them takes place most
naturally using bitmap operations.

Notice that, on WNT/2k/XP, you might also need to modify the access
control list to give users permissions, especially if the files are on
NTFS. Access and modification of the ACL is not supported in standard
Python.

Regards,
Martin


From franke at ableton.com  Mon Sep  2 11:30:48 2002
From: franke at ableton.com (Stefan Franke)
Date: Mon, 02 Sep 2002 17:30:48 +0200
Subject: How to parse over a Python expr?
References: <8sjvmugochmuvo4fnfaelq4m59plj6lg0r@4ax.com> <3D6FD585.111E32AE@alcyone.com> <s7tvmukn21cpi99ecksvr2up7u4aurm130@4ax.com> <3D7002C4.22DD2AB3@alcyone.com>
Message-ID: <hn07nucub4elvt0dkauff7ee2jh8n8l2pa@4ax.com>

On Fri, 30 Aug 2002 16:41:56 -0700, Erik Max Francis <max at alcyone.com> wrote:

>Hmm, you're right.  I didn't take into account the corner case where a
>triple-quoted string contains single-quoted strings.  Thanks for
>catching that.  (Fortunately, it's a rare occurrence that has certainly
>never happened in the field -- you'd have to do a triple-quoted string,
>then a single quoted one, _then_ an unbalanced close parenthesis.)

Are you sure? I think a TQS with a single quote inside would be
sufficient to miss the right TQS end:

'''That's all'''

These a 7 single quotes. Parsed as single-quote strings, the last one 
opens a string until EOF.



From martin at v.loewis.de  Thu Sep 12 02:51:57 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 12 Sep 2002 08:51:57 +0200
Subject: how can i make a binary executable program with a python2's script
References: <dd083388.0209111822.7598320d@posting.google.com>
Message-ID: <m3ptvj66pe.fsf@mira.informatik.hu-berlin.de>

jingguoming at yahoo.com.cn (jgm2163) writes:

> how can i make a binary executable program with a python2's script

The easiest way is to put 

#! /usr/bin/env python2

into the first line of script.py, then do

chmod +x script.py

HTH,
Martin


From loewis at informatik.hu-berlin.de  Fri Sep 20 16:30:14 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 20 Sep 2002 22:30:14 +0200
Subject: Python Compiling
References: <amfveg$5g64k$1@ID-53055.news.dfncis.de>
Message-ID: <j4lm5w9zft.fsf@informatik.hu-berlin.de>

"Terje Johan Abrahamsen" <spoermeg at voldelig.com> writes:

> Could someone write a compiler and get Python to be a compiled
> language like C?

Yes; there have been a few demonstrations that this is possible.

> Is it not possible to write a complete compiler like the C compiler for
> Python? Or is it just that nobody has done it? 

It has been done, see John Aycock's IPC7 paper (Converting Python
Virtual Machine Code to C) for an example, and Jonathan Riehl's IPC7
paper (PyFront: Conversion of Python to C Extension Modules) for
another.

> Or must the language be constructed differently?

To merely produce a complete "machine-code" binary, no, with
Johnathan's technique. 

The question is: why do you want to have a compiler? Using it is much
more difficult than an interpreter, and more time-consuming during
development of an application. People often expect significant speed
increases from compilation, but it turns out that those don't
materialize. 

The reason for that is that any compiler would need to maintain the
semantics of the Python language, which means that certain
optimizations that a C compiler does for C code are not possible in
Python, since, at compile time, it is uncertain whether the
prerequisites for that optimization are met.

HTH,
Martin


From jim at interet.com  Mon Sep 30 09:31:27 2002
From: jim at interet.com (James C. Ahlstrom)
Date: Mon, 30 Sep 2002 09:31:27 -0400
Subject: zlib needs the python15.dll on python22 ???
References: <an3g6c$j5t$00$1@news.t-online.com>
Message-ID: <3D98522F.8080508@interet.com>

Ralf Claus wrote:
> Hello NG,
> by importing the zlib module on python 2.2 (windows)
> i get a popup that means "python15.dll not found in path 
> //winnt/system32/......and so on...
> 
> On that computer, there was never installed python 1.5 !?

Each Python release has a new version of the Python
interpreter in the PythonXX.dll.  Your Python 2.2 is using
Python22.dll, but you have an old version of zlib which
was linked using Python15.dll.

You need to find the old zlib and remove it.  Then
find the new zlib.pyd that uses Python22.dll and install it.

Jim Ahlstrom



From Padraig at Linux.ie  Fri Sep  6 07:59:33 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Fri, 06 Sep 2002 11:59:33 GMT
Subject: tree data structure
Message-ID: <FM0e9.2860$cP3.6165@news.iol.ie>

Hi,

I've a graph like stucture represented
by a list of tuples:

[(L0,S2), (S2,S0), (S2,S1), (S1,Q2), (S0,Q1), (S2,S3)]

And would like to convert it to
a tree structure like:

           S0 - Q1
         /
L0 - S2 - S3
         \
           S1 - Q2

Any tips appreciated.

Thanks,
P?draig.



From mgerrans at mindspring.com  Thu Sep 19 02:06:12 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Wed, 18 Sep 2002 23:06:12 -0700
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <bd751532.0209082142.54236307@posting.google.com> <allrrs$6hm$2@solaris.cc.vt.edu>
Message-ID: <ambpgq$dre$1@nntp9.atl.mindspring.net>

I think "the pound-end operator" is the technical term (I'll have to defer
to Guido on that).

One of the niftiest things about Python is you never have to argue about
whether you should put your curlies in the wrong place, like this:

void foo() {
   //...
}

...or whether you should place them correctly, like this:

void foo()
{
    //...
}

It saves the world from a lot of unnecessary creation and release of hot air
(so Python has probably done more to help reduce the greenhouse effect than
any other software technology!).

;-)




From jwbaxter at spamcop.net  Thu Sep 26 20:25:46 2002
From: jwbaxter at spamcop.net (John Baxter)
Date: Thu, 26 Sep 2002 17:25:46 -0700
Subject: How do you pronounce "__" (double underscore)?
References: <525eaba.0209260956.72bf6786@posting.google.com>
Message-ID: <jwbaxter-4CCF41.17254626092002@corp.supernews.com>

In article <525eaba.0209260956.72bf6786 at posting.google.com>,
 pknotz at sandia.gov (Pat Notz) wrote:

> Is there a short (one or two syllable) pronounciation for "__" (double
> underscore).

Unfortunately, we can never have the definitive answer, since Victor 
Borge is no longer with us.

(His answer would have been unspellable, of course.)

   --John


From dsavitsk at e-coli.net  Tue Sep  3 12:59:12 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Tue, 03 Sep 2002 16:59:12 GMT
Subject: PIDDLE install problems
Message-ID: <AT5d9.6490$yt3.3042205@newssrv26.news.prodigy.com>

I am trying to install PIDDLE (Win2k, 2.1.3 ).  I downloaded the recent .zip
file, unzipped it to C:\piddle-1.0.15, cd'd to that directory, typed "python
setup.py install" and received the following error.

...
copying src\piddle\python.gif -> C:\Python\piddle
Traceback (most recent call last):
  File "setup.py", line 28, in ?
    packages = ['piddle', 'piddle.piddleGTK', 'piddle.piddleSVG',
'piddle.piddle
TK2'] )
  File "C:\Python\lib\distutils\core.py", line 138, in setup
    dist.run_commands()
  File "C:\Python\lib\distutils\dist.py", line 899, in run_commands
    self.run_command(cmd)
  File "C:\Python\lib\distutils\dist.py", line 919, in run_command
    cmd_obj.run()
  File "C:\Python\lib\distutils\command\install.py", line 484, in run
    self.run_command(cmd_name)
  File "C:\Python\lib\distutils\cmd.py", line 328, in run_command
    self.distribution.run_command(command)
  File "C:\Python\lib\distutils\dist.py", line 919, in run_command
    cmd_obj.run()
  File "C:\Python\lib\distutils\command\install_data.py", line 60, in run
    dir = convert_path(f[0])
  File "C:\Python\lib\distutils\util.py", line 86, in convert_path
    if pathname[0] == '/':
IndexError: string index out of range

is this unusual (the same happened w/ 1.0.14), and does anyone know a way to
get around this, or why it is happening?

thanks,

-doug




From gerhard.haering at opus-gmbh.net  Thu Sep  5 05:51:51 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 5 Sep 2002 09:51:51 GMT
Subject: linking embedded python
References: <3D7719A9.8050804@ntlworld.nospam.com>
Message-ID: <slrnanea6i.27o.gerhard.haering@haering.opus-gmbh.net>

Keith S. wrote:
> Hi,
> 
> This might be a dumb question, but how does one link
> a C++ application in Unix with embedded python?
> [...] However, when I try and link, I get errors like:
> 
> Undefined                       first referenced
>   symbol                             in file
> dlsym 

You need to link against additional libraries. To find out which ones,
doing a ldd (or whatever the SunOS equivalent of it is) on the python
executable should tell you.

Does this work for you, too?:

    from distutils.sysconfig import get_config_var
    print get_config_var("LIBS") + " " + get_config_var("SYSLIBS")

It seems to work for some people, but not for everybody.

-- Gerhard


From lpgray at uop.com  Tue Sep 17 17:05:54 2002
From: lpgray at uop.com (Lee Gray)
Date: 17 Sep 2002 14:05:54 -0700
Subject: Python GUI app to impress the boss?
Message-ID: <96904b50.0209171305.6d65f574@posting.google.com>

I'm a fairly new programmer with experience in ASP/VBScript/DHTML.
Other than small utilities, so far I've mostly written ASP apps to
talk to an Oracle database on HP/UX, with Windows NT/2000 clients, and
I've done maintenance work on VB and Oracle Forms apps.

We're currently looking at .Net, and I'm not very impressed, for a
variety of reasons. A little research makes me feel that either Java
or Python would be a good alternative, and given my level of
experience, I'd much prefer Python over Java.  I *think* I could
actually get up to speed faster with Python than with VB and ASP.Net,
in spite of my ASP/VBS experience.

My boss is pretty open-minded, but also needs to see something working
to be convinced Python is even a viable platform (he had never heard
of it).  Otherwise, .Net is a given, whether it's any good or not,
since corporate is MS all the way.

Can someone point me to a good Python GUI app to demonstrate (other
than IDLE)?

It may very well be that .Net is best for me/us, but from what I've
seen of Python, I *really* like the language and I'd at least like to
see a good demonstration of a full-blown app before ruling it out.
(Something along the lines of what I mentioned above would be ideal,
but not a necessity.)

Thanks!

Lee Gray


From usenet at thinkspot.net  Tue Sep 17 21:11:51 2002
From: usenet at thinkspot.net (Sheila King)
Date: Tue, 17 Sep 2002 18:11:51 -0700
Subject: timeoutsocket module help
Message-ID: <am7r98.h0.1@kserver.org>

OK, I'm doing DNS lookups in a script I've written, to determine whether
certain IP addresses are listed in DNS BL lists (for blocking spammers).

Have been using the code I've written for a long time now (months) without
significant problems. However, someone has made the point to me, that my
code should technically have a default timeout on the socket that tries to
do the DNS lookup. And they are right, so I'm trying to modify my code.

So, I'm using Tim O'Malley's timeoutsocket module, or trying to. But I am
not having any success with this. In the following C&P from an interactive
session, it took 6 or 7 seconds to return the socket error, which is the
result I would expect in the case that the host name is not in the DNS list
(and in this case it is not, so this is a correct result). The problem is,
that I set the default time out (I thought???) to 0.02 seconds, but the
lookup went waaaayyyy past that and basically seemed to ignore the
instruction.

>>> import sys
>>> import socket
>>> sys.path.append(r'j:\web\thinkspotwebsite\xthinkspot\emailfilter_test')
>>> import timeoutsocket
>>> timeoutsocket.setDefaultSocketTimeout(0.02)
>>> socket.gethostbyname('226.200.3.4.bl.spamcop.net')
Traceback (most recent call last):
  File "<pyshell#5>", line 1, in ?
    socket.gethostbyname('226.200.3.4.bl.spamcop.net')
gaierror: (7, 'getaddrinfo failed')
>>> 

Now, I will note, that the comments in the timeoutsocket module indicate
that it works for TCP connections, and I'm not sure that TCP is being used
in the gethostbyname function, as it is my understanding that DNS should
usually be done by UDP instead. So, maybe in this case, using O'Malley's
timeoutsocket module is totally inappropriate and won't work.

In any case, I'm hoping to get some clues here.

Thanks,

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From claird at starbase.neosoft.com  Thu Sep 19 09:06:45 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 19 Sep 2002 08:06:45 -0500
Subject: shared mem & semaphores module
References: <YWbi9.17569$V7.4455639@twister.socal.rr.com>
Message-ID: <CC8D6E83117DF503.CA79F8C321621F29.4ECA2112DC6936EB@lp.airnews.net>

In article <YWbi9.17569$V7.4455639 at twister.socal.rr.com>,
Jeff Davis  <jdavis at empires.org> wrote:
>I found a module online by "INRIA" (I don't know who that is) that seems to 
>wrap the C functions for shared memory management and semaphores.
			.
			.
			.
French National Institute for Computing and Automation
<URL: http://www.inria.org/ >.  As dense with the usual
academic injustices as one might expect, but filled
with interesting people, many of whom you've probably
bumped into already.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From mwright at pro-ns.net  Fri Sep 20 14:19:15 2002
From: mwright at pro-ns.net (Mark Wright)
Date: 20 Sep 2002 11:19:15 -0700
Subject: Why does Tkinter do this?
Message-ID: <f9ff3f62.0209201019.81fb220@posting.google.com>

I'm trying to create a resizeable Label with an image.  bind()ing to
the configure event allows me to resize the image whenever the label
is resized, but if I set the size of the image to the new size of the
label, it is too big (which produces cascading 'configure' events and
a steadily expanding Label).  OK, so I try subtracting the borderwidth
- which is no good either.  But if I subtract the borderwidth *times
two*, it works.  Can anyone explain this to me?

Here's my code:

from Tkinter import *
from PIL import Image, ImageTk


class PhotoLabel(Label):

	def __init__(self, parent, fileName):
		self.startImage = Image.open(fileName)
		self.currentImage = self.startImage
		self.photo = ImageTk.PhotoImage(self.startImage)
		Label.__init__(self, image=self.photo, bd=5)
		self.bind('<Configure>', self.OnConfigure)
	
	def OnConfigure(self, event):
		if event.type == '22':
			# why * 2?
			borderwidth = 2*int(self['borderwidth'])
			newHeight = max(0, event.height - borderwidth)
			newWidth = max(0, event.width - borderwidth)
			self.currentImage = self.startImage.resize((newWidth, newHeight))
			self.photo = ImageTk.PhotoImage(self.currentImage)
			self.configure(image=self.photo)


From pereira at cis.upenn.edu  Thu Sep  5 04:00:41 2002
From: pereira at cis.upenn.edu (Fernando Pereira)
Date: Thu, 05 Sep 2002 04:00:41 -0400
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com> <eppstein-56D20B.14244204092002@news.service.uci.edu> <c0Dd9.6121$e5.1077322@newsb.telia.net>
Message-ID: <B99C8769.120CD%pereira@cis.upenn.edu>

On 9/5/02 2:41 AM, in article c0Dd9.6121$e5.1077322 at newsb.telia.net,
"Fredrik Lundh" <fredrik at pythonware.com> wrote:

> David Eppstein wrote:
> 
>> I'm sure there were good reasons for this decision, but it leads to ugly
>> workarounds like making each nonlocal variable into a list and assigning
>> to x[0] instead of x...
> 
> when you find yourself doing that, it's a sure sign that you're
> no longer using python to write python code...
There is a reason why this may be puzzling to some. Other languages with
non-local lexically scoped variables seem to come in two varieties: those
like Pascal or Common Lisp in which all variables name assignable locations,
and those like ML or Haskell in which all variables name values, and no
bindings be changed. Python is a hybrid, in that variables name values,
local and global bindings can be changed dynamically, but bindings
established at intermediate scopes cannot. This is not a criticism of Python
--- in my experience with languages like Pascal and Common Lisp, assignments
to non-local, non-global variables are more confusing than they are worth
--- but a suggestion of why it may seem strange at first.

-- F



From theoryboy at my-deja.com  Tue Sep  3 14:12:38 2002
From: theoryboy at my-deja.com (Peter Saffrey)
Date: 3 Sep 2002 11:12:38 -0700
Subject: efficient list merging
Message-ID: <ced73313.0209031012.13624f3@posting.google.com>

I have two lists of lists of strings which I would like to merge
efficiently without repeats. I assume that each list currently
contains no repeats. At the moment, I have two possible
implementations:

for item in list2:
   if item not in list1:
      list1.append(item)

which is obvious, and I suspect slow and

list1hash = {}
for item in list1:
   itemtxt = '%s' % item
   list1hash[itemtxt] = 1

for item in list2:
   itemtxt = '%s' % item
   if not list1hash.has_key(itemtxt)
   list1hash.append(item)

which I thought would be quicker because it uses a hash, but it
doesn't seem to be from my crude tests. Maybe it's the text conversion
which is slowing it down - the strings stored in the inner lists are
quite long (about 60 characters, and each inner list has up to 8
items) is there a way I can index the hash by the address of the
lists? I converted to a string just as a "this is the first thing I
can think of" approach.

Or is there a better way?

Thanks,

Peter


From remi at cherrypy.org  Mon Sep  2 13:40:06 2002
From: remi at cherrypy.org (Remi Delon)
Date: Mon, 2 Sep 2002 19:40:06 +0200
Subject: [ANN] CherryPy-0.6 released (SSL support)
Message-ID: <al07pl$mbu$1@wanadoo.fr>

We're pleased to announce the release of CherryPy-0.6.

This release adds SSL support to CherryPy (based on PyOpenSSL).

Many thanks to Martin Sj?gren for his PyOpenSSL package and to Mark Wormgoor
for his contribution to the CherryPy SSL code.

------------------------------------------
About CherryPy:

CherryPy is a Python-based tool for developing dynamic websites. It
uses many powerful concepts together, which makes it unique in its
approach to website development.
CherryPy sits between an application server and a compiler. You write
source files, compile them with CherryPy and CherryPy generates an
executable containing everything to run the website (including an HTTP
server).

Key properties/features of CherryPy are:
   - Based exclusively on Python (runs everywhere Python runs)
   - Delivers fast, robust, and scalable websites
   - Developers can use OOP as well as AOP (Aspect Oriented
Programming) concepts to develop websites
   - True separation of content and presentation
   - Simple but powerful templating language
   - "HTML editor safe" templating language (templates can go back and
forth between designers and developers)
   - Powerful standard libraries to make your life easy

Other properties/features are:
   - Can be linked to many databases (Oracle, Sybase, MySql,
PostgreSql, ...)
   - Can run behind another webserver (Apache, ...)
   - Easy clustering and load-balancing set up for high-traffic
websites
   - Built-in caching capability
   - SSL support

Remi.

http://www.cherrypy.org








From michaels at one.net  Wed Sep 11 22:17:05 2002
From: michaels at one.net (Michael Schneider)
Date: Wed, 11 Sep 2002 22:17:05 -0400
Subject: iterators and generators, is this the Python Way???
Message-ID: <3D7FF921.2040305@one.net>

Hello All.

I just updated my python to 2.2 and noticed that generators, and 
iterators were in there.

I often parse through many files (one at a time), and wanted the speed 
of read,  and
the convience of   for x in myGen:


The code below works just fine,  could you review this code against the 
intent of
iterators and generators.

Thanks,
Mike

------------------------------------------------------------------------------------------------
# Goal:
#
#      Process files quickly, by reading the file in to memory in one 
blast,
#      the adding iterator behavior
#
#
#      Michael Schneider

from __future__ import generators

import sys
import os
from stat import *


class FastFile:
    def __init__(self, filename):
        """ Read in entire file into memory"""
        f = open(filename, 'r')
        fileContents = f.read();
        self.lines = fileContents.split('\n')
        f.close()
   
    def  __iter__(self):
        return self.fastFileGen()

    def fastFileGen(self):
        for line in self.lines:
                yield line
       

# usage
.....

    try:
        myFile = FastFile(filename)
    except IOError, msg:
        print, filename , " could not open "
        return 1

   
    # Process each line in the file
    for line in myFile:
        fixline(line)       




From wganz at imacs.com  Thu Sep  5 12:17:39 2002
From: wganz at imacs.com (Ganz, Will)
Date: Thu, 5 Sep 2002 11:17:39 -0500 
Subject: What database should I use
Message-ID: <F3EF14447A98D311B981009027B0FB380172469E@IMACS_EXCHANGE>

I grant that FoxPro is still a formidable competitor in the marketplace but
given the size that the original poster stated of 2K-3k records; the 2gig
Access limitation shouldn't come into play. In this instance, Visual FoxPro
would be like using a supersonic fighter jet as a crop duster. 8-)

-----Original Message-----
From: Ron Darling [mailto:rond at ixsys.com] 
Sent: Thursday, September 05, 2002 10:30 AM
To: Sathish Chandrasekaran (S); printers at sendme.cz; python-list at python.org;
tutor at python.org; activepython at listserv.ActiveState.com;
python-help at python.org
Cc: db-sig at python.org; wx-users at lists.wxwindows.org
Subject: RE: What database should I use


>> (though it's it considered obsolete now-a-days).

This is only a consideration for the uni-formed or those that have never
really used it.  MS is still dumping lots of money into it.  Just shipped
v7.0 will be going into Beta with 8.0 in a couple of months.

Amazing String capabilties.  As a matter of fact there is now a VFP style
.net class that moves a lot of the wonderful commands internal to VFP to
.NET.  I think the Fox Dev team created.

Cursor Technology, Rushmore Speed and many other things have moved there way
into not only Access but SQL Server and ADO.

Due to the fact that each part of a table is its own file.  .DBF is the
table.  .FPT is the Memo field  .CDX is the index.  The file size limitation
of the OS of 2 gig is at the individual file level.  In Access everything is
stored in the same file (you could break up but not really meant to) so the
2 gig file size is limited to the app and data which is more than likely in
the same file.  I have dealt with lots of situations where the .DBF was 1.8
gig, the index was 1.2 gig and the FPT was another 1+ gig size.

Indexes are stored in a .CDX file.  A .CDX file can have as many as 255
indexes (take into account 2 gig file size).  Each open table can have 10
open CDX's so you can have 2550 indexes on a table.

Very Robust SQL pass Through for back end databases.

There is nothing out there as fast as VFP for pure raw speed.  SQL 2000 gets
close but not quite there, the advantage to SQL as far speed goes is it does
not bog down under load.  Load means 25 + heavy users.

As far as speed goes I have seen the return of queries on tables as large as
21 million records within seconds and that was pulling data across a 10 meg
network with Foxpro For DOS.  United We Stand (remember Ross Perot)
member/prospect database.  Sounds a little dated but the point is its been
that way in Fox for a long time.

Fox and VFP managed/tracked the military assets to the Persian Gulf and is
till in use. Manages the flow of info for Chunnel. It is far more prevalent
than most realize.

Very good OOP.  It seems that VFP is becoming known for the best language
that can really implement the "Abstract Factory Pattern"

Very good interface. Rich Native Controls (No OCX) and does very well with
OCX's.

Easy Distribution.  (Drop a couple of DLLS into system directory not reg)

Rich Debugger.

Can access any data ODBC, ADO or ADO.Net can access.

Internal XML.  Select Result set from SQL into a local cursor then use
CURSORTOXML().  With the XMLTOCCURSOR() and
CURSORTOXML() commands it is very easy to build thin wire apps.  Bind the
interface to the cursor.

I can try any and all commands from the command prompt and even from within
a debugging suspend.  Pretty much interrogate the entire system at hand very
easily from a suspend.

Many options for hanging VFP behind IIS and ISAPI.

You can create com objects and run behind other languages even from within
SQL Server or visa versa.

One of the most robust SOUNDEX/Fuzzy Search add-ons available with the
PHDbase Library.  You can index memo fields and do 2 - 3 second searches on
huge amounts of memo data.

Best Table Repair Libraries that I am aware of are available.

Error trapping has been far superior to VB and a lot of languages since the
Fox 2.0 days (1992).

Last but not least http://www.universalthread.com.  Great support from all
over the world.


obsolete...hmmmmm maybe lack of exposure or conflicting marketing problems
with MS owning Access, SQL and Fox but certainly not obsolete.

-----Original Message-----
From: activepython-admin at listserv.ActiveState.com
[mailto:activepython-admin at listserv.ActiveState.com]On Behalf Of Sathish
Chandrasekaran (S)
Sent: Thursday, September 05, 2002 9:24 AM
To: printers at sendme.cz; python-list at python.org; tutor at python.org;
activepython at listserv.activestate.com; python-help at python.org
Cc: db-sig at python.org; wx-users at lists.wxwindows.org
Subject: RE: What database should I use


In that case MS-access will be better on the size issue. But on the other
hand if u want ur application to be flexible and ease of use I would rather
suggest FoxPro (though it's it considered obsolete now-a-days).


Here Is A Test To Know Whether The Mission Of Your Life Is Over . If You Are
Alive It Isn't .

Sathish C
Infra-Amdocs


-----Original Message-----
From: A [mailto:printers at sendme.cz]
Sent: 05 September 2002 15:21
To: python-list at python.org; tutor at python.org;
activepython at listserv.ActiveState.com; python-help at python.org
Cc: db-sig at python.org; wx-users at lists.wxwindows.org
Subject: What database should I use


Hello,
I am going to program a small application (in Python and wxPython) which
consists about 2000 - 3000 records . Can you please recommend which database
is the best for that. I use Windows. Maybe MS Access  or FoxPro or MySQL or
better? I want to keep the program as small as possible. Thanks Ladislav

_______________________________________________
ActivePython mailing list
ActivePython at listserv.ActiveState.com
To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
Other options: http://listserv.ActiveState.com/mailman/listinfo/ActivePython

_______________________________________________
ActivePython mailing list
ActivePython at listserv.ActiveState.com
To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
Other options: http://listserv.ActiveState.com/mailman/listinfo/ActivePython


_______________________________________________
ActivePython mailing list
ActivePython at listserv.ActiveState.com
To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
Other options: http://listserv.ActiveState.com/mailman/listinfo/ActivePython



From paul at boddie.net  Fri Sep 20 11:51:15 2002
From: paul at boddie.net (Paul Boddie)
Date: 20 Sep 2002 08:51:15 -0700
Subject: How to access XML passed from an XForm
References: <mailman.1032506367.3820.python-list@python.org>
Message-ID: <23891c90.0209200751.335a84f2@posting.google.com>

Jeremy Rew <huxley1 at westnet.com.au> wrote in message news:<mailman.1032506367.3820.python-list at python.org>...
> 
> I have an XForm that passes a serialised XML document to a URI ( a python 
> script). Im not exactly sure of the internal representation of what it 
> sends, but I am having trouble just finding out how to access the XML from 
> within the python script.  I just need to know if theres a way to store the 
> XML in a string or find out where I can access it from, as I already know 
> how to process/manipulate the XML once I can get at it. Any ideas (Im 
> rather new to python)?

I imagine that what you want to know is documented here:

  http://www.w3.org/TR/2002/WD-xforms-20020821/slice11.html

I suspect that you can handle the incoming data in a similar way to
HTML form "file upload" information, but at the most basic level,
you're most likely to be getting a generic HTTP request with certain
parts of the message body containing the information you require. I
wonder if some of the work around this hasn't already been done in
certain XML-RPC or SOAP implementations...

Paul


From aahz at pythoncraft.com  Tue Sep 17 00:52:09 2002
From: aahz at pythoncraft.com (Aahz)
Date: 17 Sep 2002 00:52:09 -0400
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <c%qh9.102794$ub2.2120685@news1.tin.it> <mailman.1032212180.11973.python-list@python.org> <3d868ccc$1@news.sentex.net>
Message-ID: <am6cdp$2i3$1@panix1.panix.com>

In article <3d868ccc$1 at news.sentex.net>,
Peter Hansen  <peter at engcorp.com> wrote:
>
>Copyright is actually a passive thing now, as opposed to active.
>That is, you don't have "to copyright your work", because unless
>you explicitly disclaim the copyright, you automatically get it.

True.

>As I recall, this was not always the case in at least the U.S., but
>has been the case for a number of years now.  In other words, if you
>don't write something like "placed in the public domain" then you still
>have copyright on it and could take steps in courts of law to recover
>lost income if someone else attempts to profit from your work without
>permission.
>
>There *are* steps you can take that are considered by some to increase
>the strength of the copyright (registering it officially and so forth),
>but others believe this is mostly a waste of time.

False (mostly).  It's not the strength of the copyright, it's the amount
of money that one can gain from a copyright infringement lawsuit.
Registering copyright permits punitive damages; unregistered copyright
permits only "actual" damages, and you have to prove every penny of it.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From aahz at pythoncraft.com  Sun Sep 29 11:14:13 2002
From: aahz at pythoncraft.com (Aahz)
Date: 29 Sep 2002 11:14:13 -0400
Subject: A simpler singleton pattern?
References: <mailman.1033259618.14506.python-list@python.org>
Message-ID: <an75c5$hc0$1@panix1.panix.com>

In article <mailman.1033259618.14506.python-list at python.org>,
lee lion <dance_code at hotmail.com> wrote:
>
>     singleton pattern is a way to provide one and only one object of a 
>particular type. There are many ways to implement a singleton pattern, is 
>the below one?

Dunno, but the simplest way to create a singleton is to use a class
(*not* an instance) or a module and set attributes on the class/module.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From bloke at ii.net  Fri Sep 27 10:37:19 2002
From: bloke at ii.net (Rob Hall)
Date: Fri, 27 Sep 2002 22:37:19 +0800
Subject: Problem with ClientCookie
Message-ID: <3d946d0f$0$10364@echo-01.iinet.net.au>

I am trying to use ClientCookie to access a HTTPS website under Win98.  I
have installed the W32 SSL wrappers.

The code I have is similar to as follows (have removed authorisation
detail):

import urllib
import ClientCookie

if __name__ == '__main__':
    urllib.URLopener.version = 'Mozilla/4.0 (compatible; MSIE 5.5; Windows
NT 5.0)'
    print 'Logging on...'
    request =
urllib2.Request('https://www.quicktrade.com.au/cgi-bin/login?USERNAME=myUser
Name&PASSWORD=myPassword')
    result = ClientCookie.urlopen(request)

I have set HTTP_DEBUG = 1 in _ClientCookie.py

I get the following response...


Logging on...
Page requested...
send: 'GET /cgi-bin/login?USERNAME=myUserName&PASSWORD=myPassword
HTTP/1.0\r\n'
send: 'Host: www.quicktrade.com.au\r\n'
send: 'User-agent: Python-urllib/2.0a1\r\n'
send: '\r\n'
reply: 'HTTP/1.1 200 OK\r\n'
header: Date: Fri, 27 Sep 2002 14:21:31 GMT

header: Server: Stronghold/2.4.2 Apache/1.3.6 C2NetEU/2410 (Unix)

header: set-cookie:
key=Kpfz5sbkkPrwsM7eng5ihtfuKq6sAlSdwTXhn/jUiEOSOCYJAC5r/Vlq+5uS9wqkpHJzOWHi
m4v+^Va7SldrYP9Vqea2vDKtZWnOiP7Ff5ckrrM61DDM+tIGO44jUyKPt2yof5Uz7lvUu0ErVqJk
DWAYX^7x66TxqZ2esB823pOue6zNxcmhY1SiS7pPTd5w3X/5ncX3njMZs_^; Path=/; secure
header: set-cookie: acct=myUserName STGBNET; Path=/; secure
header: set-cookie: level=1; Path=/; Domain=.quicktrade.com.au; secure
header: set-cookie: ID=1802009989; Path=/; Domain=.quicktrade.com.au; secure
header: set-cookie: w1=None; Path=/; secure
header: P3P: CP="NON DSP ADMa OUR IND UNI FIN"
header: Connection: close
header: Content-Type: text/html

Traceback (most recent call last):
  File
"C:\WINDOWS\Desktop\backup\uni\project\development\htmlProcessor\htmlProcess
or.py", line 22, in ?
    result = ClientCookie.urlopen(request)
  File "C:\PYTHON22\lib\site-packages\ClientCookie\_ClientCookie.py", line
1387, in urlopen
    return _opener.open(url, data)
  File "C:\PYTHON22\lib\urllib2.py", line 322, in open
    '_open', req)
  File "C:\PYTHON22\lib\urllib2.py", line 301, in _call_chain
    result = func(*args)
  File "C:\PYTHON22\lib\site-packages\ClientCookie\_ClientCookie.py", line
1334, in https_open
    return self.do_open(httplib.HTTPS, req)
  File "C:\PYTHON22\lib\site-packages\ClientCookie\_ClientCookie.py", line
1319, in do_open
    self.c.extract_cookies(result, req, redirect=redirect)
  File "C:\PYTHON22\lib\site-packages\ClientCookie\_ClientCookie.py", line
627, in extract_cookies
    k, v = re.split(r"\s*=\s*", param, 1)
ValueError: unpack list of wrong size

I can't find anything to help me debug this.  Is the error because the web
server is returning more than one cookie?  or is it just returning a bad
format?

Any help is greatly appreciated.

Rob




From rob at zoism.org  Mon Sep 16 20:39:30 2002
From: rob at zoism.org (Rob Brown-Bayliss)
Date: 17 Sep 2002 12:39:30 +1200
Subject: Socket server and threads
In-Reply-To: <1032217212.2735.21.camel@caspian.everglade>
References: <1032217212.2735.21.camel@caspian.everglade>
Message-ID: <1032223169.2736.24.camel@caspian.everglade>

Hi, I am trying to write an app. it Has a gui, and I am using xmlrpc via
socket server to communicate with other apps on other machines.

I currently have the socket server running as a thread like so:

args = (('', 48002), BHH_Communication.BHH_DMHandler, self)
	self.server = None
	self.tserver = thread.start_new_thread(BHH_Communication.BHH_TCPServer,
args)
	while (self.server == None):
		pass

where BHH_TPCServer is:

class BHH_TCPServer(SocketServer.TCPServer):
	def __init__(self, server_address, RequestHandlerClass, parent):
		SocketServer.TCPServer.__init__(self, server_address,
RequestHandlerClass)
		self.BHH = parent
		self.BHH.server = self   # so handlers have access to 					 # GUI app
		self.in_comms = []
		self.in_comms.append(self)
		self.out_comms = []
		
		while 1:
			rtoread, rtowrite, err = select.select(self.in_comms ,
self.out_comms, [], 5.0)
			if (len(rtoread) != 0):
				rtoread[0].handle_request()


So with this my handler class can call functions in the main app via
self.server.BHH.some_function()


My question is what do I need to do to make sure it is safe, what
happens if the thread and the parent app are trying to call the same
function at  the same time?  What about if two or more request handlers
are running and want the same function at the same time (via a threaded
TCP handler class?

-- 

*
*  Rob Brown-Bayliss
*



From erikprice at mac.com  Tue Sep 10 08:41:45 2002
From: erikprice at mac.com (Erik Price)
Date: Tue, 10 Sep 2002 08:41:45 -0400
Subject: Pickle problem
In-Reply-To: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220D0E@hendrix.empolisuk.com>
Message-ID: <AA7834D0-C4BA-11D6-A54B-00039351FE6A@mac.com>

On Tuesday, September 10, 2002, at 06:13  AM, Harvey Thomas wrote:

> I'm trying to pickle a complicated data structure which includes a 
> dictionary whose entries are created by (simplified):
>
> a[x] = re.compile(y).match
>
> The pickle fails, however, with the error:
>
> PicklingError: Can't pickle <built-in method match of _sre.SRE_Pattern 
> object at
>  0x00838700>: it's not found as __main__.match
>
> I'd be grateful if someone could tell me if it is possible to pickle 
> the data structure and what I'm doing wrongly.

Are you trying to store the MatchObject returned by a call to a regular 
expression object's match() method?  If so, it looks like you're 
forgetting the parens at the end of "re.compile(y).match".



Erik





--
Erik Price                                                  (zombies 
roam)

email: erikprice at mac.com
jabber: erikprice at jabber.org




From kodokuchef at yahoo.com  Fri Sep 20 01:37:12 2002
From: kodokuchef at yahoo.com (akirasugiura)
Date: 19 Sep 2002 22:37:12 -0700
Subject: newbie seeks advise.
Message-ID: <77c84db2.0209192137.51167ebb@posting.google.com>

Hi,

I've been learning python for about a month. I had no prior knowledge
of programming.

My question is, I have a tuple which is, data = (69, 40, 74, 41, 56,
57, 38, 58, 26, 55). How I can make a function that picks up a largest
number
 from this "data" tuple. If "data" was a  'List' I would be able to
use "sort" but since sort doesn't  work with tuple. I am stuck here.

Thanks in advance

Sincerely,


From robin at jessikat.fsnet.co.uk  Wed Sep 18 15:13:12 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 18 Sep 2002 20:13:12 +0100
Subject: cgi binary upload
Message-ID: <d5t2JzAIBNi9EwtI@jessikat.demon.co.uk>

I know there's a lot of skilled people using python so I thought I'd ask
how one can get an HTML page to do a proper binary upload using the 
<input type=file> form tag. We want to do this for PDF documents, but it
seems as though IE/Mozilla/Netscape clients insist on putting 

Content-Type: data/pdf

or similar headers into the multi part file section and the files have
their line endings corrupted.

On the receiving side using cgi.py we are certainly seeing some
filetypes OK eg zip files, but surely we should be able to fix this more
reliably. I should note that experimentally renaming the file as
xxx.xpdf made no difference as the mime type was being correctly
determined by the browser. 
-- 
Robin Becker


From jtk at yahoo.com  Mon Sep 30 16:15:17 2002
From: jtk at yahoo.com (Jeff Kowalczyk)
Date: Mon, 30 Sep 2002 16:15:17 -0400
Subject: Using __import__ to access a local module attribute
Message-ID: <anac43$vqb$1@main.gmane.org>

I'm trying to use __import__ to dynamically import a module in a customized
Zope Formulator python product, and thereby access a singleton instance attribute
of the primary class wtih a call to getattr. I can get it to work with
hardcoding the module name, but I don't see what I'm missing when I try to
use __import__. The code below is just an early version of a means to an
end, and I'm wide open to better ways accessing the instance attribute.
Metaprogramming babysteps, I guess...

The MyExtClassInstance attribute in MyExtModule.py that I'm trying to access
is defined like this:
MyExtClassInstance = MyExtClass()

just like the widget and validator formulator classes are. MyExtModule.py
is in the same product directory with my customized formulator filesystem
product, call it 'FormulatorMyExt'.

--- FieldRegistry.py ------------------------------------------------------
def initializeFieldForm(field_class, components=('widget','validator')):
    """ """
    from Form import BasicForm
    from DummyField import fields

    form = BasicForm()
    override_form = BasicForm()
    tales_form = BasicForm()
    for component in components:
        if component not in ('widget','validator'):
           # when the seq is ('widget','validator','myExtModule')
            import MyExtModule # This works ...
           # but what I intend is a working form of either of these next lines....
           # __import__(component.capitalize(), globals(), locals())
           #__import__(getattr(Products.FormulatorMyExt,component.capitalize()),
                                globals(), locals())
            setattr(field_class, component,
getattr(MyExtModule,component.capitalize()+'Instance'))

I get errors like:
NameError: global name 'Layout' is not defined (in the first case, and)
NameError: global name 'Products' is not defined (in the second case)

Does anyone see my mistakes here?

And obviously, I want to get that last line of hardcoded MyExtModule name
out of there, but it obfuscated the problem to add that here. I can do it
with a getattr call, I think, contingent on solving the former problem.

Thanks for any help with using __import__, including the from list, or
suggestion of a completely different way to do it as appropriate.







From mail at peterbe.com  Thu Sep 26 18:44:07 2002
From: mail at peterbe.com (Peter Bengtsson)
Date: 26 Sep 2002 15:44:07 -0700
Subject: Integrate with Flash MX
Message-ID: <322e5959.0209261444.797a936e@posting.google.com>

I want to use Flash MX to build a GUI for use on the desktop on
windows, not embedded in a webpage.
ActionScripts don't do it for me so I want to use Python for the
business logic.

Is there some way to enable this connection? Maybe this is more of a
Flash question.
I suppose my python module could spit and swallow XML that the Flash
app could pick up, but wouldn't want to use HTTP for that.

Anybody got any experience with this?

Cheers,
Peter


From duncan-news at grisby.org  Sun Sep  1 18:25:17 2002
From: duncan-news at grisby.org (Duncan Grisby)
Date: Sun, 01 Sep 2002 22:25:17 GMT
Subject: Python IPC options
References: <1030829572.943495@irys.nyx.net> <ddc19db7.0209010836.50242309@posting.google.com> <mailman.1030910102.14267.python-list@python.org>
Message-ID: <htwc9.9930$5g6.230732@newsfep2-win.server.ntli.net>

In article <mailman.1030910102.14267.python-list at python.org>,
 holger krekel  <pyth at devel.trillke.net> wrote:

>> Corba 2.2 -> http://www.fnorb.org 
>
>CORBA 2.0 AFAIK.

Fnorb is CORBA 2.0, and does not (yet) follow the official standard
Python language mapping. omniORBpy is CORBA 2.6, and does follow the
standard language mapping:

  http://omniorb.sourceforge.net/

Cheers,

Duncan. (omniORBpy author)

-- 
 -- Duncan Grisby         --
  -- duncan at grisby.org     --
   -- http://www.grisby.org --


From duncan at NOSPAMrcp.co.uk  Fri Sep 20 04:50:29 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Fri, 20 Sep 2002 08:50:29 +0000 (UTC)
Subject: Avoiding referencing (was: Python sequences reference - problem?)
References: <mailman.1032443760.5578.python-list@python.org> <YZli9.114624$ub2.2439313@news1.tin.it> <amejvb$55qsi$1@ID-77524.news.dfncis.de>
Message-ID: <Xns928F61C6BEE5Cduncanrcpcouk@127.0.0.1>

Thorsten Kampe <thorsten at thorstenkampe.de> wrote in news:amejvb$55qsi$1 at ID-
77524.news.dfncis.de:

> Obviously I did not pass 'foo' "by value" to function 'part' but by 
> reference and obviously not only local variable 'seq' was modified in 
> the function but "outer" variable 'foo', too.
> 
> So how to deal with that?
> 
You missed out option number 4:

Don't write code that modifies its arguments unless you always want the 
arguments modified.

In general, if you want to modify an argument, but don't want the change 
reflected back to the caller, make a copy of the argument. If you want the 
change reflected back to the caller always, then go ahead and modify the 
argument, but make it clear from the function name that it does that. If 
you want the function to sometimes modify its argument, then write it so 
that it never modifies the argument but returns a modified value as a 
result. Then the caller can decide whether to use that result or EXPLICITLY 
ignore it.

For the specific example you gave, there was no reason to modify the 
argument, the code is just as simple without either copying or modifying:

>>> def part(seq, indices):
	partition = []
	start = 0
	for slice in indices:
		partition.append(seq[start:slice])
		start = slice
	partition.append(seq[start:])
	return partition

>>> foo = [11, 22, 33, 44, 55, 66, 77]
>>> part(foo, [2, 3])
[[11, 22], [33], [44, 55, 66, 77]]
>>> foo
[11, 22, 33, 44, 55, 66, 77]
>>> 

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From hwlgw at hotmail.com  Sun Sep 29 16:23:55 2002
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 29 Sep 2002 13:23:55 -0700
Subject: Tkinter: bind Alt + mouse click
References: <i0kznu0j4zx.fsf@Thales.MI.Uni-Koeln.DE>
Message-ID: <cb035744.0209291223.4f03a9ff@posting.google.com>

[Ulrich Goertz]
> I am puzzled why I cannot bind events of the type "Alt" button press plus
> mouse click. 

Those Alt-keycombos are probably catched by a master widget.  It could
help if you do

self.master.unbind('<Alt>-s')

this assumes self is your widget and you have the master widget saved
in a .master attribute, as many widgets do in their __init__
constructor.


'''
A major, with wonderful force,
Called out in Hyde Park for a horse.
        All the flowers looked round,
        But no horse could be found;
So he just rhododendron, of course.
'''


From chris.gonnerman at newcenturycomputers.net  Thu Sep 19 23:00:49 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Thu, 19 Sep 2002 22:00:49 -0500
Subject: Nokia 3310 SMS Application, done:connecting to serial port(linux/win), done:reading some hex like \x00 . . ., problem:write/send command! noresponse
References: <52facf1d.0209190533.3048e62c@posting.google.com>
Message-ID: <004701c26051$f13fc160$0101010a@local>

----- Original Message ----- 
From: "Polerio Babao Jr.II" <ptbjrii at hotmail.com>

<< hack hack hack >>

> Now I've tried sending commands using the Nokia AT Command 
> which was taken from this link here -> 
> http://3ton.com/besik/ATNOKIA.pdf
> 
> The problem is, this still gives no response, when I type
> ser.write('ATCGMI') or any other command, I get no response. 

<< hack hack hack >>

Ah... shouldn't that be:

    ser.write('ATCGMI\r')

... as in, don't you need the command-line-terminator?

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From gregory.p.green at boeing.com  Thu Sep 12 15:14:36 2002
From: gregory.p.green at boeing.com (Greg Green)
Date: Thu, 12 Sep 2002 19:14:36 GMT
Subject: Extending a C++ App
References: <996eade8.0209120912.3a653d7@posting.google.com>
Message-ID: <n0qnhvfn.fsf@boeing.com>

Embedding python inside an application can be simple. The basic step
is to call Py_Initialize(). After that you can execute python code
with the PyRun_* functions. Of course, in practice, there is a lot
more to worry about. Threads and the GIL, if you are using a GUI, you
have to worry about the event loop, etc.

Accessing your c++ object model requires you to wrap them just as if
you were going to use the model in a python program. The normal
answers are swig, sip, boost and some others. I've used both swig,
sip, and writing the wrapper by hand. All will work. Once the api is
wrapped, the resulting code is either statically or dynamically linked to
your exe.

-- 

Greg Green



From pyth at devel.trillke.net  Fri Sep 27 16:57:04 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 27 Sep 2002 22:57:04 +0200
Subject: ENOSPC doesn't cause exception?
In-Reply-To: <20020927195418.GF5941@unpythonic.net>; from jepler@unpythonic.net on Fri, Sep 27, 2002 at 02:54:19PM -0500
References: <20020927195418.GF5941@unpythonic.net>
Message-ID: <20020927225704.S30315@prim.han.de>

Jeff Epler wrote:
> Can someone tell me why this doesn't cause an exception?
> $ python -c print > /dev/full
> 
> (/dev/full is a special device found on at least some kinds of Linux,
> which always returns -ENOSPC when written)
> 
> if I use 'strace', I can see that -ENOSPC is returned from write():
> $ strace -e write python -c print > /dev/full
> write(1, "\n", 1)                       = -1 ENOSPC (No space left on device)
> 
> I would expect this to print something like
> Traceback (most recent call last):
>   File "<string>", line 1, in ?
> OSError: [Errno 28] No space left on device
> 
> I tested this on Python 1.5.2 and Python 2.3a0 CVS, so it doesn't appear
> to be anything new...

very interesting.  If i have a file which contains:

    f = open('/dev/full','w')
    f.write('hallo')
    f.close()

and do 

    python /tmp/q 

i get the expected behaviour (no space ...).  If you leave out 
'f.close()' you don't get the exception any more. 

But even running with 'python -cu print >/dev/full' (stdout is unbuffered)
this doesn't throw an exception. 

hmmm. Wildly guessing, during interpreter shutdown this exception 
(and possibly others) seem to be ignored?!

At least the following code

    class a:
    def __del__(self):
        raise OSError, "hello"
    a1=a()

produces  

[hpk at cobra /tmp]$ python q
Exception exceptions.OSError: <exceptions.OSError instance at 0x813a8b4>
in <bound method a.__del__ of <__main__.a instance at 0x81344fc>> ignored
[hpk at cobra /tmp]$

Couldn't easily find the place in the C-source where this happens, though.

regards,

    holger



From support at internetdiscovery.com  Mon Sep 30 13:56:48 2002
From: support at internetdiscovery.com (Mike Clarkson)
Date: Mon, 30 Sep 2002 17:56:48 GMT
Subject: Doc for Tix / Python ?
References: <3D9382F9.7040407@novagrid.com>
Message-ID: <3d988f49.6335208@nntp.ix.netcom.com>

On Thu, 26 Sep 2002 23:58:17 +0200, olivier Ravard
<olivier.ravard at novagrid.com> wrote:

>Hi,
>
>Is someone know where we can find some documentation
>or tutorial about using Tix with Python ?

http://www.python.org/doc/current/lib/module-Tix.html
and the references contained therein, and the library documentation
distributed with Python.

>I try to use PanedWindow without succes...

What exactly is the problem?  Have your tried the demos in
the source distribution (Demo/tix). I think there's one that uses
PanedWindow.

Mike.


From martin at v.loewis.de  Thu Sep  5 14:46:44 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 05 Sep 2002 20:46:44 +0200
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com> <j4it1k4ufy.fsf@informatik.hu-berlin.de> <al886p$lfj$0@216.39.172.122>
Message-ID: <m3d6rs46ln.fsf@mira.informatik.hu-berlin.de>

bokr at oz.net (Bengt Richter) writes:

> How about introducing an optional "degree of globalness" for the
> global declaration? E.g.,

> 
>     x = 1
>     def foo():
>       x = 2
>       def bar():
>         x = 3
>         def baz():
>           global(1) x
>           x = 4
> 
> would go outwards 1 nested scope and thus target the x in bar,
> rebinding it from 3 to 4.

That would work - although I'd claim that some users will find that
notation plain ugly.

It will also have confusing results:

x = 1
def foo():
  print x
  def bar():
    global(1) x
    x = 1

At the moment, the "print x" in foo prints the global x, since there
is no assignment to x, and hence there is no local variable foo. With
the global statement, you declare that there is a local variable
foo::x. As a result, since there is no assignment to foo::x, this
would raise an UnboundLocal error.

Regards,
Martin



From r.l.richardson at verizon.net  Mon Sep  9 21:33:55 2002
From: r.l.richardson at verizon.net (Rick Richardson)
Date: Mon, 9 Sep 2002 18:33:55 -0700
Subject: IDLE
In-Reply-To: <wpbf9.264320$aA.45691@sccrnsc02>
References: <wpbf9.264320$aA.45691@sccrnsc02>
Message-ID: <20020910013604.SLEV16391.out018.verizon.net@there>

On Monday 09 September 2002 05:54 pm, you wrote:
> Why won't my IDLE work?

well. given your amount of included information, I can only assume that 
either 
a: Your computer is off. 
b: You don't have the proper tk libs.
c: You don't have python installed.
d: Your computer is an atari 2600.
e: The planets are misaligned.
f:  your dog snuck in in the middle of the night and rewrote the IDLE code so 
it only works when he/she is logged in.



Hope that helps. 

Rick Richardson.



From missive at frontiernet.net  Thu Sep 26 18:27:48 2002
From: missive at frontiernet.net (Lee Harr)
Date: Thu, 26 Sep 2002 22:27:48 -0000
Subject: curses troubles
References: <mailman.1033018354.14703.python-list@python.org>
Message-ID: <up72f4de73am89@corp.supernews.com>

> I'm trying to take user entered data input and store it as a var.
> 
> inside an __init__ I have curses set up a window, and display some 
> options. Then I use:
> while 1:
>     self.k = self.stdscr.ch()
>     if self.k == ord('G'): do_this()
>     elif self.k == ord('P'): do_another()
>     elif self.k == ord('Q'): break
> 
> Now, when running this, if I hit either G or P, the applications 
closes as if I'd hit Q
> 


Hi;

I made a pretty large application which uses curses.

The trouble is that it can be tough to debug, since all the screen
output is going through curses.

What i ended up doing is creating a log function which i can
call with various strings and variables and which writes its
output to a file... then on another terminal I can just tail
that file to see what is going on.

So what i would do is create such a function, then log the
value of self.k in that loop so you can see exactly what
is going on.



From aleax at aleax.it  Thu Sep 26 09:55:58 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 26 Sep 2002 13:55:58 GMT
Subject: comparing all values of a list to regex
References: <mailman.1032997118.7138.python-list@python.org>
Message-ID: <OlEk9.170909$pX1.6083814@news2.tin.it>

Delaney, Timothy wrote:
        ...
>> num_matches = [ there.match(item) is not None for item in
>> alist ].count(1)
        ...
> # Expanded out for clarity - could be one line
> matches = map(regexp.match, alist)
> matches = filter(None, matches)
> matches = len(matches)

The list-comprehension equivalent of this would be:

matches = len([ item for item in alist if there.match(item) ])

and may indeed be clearer than the .count approach I originally
proposed (except that I think that naming the variable matches
is ambiguous -- num_matches or numMatches & being preferable).

A more concise expression of your approach would of course be:

matches = len(filter(there.match, alist))

I don't see exactly what value the call to map adds in your case.


> # Find out which list to append to
> 
> match_list = match_all
> 
> if not matches:
>     match_list = match_none
> elif matches == 1:
>     match_list = match_one
> elif matches != len(alist):
>     match_list = match_some
> 
> match_list.append(alist)

Yes, choosing a list first and then appending to the chosen
list is nicer, but you cannot do that within the original
specs (for all I know the match_some list might grow
unbearably large over time, in this case).

> Note that I added an additional case - where more than one matched, but
> not all.

Yep, that's the part that's out of the original specs.

The concise equivalent alternative here might be (e.g.):

{0:match_none, 1:match_one, len(alist):match_all}.get(
    matches, match_some).append(alist)

or to get back in the original specs:

match_list = {0:match_none, 1:match_one, len(alist):match_all}.get(matches)
if match_list is not None: match_list.append(alist)


Alex



From dhein at acm.org  Sun Sep 15 16:20:46 2002
From: dhein at acm.org (Coot)
Date: 15 Sep 2002 13:20:46 -0700
Subject: Popup menu won't hide on Linux
Message-ID: <a3a96207.0209151220.5228a262@posting.google.com>

When a popup menu is displayed, I expect that it should goes away when
one of the menu items is clicked (button <1>) or when I click anywhere
else (such that the focus is no longer on the popup menu).

However, on Linux, the latter behavior doesn't happen.  The popup
remains on top of the main app window and the only way to get rid of
it is to click one of its menu command items or close the application.

This behavior isn't there on Windows.

Can anyone tell me how to solve, or work around, the problem???

Here is some example code you can use to demonstrate this problem for
yourself (taken from
http://www.pythonware.com/library/tkinter/introduction/x5819-patterns.htm):

# menu-example-4.py

from Tkinter import *

root = Tk()

def hello():
    print "hello!"

# create a popup menu
menu = Menu(root, tearoff=0)
menu.add_command(label="Undo", command=hello)
menu.add_command(label="Redo", command=hello)

# create a canvas
frame = Frame(root, width=512, height=512)
frame.pack()

def popup(event):
    menu.post(event.x_root, event.y_root)

# attach popup to canvas
frame.bind("<Button-3>", popup)

mainloop()


From dgallion1 at yahoo.com  Wed Sep 18 22:31:57 2002
From: dgallion1 at yahoo.com (darrell)
Date: Thu, 19 Sep 2002 02:31:57 GMT
Subject: RegEx editors
References: <amaq0d$4aton$1@ID-144898.news.dfncis.de>
Message-ID: <xGai9.1630$k92.392818@twister.nyroc.rr.com>

import re

while 1:
    buf=eval('"""%s"""'%raw_input("Enter string to search:"))
    print buf
    while 1:
        pat=raw_input("Enter pattern:")
        if len(pat)==0:
            break
        mo=re.search(pat, buf)
        if mo:
            print mo.groups()
        else:
            print ""
            
##########################################
Enter string to search:now is the \n time for
now is the
 time for
Enter pattern:(?s)now(.*?)for
(' is the \n time ',)


--Darrell

Eric Arnold wrote:

> Does anybody know of any good graphical regex editors out there?
> 
> Eric



From emile at fenx.com  Sun Sep 22 13:25:29 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 22 Sep 2002 10:25:29 -0700
Subject: How to get memory usage on Windows?
References: <amilhk$bi8$1@reader1.tiscali.nl>
Message-ID: <amkubm$ahm$1@main.gmane.org>

GerritM:
[snip]
> how much memory is used per phase and/or per function. Is there any
function
> returning this information. I noted that on Unix the resource module
gives
> access to this kind of information, but I could not find any Windows
> alternative.
>

You could try pslist from the pstools package at www.sysinternals.com to
get an idea of total memory in use by process.

--

Emile van Sebille
emile at fenx.com

---------






From erikprice at mac.com  Sun Sep  1 11:57:06 2002
From: erikprice at mac.com (Erik Price)
Date: Sun, 1 Sep 2002 11:57:06 -0400
Subject: yield keyword reserved
Message-ID: <770604FC-BDC3-11D6-8A4A-00039351FE6A@mac.com>

I'm sorry if this is an easily answered-question.  I have Python2.2 
running on my system (it is installed on Darwin by default in Mac OS X 
10.2), and am learning about iterators from 
<http://python.org/doc/current/whatsnew/node5.html>, but the keyword 
doesn't appear to be present on my system:

Python 2.2 (#1, 07/14/02, 23:25:09)
[GCC Apple cpp-precomp 6.14] on darwin
Type "help", "copyright", "credits" or "license" for more information.
 >>> def generate_ints(N):
...  for i in range(N):
...   yield i
<string>:3: Warning: 'yield' will become a reserved keyword in the 
future
   File "<string>", line 3
     yield i
           ^
SyntaxError: invalid syntax
 >>>


Is my Python not a true 2.2 release, or do I have something wrong in my 
syntax?
For that matter a recompile wouldn't be so bad since this build doesn't 
have readline :(


Erik






--
Erik Price

email: erikprice at mac.com
jabber: erikprice at jabber.org




From mcherm at destiny.com  Wed Sep 11 09:45:10 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Wed, 11 Sep 2002 09:45:10 -0400
Subject: Multi-threading python
References: <almvvt$34p$1@news.netpower.no>
Message-ID: <alnhcu$40b$1@main.gmane.org>

David Brown wrote:
> I've just been reading a little about the GIL and threading in python, and I
> have a couple of questions that I hope someone can answer.

I'm not an expert, but I'll try.

> Am I right in thinking that threading in Python does not use the underlying
> OS's threads, but effectively implements its own scheduler?  Or does it use
> the OS's threads, but uses the GIL to ensure that only one is runable at a
> time?

It uses the OS's threads, and then uses the GIL to ensure that only one 
is executing Python bytecodes at a time. However, other threads may be 
runable... they could be blocked waiting on I/O, or executing some C 
code from a module that releases the GIL before performing a 
long-running computation.

> On a dual-processor system, does this mean that only one thread of a Python
> program can run at a time, even when both CPUs are free, so that two
> seperate python interpreter processes are needed to take advantage of the
> two CPUs?

Yes.

> What about when a Python thread is blocked for some other reason, such as
> while waiting for a file read - will that block other threads, or will the
> waiting thread definitely release the GIL and allow other threads to run?

Depends on what it's blocked for. If it's blocked for I/O, then the GIL 
is released and the other threads will run. I'm not sure what else it 
would block for that might be a problem. But you could certainly CREATE 
a problem if you tried... simplest way would be to write C code which 
ran for a long time WITHOUT releasing the GIL.

It would be much more difficult to block things through badly written 
Python code, since after each bytecode the interpretor can switch to a 
different thread.

> If the GIL does block all threads in this manner, are there any plans to
> introduce more fine-grained locking to improve scalability, analagous to
> locking in the Linux kernel?

No. The general consensus is that eliminating the GIL would be quite 
difficult and that there's not really all that much to gain. (Most 
casual users don't have multi-processor systems; serious users of this 
sort probably use Numeric and/or put the tight loops into C anyhow.) 
However, if you know someone who's good with this stuff and wants to 
volunteer, I'm sure they wouldn't turn down working code (unless it 
introduced other problems).

-- Michael Chermside






From jp at NOSPAMdemonseed.net  Wed Sep 18 12:04:10 2002
From: jp at NOSPAMdemonseed.net (jason petrone)
Date: Wed, 18 Sep 2002 16:04:10 +0000
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com>
Message-ID: <r58ama.lkf.ln@demonseed.net>

Lee Gray <lpgray at uop.com> wrote:
> Can someone point me to a good Python GUI app to demonstrate (other
> than IDLE)?

You could bring him the Star Wars: Episode II DVD:
  http://www.python.org/Quotes.html

A few other polished programs written in python:

  Visual Python: http://www.activestate.com/Products/Visual_Python/
  PySol: http://www.oberhumer.com/opensource/pysol/
  Sketch: http://sketch.sourceforge.net/
  Grail: http://grail.sourceforge.net/

Visual Python probably has the prettiest screenshot.

Jason


From jdavis at empires.org  Thu Sep 19 00:26:49 2002
From: jdavis at empires.org (Jeff Davis)
Date: Thu, 19 Sep 2002 04:26:49 GMT
Subject: Problem with urllib
References: <B9AEA35F.7FF7%jpmieville@bluewin.ch>
Message-ID: <dmci9.43613$U7.15655300@twister.socal.rr.com>

I can't reproduce the error either way (interactive or otherwise). Are you 
use that you're using the same python interpreter version for both? 

Regards,
        Jeff

Jean-Paul Mi?ville wrote:

> I don't understand why with some URL I have got [Errno 57] Socket is not
> connected.
> 
> The code is :
> 
> import urllib2
> 
> page = urllib2.urlopen(' http://www.python.org/doc/NonEnglish.html')
> print page.read()
> 
> If I do the same thing with python interactive, I can't reproduce the
> error.
> 
> Thanks in advance,
> Jean-Paul



From stevena at neosynapse.net  Sat Sep  7 16:29:15 2002
From: stevena at neosynapse.net (Steven D. Arnold)
Date: Sat, 7 Sep 2002 16:29:15 -0400
Subject: Possible bug in Email and Multifile Modules ?
Message-ID: <180174913421.20020907162915@neosynapse.net>

I'm picking up on a very old thread because I am only now encountering
issues related to it.

> On Mon, Feb 04, 2002 at 01:01:43AM -0800, Sheila King wrote:
>> >>> msg = email.message_from_file(open('virus.txt'))
>> Traceback (most recent call last):
>>   File "<interactive input>", line 1, in ?
>>   File "E:\PYTHON\PYTHON22\lib\email\__init__.py", line 35, in
>> message_from_file
>>     return _Parser(_class).parse(fp)
>>   File "E:\PYTHON\PYTHON22\lib\email\Parser.py", line 40, in parse
>>     self._parsebody(root, fp)
>>   File "E:\PYTHON\PYTHON22\lib\email\Parser.py", line 116, in _parsebody
>>     raise Errors.BoundaryError(
>> BoundaryError: Couldn't find terminating boundary:
>> ====_ABC1234567890DEF_====
>> =============(end interpreter session)================
>> 
>> Now, it turns out that this message is not well formed. Instead of the
>> correct closing boundary at the end of the message, it has a boundary which
>> indicates that another part should follow.
>> 
>> >  No parts were removed, it's 100% intact. And you're right, it's not
>> >  well-formed, but that's the whole idea behind it. It tricks Outlook
>> >  into executing the attachment.
>> > But not being well-formed shouldn't break anything...
> 
>    GIF viewer should not display EXE-files. PDF viewer should not print
> broken PDFs. And email scanner should not scan files that are not emails.
> Just drop it on the flor and forget.

I disagree.  Excellent GIF viewers should do their best to display
even broken GIF files.  Same goes for PDF, etc.  Likewise, an
excellent Python email library will do its best to handle and
correctly process broken emails.  I would say the above represents an
unfortunate prima donna attitude which ultimately reduces the utility
of Python's libraries.

In my application, I'm trying to process a message based on its
header, but I can't even get the header information, because instead
of trying to intelligently handle a broken MIME message, an exception
is raised.  Since I have no alternative but to find a way to
automatically handle even broken messages, I either have to patch the
email module to do what I want or reproduce much of the same logic in
my own module.  I'd like to patch the email module and submit the
patch for inclusion in the standard library.  What is the process for
this?  Any other suggestions or comments would be appreciated as well.

--------------------------------------------------------------
Steven D. Arnold                                    Neosynapse
stevena at neosynapse.net                        Managing Partner
AIM: abraxan                       MSN: neosynapse at hotmail.com





From op73418 at mail.telepac.pt  Thu Sep 26 11:58:44 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Thu, 26 Sep 2002 16:58:44 +0100
Subject: design question
Message-ID: <1ha6pukqdqell8f31q1k7rtor2rv5ogjs0@4ax.com>

Hi,

My problem is the following: I have a module to which I want to build a
more useful interface. An example would be a third-party extension
module where many of the classes are begging to be dressed in Python's
specific protocols (as sequences, mappings, etc.).

To make matters even more concrete, suppose that among the classes to be
wrapped we have

  Class1
     |
     |
  Class2

Where the dashed line means a parent-child relationship. Now what I want
is something like

  Wrapper1
     |
     |
  Wrapper2

where Wrapper1 (Wrapper2) wraps Class1(Class2) and the inheritance
reltionships between the several wrappers mirror those of the classes
they wrap -- but as I said, the protocol that the wrappers have can be
substantially different from the classes they wrap.

Leaving aside the question of the usefulness of doing all this, I would
like to ask how would you proceed? Making Wrapper derive from Class is
stupid since their protocols can differ widely and I would end up with
lots of diamonds in the inheritance graph. On the other hand I could
just do:

class Wrapper1(object)
    def __init__(self, <more args>):
        self.__wrap = Class1()
        <more initialization code>

But then if I did the same with Wrapper2, it would be at least quirky
since it inherits from Wrapper1 but it wraps Class1 <scratches head in
puzzled bewilderment>

I know I am being a little vague in my question, but does anyone have
any ideas? Open-source code where this "same problem" has been tackled?

With my best regards,
Gon?alo Rodrigues




From gerson.kurz at t-online.de  Sat Sep 28 03:28:16 2002
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Sat, 28 Sep 2002 07:28:16 GMT
Subject: Lambdaizing Python Code
Message-ID: <3d9559c4.14281625@news.t-online.de>

A really usefull study ... NOT. 

http://p-nand-q.com/e/quicksort_lambda.html

also known as "making Guido regret he put lambda in THE BEST
PROGRAMMING LANGUAGE IN THE WORLD(TM)".



From peter at engcorp.com  Wed Sep 11 08:33:06 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 11 Sep 2002 08:33:06 -0400
Subject: deleting the first and the last character of a string
References: <mailman.1031741971.20624.python-list@python.org> <3D7F2960.20507@Linux.ie>
Message-ID: <3d7f37ee$1@news.sentex.net>

Padraig Brady wrote:
> Manuel Hendel wrote:
> 
>> Is there anything which says delete the first or last
>> character of a string? 
[snip]
>> for line in inputfile.readlines():
>>         fields = string.split(line, "|")
>>         lineN = string.join(fields[1:-1], "|")
>>
>> Isn't there any better solution? ;-)
>>
> How about: line = line[1:-1]

Which doesn't quite work as-is for the OP, since he's using readlines()
to bring in the lines and this will delete only the trailing newline,
not the final | character.  The basic idea is sound though, but
susceptible to problems if there is any leading or trailing whitespace
as well.  The OP didn't specify whether that was possible.

-Peter



From marklists at mceahern.com  Wed Sep  4 10:48:34 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 4 Sep 2002 09:48:34 -0500
Subject: efficient list merging
In-Reply-To: <1031150751.2615.40.camel@tass1.localdomain>
Message-ID: <JHEOKEOOLIGLDHCMAHMOAEEODMAA.marklists@mceahern.com>

Question:  What if lol1 had ['g', 'f']?  Would that be considered the same
as ['f', 'g']?

Anyway, here's yet another approach:

#! /usr/bin/env python

lol1 = [['a'], ['a', 'b', 'c'], ['d', 'e', 'f'], ['g', 'h', 'i', 'j']]
lol2 = [['d', 'e', 'f'], ['k', 'l', 'm'], ['d'], ['f', 'g'], ['a']]

def mergelists(*lists):
    d = {}
    for each_list in lists:
        for item in each_list:
            key = str(item)
            if key not in d:
                d[key] = item
    ret = d.values()
    ret.sort()
    return ret

mergelol = mergelists(lol1, lol2)

expected = [['a'], ['a', 'b', 'c'], ['d'], ['d', 'e', 'f'], ['f', 'g'],
            ['g', 'h', 'i', 'j'], ['k', 'l', 'm']]

assert mergelol == expected

// m

-




From dougfort at dougfort.net  Mon Sep 23 15:07:47 2002
From: dougfort at dougfort.net (dougfort)
Date: Mon, 23 Sep 2002 12:07:47 -0700 (PDT)
Subject: Dynamic loading of modules
Message-ID: <20020923120748.18158.h006.c001.wm@mail.dougfort.net.criticalpath.net>

On Mon, 23 September 2002, "Andreas Lydersen" wrote:

> 
> Hi,
> I'd like to load a set of modules from a directory.  I have tried to use the
> __import__(...) statement to do so, but with no success.
> ?
> The static approch works (of course), but the dynamic does not!
> 
> static:
> import plugins.Password
> 
> dynamic atttempt:
> __import__(&quot;plugins.Password&quot;)
> 
> I have tried 'exec &quot;plugins.Password&quot;', and that works.  However, I
need to
> get a list of pointers to the modules I import in order to work with them
> later.
> 
> Bottom line is:  I need to know how to dynamically load modules and get a
> pointer to the same module.
> 
> Any help is appreciated.
>
Here's something my friend Joe found for a project we're working on
 
Regarding loading python "modules" from a directory... check out:
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/82234

Additional info:

Title: Importing a dynamically generated module 
Submitter: Anders Hammarquist (other recipes) 
Last Updated: 2001/10/17 
Version no: 1.0 
Category: System


Doug Fort <dougfort at dougfort.net>
http://www.dougfort.net



From b.maryniuk at forbis.lt  Mon Sep  2 06:28:58 2002
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Mon, 2 Sep 2002 12:28:58 +0200
Subject: Why it is so dramatical?
In-Reply-To: <Xns927D6DBDE418Cduncanrcpcouk@127.0.0.1>
References: <mailman.1030957743.4021.python-list@python.org> <Xns927D6DBDE418Cduncanrcpcouk@127.0.0.1>
Message-ID: <200209021228.58453.b.maryniuk@forbis.lt>

On Monday 02 September 2002 11:57, Duncan Booth wrote:
> None of this is specific to Python, the same issues arise in almost any
> programming language (although not always the same solutions).
Almost... Keep in mind, that here is the range 0-0xffffff, what is quite
bigger that 0-0xfff:
----------------------8<---------------------------
#!/usr/bin/perl

$b = '';
$data = 'AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA';
for $a (0..0xffffff) {
    $b += $data;
}
----------------------8<---------------------------

Since time() always show me zero, I've used UNIX time command:
----------------------8<---------------------------
*bo at oberon:(~/work/spike/benchmarks) time perl str_vs_list.pl 

real    0m8.734s
user    0m8.710s
sys     0m0.010s
----------------------8<---------------------------

As you see, adding string to the string isn't a huge problem for Perl as for Python.
What's "wrong" with Perl?

-- 
Regards, Bogdan

Un*x admins know what they are doing by definition.
	-- Bernd Petrovitsch




From mstenner at phy.duke.edu  Fri Sep 27 16:35:03 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Fri, 27 Sep 2002 16:35:03 -0400
Subject: watching mutables?
In-Reply-To: <an2doh$ajo3p$1@ID-59885.news.dfncis.de>; from theller@python.net on Fri, Sep 27, 2002 at 10:06:41PM +0200
References: <mailman.1033137396.14482.python-list@python.org> <an211j$oi$1@news.hccnet.nl> <Xns9296C58C9ADFCcliechtigmxnet@62.2.16.82> <k7l7qpvf.fsf@ion-tof.com> <Xns9296D3AB53777cliechtigmxnet@62.2.16.82> <mailman.1033154856.20758.python-list@python.org> <an2doh$ajo3p$1@ID-59885.news.dfncis.de>
Message-ID: <20020927163503.A17848@phy.duke.edu>

On Fri, Sep 27, 2002 at 10:06:41PM +0200, Thomas Heller wrote:
> "Michael Stenner" <mstenner at phy.duke.edu> wrote ...
> > I am also interested in this module (curiosity, mostly) and would
> > appreciate a working link if you have one.
> >
> Here's what seems to be available:
> http://oomadness.tuxfamily.org/downloads/
> and this is obviously where the EventObject is contained in:
> http://oomadness.tuxfamily.org/downloads/EditObj-0.3.tar.gz

Well, I'm not sure what makes it obvious, but it is correct :)
Thanks!
				-Michael
-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From rjones at ekit-inc.com  Wed Sep 18 20:22:40 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Thu, 19 Sep 2002 10:22:40 +1000
Subject: Python doesn't suck!
In-Reply-To: <ama9s2$e24$1@mailgate2.lexis-nexis.com>
References: <e34e8ce6.0209180737.49868d9f@posting.google.com> <ama9s2$e24$1@mailgate2.lexis-nexis.com>
Message-ID: <200209191022.40648.rjones@ekit-inc.com>

On Thu, 19 Sep 2002 2:31 am, rzed wrote:
> "Domenico" <grayshade at libero.it> wrote in message
> news:e34e8ce6.0209180737.49868d9f at posting.google.com...
>
> > http://www.theperlreview.com/at_a_glance.html
> >
> > According to the most accurate algorithm, Python is the second
> > less-sucking language.
>
> Behind PHP (?!) if I read the charts right. Says a lot about the
> languages.

Actually, that page says stuff-all. A friend wrote another version of the 
sucks-rules-ometer that filtered out a bunch of false positives and negatives 
for a bunch of different languages (Monty Python hits, for example, were 
filtered out of the Python results). IIRC Python had a significantly better 
standing than the results on the perl review's site :)


   Richard




From r.kalbermatter at hccnet.nl  Wed Sep 18 16:48:58 2002
From: r.kalbermatter at hccnet.nl (Rolf Kalbermatter)
Date: Wed, 18 Sep 2002 22:48:58 +0200
Subject: Python as shared library on Unix
Message-ID: <EKEJLGEFHHFGJOBEGJECGEIICFAA.r.kalbermatter@hccnet.nl>

I'm currently in the process of porting a Python embedding from
Windows to Linux and have come across an appearent problem.

Under Windows I have linked my app dynamically to the pythonXX.dll using
LoadLibrary/GetProcAddress and that works very nicely. Under Linux I see
no option to build a python.so.2.2.1 or similar to link my application
to with the dl* functions.

My question is if I have overlooked some option in ./configure or
somewhere else or if this is not possible out of the box? If it is
not available as a standard, what are the changes to get it working
by modifying the ./configure and or Makefile scripts?

I read in the documentation that under Windows one needs to link to
the DLL in order to have dynamic module loading available, for the
Mac OS X part there where some adjustments in the past to make sure
that sys.executable was pointing to the user executable instead of
the shared module. I could think of several such pitfals which could
show up when creating a Linux shared library.

I can of course always link the Python core into my application which
by the way is a shared library itself which is loaded dynamically
into the main application but coming from the Windows world, this
seems kind of unelegant to me.

Rolf Kalbermatter




From member at dbforums.com  Fri Sep 13 23:20:26 2002
From: member at dbforums.com (hpyhpy)
Date: Sat, 14 Sep 2002 03:20:26 +0000
Subject: Anyone can tell me why?
Message-ID: <1815200.1031973626@dbforums.com>

hi,everybody:

I'm puzzled about SWIG.

On sco openserver5,I try swig's example:

/****************************/
example.c

/* File : example.c */

/* A global variable */
double Foo = 3.0;

/* Compute the greatest common divisor of positive integers */
int gcd(int x, int y) {
int g;
g = y;
while (x > 0) {
g = x;
x = y % x;
y = g;
}
return g;
}
/****************************/
example.i

/* File : example.i */
%module example

extern int gcd(int x, int y);
extern double Foo;
/****************************/
And I do :
#gcc -c example.c
#swig -python example.i
#gcc -c example_wrap.c -I/usr/local/include/python2.2/
#gcc -G -Wl example.o example_wrap.o -o _examplecmodule.so

Then it has these files:
example.c
example.i
example.o
example_wrap.c
example_wrap.doc
example_wrap.o
_examplecmodule.so

So I wonder where is the "example.py"???
So I can't "import example"

Traceback (most recent call last):
File "", line 1, in ?
ImportError: No module named example

*_* *_* *_*

Who can tell me why,please help me,I'm a tyro^_^
Thank!

--
Posted via http://dbforums.com



From vze3myxh at verizon.net  Fri Sep 20 19:29:46 2002
From: vze3myxh at verizon.net (Satheesh Babu)
Date: Fri, 20 Sep 2002 23:29:46 GMT
Subject: pyrhon and oracle
References: <5b8834c2.0209201412.50e77861@posting.google.com>
Message-ID: <pan.2002.09.20.23.29.54.67293.2172@verizon.net>

cx_Oracle might be the *easiest* to use. I use DCOracle while working
with Zope, but for standalone development, cx_Oracle is tops.

Download the driver from
http://www.computronix.com/utilities.shtml

go to Python DB-API specs
http://www.python.org/topics/database/DatabaseAPI-2.0.html

Here's an example using DB-API
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/81189

vsb
http://vsbabu.org/

> hi , i just want to have some detail about python and oracle what ca i
> use to query oracle with python ?
> 
> i jsut want a little exemple
> 
> thanks


From gua81 at XXXyahoo.com  Wed Sep 25 08:12:24 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Wed, 25 Sep 2002 22:12:24 +1000
Subject: Strange python and cgi!
References: <amquos$rcu$1@lust.ihug.co.nz> <ajk2pukipmnmqtc662o36gnngiir552566@4ax.com> <3D9158F4.F826F039@alcyone.com>
Message-ID: <ams96h$v3i$1@lust.ihug.co.nz>

yeah in my uni's it's
/web/log/apach1.3.0/error_log
or something like that.......

anyway thanks for all who responded.

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3D9158F4.F826F039 at alcyone.com...
> Tim Roberts wrote:
>
> > If you are running Linux, check the log file at
> > /var/log/httpd/error-log.
>
> The problem is that the log location depends entirely on how Apache is
> configured, which in turn depends upon the version, the distribution
> (which often choses different locations for different log files), and
> the local configuration.  It could be /var/log/error_log,
> /var/log/errors, /var/log/apache/errors, etc.
>
> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ Music washes away from the soul the dust of everyday life.
> \__/ Berthold Auerbach
>     7 Sisters Productions / http://www.7sisters.com/
>  Web design for the future.




From fredp at mygale.org.nospam  Wed Sep 18 03:20:00 2002
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 18 Sep 2002 07:20:00 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com>
Message-ID: <Xns928D5F77DC414PaCmAnRDLM@161.48.128.20>

lpgray at uop.com (Lee Gray) said :

> I'm a fairly new programmer with experience in ASP/VBScript/DHTML.
> Other than small utilities, so far I've mostly written ASP apps to
> talk to an Oracle database on HP/UX, with Windows NT/2000 clients, and
> I've done maintenance work on VB and Oracle Forms apps.
> We're currently looking at .Net, and I'm not very impressed, for a
> variety of reasons. A little research makes me feel that either Java
> or Python would be a good alternative, and given my level of
> experience, I'd much prefer Python over Java.  I *think* I could
> actually get up to speed faster with Python than with VB and ASP.Net,
> in spite of my ASP/VBS experience.
> My boss is pretty open-minded, but also needs to see something working
> to be convinced Python is even a viable platform (he had never heard
> of it).  Otherwise, .Net is a given, whether it's any good or not,
> since corporate is MS all the way.
> Can someone point me to a good Python GUI app to demonstrate (other
> than IDLE)?

In addition to what's already been suggested, you may want to look at the 
"samples" (demos) in the PythonCard distribution. They're small apps, but 
quite demonstrative, especially considering the size of the code and ease 
of use :

http://pythoncard.sourceforge.net/

Of course, PythonCard is based on wxPython, and that has a nice showcase 
demo suite as well.

Finally, a nice (very VB-like) IDE for wxPython is Boa Constructor, 
although you'll need to pull the matest version from CVS :

http://boa-constructor.sourceforge.net/

HTH,
fp

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


From jubafre at brturbo.com  Fri Sep 13 14:14:45 2002
From: jubafre at brturbo.com (jubafre at brturbo.com)
Date: Fri, 13 Sep 2002 15:14:45 -0300 (GMT-03:00)
Subject: transform list of int in a list of string??
Message-ID: <-1483717605.1031940885180.JavaMail.nobody@webmail2.brturbo.com>

How can i transform a list of integer numbers in a list of string numbers, like that:

x=[1,2,3,4] to y=['1', '2'...]????????
	
Juliano Freitas
www.gebrasil.hpg.com.br 

From gerhard.haering at opus-gmbh.net  Tue Sep  3 09:17:44 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 3 Sep 2002 13:17:44 GMT
Subject: How to avoid freeing memory on exit?
References: <roy-54DDA9.09043103092002@reader1.panix.com>
Message-ID: <slrnan9dhv.1vs.gerhard.haering@haering.opus-gmbh.net>

Roy Smith wrote:
> [...] Is there any way to get the interpreter to just exit without doing all
> that memory cleanup?

os._exit

-- Gerhard


From Oschler at earthlink.net  Mon Sep  2 07:24:57 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Mon, 2 Sep 2002 07:24:57 -0400
Subject: Must have Python modules?
Message-ID: <_THc9.42153$vY2.1039887@e3500-atl2.usenetserver.com>

I've scanned Vaults of Parnassus and several other hot Python sites and was
overwhelmed by the amount of stuff that is available.  So I'm asking you,
what is your favorite, absolute must-have Python module?

thx






From marklists at mceahern.com  Mon Sep 30 18:22:02 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 30 Sep 2002 17:22:02 -0500
Subject: Question about __getattribute__/__getattr__
In-Reply-To: <Pine.LNX.4.44.0209301746400.12980-100000@mercury.sabren.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOGELBECAA.marklists@mceahern.com>

[Joao Prado Maia]
> My question is, how do I get the actual argument list from inside
> __getattr__, so I can create an unique method/argument list signature for
> the cached files ? :)

__getattr__ may not be the ticket.  I can sketch one possibility, but sadly
I don't have time to flesh it out fully:

When you init your Cache class, dynamically build proxy methods for each
method in the cached class.  Something like this:

  import inspect

  class Foo(object):

    def do_something(self, *args, **kwargs):
      # whatever

  class Cache(object):

    def __init__(self, cached_class):
      self._inst = cached_class()

      for m in inspect.getmembers(self._inst, inspect.ismethod):
        # m is a tuple, m[0] is the method name, m[1] is the method
        setattr(self, m[0], MemoizedFunction(m[1]))

Cheers,

// m




From marco at reimeika.ca  Wed Sep 25 19:33:06 2002
From: marco at reimeika.ca (marco)
Date: 25 Sep 2002 19:33:06 -0400
Subject: "which" command?
Message-ID: <yibznu5fxvx.fsf@coxeter.math.toronto.edu>

Hi,

A few days ago there was a discussion here about a
"which" written in python (as opposed to running
the OS "which" command). Any pointers to that thread?
As you can imagine it's been difficult to make a
sensible google groups query since I can't think
of any reasonably specific terms to search for...

Cheers,

--
marco at reimeika.ca
Gunnm: Broken Angel
http://reimeika.ca/


From duncan at NOSPAMrcp.co.uk  Thu Sep  5 04:42:24 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Thu, 5 Sep 2002 08:42:24 +0000 (UTC)
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com> <yu99u1l5bkra.fsf@europa.research.att.com> <al5t94$vse$0@216.39.172.122> <m3bs7da0zs.fsf@mira.informatik.hu-berlin.de>
Message-ID: <Xns928060E75E42Fduncanrcpcouk@127.0.0.1>

martin at v.loewis.de (Martin v. Loewis) wrote in 
news:m3bs7da0zs.fsf at mira.informatik.hu-berlin.de:

>> OTOH, if "everything is an object," then the namespace of
>> an enclosing scope could also be an object, and as such
>> could support rebinding of names therein ;-)
> 
> It is an object - the frame object. There is just no easy way to refer
> to that object.

And don't forget that when the nested function is called the frame object 
may no longer exist: nested scope variables can continue to exist after the 
frame in which they were created has been destroyed.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From david.abrahams at rcn.com  Sun Sep  8 09:42:36 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Sun, 8 Sep 2002 09:42:36 -0400
Subject: C++ and Python
References: <mailman.1031460196.14007.python-list@python.org>
Message-ID: <alfk9e$cj7$1@bob.news.rcn.net>


"Tony Leotta" <tonyl at pillarsoftware.com> wrote in message
news:mailman.1031460196.14007.python-list at python.org...
> Does any one know of any examples written in C++ that use Python as a
> scripting language?
>
> I have been looking at VBA.but I don't know all that COM interfacing
> gets me down. :(
>

You might try posting your question to c++-sig at python.org
(http://www.python.org/sigs/c++-sig/ to subscribe). I know other people on
that list have been using Python to add scripting to C++ apps.

--
-----------------------------------------------------------
           David Abrahams * Boost Consulting
dave at boost-consulting.com * http://www.boost-consulting.com




From robin at alldunn.com  Thu Sep 19 21:42:24 2002
From: robin at alldunn.com (Robin Dunn)
Date: Thu, 19 Sep 2002 18:42:24 -0700
Subject: ANNOUNCE: wxPython 2.3.3.1
Message-ID: <3D8A7D00.8030002@alldunn.com>

wxPython 2.3.3.1 has just been released!  You can get binaries for
Win32, Mac OS X or Linux, as well as sources, documentaion, etc. from
http://wxPython.org/ or directly from SourceForge at

http://sf.net/project/showfiles.php?group_id=10718&release_id=111669


wxPython is an extension module for Python that provides access to many
of the wxWindows GUI C++ classes from applications written in the Python
programming language.  It uses native controls wherever possible
providing native Look and Feel as well as the speed you would expect
from natvie compiled applications.

Major changes from the previous release are listed below.

-- 
Robin Dunn
Software Craftsman
http://wxPython.org  Java give you jitters?  Relax with wxPython!


Changes in 2.3.3.1
------------------

Added wxSplashScreen.

Added wxGenericDirCtrl.

Added wxMultiChoiceDialog.

The calltip window and autocomplete window in wxSTC will now use a
wxPopupWindow if available on the platform (and functioning correctly)
so they can extend beyond the client area of the STC if needed.

Finished wrapping and providing typemaps for wxInputStream and also
added the stream ctor and other methods for wxImage so images can now
be loaded from any Python "file-like" object.

Changed the img2py tool to use PNG instead of XPM for embedding image
data in Python source code, and the generated code now uses streams to
convert the image data to wxImage, wxBitmap, or wxIcon.

Added the wxPython.lib.rcsizer module which contains RowColSizer.
This sizer is based on code from Niki Spahiev and lets you specify a
row and column for each item, as well as optional column or row
spanning.  Cells with no item assigned to it are just left blank.
Stretchable rows or columns are specified and work the same as in
wxFlexGridSizer.

Updated XRCed from Roman Rolinsky

Added wxBufferedDC.

Upgraded wxSTC from Scintilla 1.40 to Scintilla 1.45, and then again
to version 1.47, and one more time to 1.48! <wink>

UNICODE!
     wxWindows/wxPython can be compiled with unicode support enabled or
     disabled.  Previous to wxPython 2.3.3 non-unicode mode was always
     used.  Starting with 2.3.3 either mode is supported, but only if
     it is also available in wxWindows on the platform.  Currently
     wxWindows only supports unicode on MS Windows platforms, but with
     the recent release of GTK+ 2.0 it is only a matter of time until
     it can be done on wxGTK (Linux and other unixes) as well.

     Unicode works best on platforms in the NT branch of the Windows
     family tree (NT, win2k, XP) but it is now also possible to use the
     same unicode binaries on win95/98/ME platforms as well!  This is
     done by using a special library and DLL with the application
     called MSLU, (Microsoft Layer for Unicode).  It simply gets out of
     the way if the app is run on an NT box, otherwise if run on a
     win9x box it loads a special DLL that provides the unicode
     versions of the windows API.  So far I have not been able to get
     this to work perfectly on win9x.  Most things work fine but
     wxTaskBarIcon for example will cause a crash if used with the
     unicode build on win95.

     So how do you use it?  It's very simple.  When unicode is enabled,
     then all functions and methods in wxPython that return a wxString
     from the C++ function will return a Python unicode object, and
     parameters to C++ functions/methods that expect a wxString can
     accept either a Python string or unicode object.  If a string
     object is passed then it will be decoded into unicode using the
     converter pointed to by wxConvCurrent, which will use the default
     system encoding.  If you need to use a string in some other
     encoding then you should convert it to unicode using the Python
     codecs first and then pass the unicode string to the wxPython
     method.

Added wxListCtrlAutoWidthMixin from Erik Westra.

Added wxIconBundle and wxTopLevelWindow.SetIcons.

Added wxLocale and wxEncodingConverter.

A little black magic...  When the C++ object (for a window or
whatever) is deleted there is no way to force the Python shadow object
to also be destroyed and clean up all references to it.  This leads to
crashes if the shadow object tries to call a method with the old C++
pointer.  The black magic I've done is to replace the __class__ in the
Python instance object with a class that raises an exception whenever
a method call (or other attribute access) is attempted.  This works
for any class that is OOR aware.

Added OOR support for wxGridCellRenderer, wxGridCellEditor,
wxGridCellAttr, wxGridCellAttrProvider, wxGridTableBase and their
derived classes.

Added wxImage.GetDataBuffer which returns an in-place edit buffer of
the image data.  (Patch #546009)

Added a sample that shows how to embed wxPython in a wxWindows C++
application.

Added wxPyWindow, wxPyPanel and wxPyControl which are just like their
wx counterparts except they allow some of the more common C++ virtual
methods to be overridden in Python derived classes.  The methods
supported are:

     DoMoveWindow
     DoSetSize
     DoSetClientSize
     DoSetVirtualSize
     DoGetSize
     DoGetClientSize
     DoGetPosition
     DoGetVirtualSize
     DoGetBestSize
     InitDialog
     TransferDataFromWindow
     TransferDataToWindow
     Validate
     AcceptsFocus
     AcceptsFocusFromKeyboard
     GetMaxSize
     AddChild
     RemoveChild

     If there are other methods that you think should be supported
     please let me know.

Changed wxGenButton to derive from wxPyControl and overload
DoGetBestSize and AcceptsFocus.

Added wxArtProvider.

Added wxCallAfter which is a helper function that registers a function
(or any callable Python object) to be called once the next time there
are no pending events.  This is useful for when you need to do
something but it can't be done during the current event handler.  The
implementation is very simple, see wxPython/wx.py.

Fixed a boatload of reference leaks.

Added a demo of using a sizer in a wxScrolledWindow, in effect
creating a ScrolledPanel.

Added a sample to the demo that shows how to use radio menu items, and
other menu stuff.

Added wxIEHtmlWin.  This is essentially the same as using IE with the
ActiveXWrapper already in the library, but it is implemented all in
C++ and therefore does not need any of the modules from win32all and
so it is less fragile in the face of changes.

Fixed the ActiveXWrapper problem.  Looks like when the win32com
modules make a "callback" that they (incorrectly, IMHO) allocate a
transient thread state structure.  Since wxPython is now saving
tstates for it's own callbacks it ended up using garbage after
win32com got rid of the temporary tstate...

Added a generic static text control to wxPython.lib.stattext.  This is
so things like Boa and PythonCard can have a static text that can
respond to mouse events and etc.

Changed the wxDateTime.Parse* methods to return an int that will be -1
on failure, and the index where parsing stopped otherwise.

Moved tools to be a Python package in wxPython.tools, added scripts to
import and launch each tool.  This will let you import and use the
tools in your own scripts or apps as needed.  On Linux and OS X the
tool scripts are installed to {prefix}/bin so you should be able to
easily launch them from the command line.  For example, PyCrust can be
started with just the "pycrust" command.

Added a sample to the demo that catches various key events and
displays the details of the event.

Added wxWizard, wxWizardPage, wxWizardPageSimple and wxPyWizardPage.

Added wxXmlResourceHandler which allows you to create custom handlers
for nonstandard class types in XRC resources.  See the demo for an
example.

Added wxPython.lib.mixins.rubberband module from Robb Shecter.

Added wxTimeCtrl from Will Sadkin.






From ianholmes01 at lycos.com  Thu Sep 19 10:39:12 2002
From: ianholmes01 at lycos.com (Ian Holmes)
Date: 19 Sep 2002 07:39:12 -0700
Subject: monetary applications (et al)
References: <mailman.1032383305.18515.python-list@python.org>
Message-ID: <b31eb63a.0209190639.98dd1ab@posting.google.com>

terry <tg5027 at citlink.net> wrote in message news:<mailman.1032383305.18515.python-list at python.org>...
> >>>>> Chris wrote:
> -------------------------------------------------
> class money:
> 	def __init__(self,name):
> 		self.name = name
> 
> unitprice = money(100)
> qty = 2
> amount = qty * unitprice
> 
> terry at terry:~> python money.py
> Traceback (most recent call last):
>   File "money.py", line 7, in ?
>     amount = qty * unitprice
> TypeError: unsupported operand type(s) for *: 'int' and 'instance'
> -------------------------------------------------

not that I want to do all your work for you but

>>> unitprice = money(100)
>>> qty = 2
>>> amount = qty * unitprice.name #as .name as the var that was assigned a value
>>> amount
200
>>> taxrate = 0.06
>>> taxedamount = qty * unitprice.name + (qty*unitprice.name)*taxrate
>>> taxedamount
212.0
>>>


From ben at remove.Stanford.EDU  Tue Sep 17 01:58:21 2002
From: ben at remove.Stanford.EDU (Ben Escoto)
Date: Mon, 16 Sep 2002 22:58:21 -0700
Subject: Large file support and stat vs stat64 in extension modules
References: <pan.2002.09.10.16.03.39.756929.4320@remove.Stanford.EDU>
Message-ID: <pan.2002.09.16.22.58.21.620527.4504@remove.Stanford.EDU>

On Tue, 10 Sep 2002 16:03:40 -0700, Ben Escoto wrote:

> Does anyone know how to correctly stat a file within an extension
> module? Why does:

Someone (Russ Allbery) recently helped me with this, so I thought I might
as well try answering my own question in case anyone else has this
problem.  The basic answer seems to be to define _LARGEFILE_SOURCE and
_FILE_OFFSET_BITS=64.  In fact, looking at Python's configure.in we see:

AC_DEFINE(_LARGEFILE_SOURCE)
AC_DEFINE(_FILE_OFFSET_BITS, 64)

so just compile everything with -D_LARGEFILE_SOURCE and
-D_FILE_OFFSET_BITS=64.  If using distutils, adding

define_macros=[("_LARGEFILE_SOURCE", 1), ("_FILE_OFFSET_BITS", 64)]

when declaring the ext_modules does something similar and seems to work.

    Again, this is a bit ad hoc, but seems to work, and I don't know of a
better way.

-- 
Ben Escoto


From lordmacro at hotmail.com  Mon Sep 30 04:09:54 2002
From: lordmacro at hotmail.com (Tom Widrick)
Date: Mon, 30 Sep 2002 03:09:54 -0500
Subject: sizeof PyTypeObject difference
References: <j5Rl9.29807$IO5.4027049@e3500-atl1.usenetserver.com> <m34rc8qamo.fsf@mira.informatik.hu-berlin.de>
Message-ID: <mrTl9.30085$IO5.4052423@e3500-atl1.usenetserver.com>

> > I'm trying to make a metatype in C, but there seems to be some
difference
> > between the size of the type object.
> >
> > typedef struct {
> >  PyTypeObject type;
> > } MyMetaObject;
> >
> >
> > and in the tp_basicsize field I place sizeof(MyMetaObject) the value is
only
> > 196.
>
> This is a type declararation. In C, you cannot put values into a type
> declaration. So I assume you are creating a value (an object) of
> struct MyMetaObject, and you put 196 into this value.

Well, I didn't specify but I figured it obvious the sizeof(MyMetaObject)
goes into the PyTypeObject of the metatype, i.e.

PyTypeObject MyMetaType_Type = {
 PyObject_HEAD_INIT(NULL)
 0,          /*ob_size*/
 "MyMetaType",       /*tp_name*/
 sizeof(MyMetaObject),          /*tp_basicsize*/

> If so, you misunderstand the meaning of tp_basicsize: It does not
> describe the size of the type object, but the size of instances of the
> type.

Well, aren't instances of a metatype types?  I know this statement was
probably made from invalid assumptions, but just as a point.
Since types are variable sized, as you say, how does this tie in with
tp_basicsize and metatypes?

> That is hard to believe. Declaring the typedef alone can't cause a
> core dump. Defining an object of that type, and putting the value of
> 196 into it can't cause a core dump, either. You must be doing other
> things, like using the resulting object in some way, to cause a core
> dump.

Trying to use the metatype as a metaclass for a python class.

> > What is the cause of this difference?
>
> The type type is really a variable-sized object. You are seeing the
> size of struct etype, which is hidden in typeobject.c.
>

> You can't currently declare a type with additional C fields, since
> type objects are variable-sized. One approach is to put additional
> things into the type's __dict__. Another approach is to use Christian
> Tismer's patches to overcome this limitation; I recommend that you
> contact Christian.

Thanks for the info on that one. Very helpful. Wish that was stuck in PEP
253.

As an aside, are there any other quarks about making metatypes in C? the
xxsubtype.c example in the distrib is helpful, along with the PEPs, but I'm
yet to find a good example of a metatype implemented in C (besides the 2
builtin metatypes, but I don't think those serve as  great minimalistic
examples).

Tom








From erikprice at mac.com  Thu Sep 12 23:31:12 2002
From: erikprice at mac.com (Erik Price)
Date: Thu, 12 Sep 2002 23:31:12 -0400
Subject: ANN: NINZ 1.2.1
In-Reply-To: <20020912191701.A20774@ibook.distro.conectiva>
Message-ID: <40197710-C6C9-11D6-B141-00039351FE6A@mac.com>

On Thursday, September 12, 2002, at 06:17  PM, Gustavo Niemeyer wrote:

> In particular, ZSI parses and generates SOAP messages,
> and converts between native Python datatypes and SOAP syntax.

This is sort of OT, but is this what is sometimes referred to as "data 
marshalling"?



Erik





--
Erik Price                                                  (zombies 
roam)

email: erikprice at mac.com
jabber: erikprice at jabber.org




From eric.brunel at pragmadev.com  Mon Sep  9 04:38:59 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 9 Sep 2002 10:38:59 +0200
Subject: Creating a Text widget in a Dialog  (Tkinter)
References: <b714b8de.0209081227.448fd2f2@posting.google.com>
Message-ID: <alhm8j$rd$1@wanadoo.fr>

Aki Niimura wrote:

> Hi,
> 
> I'm trying to create a simple dialog which has a Text with
> scrollbars(H and V)
> in a dialog. (I'm inherting Tkinter Dialog class)

BTW, if it's for a real application, have you considered the Pmw package? 
See:
http://pmw.sourceforge.net/doc/refindex.html
The TextDialog class already does exactly what you want.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From duncan at NOSPAMrcp.co.uk  Fri Sep 13 04:26:51 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Fri, 13 Sep 2002 08:26:51 +0000 (UTC)
Subject: wrapping yield ?
References: <alqehl$9qv$1@nntp0.reith.bbc.co.uk>
Message-ID: <Xns92885F082A21Cduncanrcpcouk@127.0.0.1>

"Michael Sparks" <Michael.Sparks at rd.bbc.co.uk> wrote in 
news:alqehl$9qv$1 at nntp0.reith.bbc.co.uk:

> I'd like to be able to wrap "yield"  in a function - is that
> possible, and if not (my eading of stuff  I've seen so far
> says not) are there any plans to make it possible?
> 
> Whilst it may seem odd, I'd like to be able to do this:
> 
> def my_generator(arg):
>     dostuff....
>     while(1):
>        waitForSomethingCheckingPeriodically()
>        processit
> 
> where
> 
> def waitForSomethingCheckingPeriodically():
>    found = 0
>    while(not found):
>        x = checkTheThing()
>        if foundRightThing(x):
>            found = 1
>       else:
>          yield "foo"
>    return x
> 

You have to propogate the yield up to the top level by yielding from each 
intermediate generator, and your generator cannot return a value so you 
should yield the returned results. So something like this should work:

def my_generator(arg):
   dostuff....
   for item in waitForSomethingCheckingPeriodically():
      if item is None:
          yield None
      else:
          processit(item)

def waitForSomethingCheckingPeriodically():
   while 1:
       x = checkTheThing()
       if foundRightThing(x):
           yield x
       else:
           yield None
 

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From jepler at unpythonic.net  Wed Sep 11 21:57:58 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Wed, 11 Sep 2002 20:57:58 -0500
Subject: ntpath bug or expected?
In-Reply-To: <xXRf9.7957$yt3.4120228@newssrv26.news.prodigy.com>
References: <xXRf9.7957$yt3.4120228@newssrv26.news.prodigy.com>
Message-ID: <20020911205757.C1555@unpythonic.net>

On Thu, Sep 12, 2002 at 01:18:21AM +0000, dsavitsk wrote:
> python 2.1.3.  this is pretty minor ... os.path.join() on win32 fails if the
> first path component is C: (or any other letter).  that is, if the first
> component is a drive letter, it must contain either '\\' or '/' or else the
> path is wrong.  this makes it the only component which must have slashes.
> 
> example
> >>> os.path.join('C:', 'path', 'to', 'file.txt')
> C:path\\to\\file.txt

Sure.  This is the file on drive C which is in the relative path
"path\\to\\file.txt" from the current directory of drive C.

Remember, DOS maintains a separate working directory for each drive.

A first element of C:/ would make the path absolute, and a first element of
C: would make the path relative to the current directory of C.  I think
that is why
    >>> ntpath.join("C:", "/path", "to", "file.txt")
    'C:/path\\to\\file.txt'
is "correct".  This is the one I wonder about, though:
    >>> ntpath.join("C:", "/a", "/b", "x.txt")
    '/b\\x.txt'
    >>> ntpath.join("C:/a", "/b", "x.txt")
    '/b\\x.txt'
if the drive letter is preserved in my first example, why isn't it in the
last two?

Jeff



From bloke at ii.net  Thu Sep 19 22:31:01 2002
From: bloke at ii.net (Rob Hall)
Date: Fri, 20 Sep 2002 10:31:01 +0800
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <96904b50.0209181149.429e8533@posting.google.com> <mailman.1032380064.3031.python-list@python.org>
Message-ID: <3d8a887f$0$31798@echo-01.iinet.net.au>

Hi.

> I've written a quite big wizard for 3D image processing
> with wxPython, PIL, Stackless.

I wouldn't mind a look - whats the address for your wizzard?

Rob




From pinard at iro.umontreal.ca  Sun Sep  1 20:45:58 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Sun, 01 Sep 2002 20:45:58 -0400
Subject: Making regex suck less
In-Reply-To: <aku8j5$3i2$2@solaris.cc.vt.edu> (Carl Banks's message of "Sun,
 1 Sep 2002 19:17:19 -0400")
References: <3d725881.345921@news.t-online.de>
	<aku8j5$3i2$2@solaris.cc.vt.edu>
Message-ID: <oqu1l92p89.fsf@titan.progiciels-bpi.ca>

[Carl Banks]

> You might want to look at the Plex package.  [...]
>
> First, this will certainly be slower than regular expressions, since
> there are many Python calls needed to build the structure.  (Of course,
> after you've compiled it, it can be as fast as regexps.)

The Plex matching engine does not backtrack, which might be an advantage over
Python regexps at matching time, at least theoretically for some regexps.
Building the matching tables, however, consumes a lot of time.  So I guess
that for most usual regexps, Python regexps are quite OK.

> Second, even if you use re module, it is still nowhere near Perl-ish
> ugliness.  You still have Python's clean syntax outside of the pattern.

A clear and definite advantage! :-) :-)

> Third, readability is not a unilateral good thing; conciseness is also
> important [...]

Python regexps, given some flag at compilation time, may allow for embedded
whitespace and comments.  With proper care, difficult regexps could be made
less compact and more readable, without changing at all how they behave at
run-time.  Conciseness is a quality for short regexps.  Horrid regexps are
more advantageously written non-compactly.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



From aleax at aleax.it  Thu Sep 26 10:03:18 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 26 Sep 2002 14:03:18 GMT
Subject: object changing itself to another object
References: <amu0lv$6ac$1@otis.netspace.net.au> <oFDk9.145692$ub2.3210552@news1.tin.it> <E9Ek9.23498$6g7.74155@news-server.bigpond.net.au>
Message-ID: <GsEk9.170959$pX1.6084628@news2.tin.it>

Glen Murphy wrote:

>> You have not checked O'Reilly's Python Cookbook, I think, because
>> this particular idea of changing an object's __class__ is exemplified
>> there.  http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/68429
>> in the online version, but I think you'll find the discussion more
>> complete, &c, in the printed version.
> 
> Thanks Alex, I have actually subscribed to the book on Safari, but it's
> kinda hard to flick through there :D

Yep -- I love Safari for certain things, but I *still* stock up on
the paper versions too:-).


Alex



From duncan at NOSPAMrcp.co.uk  Wed Sep  4 11:34:23 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Wed, 4 Sep 2002 15:34:23 +0000 (UTC)
Subject: newbie-question: difference between classmethod and staticmethod in Python2.2
References: <3D7619AE.3050605@snafu.de>
Message-ID: <Xns927FA7CBEDFB1duncanrcpcouk@127.0.0.1>

"T. Kaufmann" <merman at snafu.de> wrote in news:3D7619AE.3050605 at snafu.de:

> 
> When should I use staticmethod or classmethod in classes?
> 
> I thought a staticmethod is a classmethod like in C++/ Java?

A staticmethod is like a static method in C++/Java. A classmethod has no 
direct equivalent in C++/Java.

Use a staticmethod when you know which class you want to access as you are 
writing the code.

class C:
    def smethod(newX):
         C.x = newX    	# Update class variable
    smethod = staticmethod(smethod)

Use a classmethod if you have a class hierarchy and want the method to 
operate on the actual class used in the call rather than the class where it 
was defined:

>>> class C:
    def cmethod(cls, newX):
        cls.x = newX
    cmethod = classmethod(cmethod)

>>> class D(C): pass

>>> c = C()
>>> d = D()
>>> c.cmethod(5) # or C.cmethod(5)
>>> print c.x
5
>>> print d.x
5
>>> d.cmethod(42) # or D.cmethod(42) has the same effect.
>>> print c.x
5
>>> print d.x
42

So here we have two separate class variables, one in C and one in D but 
both are set by the same method. It is a fairly specialised and not very 
common requirement to do this. Mostly you probably want a staticmethod 
instead.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From gerhard.haering at gmx.de  Fri Sep 13 23:34:43 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 14 Sep 2002 03:34:43 GMT
Subject: Question regarding mysqldb module and variable table names
References: <45ac92e5.0209131920.23b1c22a@posting.google.com>
Message-ID: <slrnao5bik.8ni.gerhard.haering@lilith.my-fqdn.de>

Glenn Scott wrote in comp.lang.python:
>  c.execute("""INSERT INTO %s \
>         SET playerkey=%s, link=%s, snippet=%s, time_created=%s, \
>         last_seen=%s, relevancy=%s, doc_title=%s """, \
>          (infoTable, playerKey, urlLink, snippetText,
> self.getCurrentTime(), self.getCurrentTime(),\
>          noOfMentions, re.escape(str(self.doc_title))) )

sql = """INSERT INTO %s (playerkey, link, snippet, time_created,
         last_seen, relevancy, doc_title)
         VALUES (%%s, %%s, %%s, %%s, %%s, %%s, %%s)""" % infoTable

c.execute(sql, (...))

The reason why your original approach didn't work is that DB-API
quoting is different from Python-quoting. So MySQLdb quotes your table
name for SQL insertion, and adds apostrophes around it. You can use
Python quoting as above to work around this.

Also, forget about re.escape. The DB-API module will quote the values
appropriately for insertion into SQL.

-- Gerhard


From cliechti at gmx.net  Sun Sep  1 17:47:01 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 1 Sep 2002 23:47:01 +0200
Subject: Add methods to a class at runtime?
References: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com> <slrnan4dl3.18h.gerhard.haering@lilith.my-fqdn.de> <wmvc9.21757$FJ1.3181934@e3500-atl1.usenetserver.com>
Message-ID: <Xns927CF1E86FFE8cliechtigmxnet@62.2.16.82>

"Robert Oschler" <Oschler at earthlink.net> wrote in
news:wmvc9.21757$FJ1.3181934 at e3500-atl1.usenetserver.com: 
> Some interesting tidbits I
> found while trying out your suggestions,  as you will see in the two
> ".py" files that follow, I discovered that to create a method you can
> dynamically add to a class _instance_, you need to omit the 'self'
> parameter in the method declaration.  For a method you add dynamically
> to a class definition itself, then you must include the 'self' 
> parameter in the method declaration.  If not, you will get an argument
> count mismatch complaint from the interpreter.  Fascinating stuff.

that's just the differece between a "function" and a "method". a method 
works on an object and that is passed explicit as first argument. "explicit 
is better than implicit" (you can read this and some other design ideas by 
typing ">>> import this" ;-)

> I'm running 2.1.1, anybody know if 2.2.2 is going "disable" either of
> these abilities? (class or class instance dynamic method addition).  I
> saw a note about such activities requiring a "dynamic" keyword, which
> is fine.  I'd really miss either of these abilities.

nah. "classic classes" stay the way they are and are available until 3.0
new style classes, those that inherit from "object" or "type", will need 
the marking as "dynamic".

chris
-- 
Chris <cliechti at gmx.net>



From sholden at holdenweb.com  Mon Sep 30 07:58:50 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 30 Sep 2002 07:58:50 -0400
Subject: Python's import: why doesn't it work?
References: <mailman.1033322378.3697.python-list@python.org> <an83an$2qun$1@norfair.nerim.net>
Message-ID: <%VWl9.23638$F21.9224@fe02>

"Paul Pichaureau" <paul.pichaureau at nospam.laposte.net> wrote in ...
>
> "Tim Peters" <tim.one at comcast.net> a ?crit dans le message de news:
> mailman.1033322378.3697.python-list at python.org...
>
> > The docs are exactly as good as volunteers have made them (well,
actually
> > better, thanks to Fred <wink>).  When you find an explanation inadequate
> to
> > your needs, suggesting specific improvements in the right place is the
> only
> > way it will improve.
>
>   Then, these are my suggestions to improve documentation on installing
and
> managing module :
>
[reasonable suggestions]

And one of the reasons the documentation doesn't improve faster is that
apparently intelligent people, confronted with an explicit quotation of
explicit adice about how to improve it *still* appear to think that a post
of c.l.py will suffice to make a change.

It can't be stated often enough: if you want a change made, SourceForge is
the appropriate place to suggest it. In the case of the documentation, Fred
Drake is very responsive, and emailing him via python-docs at python.org
usually works (but this makes extra work for Fred, who's probably busy
enough).

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From zopestoller at thomas-guettler.de  Mon Sep 16 04:01:16 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Mon, 16 Sep 2002 10:01:16 +0200
Subject: Book for win32?
Message-ID: <am430k$9l3$03$1@news.t-online.com>

Hi!

I am interested in book about Python on win32.

I found one from o'reilly: "Python on win32"

Are there better books or alternatives?

I am *not* looking for a python introduction, I am
programming more than one year with python, but up to
now mostly on unix.

Now I am intersted in win32 stuff:
  - Opening a file-dialog
  - COM
  - Accessing the registry
  - Maybe GUI Programming

   ....

  thomas



From theller at python.net  Wed Sep  4 09:05:51 2002
From: theller at python.net (Thomas Heller)
Date: Wed, 4 Sep 2002 15:05:51 +0200
Subject: newsgroup for zope
References: <al4tru$o3j$03$1@news.t-online.com> <3d75fddc@news.sentex.net>
Message-ID: <al50fg$1nok5v$1@ID-59885.news.dfncis.de>

"Peter Hansen" <peter at engcorp.com> wrote in message news:3d75fddc at news.sentex.net...
> stephan wrote:
> > Hi perhaps a stupid question,
> >
> > but does there exist a newsgrup for zope centric items?
>                           ^^^^^^^^
> You spelled "newsgroup" wrong.  It's spelled "mailing list" and
> you can find out about them here:
>
>     http://www.zope.org/Resources/MailingLists
>
> :-)
>
> -Peter
>
> (There's no actual newsgroup though.)
>
But then there's news.gmane.org.
Read mailing lists with your newsgroup reader...

Thomas




From dsavitsk at e-coli.net  Sun Sep 15 12:51:05 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Sun, 15 Sep 2002 16:51:05 GMT
Subject: win32security.LogonUser not working
References: <x6Pg9.8439$yt3.4473830@newssrv26.news.prodigy.com> <L9Pg9.8440$yt3.4474038@newssrv26.news.prodigy.com>
Message-ID: <ZT2h9.8834$yt3.4540499@newssrv26.news.prodigy.com>

I have gone so far as to assign every possible right in dir(ntsecuritycon),
but i am still getting the same error. I have also assigned nearly everyone
and every thing on my network the right to act as the os. any ideas?

-d


"dsavitsk" <dsavitsk at e-coli.net> wrote in message
news:L9Pg9.8440$yt3.4474038 at newssrv26.news.prodigy.com...
> should have mentioned
> python:   2.1.3
> win32all: 145
> win2k:    sp2
>
>
> "dsavitsk" <dsavitsk at e-coli.net> wrote in message
> news:x6Pg9.8439$yt3.4473830 at newssrv26.news.prodigy.com...
> > I am trying to run code as another user. I copied the code below from
> ASPN,
> > so it would seem that someone has gotten this to work. I keep getting
the
> > error
> >
> > pywintypes.api_error (1314, 'LogonUser', 'A required privilege is not
held
> > by the client.')
> >
> > Google has revealed that others gave had the same problem.  does anyone
> know
> > a solution, or even a different way to do this?
> >
> > -doug
> >
> > link to the original of code below
> >
>
http://www.google.com/search?q=cache:Jnbt2XVyGfUC:aspn.activestate.com/ASPN/
> >
>
Python/Reference/Products/ActivePython/PythonWin32Extensions/Windows_NT_Secu
> >
>
rity_.2d.2d_Impersonation.html+self.handel%3Dwin32security.LogonUser&hl=en&i
> > e=UTF-8
> >
> > ---------------------------------------------
> > import win32security
> > import win32con
> > import win32api
> > import sys
> >
> >
> > class Impersonate:
> >
> >     def __init__(self):
> >         self.domain   = '.'
> >         self.login    = 'myusername'
> >         self.password = '**********'
> >
> >     def logon(self):
> >         self.handel=win32security.LogonUser(self.login,
> >                                             self.domain,
> >                                             self.password,
> >
> > win32con.LOGON32_LOGON_INTERACTIVE,
> >
> > win32con.LOGON32_PROVIDER_DEFAULT
> >                                             )
> >         win32security.ImpersonateLoggedOnUser(self.handel)
> >
> >     def logoff(self):
> >         win32security.RevertToSelf() #terminates impersonation
> >         self.handel.Close() #guarantees cleanup
> >
> >
> >
> >
> > if __name__ == '__main__':
> >     a=Impersonate()
> >     try:
> >         a.logon() #become the user
> >         print win32api.GetUserName() #show you're someone else
> >         a.logoff() #return to normal
> >     except:
> >         print sys.exc_type , sys.exc_value
> >
> >
>
>




From lordmacro at hotmail.com  Fri Sep 20 18:37:33 2002
From: lordmacro at hotmail.com (Tom Widrick)
Date: Fri, 20 Sep 2002 17:37:33 -0500
Subject: implementing reduce protocol in C
References: <3D8A7DA1.7090506@hotmail.com> <62Ai9.136323$pX1.4852975@news2.tin.it> <3D8AF51D.5040305@hotmail.com> <dkGi9.138788$pX1.4922161@news2.tin.it>
Message-ID: <UlNi9.250007$2L.9285322@e3500-atl2.usenetserver.com>

"Alex Martelli"
> I recommend you use "ANSI C" style as a habit:
>
> static PyObject*
> proto_reduce(HyProtoObject* self)

That's what I get for copy/pasting from the python source code.

> Rather, you probably want to expose a factory function to be
> passed as the first item in the tuple returned by __reduce__
> (you may then want to return a 2-items tuple rather than a
> 3-items one from __reduce__, perhaps).  Just make sure said
> first item exposes an attribute named __safe_for_unpickling__
> with a true value, or register it as a "safe constructor" with
> module copy_reg.

If it's not one thing, it's another. I decided to try this route but have
been barricaded again. I stuck a _build function in the module
method table and used copy_reg in the init function to make it a
"safe constructor."

 module = PyImport_ImportModule("copy_reg");
 if(module == NULL)
  return;

 method = PyDict_GetItemString(PyModule_GetDict(module), "constructor");
 if(method == NULL)
  return;

 func = Py_FindMethod(mod_methodlist, NULL, "_build");
 if(func == NULL)
  return;

 args = PyObject_CallFunction(method, "(O)", func);
 Py_DECREF(func);
 if(args == NULL)
  return;
 else
  Py_DECREF(args);

Now cPickle is complaining that _build is not
found as __main__._build. I tried, in my python script, to directly
stick it in the namespace, but it says it's not the same.

from proto import proto
import cPickle

Root = proto('Root')
Root.spam = 'spam'
d = cPickle.dumps(Root)

cPickle.PicklingError: Can't pickle <built-in function _build>: it's not
found a
s __main__._build

and when I put
from proto import proto, _build

cPickle.PicklingError: Can't pickle <built-in function _build>: it's not the
sam
e object as __main__._build

I was easily able to do this in pure python using copy_reg, but the reason
why
I even started to port this type to C was because it was pickling/unpickling
too slowly.

I have a feeling this has something to do with wrapping the function object
from C into python.

My next attempt will be to create a factory type, like _protobuilder, that
will
be the first element of the tuple returned by __reduce__. I have the
pickling
working, now I just have to figure out how to make _protobuilder return a
proto object.

Tom






From mertz at gnosis.cx  Mon Sep 23 14:18:47 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Mon, 23 Sep 2002 14:18:47 -0400
Subject: conditional expressions
References: <mailman.1032762750.13345.python-list@python.org>
Message-ID: <Hs1j9kKkXQhW092yn@gnosis.cx>         

"Delaney, Timothy" <tdelaney at avaya.com> wrote previously:
|This type of thinking is one very good reason why bool() is a Good Thing
|(TM). A large number (my guess is the majority) of Python users do not know
|about operator.truth().

Well...  Python users won't necessarily innately know about bool() when
it is available in 2.3 (or 2.2.x).  In either case, users will gain the
knowledge from documentation/discussion.  While I'm not against the new
boolean type, or built-in bool(), it's not impossible simply to let
users know they can write:

    from operator import truth as bool

at the top of their module.

For that matter, they could also write:

    bool = lambda x: not not x

if you really wanted to save 8 characters :-).

Yours, Lulu...

--
 mertz@  _/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: \_\_\_\_    n o
gnosis  _/_/             Postmodern Enterprises            \_\_
.cx    _/_/                                                 \_\_  d o
      _/_/_/ IN A WORLD W/O WALLS, THERE WOULD BE NO GATES \_\_\_ z e




From mdketj at dwdddaae.com  Thu Sep  5 19:45:35 2002
From: mdketj at dwdddaae.com (mdketj at dwdddaae.com)
Date: Thu, 5 Sep 2002 23:45:35 +0000 (UTC)
Subject: empresario frustrado e impotente
Message-ID: <1031269091.522331@proxynews1>

############     http://burlasleo1.cjb.net     #############

- ACT TO STOP CRIMINALS AND ABUSE OF POWER IN ENTERPRISES
- HAJA PARA PARAR COM CRIMINOSOS E ABUSO DE PODER EM EMPRESAS

############     http://burlasleo1.cjb.net     #############
sgzzkq



From david at no.westcontrol.spam.com  Fri Sep 27 08:21:51 2002
From: david at no.westcontrol.spam.com (David Brown)
Date: Fri, 27 Sep 2002 14:21:51 +0200
Subject: a python puzzle
References: <Pine.LNX.4.44.0209251736470.16806-100000@hydrogen.sabren.com> <mailman.1032991473.7292.python-list@python.org>
Message-ID: <an1i31$1r3$1@news.netpower.no>

I'm not too interested in the crypogram puzzle, but your solution here has
just given me a solution to something that has been bugging me for some
time - finding an easy way to convert a list of characters into a string:

>>> a = "Hello"
>>> b = [x for x in a]
>>> b
['H', 'e', 'l', 'l', 'o']
>>> "".join(b)
'Hello'
>>>

Thanks !

>
> def decrypt(yourstring):
>     dedict = {'F': 'S', 'I': 'C', 'H': 'I', 'a': 't', 'c': 'f', 'b': 'v',
'e': 'n',
>     'd': 'o', 'f': 's', 'i': 'c', 'h': 'i', 'k': 'p', 'j': 'h', 'm': 'l',
>     'l': 'j', 'o': 'm', 'q': 'u', 'p': 'w', 's': 'b', 'u': 'z', 't': 'd',
>     'w': 'y', 'v': 'x', 'y': 'e', 'x': 'a', 'z': 'g', 'g': 'k'}
>     return "".join([dedict.get(i,i) for i in yourstring])
>





From syver-en+usenet at online.no  Thu Sep 12 16:13:40 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Thu, 12 Sep 2002 20:13:40 GMT
Subject: embedded Python sample
References: <2a0f4cc4.0209100520.1d3a723b@posting.google.com> <mailman.1031676313.8800.python-list@python.org>
Message-ID: <uhegvx8yq.fsf@online.no>

Michael Chermside <mcherm at destiny.com> writes:

> bruno Monnet wrote:
> > Hi,
> > I'm trying to call a python script from my c++ code. All I need is
> to
> 
> > invoke the script and check the return status.
> > does anyone has a simple example ? All I found until now is very
> > complex and ugly code :-((
> > Regards,
> > Bruno
> 
> Why not just execute as a sub-process... invoke it as a shell command
> just as you'd do with "ls".
> 
> -- Michael Chermside
> 

Yes that is also an option, use 

system("python your_script.py");

-- 

Vennlig hilsen 

Syver Enstad


From aleax at aleax.it  Tue Sep 17 08:27:46 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 17 Sep 2002 12:27:46 GMT
Subject: threads or queue for this task
References: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com> <mailman.1032241763.6933.python-list@python.org> <d7Ah9.103593$ub2.2156613@news1.tin.it> <mailman.1032260122.29578.python-list@python.org>
Message-ID: <6dFh9.120102$pX1.4321019@news2.tin.it>

James J. Besemer wrote:

> I also sense that I've somehow done something to piss you off.  I'm
> sorry if I did but I can't think what it would have been.

I apologize in turn for giving that unintended impression.  My
reaction to your previous post can be well summarized with the words 
of Bjorn Pettersen's post (sent 3 minutes before mine, but I had
not seen it as I composed mine): "You must be kidding (or you can't 
have written any significant amount of multithreaded code)".

Taking it for granted that one wouldn't pontificate without vast
relevant experience, I further surmised the "You must be kidding"
part had to be correct -- e.g., that having posted an over-general
assertion (something that happens to all) you were now bloody-
mindedly defending the indefensible.  Now it appears that you're
substantially convinced of that assertion -- even though it appears
so patently false as to be indefensible to both me and Mr Pettersen --
and am left to wonder about how our experiences can be so vastly
different.  Perhaps the hint is in Mr Pettersen's mention of
"processing too many 16-GB files" vs your insistence on "capacity
to handle the expected traffic", "time-critical requests", "the
work requests may be created on a different machine", etc.

I.e., you're focusing (perhaps because of how life shaped your
experience of threading) on a "server" process or machine that
is just responding to traffic originating elsewhere, while I'm
giving at least equal weight to a process that's "just" doing a
lot of batch-like work, e.g. generated by splitting up the task
of processing a "16-GB file" or the like, and using threading
to ensure the best throughput (implies that substantial parts of
processing each work request happen in C code that release the
GIL, be it within the Python core or in extensions).  Thus, work
requests can typically be generated quite fast (just a matter
e.g. of sucking in some more of the "16-GB file" and finding
the right boundaries therein for each work-unit to request), it
IS crucial to consider the ratios of WU-preparation to
WU-processing times, the total number of requests (say 1G
requests if the average work-request deals with about 16
bytes of the 16-GB file...:-), and other such issues that you
dismiss from "the general case".

In such cases, the number of work requests is unbounded when
generating a work-request is a matter of "rolling dice" a
few times.  For example, generating a work request might mean
dealing and selecting a "suitable" bridge deal, and processing
the request might mean simulating double-dummy play over such
a deal and posting the result.  The number of possibilities
is NOT truly unlimited in such cases (there are, after all,
just 5.3E28 possible bridge deals:-) but it's generally simplest
to consider it unlimited, as it's large enough;-).

If this is the root of our disagreement, I suggest you reconsider
whether "the general case" need ONLY include "networked"
situations, which seem to be what you istinctively think of
in this context, or "threading to slice up a big batch of
work" too.  I think it definitely must cover both scenarios.
In this case, we DO want to include bounded queues as a part
of the "general case" -- and I think we should.  "Optional
embellisment": if e.g. the total number of requests won't
cause the machine to trash, bounding the queues doesn't matter.
But it's such a simple thing to do, that it can and should
be considered in most such cases.


>>>>>Generally, Queues never get "full" and you generally don't care if
        ...
> I took your point to be that a bounded queue also is sometimes useful,
> which I do not and did not dispute.

Right: therefore, your here-re-quoted assertion is false.  Queues
DO get full when you set them to be bounded, and this need not
cause substantial complication.

> It goes without saying that if the designer knows in advance that he
> doesn't have the throughput to handle incoming requests, then a

...if requests are "incoming" independently from what the
designer can do, which is not the case when the process is
dealing with a big batch of predefined work, e.g. processing
a 16-GB file...

> significantly more complicated solution is called for.  I alluded to

...but setting a Queue to bounded, and implicitly sleeping when
trying to add one more item to a full Queue, is NOT significantly
more complicated.  It's strongly parallel to sleeping just as
implicitly when trying to peel an item from an empty Queue.  The
symmetry, when applicable, may be considered more elegant than
an *asymmetric* solution, after all.

> So at a high level I am at a loss, as you seem to be picking a fight I
> did not intend to participate in.  Like I said, I thought we were
> largely in agreement, while expanding upon different scenarios.

Again I apologize for giving the impression of "picking a fight".
We can't be "largely in agreement" if you insist that your assertion:
        Generally, Queues never get "full"
was correct and helpful to the OP, while I keep believing it was
a dangerous over-simplification.  But at least we can agree to
disagree civilly, whence my apologies for appearing unwilling to
do so in my earlier post.

> In this, I am "coming from" the standpoint of Extreme Programming -- you
> may recall -- that dictates you start by writing the simplest solution
> you don't know won't work.  Performance optimization and other
> "embellishments" are always deferred unless/until you can't avoid it.

But from my POV this generally goes for *threading in general*!  I
always consider a first-cut solution WITHOUT multi-processing, which
is a huge complication in itself, more often than not.  Therefore
when I consider how to refactor an architecture for processing a
batch of work into a threaded one I'm *ALREADY* dealing with
performance optimization -- if my performance was good enough
without threading, I wouldn't be threading, I'd be using some
simpler approach (generally event-driven, if I must appear to
be doing several things at once, e.g with GUIs, networks, etc).


> That simplest, initial solution IMHO, is an unbounded queue.  I further
> characterize it as "elegant" as it's fully functional for many
> applications and it's as simple as it gets.    To me, real world
> complications, however realistic, tend to result in less "elegant"
> solutions but perhaps you disagree.

I do disagree, and this may be a philosophical point in part.  E.g.,
some Renaissance architects dreamed of "ideal cities" and went so
far as to design and even build some.  MY idea of the most elegant
city-plan produced during the Renaissance is Ferrara's, where the
architect (Rossetti) brilliantly *integrated* real-world complications
(there's a river here, a marsh there, a hill yonder, a major
thoroughfare from X to Y, ...) rather than brushing them aside or
trying to ignore them.  Indeed, Ferrara is the only "designed" (as
opposed to "organically grown") city I know where I wouldn't mind
living.  Of course, not everybody can have Rossetti's genius, but
to my mind THAT is the goal to aim for.  And in fact, in their
everyday professional practice, even Alberti and even more
strikingly Palladio *integrated* "real-world complications" to
produce their most elegant work, whatever they were saying in
their theoretical treatises (I'll never forgive them for placing
"venustas" BEFORE "utilitas", thus betraying Vitruvius' real
heritage and helping start the split between architecture and
engineering that broke the single concept of "techne" into
artificial contrasts between "art" and "technique"... but when
I consider what they actually BUILT rather than what they WROTE,
I see great engineers at work, not "would-be-pure artists"
frustrated in an abstract quest for elegance by the down-and-
dirty "real-world complications"...!).

But that's by the by.  In practice, the amount of "complication"
in setting a Python Queue instance to be bounded is trifling,
anyway.


> Yes.  If you make them too small they will overflow.

Nope!  Rather, the thread that is trying to put one more
item on an already-full Queue will implicitly and simply
sleep until a slot becomes free, exactly the same way as
one trying to pull one more ifem off an already-empty
Queue will implicitly and simply sleep until an item is
placed there.  No intrinsic complications here.

> Bounded queues are A solution to some problems but not The solution for
> all situations.

Of course not!  Never did I assert nor imply that you should
make ALL of your Queue instances bounded -- that would be
silly indeed.  But the ability to make SOME Queue's bounded,
and still handle them without fuss, IS important in a vast
enough number of cases that ruling this possibility out is
a serious didactic mistake, IMHO.

> And as I said, they push part of the problem to some other place in the
> design, which may or may not be a good thing.

In particular, when preparing work-requests is part of the
design's task (e.g. by analyzing pieces of a huge existing
file), a bounded Queue can simply and without problems do
some appropriate balancing within the system between the
amount of resources devoted to preparing work requests,
and the amount devoted to processing them.  It's as simple
as that!


>>bad way to teach: everything must be kept as simple as possible,
>>*but no simpler than that*.
> 
> Is it my turn to retort with a Monsieur de La Palice allusion?

Perhaps, but this is a straight quote from Albert Einstein, of
course, therefore you should take the matter up with him...


Alex



From aleax at aleax.it  Tue Sep 24 03:40:05 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 24 Sep 2002 07:40:05 GMT
Subject: confusion about opening files
References: <mailman.1032838112.4941.python-list@python.org>
Message-ID: <pFUj9.156068$pX1.5631353@news2.tin.it>

Arief wrote:

[ I'm putting last things first here...:-) ]
> I hope it would help you. Any correction is welcomed ...

No corrections -- just a few extra tidbits...

> The os.open is os syscall (system call). It is used to open file, fifo,
> device, etc in UNIX or Linux environment (I don't know about MS Windows).

On Windows, Visual C++'s runtime libraries simulate the Unix/Linux/BSD
system calls open / read / write / etc to some extent, basically to
ease porting programs originally written for Unix &c.


> The built in function: "open" uses c function "fopen". It is a buffered
> version of os syscall "open" implementation. You should use appropriate

You can, if you wish, open a file in unbuffered mode, too.

> function for further operation too, for example:
> 
> fname = '/tmp/a.txt'
> f = open(fname, 'w')

f = open(fname, 'w', 0)

gives you an unbuffered file object f.

> It is "not guaranteed" that as soon as you complete the f.write call, the
> file is written with data you've feed. Because it is a buffered operation.
> So to assure the OS to write your text immediately, you should do this:
> 
> fname = '/tmp/a.txt'
> f = open(fname, 'w')
> f.write('this is only example\n')
> f.flush()             # flush buffered data for previous write operation
> f.close()

No need to flush right before closing, ever -- close flushes all that
needs to be flushed.

No need to flush if you open the file in unbuffered mode.  With unbuffered
files, or frequent flushing, I/O performance may degrade.

Flush, and unbuffered files, ASK the OS to please, please write the
data out -- but not all OS's comply.  A traditional problem with DOS
and related systems (I'm pretty sure it still was in Windows/98) is
that the OS _doesn't_ comply with your kind request: if your program
crashes without closing files, the files may end up empty even though
you wrote and flushed repeatedly.  The only "cure" (HA!) to this
horrid situation was to *CLOSE AND REOPEN FOR APPEND* such files
periodically (talk about performance being degraded...!!!).  I hope
(but don't know for a fact) that there are no such horrid issues
with current OS'es such as Win/XP.


Alex



From user at domain.invalid  Fri Sep  6 18:35:55 2002
From: user at domain.invalid (ryan)
Date: Fri, 06 Sep 2002 22:35:55 GMT
Subject: old ODBC module
Message-ID: <f5ae9.2146$Qx1.35346@news1.telusplanet.net>

Does anyone know where I can find the old free odbc module? Not the 
mxODBC module, but the other (apprently not very well supported) one. 
I've been looking for a while without success. Thanks!

ryan.



From Chris.Barker at noaa.gov  Wed Sep 18 14:42:11 2002
From: Chris.Barker at noaa.gov (ChrisBarker)
Date: Wed, 18 Sep 2002 11:42:11 -0700
Subject: don't understand error message
References: <mailman.1032251060.4906.python-list@python.org> <Xns928C66E3994D7duncanrcpcouk@127.0.0.1> <mailman.1032255920.1725.python-list@python.org>
Message-ID: <amahn1$8s8$1@news.nems.noaa.gov>

Manuel Hendel wrote:
> if domainfield in domain:
>     domainlines.append(line)
> [....]
> 
> This should work if the domainfield and domain got the following
> values, shouldn't it?
> 
> domainfield = 1234.ab.cdef.de
> damain = 1234.ab.cdef.de

Nope. the "in" operator looks for an itme in a sequence. a string is a 
sequence of charactors, so you can only do something like:

if 'f' in 'a long string'

but not:

if 'this' in 'a long string'

what you want is one of the string methods:

if domain.find(domainfield) >= 0 :
	domainlines.append(line)

string.find() returns the index of the substring if it finds it, and -1 
otherwise. Since 0 is a valid index (and so is -1) you have to check if 
the index >= 0 to see if it is found.


-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov



From gerhard.haering at gmx.de  Sun Sep  1 11:47:48 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 1 Sep 2002 15:47:48 GMT
Subject: Add methods to a class at runtime?
References: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com>
Message-ID: <slrnan4dl3.18h.gerhard.haering@lilith.my-fqdn.de>

Robert Oschler wrote in comp.lang.python:
> [Glyph said:]
> "One of the eye-popping cool features of Python is the ability to
> change code and have existing instances update to use the new
> methods automatically. "
> 
> Can someone point me to a code snippet shows how to do this?

>>> class Foo: pass
... 
>>> foo = Foo()
>>> foo.bar()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: Foo instance has no attribute 'bar'
>>> def bar(self): print "bar"
... 
>>> Foo.bar = bar
>>> foo.bar()
bar
>>> 
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From skip at pobox.com  Tue Sep  3 18:36:41 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 3 Sep 2002 17:36:41 -0500
Subject: Would you like a Snobol pattern matcher? (Re: Making regex suck less)
In-Reply-To: <unad783a7o5oa7@news.supernews.com>
References: <3d725881.345921@news.t-online.de>
        <un7n06qrl1gm93@news.supernews.com>
        <3D740DFC.1040009@something.invalid>
        <un9bkdrqeh305c@news.supernews.com>
        <al30tq$aud$1@solaris.cc.vt.edu>
        <unad783a7o5oa7@news.supernews.com>
Message-ID: <15733.14713.557144.930672@12-248-11-90.client.attbi.com>

    John> One of the things Perl has right now is the ability to use spaces
    John> within regular expressions via a suffix flag. In Perl 6, this will
    John> become standard, and will include the ability to include comments!

Python also supports that with the re.VERBOSE flag.

Skip



From neal at metaslash.com  Tue Sep  3 22:27:47 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Wed, 04 Sep 2002 02:27:47 GMT
Subject: how do i know i built python --with-threads
References: <mailman.1031098984.2371.python-list@python.org>
Message-ID: <pan.2002.09.04.02.27.41.270256.2752@metaslash.com>

On Tue, 03 Sep 2002 20:22:09 -0400, Mark McEahern wrote:

> Suppose somebody else installed python on my system--e.g., my evil
> twin--and I wanted to determine whether the python was built like so:
> 
>   ./configure --with-threads --without-pymalloc
> 
> How would I do it?

For threads, you can probably do:

  >>> import sys
  >>> print 'thread' in sys.builtin_module_names

pymalloc is off by default.  The only way I can think of is to do
an nm on the executable and see if there are any pymalloc symbols:

	nm python | grep -i arenabase

Neal


From eagle at unicode.cz  Wed Sep  4 06:11:49 2002
From: eagle at unicode.cz (Jirka 'Eagle' Novak)
Date: Wed, 04 Sep 2002 12:11:49 +0200
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com> <Pine.LNX.4.44.0209041047500.13804-100000@localhost.localdomain>
Message-ID: <al4m9q$2mfh$1@ns.felk.cvut.cz>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

| I.e.: first try to make your programming needs clear, and then you can
| evaluate advice on which programming language to use.
|
| Cheers,

I'm afraid one cannot solve everything with only one language.

- --
Jirka 'Eagle' Novak
registered linux user #173581

E-Mail:   eagle at unicode.cz
ICQ:      46044619
HomePage: http://seal.unicode.cz/~eagle/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6-2 (MingW32)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAj113GQACgkQ3C/LHJbMZlKVugCgpHuScHt9IhJiljSPj3dTrMDH
Gz0An1tn20rWUGyXV+zKC0bDKjVguxk0
=ME+3
-----END PGP SIGNATURE-----



From mertz at gnosis.cx  Thu Sep 26 00:20:33 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Thu, 26 Sep 2002 00:20:33 -0400
Subject: a python puzzle
References: <mailman.1033002153.8360.python-list@python.org>
 <3D9284CF.2B1161F2@alcyone.com>
Message-ID: <Rsok9kKkXcpO092yn@gnosis.cx>         

|the fact that it was mostly Python code would have
|likely skewed the letter frequencies, since Python keywords, modules,
|and builtin names appear more frequently in Python code than general
|user-chosen identifiers; the letter frequency of the code would be
|biased against the letter frequency of the common "words" in Python,
|which is likely to be somewhat different from English as a whole.

I wonder about that.  Python reserved words (and pseudo-reserved names)
are all rather ordinary English words.  I have a hunch that their letter
distribution falls pretty close to that of English prose.

Of course, you'd have to decide how to weight things.  If you merely did
a histogram on a list of keywords, you might get a somewhat different
pattern than if you checked actual scripts (with the comments and
variable names removed).  For example, most scripts have just a few
'import's at the top, but a whole bunch of 'if's 'for's and 'in's
scattered throughout the body.

Maybe I'll try an experiment.

Yours, Lulu...

--
---[ to our friends at TLAs (spread the word) ]--------------------------
Echelon North Korea Nazi cracking spy smuggle Columbia fissionable Stego
White Water strategic Clinton Delta Force militia TEMPEST Libya Mossad
---[ Postmodern Enterprises <mertz at gnosis.cx> ]--------------------------




From jepler at unpythonic.net  Sun Sep 15 18:06:25 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Sun, 15 Sep 2002 17:06:25 -0500
Subject: Popup menu won't hide on Linux
In-Reply-To: <a3a96207.0209151220.5228a262@posting.google.com>
References: <a3a96207.0209151220.5228a262@posting.google.com>
Message-ID: <20020915170616.A8221@unpythonic.net>

Using menu.tk_popup() instead of menu.post() takes care of grab & other
details.  However, this seems to unpost the menu as soon as the button is
released.

Jeff



From manuel at hendel.net  Tue Sep 17 05:44:10 2002
From: manuel at hendel.net (Manuel Hendel)
Date: Tue, 17 Sep 2002 11:44:10 +0200
Subject: don't understand error message
In-Reply-To: <Xns928C66E3994D7duncanrcpcouk@127.0.0.1>
References: <mailman.1032251060.4906.python-list@python.org> <Xns928C66E3994D7duncanrcpcouk@127.0.0.1>
Message-ID: <20020917094410.GF23629@partagas.as.de.cw.net>

On Tue, Sep 17, 2002 at 09:12:06AM +0000, Duncan Booth wrote:
> It simply means that at least one line of your file doesn't have 4 fields 
> after you have finished your processing. If the line only has two "|" 
> characters in it, then you will only get three fields which you can access 
> with indices 0 to 2. Trying to use an index out of range for a list will 
> produce this error.
> 
> Try sticking a "print fields" line just before the if statement, the last 
> line printed will show you which line didn't match.

My script already changed a little bit, I tried to fix it by my self,
but it seams, that I just moved the error. How would you check the
lines for the number of fields?
Actually, I think I found that problem. I forgot some "---|--|--" at
the end.
But I already got a new one.

[....]
if domainfield in domain:
    domainlines.append(line)
[....]

This should work if the domainfield and domain got the following
values, shouldn't it?

domainfield = 1234.ab.cdef.de
damain = 1234.ab.cdef.de

But I get this error message:

TypeError: 'in <string>' requires character as left operand

Manuel



-- 
A beautiful man is paradise for the eyes, hell for the soul, and purgatory for 
the purse.  



From kingprad at mail.com  Wed Sep 25 11:47:40 2002
From: kingprad at mail.com (Carl)
Date: 25 Sep 2002 08:47:40 -0700
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com>
Message-ID: <668b76ce.0209250747.1b764286@posting.google.com>

I've had the same problem from time to time in my programming
explorations of the last few years. One of my interests is language,
which you mentioned as your field of study. I played for a long time
with some fun language stuff - no real scientific value, but
interesting, fun, not difficult to understand, and let me explore lots
of ways of doing things.

I decided to work on the idea that given enough time, monkeys could
randomly reproduce Shakespeare. I thought it would be fun to simulate
the monkeys statistically. First I wrote a program to read in a text
file and calculate the percentage frequency of each character, then
use random number generator to generate text. The larger the text, the
'more correct' distribution of letters (28 characters from 26 letters,
space, and apostrophe), so it would read in e-books by Poe,
Dostoevsky, etc (Shakespeare after removing plays' character-cue lines
like 'HAMLET: blah blah' so the cues wouldn't warp the frequencies.

Then I went to a second-order analysis, calculating frequencies for
each 2-letter group, 'aa, ab, ac, ..., th, ... zz' and generating from
that. That produced much more valid output, as in more english words,
better spacing and such. At third level groups (called it third order)
it produces mostly words, occasional valid short phrases and such.

Of course the interesting thing is trying different authors because
the output is recognizable as a particular author's. Similarly other
languages (german, french, etc) are recognizable in first order, but
in higher orders produce just as interesting output as teh english -
language is irrelevant, because words are built on letter frequencies.

Of course, at higher order you get sentences and all sorts of
interesting things out. The statistics can be used for
author-identification, encryption, test compression, and all sorts of
things. There is a measure of entropy in language (not real sure what
it means) that can be taken and compared as well. Really neat project
and fun programming, because it produces real results, encourages
repeated optimization tests. After I had the basic operations going I
had fun building a menu application around it with all sorts of tools,
results display, automation, etc. great stuff. this was all in QBasic,
but I'm eventually going to move it to Python. Also I have the
analysis engine written in C that I will probably call, as it is much
faster than QBasic or python for the actual text processing. but
python is fast enough for up through 4th order.

I bet you'd like something like this. It will probably give you ideas
for other projects as well.

Carl


From ruediger.maehl at web.de  Tue Sep 10 03:47:43 2002
From: ruediger.maehl at web.de (Rüdiger Mähl)
Date: 10 Sep 2002 07:47:43 GMT
Subject: Interrupting Python
References: <yz5e9.4237$rU2.145619@news4.srv.hcvlny.cv.net>
Message-ID: <Xns928563A556188ruedigermaehlwebde@193.101.67.2>

"Bob Easton" <bob at eleaston.com> wrote in
news:yz5e9.4237$rU2.145619 at news4.srv.hcvlny.cv.net: 

> I have a script that can run, accessing network resources, for several
> days. Since the script does not normally need keyed input, exception
> processing does not raise the keyboard exception until after the
> program ends normally. I would like to be able to interrupt if from
> the keyboard, but have not learned the trick.  How can I do this?

Hi Bob, 

you are perhaps looking for this:

# =====
import time
import atexit
import signal

CONTROL_C = False

def program_exit():
    # clean up before exiting if necessary
    print "Exiting application"

def ctrlCHandler(*whatever):
    global CONTROL_C
    CONTROL_C = True
    print
    print "Interrupt caught! Please wait..."
    print
    # You must check CONTROL_C in your program

signal.signal(signal.SIGINT, ctrlCHandler)

# optionally, if you need clean up actions
atexit.register(program_exit)

# do your task
while 1:
    print "Waiting for Control-C"
    time.sleep(2)
    if CONTROL_C: break
    
print "Finished"    
# =====

HTH, 

R?diger


From david.abrahams at rcn.com  Fri Sep 20 11:29:36 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Fri, 20 Sep 2002 11:29:36 -0400
Subject: [newbie] Calling Python from C++
References: <8L5i9.841$Ew4.62552@newsfep2-gui> <3D8B3EA0.6050500@thinkware.se>
Message-ID: <amffjq$ofg$1@bob.news.rcn.net>

"Magnus Lyck?" <magnus at thinkware.se> wrote in message
news:3D8B3EA0.6050500 at thinkware.se...
> Morfeas wrote:

> I'm really as clueless as you are about this, but as far
> as I understand, there are better tools than SWIG to
> connect Python with C++. SWIG is really a C thing rather
> than a C++ thing, right?
>
> Boost and CXX comes to mind. See
> http://www.thinkware.se/cgi-bin/thinki.cgi/UsingPythonWithOtherLanguages

While I appreciate the plug and I do think Boost.Python is a great solution
for C++, SWIG has evolved since I first did my analysis and it understands a
lot more of C++ than it used to, though it is still missing the ability to
instantiate templates. The Boost comparisons page is woefully outdated and
I've been waiting for David Beazley to write a new blurb for SWIG.

--
-----------------------------------------------------------
           David Abrahams * Boost Consulting
dave at boost-consulting.com * http://www.boost-consulting.com





From rgkindya at yahoo.co.in  Sat Sep  7 21:18:10 2002
From: rgkindya at yahoo.co.in (rgkindya)
Date: 7 Sep 2002 18:18:10 -0700
Subject: gettting input in a python embeded GUI app ...
Message-ID: <b2df190a.0209071718.408284f7@posting.google.com>

Dear Newser,

	I'm a newbie for python. I wish to embed Python interpreter into my
vc++ GUI app(for ms-windows only). when executing input statements (eg
raw_input()) , in pythonwin environment (activepython) , it gets input
using a new small input window, in IDLE , it gets input in its own
window (next line). please help me to do these two type of inputs in
my embeded GUI app. my app is only GUI , there is no console mode.

please debug my engliz bugs ...


From jwold at min.midco.net  Tue Sep 24 20:53:08 2002
From: jwold at min.midco.net (Jarrett L Wold)
Date: Tue, 24 Sep 2002 19:53:08 -0500
Subject: PyXML Building on OS X?
References: <B9B3CED8.3B9%jwold@min.midco.net> <j4k7ldghtp.fsf@informatik.hu-berlin.de> <B9B55A89.6D8%jwold@min.midco.net> <m3r8fk545j.fsf@mira.informatik.hu-berlin.de>
Message-ID: <up227fj7rfhs62@corp.supernews.com>

Thanks Martin!

I just opted to install YDL 2.3, I have been meaning to move to Linux for 
some time now.

I apologize for not testing that out for you, however I'm liking Linux much 
better ;)


Martin v. Loewis wrote:

> Jarrett Wold <jwold at min.midco.net> writes:
> 
>> extensions/expat/lib/xmlparse.c:3056: warning: `regparm' attribute
>> directive ignored
> 
> I see. Please edit internal.h to replace
> 
> #if defined(__GNUC__)
> #define FASTCALL __attribute__((stdcall, regparm(3)))
> #elif defined(WIN32)
> #define FASTCALL __fastcall
> #else
> #define FASTCALL
> #endif
> 
> with
> 
> #define FASTCALL
> 
>> cc -L/sw/lib -bundle -flat_namespace -undefined suppress
>> build/temp.darwin-5.5-Power Macintosh-2.2/pyexpat.o
>> build/temp.darwin-5.5-Power Macintosh-2.2/xmlparse.o
>> build/temp.darwin-5.5-Power Macintosh-2.2/xmlrole.o
>> build/temp.darwin-5.5-Power Macintosh-2.2/xmltok.o -o
>> build/lib.darwin-5.5-Power Macintosh-2.2/_xmlplus/parsers/pyexpat.so
>> -flat_namespace
>> /usr/bin/ld: -undefined: unknown argument: -lbundle1.o
>> error: command 'cc' failed with exit status 1
> 
> Very strange. Can you please invoke
> 
> cc -L/sw/lib -bundle -flat_namespace -undefined suppress
> build/temp.darwin-5.5-Power Macintosh-2.2/pyexpat.o
> build/temp.darwin-5.5-Power Macintosh-2.2/xmlparse.o
> build/temp.darwin-5.5-Power Macintosh-2.2/xmlrole.o
> build/temp.darwin-5.5-Power Macintosh-2.2/xmltok.o -o
> build/lib.darwin-5.5-Power Macintosh-2.2/_xmlplus/parsers/pyexpat.so -v
> 
> on the command line, and report the output?
> 
> Regards,
> Martin



From cbbrowne at acm.org  Sat Sep 21 09:40:34 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: 21 Sep 2002 13:40:34 GMT
Subject: Decimal arithmetic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amabpb$5rm$1@news.nems.noaa.gov> <3D8B1537.4010803@thinkware.se> <ad052e5c.0209201128.1a53108@posting.google.com> <amgcgf$5e1fd$1@ID-125932.news.dfncis.de> <7xn0qcfafp.fsf@ruckus.brouhaha.com> <amgpm4$5si22$1@ID-125932.news.dfncis.de> <7xwupf52jx.fsf@ruckus.brouhaha.com>
Message-ID: <amhssh$60681$1@ID-125932.news.dfncis.de>

A long time ago, in a galaxy far, far away, Paul Rubin <phr-n2002b at NOSPAMnightsong.com> wrote:
> Christopher Browne <cbbrowne at acm.org> writes:
>> REALITY CHECK for a moment.
>> 
>> What do you think A Real Live bank uses to compute interest?
>
> If I had to guess, I'd guess they use floating point.
>
> I do know that financial modelling software I worked on that was
> sold to banks used floating point.  Also, spreadsheets use floating
> point.
>
> But it's possible (likely even) that some bookkeeping applications
> use BCD.  One that I worked on actually used exact, multi-precision
> rational arithmetic, but I think that's unusual.

Ask yourself:

Q:  When did banks start automating their operations?

(A:  Roughly the 1960s.)

Q:  What were the major language choices at that time?

(A:  COBOL, FORTRAN, PL/1)

Q:  Did they use FORTRAN for all their applications?

(A: No, that's for the science people.  Who would do financial apps in
    FORTRAN?.)

Q:  What sorts of programmers were they looking for to do 
    Y2K remediation work about three years ago, in great
    hordes?

(A:  COBOL, PL/1)

Q: Might these questions suggest any reason to expect that they might
   have been using languages where money, which is their raison
   d'etre, might have been expressed in BCD?

Can you suggest languages that would be widely used in an MVS
environment that WOULDN'T include BCD arithmetic?  

I would expect that even C for S/390 would include a BCD arithmetic
language extension, because BCD gets used /everywhere/ on MVS-based
systems, and that is the classic OS platform of banking.
-- 
(concatenate 'string "aa454" "@freenet.carleton.ca")
http://cbbrowne.com/info/advocacy.html
"Not  only do  I  strongly recommend  against  using PC  NFS, but  I'm
willing to  bet you won't  find anyone willing  to code a free  PC NFS
client.   I  bet even  Richard  Stallman would  charge  for  a PC  NFS
client."  -- Steve Conley <sconley at regents.state.oh.us>


From news at nospam.eml.cc  Wed Sep 25 10:24:23 2002
From: news at nospam.eml.cc (Mahesh Padmanabhan)
Date: Wed, 25 Sep 2002 14:24:23 GMT
Subject: List/Tuple bug or feature?
References: <mailman.1032906212.8057.python-list@python.org>
Message-ID: <87ofamgnbh.fsf@nospam.eml.cc>

"Mark McEahern" <marklists at mceahern.com> writes:


> You don't really say much about how you generated this large list, but one
> idea would be to automate the process of generating it and/or fix whatever
> bug led to the omission of the comma.
>
> I don't have anything more clever than that to offer; perhaps someone else
> does.
>

It is simply a list of constants so I could not generate it
automatically. I then use this list to filter out one element of the
list using the builtin filter function. I don't think I can break down
the list so that I can view missing commas easier because all
constants seem to logically group together. Something like this:

constantList = [1,2,3,4,5]

filteredList = filter(some_function, list)

Thanks.

-- 
Mahesh Padmanabhan

-------

Replace nospam with prana to email me.


From marklists at mceahern.com  Sat Sep 14 16:07:24 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sat, 14 Sep 2002 15:07:24 -0500
Subject: newbie-question: abstract classes
In-Reply-To: <3D839852.2060405@snafu.de>
Message-ID: <JHEOKEOOLIGLDHCMAHMOEELHDOAA.marklists@mceahern.com>

> I want to create an abstract class but it doesn't work.

__init__ doesn't return.

// m
-




From dsavitsk at e-coli.net  Fri Sep 13 10:41:27 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Fri, 13 Sep 2002 14:41:27 GMT
Subject: Newbie Looking for direction
References: <uo23o95eb8tg87@corp.supernews.com> <b16e4ef7.0209130506.457b1db3@posting.google.com>
Message-ID: <rOmg9.8231$yt3.4316217@newssrv26.news.prodigy.com>

"Stephen Ferg" <steve at ferg.org> wrote in message
news:b16e4ef7.0209130506.457b1db3 at posting.google.com...
> > So far it seems like I'm a long way away from being able to be
productive
> > with Python...
>
> This might help in getting you productive quickly:
>
> http://www.ferg.org/easygui/

is there a license associated with this? is it public domain?





From chris.gonnerman at newcenturycomputers.net  Mon Sep 30 09:10:41 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Mon, 30 Sep 2002 08:10:41 -0500
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05> <an7bvd$burft$1@ID-125932.news.dfncis.de> <an7p7t$cpg$0@216.39.172.122> <mailman.1033338759.6775.python-list@python.org> <7xd6qwqtgz.fsf@ruckus.brouhaha.com> <mailman.1033352259.30842.python-list@python.org> <7xelbcgo56.fsf@ruckus.brouhaha.com> <mailman.1033357298.16341.python-list@python.org> <7x65woks0i.fsf@ruckus.brouhaha.com> <mailman.1033363178.31164.python-list@python.org> <7xit0oj859.fsf@ruckus.brouhaha.com>
Message-ID: <001301c26882$c8638e00$ba01010a@local>

----- Original Message ----- 
From: "Paul Rubin" <phr-n2002b at NOSPAMnightsong.com>


> "Chris Gonnerman" <chris.gonnerman at newcenturycomputers.net> writes:
> > > With decimal arithmetic, you get $2.97 the first way and
> > > $3.00 the second way.
> > 
> > Excellent example, wrong focus.  The business user (in this
> > case, the retail store owner/manager) would define which
> > method, and which results, are right.  The second mode is
> > the "canonical" way in the manual system, but lately I've
> > realized that many (most?) retail POS systems can do it 
> > "both" ways, handling single items by a variation of method 
> > 1 which gives answers consistent with method 2.
> > 
> > I'm glad I didn't have to write and debug that code.
> 
> I'm really skeptical that any "business user" can specify a 
> set of rules that gets all this stuff straight for a complex 
> business involving cents-off coupons, percentage advertising 
> discounts, etc.

Usually, in the "canned" systems I've seen, those 3 for a 
buck deals are supported by allowing three to five decimal
places rather than just two for amounts.  Results come out
much as you are expecting there.

It seems that what you are finding fault with (IMHO) is not
decimal arithmetic, it's *fixed* decimal arithmetic.  What
I am finding fault with isn't floating point arithmetic, it's
floating *binary* point arithmetic.  Later today (tonight 
probably) I plan to post a test module comparing the very 
simple "tax" calculation I've been using as an example.  My
code isn't quite clean enough to show yet, and given the 
nature of this discussion I want it real clean.

However, here are some interesting results so far:

    5% of $0.50 = 0.02 decimal, 0.03 float
    5% of $0.70 = 0.04 decimal, 0.03 float
    6% of $0.75 = 0.04 decimal, 0.05 float
    9% of $0.50 = 0.04 decimal, 0.05 float

Sell a lot of fifty-cent items in a 5% tax district, and you'll
wind up with a large discrepancy in your sales tax.  Of course,
it'll be long rather than short...  just watch out for those
seventy-five cent items.

If you sell exactly as many $0.75 as $0.50, you're OK...

In other words, it's just not good enough.

Now, if you can point me at a *floating decimal* math library,
THAT would be COOL.


Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From stephen at theboulets.net  Sat Sep  7 02:53:31 2002
From: stephen at theboulets.net (Stephen Boulet)
Date: Sat, 07 Sep 2002 01:53:31 -0500
Subject: py2exe and a single exe file?
References: <3D792351.3070008@motorola.com>
Message-ID: <unj8uoav1op87e@corp.supernews.com>

Thanks for both your replies. After thinking about it, I can live with the 
extra files. ;)

-- Stephen

I wrote:

> I'm trying to get py2exe to work for me, and I must be doing something
> wrong.
> 
> I have a script that uses easygui (which uses Tkinter) and Numeric. It
> also imports a small module from the same directory that the script
> resides in.
> 
> I try building it like this:
> 
>     python setup.py py2exe -d N:\ -i easygui
> 
> It builds and runs, but only with the other stuff in the build directory
> with the generaged exe. The other stuff is:
> 
>    tcl directory containing tcl8.3 and tk8.3 directories
>    _numpy.pyd
>    _sre.pyd
>    _tkinter.pyd
>    multiarray.pyd
>    umath.pyd
> 
> What I would like to accomplish is just having a single exe file lumping
> everything together. Is this possible?
> 
> -- Stephen



From mikkelfj-anti-spam at bigfoot.com  Sun Sep 29 07:46:55 2002
From: mikkelfj-anti-spam at bigfoot.com (MikkelFJ)
Date: Sun, 29 Sep 2002 13:46:55 +0200
Subject: 'File name' is not recognized as an internal or external command, operable program
References: <576c1752.0209280718.331fe456@posting.google.com>
Message-ID: <3d96e77f$0$69067$edfadb0f@dspool01.news.tele.dk>

"mike henley" <rsrchstr at hotmail.com> wrote in message
news:576c1752.0209280718.331fe456 at posting.google.com...
> rsrchstr at hotmail.com (mike henley) wrote in message
news:<576c1752.0209271443.1ea61518 at posting.google.com>...
> (p.s. advising the poster to start a new command
> windows for the environmental effects to take place is not a good
> reply when the poster said he already restarted the machine, but
> thanks anyway to those who suggested it for their wish to help :):)  )

Actually that wasn't stated, on the contrary it was hinted that it might
help to restart.

Mikkel




From skip at pobox.com  Tue Sep 17 10:53:18 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 17 Sep 2002 09:53:18 -0500
Subject: Modify AD key fields programmatically via ADSI or VB Script
In-Reply-To: <1824301.1032265587@dbforums.com>
References: <1824301.1032265587@dbforums.com>
Message-ID: <15751.16862.454609.277008@12-248-11-90.client.attbi.com>

    MPS> I am new to ADSI and was wondering if anybody knows a way to modify
    MPS> AD key fields (i.e./ DN) programmatically via ADSI or VB Script?
    MPS> Here is the exact scenario:

    ... [ snip ] ...

    MPS> Thanks a million for any help - MPS

Maybe post to an ADSI or VB Script group of some sort?  It's not obvious to
me that this has anything to do with Python.

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From timr at probo.com  Wed Sep 25 02:07:11 2002
From: timr at probo.com (Tim Roberts)
Date: Tue, 24 Sep 2002 23:07:11 -0700
Subject: Strange python and cgi!
References: <amquos$rcu$1@lust.ihug.co.nz>
Message-ID: <ajk2pukipmnmqtc662o36gnngiir552566@4ax.com>

"CheapSkate" <gua81 at XXXyahoo.com> wrote:

>Hi I made a cgi script in python and successfully compiled it.
>
>the script is some thing like this
>
>#! usr/lib.../python
>
>import ...
>def kickass(xxx)
>...
>
>...
>print "<pre>"
>kickass(xxx)
>print "</pre>
>...
>
>when I go to the webpage, it stops halfway and in "view source: it only show
>the code till the open "pre".
>however when I output it to an html file:
>python nameofscript.cgi >app.html
>....
>app.html works fine just as I want it to be.
>
>What seems to be the problem here.

Does "kickass" read or write any files?  Remember that the web server runs
as a special and unprivileged user.

If you are running Linux, check the log file at /var/log/httpd/error-log.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From bokr at oz.net  Mon Sep  9 19:26:48 2002
From: bokr at oz.net (Bengt Richter)
Date: 9 Sep 2002 23:26:48 GMT
Subject: a re question
References: <pan.2002.09.09.16.10.38.482129.20940@presidency.com>
Message-ID: <aljano$llk$0@216.39.172.122>

On Mon, 09 Sep 2002 16:10:39 -0400, Rajarshi Guha <rajarshi at presidency.com> wrote:

>Hi,
>  I have a file with lines of the format:
>
>001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23
>001 Xya FGh   143557789   7 100 09/05/2002 20:23:23
>
>I am trying to extract the 9 digit field and the single digit field
>immediatley after that.
>
>When I use Visual Regexp to try out the regexp 
>
>(\d{9,} {3,}\d)
>
>it highlights the 2 fields exactly. 
>
>But when I use the following Python code I get None:
>
>>> s='001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23'
>>> p = re.compile(r'(\d{9,} {3,}\d)')
>>> print p.match(s)
>>> None
>
>Could anybody point out where I'm going wrong?
>
>Thanks,

 >>> import re
 >>> s='001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23'
 >>> p = re.compile(r'(\d{9,} {3,}\d)')
 >>> print p.match(s)
 None
 >>> print p.search(s).groups()
 ('123456789   7',)

But if you want them separately,

 >>> p = re.compile(r'(\d{9,}) {3,}(\d)')
 >>> print p.search(s).groups()
 ('123456789', '7')

Or as actual integers,

 >>> map(int, p.search(s).groups())
 [123456789, 7]

match starts at the beginning of the string. See

    http://www.python.org/doc/current/lib/matching-searching.html

so for your pattern you might be able to prefix ".* " (i.e., anything ending in
a space before your 9 or more digits etc), e.g.,

 >>> p = re.compile(r'.* (\d{9,}) {3,}(\d)')
 >>> print p.match(s).groups()
 ('123456789', '7')

where s is still

 >>> s
 '001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23'

BTW, you did want to get extra digits beyond 9 and *no* extra digits
in the second single digit number, right? E.g.,

 >>> s='001 Abc D Efg 1234567890   70 100 09/05/2002 20:23:23'
                         incl--^    ^--not incl
 >>> p = re.compile(r'(\d{9,}) {3,}(\d)')
 >>> print p.search(s).groups()
 ('1234567890', '7')
            ^    ^--single digit guaranteed irresp of next

Regards,
Bengt Richter


From gtasker at allenbrook.com  Fri Sep  6 08:24:59 2002
From: gtasker at allenbrook.com (Tasker,George)
Date: Fri, 6 Sep 2002 07:24:59 -0500 
Subject: What database should I use
Message-ID: <DC65C541627ED41191FF0090279A2801017FBA2D@LOW-EXCHANGE>

MS Access works.  Fox Pro has not been tested that I know of.  If you are
using VC6 Professional/Enterprise, along with an MSDN subscription, you can
get MSDE (the stripped down version of SQL Server 7).

g

-----Original Message-----
From: A [mailto:printers at sendme.cz] 
Sent: Thursday, September 05, 2002 9:21 AM
To: python-list at python.org; tutor at python.org;
activepython at listserv.ActiveState.com; python-help at python.org
Cc: db-sig at python.org; wx-users at lists.wxwindows.org
Subject: What database should I use


Hello,
I am going to program a small application (in Python and wxPython) which
consists about 
2000 - 3000 records . Can you please recommend which database is the best
for that. I 
use Windows.
Maybe MS Access  or FoxPro or MySQL or better?
I want to keep the program as small as possible.
Thanks
Ladislav


_______________________________________________
wx-users mailing list
wx-users at lists.wxwindows.org
http://lists.wxwindows.org/mailman/listinfo/wx-users



From aleax at aleax.it  Mon Sep 23 03:08:53 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 23 Sep 2002 07:08:53 GMT
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <mailman.1031377930.4225.python-list@python.org> <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com> <3D8E6353.7030303@something.invalid>
Message-ID: <96zj9.150553$pX1.5399868@news2.tin.it>

Greg Ewing wrote:

> Courageous wrote:
> 
>> One should use spaces
>> and not tabs in Python code. Otherwise, any python file where
>> tabs and spaces are mixed is also a choice of which indentation
>> level other viewers of the code MUST use in their editors.
> 
> If you use spaces only, you're also choosing how other
> people will see it in their editors (unless they're
> willing to use a tool to re-indent it).
> 
> Using *tabs* only is the only way to get truly
> preference-independent indentation.

...unless other people use tools that do NOT let them change the
ways tabs are displayed, and that way is intolerably incompatible
with your use of tabs.

Both Outlook Express and KDE's KNode (in KDE 2.2 at least --
that's what I'm using today) are good examples of such tools
that are in everyday use by many people.  The amount of space
displayed for a leading tab is ZERO.  Just imagine how neat
this makes reading Python code posted to this newsgroup, when
the poster uses tabs there (for example because they use tabs
normally in their on-disk sources, and copy and paste is the
best way to post some snippet).

That's just one example (well, two, since KNode and OE are
completely separate programs) -- anybody who prints your
source code is similarly at the mercy of their printer's
drivers (many of which don't allow tabs to be changed), etc.

Unless you KNOW all the tools that will ever be used to look
at your source code -- which basically means, you will never
distribute your source code outside of a small group under
your strict control, never post any snippet thereof to c.l.py
to ask a question nor to aswer it, etc -- I consider using
tabs to be anti-social and a massive annoyance.  Follow the
style guide and use 4 spaces per indentation level, if you
like to think of yourself as a 'good citizen', rather than a
missionary who wants to force everybody to read news with
Emacs (or whatever other highly-flexible tool of choice).


Alex



From duduca007 at yahoo.com  Tue Sep 24 18:04:07 2002
From: duduca007 at yahoo.com (Carlos Moreira)
Date: Tue, 24 Sep 2002 15:04:07 -0700 (PDT)
Subject: PyBrenda Updates and some Problems
In-Reply-To: <20020924213528.79691.qmail@web11403.mail.yahoo.com>
Message-ID: <20020924220407.2832.qmail@web11402.mail.yahoo.com>

Ok, I found the problem again...:))))

for curious people:

Brenda.py: 285 line
host = socket.gethostbyname( socket.gethostname() )

Of course, primarily, this was for my personal use....
and before show this code for people I must to talk
with PyBrenda's author....

	[]'s

Cadu Moreira
KaduSoft President
-----------------------
[......]


--- 
Carlos Moreira <duduca007 at yahoo.com> wrote:
> Hi,
> 
> 	I did some updates in PyBrenda for my particular
> use:
> 
> Brenda.py
> old: sock.connect(host, port)
> new: sock.connect(host, port)
> 
> old: sock.bind(host, port)
> new: sock.bind((host, port))
> 
> old: command = ('rsh ' +
> self._nodelist[self._nodeIndex] + ' python ' +
> tmpFile)
> new: command = ('ssh ' +
> self._nodelist[self._nodeIndex] + ' $PYTHON/python '
> +
> tmpFile)
> 
> Well, I need real parallelism (as you can see), so,
> first of all, I performed tests/test3.py, that
> result
> in a "error message" below:
> 
> :ts got conn <niftySocket.FileSocket instance at
> 0x81b3594> from ('10.16.165.38', 34043)
> :tcr: command is ssh glover /home/pgs/carlos/py
> /home/pgs/carlos/.PyBrenda.tmp/@13190.0
> :ts: <niftySocket.FileSocket instance at 0x81b3594>
> pend (in) ('workerDone',)
> carlos at glover's password: 
> Traceback (most recent call last):
>   File "/home/pgs/carlos/.PyBrenda.tmp/@13190.0",
> line
> 14, in ?
>     c = Brenda.TupleClientRsh('isere', 2012)
>   File
>
"/home/pgs/carlos/python/Python-2.2/lib/python2.2/site-packages/Brenda.py",
> line 387, in __init__
>     TupleClientCds.__init__(self, host, port)
>   File
>
"/home/pgs/carlos/python/Python-2.2/lib/python2.2/site-packages/Brenda.py",
> line 297, in __init__
>     self._conn = self._connectTo(host, port)
>   File
>
"/home/pgs/carlos/python/Python-2.2/lib/python2.2/site-packages/Brenda.py",
> line 313, in _connectTo
>     sock.connect((host, port))
> socket.gaierror: (-2, 'Name or service not known')
> 
> I need your help. Do I some mistake? What I can do
> for
> repare this and configure PyBrenda correctly?
> 
> 	[]'s
> 
> Cadu Moreira
> KaduSoft President
> -----------------------
> [......]
> 
> __________________________________________________
> Do you Yahoo!?
> New DSL Internet Access from SBC & Yahoo!
> http://sbc.yahoo.com
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list


__________________________________________________
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
http://sbc.yahoo.com



From mcfletch at rogers.com  Thu Sep 26 17:21:39 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Thu, 26 Sep 2002 17:21:39 -0400
Subject: Silly question about ungzipping a file
References: <7396d2b2.0209261311.1fdd2d3a@posting.google.com>
Message-ID: <3D937A63.90904@rogers.com>

Try something like (untested):

chunk = 4096*1024 # 4MB chunks, play with this value to suit
myFile = gzip.open( 'blah.gz' ) # this is binary if I recall correctly
outFile = open( 'blah','wb') # note wb mode for windows machines!
data = myFile.read( chunk )
while data:
	outFile.write( data )
	data = myFile.read( chunk )
outFile.close()
myFile.close()

HTH,
Mike

Lemniscate wrote:
> Hi everyone,
> 
> This may be a ridiculously easy question, but I've kind of hit a wall
> and I was wondering if I am just missing something.  I want to
> automate the retrieval and unzipping of a *.gz file.  The issue is
> that the file, when it is unzipped, usually has a size somewhere in
> the range of 128MB.  Occassionally, I get memory errors when I try to
> run it.  Here is a quick idea of what I am doing...
> 
> 
>>>>import gzip
>>>>myFile = gzip.open('LL_tmpl.gz')
>>>>file('output.txt', 'w').write(myFile.read())
>>>>
>>>
> 
> Now, is there any way to do this so that less memory is used?  I mean,
> if I wanted to do some processing on the resulting output file, I
> would use xreadlines or something like that to keep memory consumption
> to a minimum.  Is there something roughly equivalent that I am not
> noticing in the gzip documentation.  Let me also say that I have tried
> the following as well:
> 
> 
>>>>myFile = gzip.open('LL_tmpl.gz')
>>>>fout = file('output.txt', 'w')
>>>>while myFile.readline():
>>>
> ... 	fout.write(myFile.readline())
> ... 	fout.write('\n')
> ... 	
> 
>>>>fout.close()
>>>>myFile = gzip.open('LL_tmpl.gz')
>>>>fout = file('output2.txt', 'w')
>>>>while myFile.readline():
>>>
> ... 	fout.writelines(myFile.readline())
> ... 	
> 
>>>>fout.close()
>>>
> 
> 
> These do solve my memory problem, but there are other issues.  First
> of all, my CPU gets pegged and it takes FOREVER (okay, not forever,
> but about ...let me test real quick... at least 4-5 times as long, and
> my computer is pretty much useless during that time (side note:  can
> you tell I am working on a woefully underpowered machine?)).  Is there
> something in-between that anybody can think of?  The other, and much
> more immediate, issue is puzzling to me.  It seems that the resulting
> files from the code are only about 65MB (64.7 to be exact) versus
> 129MB.  I'm sure I'm just missing something simple, but why is that? 
> Thanks for your time.
> 
> Lem

-- 
_______________________________________
   Mike C. Fletcher
   Designer, VR Plumber, Coder
   http://members.rogers.com/mcfletch/





From sebastian-huber at web.de  Fri Sep 13 01:49:47 2002
From: sebastian-huber at web.de (Sebastian Huber)
Date: Thu, 12 Sep 2002 22:49:47 -0700
Subject: Callback function call from C++
Message-ID: <200209122249.47697.sebastian-huber@web.de>

Hello,
I want to install a Python callback function in a C++ extension module. 
Everything works as desired, but is it possible to get the argument list 
length and types of the function in advance. For example:

PyCallable_Check( o) != 0
	get the arglist and return 0 if not ok
install the callback function

At the moment I get an error at the first call if the arglist is not ok, and I 
want to avoid this.

Ciao



From thoa0025 at mail.usyd.edu.au  Sun Sep 15 19:12:54 2002
From: thoa0025 at mail.usyd.edu.au (:B nerdy)
Date: Sun, 15 Sep 2002 23:12:54 GMT
Subject: raw post data
Message-ID: <Wt8h9.9583$Ee4.22928@news-server.bigpond.net.au>

how would i retrieve the raw post data in cgi python scripts ?

cheers




From anthony at computronix.com  Mon Sep 30 23:27:02 2002
From: anthony at computronix.com (Anthony Tuininga)
Date: Mon, 30 Sep 2002 21:27:02 -0600
Subject: redhat8
In-Reply-To: <87ofafxocv.fsf@flibuste.net>
References: <873crr4fy6.fsf@flibuste.net>
 <mailman.1033404219.15052.python-list@python.org> <87ofafxocv.fsf@flibuste.net>
Message-ID: <1033442823.3504.0.camel@localhost>

Yes. It is /usr/bin/python; Python 1.5 is history... :-)

On Mon, 2002-09-30 at 13:35, William wrote:
> Anthony Tuininga <anthony at computronix.com> writes:
> 
> > According to the release documentation:
> > 
> > Python 2.2.1 is on the CD.
> 
> great, but i would like to know if it's the default version, linked with
> /usr/bin/python...
> 
> -- 
> William Dod? - flibuste.net
> http://wikipython.tuxfamily.org
> -- 
> http://mail.python.org/mailman/listinfo/python-list





From park997 at aol.comnotospam  Wed Sep 18 10:53:32 2002
From: park997 at aol.comnotospam (Park997)
Date: 18 Sep 2002 14:53:32 GMT
Subject: Tkinter:select directory dialog?
Message-ID: <20020918105332.00910.00007029@mb-ch.aol.com>

Hi,
I have written an application that reads in several data files located in a
single directory. Currently the directory is hard coded in a module as a string
variable. On other computers, the data may be located in other directories. Is
there a Tkinter dialog for choosing a directory, but not picking a file to open
or write to? An entry field could be OK, but that would involve typing in a
path, and probably validity checking, etc. A nice graphical "choose a
directory"  Dialog would help. Does such a thing exist?

Thanks,

Wendell Cropper


From b.maryniuk at forbis.lt  Tue Sep  3 06:24:45 2002
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Tue, 3 Sep 2002 12:24:45 +0200
Subject: code doesn't work: question
In-Reply-To: <1551859543.20020903130305@is.lg.ua>
References: <1551859543.20020903130305@is.lg.ua>
Message-ID: <200209031224.45948.b.maryniuk@forbis.lt>

On Tuesday 03 September 2002 12:03, Ruslan Spivak wrote:
> Hello python-list user,
> I want somebody to explain me why next code doesn't work:
>
> var = 5
>
> if (tmp = var) == 5:         # this is the error line
>    print "ok"
> else:
>    print "error"

You probably want other operator than '='. e.g.:

	if (tmp + var) == 5:
		pass

Because the code:

	tmp = var

..is not a logical stuff and returns nothing.

-- 
Regards, Bogdan

"While preceding your entrance with a grenade is a good tactic in
Quake, it can lead to problems if attempted at work." -- C Hacking




From st03210 at student.uib.no  Tue Sep 17 08:26:38 2002
From: st03210 at student.uib.no (Steinar Dragsnes)
Date: Tue, 17 Sep 2002 14:26:38 +0200
Subject: File upload script
Message-ID: <3D871F7E.C313ADD2@student.uib.no>

Hi!

I'm a complete newbe in python and have never written anything more
advanced then the classical Hallo World example. I have already
programmed an applet where users should be able to upload images and or
other files from their local pc. This is of course not possible due to
the applet security model (sandbox) in Java. But I can start a
python-script from the running java applet which can do the actual file
upload for me and save the file to a specific directory on the apache
server.

Since the applet that needs the information about what file has been
uploaded into what node, I need a way to send back information to the
applet calling the script. I wonder if this is possible at all? Usually
scripts generate html-pages and call applets from there on they fly,
with the right paranemer tag values. Is this possible when the applet is
already running? Would it not destroy the applet context?

If this is the case, a workaround could be to write the output from the
script to a text-file. The information I need to write is sessionID,
userID, nodeID and file name. This is a "hairy" solution, but the most
important thing for me is a solution that works right now.

My main problem is that I do not know how I can pass the parameters of
type String; sessionID, userID and nodeID, from the appleten to the
called python-script. I need to know how to write a function that will
receive these parameters. The python script below is written by a
colleague of mine, and he is now on a sick-leave....and I need to figure
this out soon....

In the java code I call the upload.py script like this:
     java.net.URL url=new
java.net.URL("http://"+ip+"/cgi-bin/upload.py");
    applet.getAppletContext().showDocument(url,"_blank");

Below is the script that I need to rewrite so that it can take the
parameters I mentioned above:
Thanks in advance,
Steinar Dragsnes.

---------------------------
#!/usr/bin/python

import cgi
form = cgi.FieldStorage()
print 'Content-type: text/html'
print

if not form:
        print """
        <html><head><title>File Upload</title></head><body>
        <form action"/cgi-bin/upload.py" method="POST"
enctype="multipart/form-data">
        <input type="file" name="fila">
        <input type="submit" name="Send_File" value="Upload">
        </form>
        <p>
        Upload an image.
        </p>
        </body></html>
        """
elif form.has_key("fila"):
        item = form["fila"]
        if item.file:
                data = item.file.read()
                filnavn = item.filename
                filen1 = filnavn.split('\\')
                filen2 = filen1[len(filen1)-1]
                #filen2 = 'd.txt'
                open('/var/www/html/bilete/' + filen2,'wb').write(data)
                print """
                <html><head><title>File Upload</title></head><body>
                File has been uploaded.
                """
                print """
                <p>
                <applet code=UploadedFile.class
                codebase="http://tordenskrall.intermedia.uib.no/applets"

                width="200" height="200">
                <PARAM NAME=filNavn VALUE=
                """
                print cgi.escape(filen2)
                print """
                >
                </applet>
                </p>
                """
                print cgi.escape(filen2)
                print '</body></html>'

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





From martin at v.loewis.de  Mon Sep 30 01:50:28 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 30 Sep 2002 07:50:28 +0200
Subject: XMLRPC calls don't work in threads, or rather, they don't play well with others ! (help)
References: <3D97D190.9030003@bcm.tmc.edu>
Message-ID: <m38z1kqb4r.fsf@mira.informatik.hu-berlin.de>

Steve Ludtke <sludtke at bcm.tmc.edu> writes:

> Hi. I've been trying to write a python app that makes somewhat
> time-consuming XMLRPC calls. I planned to have several calls going at
> once using threads, but it seems like something in the XMLRPC client
> code isn't releasing the python interpreter lock properly. 

This is unlikely. xmlrpclib does not use threads. I don't know what
xmlrpclib_auth is, so all bets are off.

If it doesn't use any threading facilities, either, blocking calls
will occur eventually only in the socket read/write calls, which do
allow other threads to run, and do release the interpreter lock
properly.

> Is there something I can do at the Python level to deal with this
> issue ?

I'd add print statements at all levels of the code, to see where
exactly it is blocking.

Regards,
Martin



From leithp at selector.com  Sun Sep 15 11:27:01 2002
From: leithp at selector.com (Leith Parkin)
Date: Sun, 15 Sep 2002 15:27:01 GMT
Subject: regular expression help [port]
References: <qPYg9.7377$Ee4.17838@news-server.bigpond.net.au>
Message-ID: <20020916012701.3e606886.leithp@selector.com>

Here is what i use

import re
def valid(email):
        x = 
re.match("^[0-9a-z]([-_.]?[0-9a-z])*@[0-9a-z]([-.]?[0-9a-z])*\\.[a-z]{2,3
}$", email);        if x == None:
                return 0;
        else:
                return 1;

Sorry about the formatting errors

Leith

On Sun, 15 Sep 2002 09:56:38 GMT
":B nerdy" <thoa0025 at mail.usyd.edu.au> wrote:

> Greetings, this is a regular expression for an email address
> "^[a-z0-9_]+@[a-z0-9\-]{2,}\.[a-z0-9\-\.]{2,}$"
> 
> i wish to valid a variable against this string. would someone be so kind
> to supply a few lines of code to do this?
> 
> cheers
> 
> 


-- 
It is Texas law that when two trains meet each other at a railroad
crossing,
each shall come to a full stop, and neither shall proceed until the other
has gone.


From flognat at flognat.myip.org  Tue Sep 24 05:44:11 2002
From: flognat at flognat.myip.org (Andrew Markebo)
Date: Tue, 24 Sep 2002 09:44:11 GMT
Subject: Downloading huge files via urllib
References: <mailman.1032822990.23973.python-list@python.org>
Message-ID: <m3ptv34t90.fsf@flognat.myip.org>

| response = urllib.urlretrieve(serverURL, 'run.log')
| 
| However, I then get a MemoryError.  Tracking down the source of the
| error, I see the offending function in httplib:
| 
|     def makefile(self, mode, bufsize=None):
|         """Return a readable file-like object with data from socket.
[...]
|         The returned object contains *all* of the file data
|         """
| 
| I think the problem is that bufsize argument that is ignored.  Does

Read the docs, and you will see "the bufsize argument is ignored" and
"The returned object contains *all* of the file data", so it tries to
squeeze in 20 megs there.. 

| anyone know if this is correct, and what I can do about it?  I would
| like to automate the process of downloading this file, but is it
| possible?

Maybe read directly with recv() from the [Fake]Socket instead of doing
makefile.. And well this requires you probably to use httplib instead
of urllib, shouldn't be any problems I think.

Or maybe throw an eye at my implementation on the http-standard,
should be able to chew https also, http://flognat.myip.org/hacks/ it
is webification.tgz that might interest you.

        /Andy

-- 
 The eye of the beholder rests on the beauty!


From skothal at versant.com  Wed Sep 25 01:39:45 2002
From: skothal at versant.com (Sujata Kothalkar)
Date: Wed, 25 Sep 2002 11:09:45 +0530
Subject: AIX5.1 python 2.1.3 download????????
Message-ID: <3D914C21.FBC5503E@versant.com>

Hello,

Where can I find the downloadable for python 2.1.3 for AIX 5.1.
If there's no direct downloadable available, what  procedure can be
followed for building the sources and where can these sources be found?

Thanx,
--Sujata Kothalkar.




From phoebe_1 at att.net  Fri Sep 20 22:44:33 2002
From: phoebe_1 at att.net (Holden Caulfield)
Date: 20 Sep 2002 19:44:33 -0700
Subject: Solaris 64 bit compilation
Message-ID: <c2595393.0209201844.2ed62265@posting.google.com>

Greetings,
 This is my first attempt at compiling Python v2.2.1 in *64bit* mode 
on a Solaris 5.9 system. I am using the Sun Compiler (the latest I do not
remember the version). I am migrating a few of my apps (all developed on
64bit platforms) to Solaris, so I *need* the 64bit mode.

But, it is failing miserably. It compiles *fine* in 32bit mode (the default),
but it does me no good. As some of the extensions library that I have are
64bit. Before, I start hacking I thought I would give a query here and see
if anyone had any success?

Here are my preliminary results:
- Flags used "-xarch=v9 -xcode=pic32".
The first thing I get is the bail-out in "pyport.h":
 Where the #LONG_BITS != (8 * sizeof(LONG)) fails!!!

So, I commented it out, first to proceed and see what happens. Most of them
compiled fine except, I got some serious errors in the:
fileobject.c file probably related to the LP64 model.

I will try to add more info, when I get around to it. If anyone has successfully
done this, I would appreciate if they can share the info.
Thanks


From fperez528 at yahoo.com  Wed Sep 18 18:14:10 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 18 Sep 2002 16:14:10 -0600
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com>
Message-ID: <amatri$4of$1@peabody.colorado.edu>

Lee Gray wrote:

> 
> Can someone point me to a good Python GUI app to demonstrate (other
> than IDLE)?

Why don't you look at MayaVi (http://mayavi.sourceforge.net)? It's very 
impressive, and runs both under Un*x and Windows. It's a great show of 
python's strenghts: use Tk for the GUI, Vtk for the heavy lifting (3d 
rendering) and python for all the logic elsewhere which isn't super intensive 
computationally.

Cheers,

f.



From narnett at mccmedia.com  Wed Sep  4 19:08:37 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Wed, 4 Sep 2002 16:08:37 -0700
Subject: Bind threads to addresses -- Windows & urllib?
In-Reply-To: <slrnancha3.24s.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <PHEDLIMPMIAGAEPEFBGCOEFFCHAA.narnett@mccmedia.com>

It's a machine with multiple IP addresses bound to one adapter.  I want each
spider thread to use a different address...

--
Nick Arnett
Phone/fax: (408) 904-7198
narnett at mccmedia.com


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Gerhard H?ring
> Sent: Wednesday, September 04, 2002 10:41 AM
> To: python-list at python.org
> Subject: Re: Bind threads to addresses -- Windows & urllib?
>
>
> Nick Arnett wrote:
> > I can't see an obvious easy way to bind threads to IP addresses on a
> > Win2K machine.  I have a multi-threaded Web spider, which uses urllib to
> > retrieve pages, and would like to bind each thread to a
> different address
> > on the machine, if possible.
>
> I'm not sure what you mean. This is a multithreaded HTTP client
> app, right?
> Do you have multiple network connections and want to use more than one?
>
> If yes, then that's an issue of configuring the routing on your computer
> appropriately (if that is at all possible, I doubt it's easy).
>
> I'd also be surprised if you could saturate a 100 MBit link with a HTTP
> spider.
>
> -- Gerhard
> --
> http://mail.python.org/mailman/listinfo/python-list




From klausman-un080902 at tuts.net  Sun Sep  8 06:13:00 2002
From: klausman-un080902 at tuts.net (Tobias Klausmann)
Date: 8 Sep 2002 10:13:00 GMT
Subject: Python plugins
Message-ID: <alf7rc$1p06l1$1@ID-41653.news.dfncis.de>

Hi there,

For an open source project that needs to be extensible by its
users, I would like to create a plugin architecture. Mind you,
I'm not trying to use Python as a plugin language only like XChat
or Gnumeric do. Let me illustrate:

I have a directory somewhere in PYTHONPATH (let's just assume
it's in the CWD) that contains lots and lots of *.py files with a
defined set of functions (or, at least one: register()). What I'd
like to do now is being able to load these plugins at program
start and reload them if I'd like to later. Problem is that the
program has to figure out by itself what *.py files are there -
and I have to make sure somehow that I get a grip on the loaded
files.

What I tried is this: I created a subdir called "pi" to hold the
files and created a file __init__.py inside it. Code follows:

from os import listdir

modlist=[]

def rel():
 for fn in listdir(__name__):
    if fn != "__init__.py": 
        if  fn[-3:] == ".py":
            mod=fn[:-3]
            if mod not in modlist:
                exec("import %s"%mod)
                modlist.append(mod)
            else:
                exec("reload (%s)"%mod)
 return(modlist)

Now I can just "import pi" and execute pi.rel() which in turn
loads all the modules that are there ('cept the ones that only
exist as pyc or pyo, but that's okay). But now I have two
problems: first, I have no idea how to call, say
pi.someplugin.dostuff(args) except constructing each and every
call as a string and exec() it - which is both error-prone and
awkward IMHO.

Second, reloading just does not seem to work right as Python is
too smart ;) for me: it knows all of them have been loaded. So I
added the reload code above but to no avail.

For testing purposes, I created a.py ... l.py containing this:

print __name__
def register():
    return ("I work for %s%s",__name__,__name__)

I get the feeling that there is something fundamentally wrong
with my approach. What would you do?

Greets,
Tobias



From lbrannma at cablespeed.com  Sat Sep 28 13:29:44 2002
From: lbrannma at cablespeed.com (Lance)
Date: Sat, 28 Sep 2002 10:29:44 -0700
Subject: How to add recently edited files to IDLE?
Message-ID: <upbpl7n39co835@corp.supernews.com>

Hi All again,

It would also be nice to add a list of recently edited files to the File
menu in IDLE. Any suggestions would be appreciated.

Thanks,
Lance




From aglyportREMOvethispart at nospam.yahoo.com  Fri Sep  6 00:21:03 2002
From: aglyportREMOvethispart at nospam.yahoo.com (Anton Graph)
Date: Thu, 05 Sep 2002 21:21:03 -0700
Subject: Coredumps (Python 1.4)
References: <3D7579F1.1090500@nospam.yahoo.com> <mailman.1031108585.7120.python-list@python.org>
Message-ID: <3D782D2F.20603@nospam.yahoo.com>

Skip Montanaro wrote:
>     Anton> ... I coredump on this line ever since I linked with Electric
>     Anton> Fence memory debugging lib. Any ideas why? 
> 
> Not off the top of my head...
> 
>     Anton> This is with Python 1.4 on SCO. Any know issues with memory
>     Anton> allocation there? I'm building 2.2 now, but it will take a while.
> 
> Give up on 1.4.  It took me awhile (an hour or so) several months ago to get
> Python 2.1 (I think) to build on SCO.  You'll be better off solving 2.2
> build problems than struggling with what might be a bug in 1.4.
> 

The problem was in my code. I've included a header before 
#include<Python.h> that had this define in it:

#define DEBUG(a) { fprintf(stderr,"DEBUG: %s",a); }

which caused the crashes, duh. Once I removed it my program stopped 
crashing.
So, I'll keep 1.4 on SCO and simply wait for SCO to die off
instead of wasting my time trying to get 2.2 to build on SCO.



From grzegorz.rumatowskiNOSPAM at chello.at  Fri Sep 13 04:14:46 2002
From: grzegorz.rumatowskiNOSPAM at chello.at (Grzegorz Rumatowski)
Date: 13 Sep 2002 08:14:46 GMT
Subject: Question about Tutorial-Unicode
Message-ID: <Xns9288683A9F87Agrzegorzrumatowski@130.133.1.4>

Hello All!

I'm new to Python. Playing with it since some days. And last time I was 
playing with unicode strings. THen I've realized a problem. Okey, I 
thought let's look into tutorial on python.org. Looked into it and was 
wondering.
In the section "3.1.3 Unicode Strings" there is a following example:

>>> u"???"
u'\xe4\xf6\xfc'

But If I'm trying the same I'm getting:

>>> u"???"
UnicodeError: ASCII encoding error: ordinal not in range(128)

This whole with Python 2.2.1 on Win

I can't find any explanation. Mayby anyone of you?

Greetingx
Grzechu
-- 
G.R.Z.E.C.H.U.: General Robotic Zombie Engineered for Calculation and 
Hazardous Utility


From shalehperry at attbi.com  Tue Sep  3 14:33:42 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 3 Sep 2002 11:33:42 -0700
Subject: trailing newline in string.splitfields()
In-Reply-To: <Gv3d9.287721$me6.36853@sccrnsc01>
References: <Gv3d9.287721$me6.36853@sccrnsc01>
Message-ID: <200209031133.42429.shalehperry@attbi.com>

On Tuesday 03 September 2002 07:17, Micah Mayo wrote:
> I was splitting up some fields in a textfile using string.split, which
> works fine, but that it usee ' ' as the delimiter. I've found that some
> of the things I want to split include a space or two, so I've gone to
> using string.splitfields, which also works fine, but for one thing.
> Unlike string.split() it also grabs the newlne('\n') at the end of each
> line.
>
> The way I've worked around this is as such:
>
> var1,var2,var3 = string.splitfields(stringtosplit,',')
> list.append(var1)
> list.append(var2)
> list.append(var3[:-1])
>
> So, I slice of the character, easy enough. And I'm generally satisfied
> with that - but I was wondering if there may be, for the cause of more
> beautiful code, a way to tell string.splitfields() to do it automatically.
>
> Thanks,
> Micah

there is always rstrip() which removes any trailing whitespace, including the 
newline.



From kroc at blueyonder.co.uk  Mon Sep 23 17:04:08 2002
From: kroc at blueyonder.co.uk (Keith)
Date: Mon, 23 Sep 2002 21:04:08 GMT
Subject: Elementary confusion
Message-ID: <878z1s1kqf.fsf@smtp.blueyonder.co.uk>

Hi,

I am working through O'Reilly's "Learning Python". I am up to page 33,
and already I have a problem. The text outlines a cli entry about math
operators, all really straight forward;

.>>>a=3
.>>>b=4
.>>>b/2+a
 5
.>>>b/(2.0+a)
 0.8

This is so straight forward that I wasn't going to type it in, but I
did anyway, and got instead of '0.8' something I cannot account for. I
got '0.80000000000000004'.

So after 32 pages, I am already flumoxed. What is going on

-- 
Keith.


From tdelaney at avaya.com  Mon Sep 23 02:31:39 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 23 Sep 2002 16:31:39 +1000
Subject: conditional expressions
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5C7@natasha.auslabs.avaya.com>

> From: Bruce Dawson 
> 
> Basically, construct a two entry list of the two items and then index 
> that with your expression, converted to bool. The bool function is 
> essential since there was previously no generic way of getting the 
> true/false value of an expression.

Ignoring everything such as short-circuiting ...

This type of thinking is one very good reason why bool() is a Good Thing
(TM). A large number (my guess is the majority) of Python users do not know
about operator.truth().

Tim Delaney



From christophe.delord at free.fr  Wed Sep 18 17:21:01 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Wed, 18 Sep 2002 23:21:01 +0200
Subject: Is there a list comprehension to do this?
References: <H2nJ0I.36A@news.boeing.com>
Message-ID: <20020918232101.0e950ad7.christophe.delord@free.fr>

This solution is just the same as yours. The same loop, with the same
expression (xx[i+1],xx[i]) but using list comprehension.

>>> xx = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10)
>>> [ (xx[i+1], xx[i]) for i in range(0,len(xx),2) ]
[(2, 1), (4, 3), (6, 5), (8, 7), (10, 9)]
>>>

On Wed, 18 Sep 2002 20:45:58 GMT
"Joe Sixpack" <joe.sixpack at whoopie.com> wrote:

> Is there a list comprehension or other 'cleaner' or 'prettier' way
> to do the following:
> 
> >>> xx = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10)
> >>> ww = []
> >>> for i in range(0, len(xx), 2):
>  ww.append((xx[i+1],xx[i]))
> 
> >>> ww
> [(2, 1), (4, 3), (6, 5), (8, 7), (10, 9)]
> >>>
> 
> Thanks in advance...
> 
> 


-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From oren-py-l at hishome.net  Wed Sep 18 01:29:20 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Wed, 18 Sep 2002 01:29:20 -0400
Subject: monetary applications [was: Python GUI app to impress the boss?]
In-Reply-To: <E17rTMz-0006XA-00@mail.python.org>
References: <E17rTMz-0006XA-00@mail.python.org>
Message-ID: <20020918052920.GA81158@hishome.net>

On Tue, Sep 17, 2002 at 09:40:34PM -0400, terry wrote:
> As long as Python refuses to admit that money is the focus of most of 
> the world's software, Python will be relagated to the same status as 
> FORTRAN, BASIC, and etc - except COBOL and VB.
> 
> Is there any way to get the Python world to allow for the processing of 
> monetary applications without some hokey 60's constructs that involve 
> in-line conversions or anticipations of calculation characteristics?

I don't understand what exactly you are complaining about but is sounds
interesting - can you elaborate?

	Oren



From sludtke at bcm.tmc.edu  Mon Sep 30 00:22:41 2002
From: sludtke at bcm.tmc.edu (Steve Ludtke)
Date: Mon, 30 Sep 2002 04:22:41 GMT
Subject: XMLRPC calls don't work in threads, or rather, they don't play well
 with others ! (help) 
Message-ID: <3D97D190.9030003@bcm.tmc.edu>

Hi. I've been trying to write a python app that makes somewhat
time-consuming XMLRPC calls. I planned to have several calls going at
once using threads, but it seems like something in the XMLRPC client
code isn't releasing the python interpreter lock properly. That is, as
soon as the XMLRPC call is made, none of the other threads do anything
until it returns. Just to make sure this was what was happening, I wrote 
a little test thread that prints out messages every half-second, and 
indeed, as soon as an xmlrpc call is made, nothing happens. This code 
needs to be cross-platform compatible, so I'm rather leery of editing 
any of the associated C source.  Anyone have any ideas. Is there 
something I can do at the Python level to deal with this issue ?  TIA

PS - here's the code for the thread that doesn't seem to release:



class getThread(threading.Thread):
   def __init__(self,queue,login):
     threading.Thread.__init__(self)
     self.queue=queue
     self.login=login
     self.transport=xmlrpclib_auth.
       BasicAuthTransport(self.login[1],self.login[2])

   def run(self):
     while (1):
       item=self.queue.get(1)
       if (item=="EXIT") : return
       server=xmlrpclib_auth.Server(self.
         login[0]+item.idpath(),self.transport)

       try:
         lst=server.do_list()

         # processing omitted

         nv=server.values()

         # more processing omitted
       except:
         #error processing



Both of the server.()  calls prevent other threads from running



From jnana4 at DELETEhotmailCAPS.com  Sun Sep  1 19:00:43 2002
From: jnana4 at DELETEhotmailCAPS.com (jano)
Date: Sun, 01 Sep 2002 23:00:43 GMT
Subject: EOFError using cPickle.load
References: <6avc9.264595$me6.34329@sccrnsc01> <mailman.1030915575.26265.python-list@python.org>
Message-ID: <v_wc9.324279$UU1.55869@sccrnsc03>

<jepler at unpythonic.net> wrote in message
news:mailman.1030915575.26265.python-list at python.org...
> either do not use Windows, or make sure you open with 'rb' and 'wb'.
>
> Jeff
>

Thank you very much!  When it said that 'load' automatically figured out if
it was a binary file or not, i forgot about specifying 'rb' in open.

thanks,

jano




From f.marani at REMOVETHIStin.it  Sat Sep 28 10:32:47 2002
From: f.marani at REMOVETHIStin.it (Federico)
Date: Sat, 28 Sep 2002 14:32:47 GMT
Subject: Making directory in cgi!
References: <an15gg$s3h$1@lust.ihug.co.nz> <an15s8$slv$1@lust.ihug.co.nz>
Message-ID: <j4jl9.182791$pX1.6454376@news2.tin.it>

use os.mkdir() in os module
es: os.mkdir("homer/simpsons");

if the directory exists, it will probably generate an exception, look at the
manual...

"CheapSkate" <gua81 at XXXyahoo.com> ha scritto nel messaggio
news:an15s8$slv$1 at lust.ihug.co.nz...
> and one more thing......
>
> what's going to happen if let say, the directory simpsons already exist
> inside homer.
> What will happen to the files inside simpsons???
>
> Thanks.
>
> "CheapSkate" <gua81 at XXXyahoo.com> wrote in message
> news:an15gg$s3h$1 at lust.ihug.co.nz...
> > HI,
> > let say my createdir.cgi is in
> > /bug/ger/createdir.cgi
> >  and doing ls in /bug/ger/ will list out
> > .
> > ..
> > (file)createdir.cgi
> > (dir)homer
> > how do I create directory inside homer?
> > like homer/simpsons/
> > ???
> > what are the library that I need to import?
> > Thanks......
> > (I know this is a simple Q, so please excuse me... im a newbiee)
>
>




From Peer.Griebel at entire.de  Mon Sep 16 09:01:38 2002
From: Peer.Griebel at entire.de (Griebel, Peer)
Date: Mon, 16 Sep 2002 15:01:38 +0200
Subject: Mutable Stings
Message-ID: <673E8D215B3BD311A9C90090279BBB8402829E77@mail.ebssoftware.de>

Hi,

I do understand that strings are immutable. This is the reason (I think)
that the class str does not define the methods e.g. __iadd__().

Now I had a look at UserString.py. It says that the class UserString
provides a wrapper class for immutable strings. Therefore it can safely
define the method hash(). (In contrast the class MutableString raises an
exception when hash() is going to be called.)

My problem now is that UserString does provide the method __iadd__ which
does an in-place modification of the string (self.data)! This would result
in a non constant hash value!

So my question: Is this correct - or did I overlook something?




connection reset by
  Peer

-- 
Dr. Peer Griebel                     Tel: +49(0) 731 / 9 74 95-0
Dipl. Inform.                       Fax: +49(0) 731 / 9 74 95-20
Entire Software AG                 mailto:Peer.Griebel at entire.de
Pfarrer-Wei?-Weg 10 - 12
D-89077 Ulm                                 http://www.entire.de




From lists at gregfortune.com  Thu Sep 12 00:10:50 2002
From: lists at gregfortune.com (Greg Fortune)
Date: Wed, 11 Sep 2002 21:10:50 -0700
Subject: I must be an idiot, please pity me!
References: <VAOf9.2572$Pa7.155732@newsfep1-gui.server.ntli.net>
Message-ID: <3d801224$0$309$39cecf19@nnrp1.twtelecom.net>

Sure, make another module named db.  Provide a couple functions in db like 
connect and query.

Then in CGI1.py
import db
import mod1
import mod2

db.connect(host, db_name, user, pass)

#do your stuff



And in mod1
import db

def functiona():
        return db.query('SELECT * FROM some_table')


The db module would need to look something like this
#assuming you are using MySQL
import MySQLdb
THE_CURSOR = None

def connect(host, db_name, user, pass):
        global THE_CURSOR
        the_db = MySQL.connect(user=user, passwd=pass, db=db_name,
 hostname=host)
        THE_CURSOR = the_db.cursor()

def query(query_string):
        return THE_CURSOR.execute(query_string)



Of course, that doesn't take into account connecting to multiple database, 
connections sharing, closing the connection, etc...  But it should work for 
what you're trying to do.  If you'd like a more powerful db module, let me 
know and I can share some of the code I've developed over the past couple 
years.

Greg Fortune
Fortune Solutions



newbie wrote:

> Hello all. I'm getting very desperate now, so I have to ask for some help.
> I'm working on a small project and have slipped into what must be
> namespace hell.
> 
> I have a few controlling CGI scripts that use functions from other modules
> to build parts of a page. Each function in these other modules contains
> the SQL processing and HTML building code grouped by related data. The
> idea being I can reuse the functions in the other scripts. All works fine,
> providing each modules connects to the database as a stand alone job, or I
> pass a database connection object thing to each and every function. I
> suspect that both of these are wrong, due to the ugliness or nasty system
> overhead of the datbase connections.
> 
> What I thought I should be able to do is:
> 
> CGI1.py:
> 
> import mod1, mod2
> 
> connect to database
> 
> create part of page using mod1.functiona
> create part of page using mod1.functionb
> create part of page using mod2.functionz
> 
> close database
> 
> 
> But it doesn't seem to work like that. I've trawled though google and deja
> for days (plus a couple of new rider books), and no one seems to hit this
> problem. I must have a HUGE misunderstanding somewhere. I know mod1 and
> mod2 cannot see CGI.py namespace, but there must be a way of sharing the
> database connection, mustn't there?
> 
> Help or a kick up the bottom would be really appreciated here!
> 
> Sammy.



From tjreedy at udel.edu  Thu Sep 19 14:52:13 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 19 Sep 2002 18:52:13 GMT
Subject: strong/weak - dynamic/static [Was: Getting started]
References: <uojinmcfapjf45@corp.supernews.com> <tyf8z1y5cp5.fsf@pcitapi22.cern.ch> <Bqli9.36669$216.1428116@bin4.nnrp.aus1.giganews.com> <Ugmi9.114712$ub2.2441284@news1.tin.it>
Message-ID: <x1pi9.41599$216.1593993@bin4.nnrp.aus1.giganews.com>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:Ugmi9.114712$ub2.2441284 at news1.tin.it...
> Terry Reedy wrote:
...
> > Python objects are typed at creation.  Is that dynamic or static?

> Judge for yourself:
<snip>
> >>> x.__class__ = Y

I specifically *excluded* this type of assignment and the changed
behavior of new-style classes from my discussion when I wrote
"[Discussion of old and new classes and modification of
instance.__class__ omitted.]"  Classically, and in the C code, x has
type 'instance' (of a user-defined-class) both before and after the
class mod.  But this gets into the lengthy and most likely
inconclusive discussion that I intentionally avoided.

> If you find a way to name this behavior "static", you'll be well
> placed for the yearly Humpty Dumpty award.

This is pretty inane.  Again, I specifically excluded 'this behavior'
from my comments.  Anyway, my question *rhetorically* asked whether
creation-time typing (which is fixed except for the exception that I
excluded) is (to quote the OP) 'dynamic or static'.  I then suggested
that this dichotomy does not fit Python very well.

Terry J. Reedy




From yves_breemersch at hotmail.com  Mon Sep  2 09:53:39 2002
From: yves_breemersch at hotmail.com (yves)
Date: 2 Sep 2002 06:53:39 -0700
Subject: call by reference from automation objects
Message-ID: <8f2ca1f9.0209020553.7646a6c3@posting.google.com>

Hello,
I have an automation object generated from VC6:

BEGIN_DISPATCH_MAP(CSNMPManager, CCmdTarget)
//{{AFX_DISPATCH_MAP(CSNMPManager)
  DISP_FUNCTION(CSNMPManager, "Get", Get, VT_I4, VTS_I4 VTS_VARIANT
VTS_VARIANT)
//}}AFX_DISPATCH_MAP
END_DISPATCH_MAP()

long CSNMPManager::Get(long RID, const VARIANT FAR& OIDs,  const
VARIANT FAR& Result)
{
  ...
  PerformGet(RID, vaSNMPObject, vaValues, lErrorCode);
  // Copy the result 
  vaDetachedValues = vaValues.Detach();
  
  VariantClear(Result.pvarVal);
  Result.pvarVal->vt = vaDetachedValues.vt;
  Result.pvarVal->parray = vaDetachedValues.parray;

  lRetVal = lErrorCode;
  return lRetVal;
}

My Python script is loaded as code file from my C++ application and
the appropriate IDispatch pointers are supplied to the script to be
able to call the object.

When I call the method from Python the result lst stays empty:

lst=()
lstMIBObjects=["1.3.6.1.2.1.2.2.1.10.1",]
lRetval=Snmp.Manager.Get(557848457, lstMIBObjects, lst)

lst is an empty tuple although the safearray contains actual values.

As I understand it Python does not support call by reference and
parameters that are changed within a method should be returned as
multiple return values.
Does this also applies to methods called from automation objects and
do I have to encapsulate multiple return values in a safearray?

kind regards
yves


From jhorneman at pobox.com  Sun Sep 29 15:19:40 2002
From: jhorneman at pobox.com (Jurie Horneman)
Date: Sun, 29 Sep 2002 21:19:40 +0200
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com> <3d9739c7$0$19818$91cee783@newsreader01.highway.telekom.at> <3D973DA4.4040600@nospam.attbi.com>
Message-ID: <3d9751ec$0$20510$91cee783@newsreader01.highway.telekom.at>

"djw" <dwelch91 at nospam.attbi.com> wrote in message
news:3D973DA4.4040600 at nospam.attbi.com...
>
> Although I believe that the Python docs are a bit on the brief side
> sometimes, I don't think this topic has this problem... if you look at
> the package docs at:
>
> http://python.org/doc/current/tut/node8.html#SECTION008400000000000000000
>
> ... it has a nicely done example of a directory layout and how the
> __init__.py files fit into the picture. What I would fault the docs with
> is the fact that (having gone through this fustration myself) this
> information is not easily discoverable. How was I supposed to know that
> I needed a "package" in order to import a module from a subdirectory?
> I'm sure many would say, go buy a Python book and look there... but I
> did that, and the books aren't any better in my opinion in this area.
> This topic of importing modules from various directories has been my
> biggest fustration with the language... definitely a "wart" in my
> opinion. Don't get me wrong, discovering Python-and-All-Its-Splendor has
> changed my whole approach to software development (and made it fun
> again), but there are still a few things that could use some polish.
>

Well, like I said (admittedly not as succinctly as I might have) I found
that (after a while), tried it, and it didn't work. It now appears that
problem was related to my Python installation: after reinstallation it
worked fine.

And for what it's worth, if I didn't like Python I would have stayed with
C++ ;)

Jurie.





From pknotz at sandia.gov  Fri Sep 27 08:25:37 2002
From: pknotz at sandia.gov (Pat Notz)
Date: 27 Sep 2002 05:25:37 -0700
Subject: Searching for and sorting files by date and compiling stats in Windows
References: <MOKk9.66578$U7.23791943@twister.socal.rr.com> <an002g$9us6o$1@ID-71831.news.dfncis.de>
Message-ID: <525eaba.0209270425.55885f9d@posting.google.com>

"James Kew" <james.kew at btinternet.com> wrote in message news:<an002g$9us6o$1 at ID-71831.news.dfncis.de>...
> 
> os.path.walk and match a regular expression against each file in names on
> each callback.
> 

There's also the fnmatch module (and glob) for shell like file name matching...


From lutz at lutz-horn.de  Tue Sep 17 04:13:29 2002
From: lutz at lutz-horn.de (Lutz Horn)
Date: 17 Sep 2002 10:13:29 +0200
Subject: Best way ( easiest ) to generate RSS/RDF on MySQL-data ??
References: <oJsh9.18870$0p1.344943@news2.ulv.nextra.no>
Message-ID: <wklm61f2za.fsf@ID-32301.user.dfncis.de>

"Thomas Weholt" <2002 at weholt.org> writes:

> Any hints or clues on how to generate RSS/RDF-files, not even
> related to this problem in particular would be highly appreciated.

Do you know about the Redland library[1]? It's written in C but has a
Python API.

Regards
Lutz

[1] http://www.redland.opensource.ac.uk/


From jb at cascade-sys.com  Fri Sep 20 08:32:36 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Fri, 20 Sep 2002 05:32:36 -0700
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <bd751532.0209082142.54236307@posting.google.com> <allrrs$6hm$2@solaris.cc.vt.edu> <ambpgq$dre$1@nntp9.atl.mindspring.net> <006501c25fd5$ecc80740$0101010a@local> <3D8AAC3C.9020300@cascade-sys.com> <003001c2609e$04d0fea0$0101010a@local>
Message-ID: <3D8B1564.5000507@cascade-sys.com>

Chris Gonnerman wrote:

>----- Original Message ----- 
>From: "James J. Besemer" <jb at cascade-sys.com>
>
>  
>
>>They are a hell of an improvement over "begin" and "end".
>>    
>>
>
>I disagree slightly... I don't have to shift for the literal
>"begin" and "end" statements, and I don't have to reach either.
>I'm left-handed; my right pinky is probably my weakest finger.
>So, while I don't really *like* "begin" and "end" as in 
>Pascal, I rarely curse while typing them.
>
Ah.  I took a typing class in high school and in retrospect it was one 
of the most useful classes from that period in my life.

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From irmen at NOSPAMREMOVETHISxs4all.nl  Tue Sep 24 16:30:23 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Tue, 24 Sep 2002 22:30:23 +0200
Subject: Pyro Error.
References: <mailman.1032889706.24130.python-list@python.org>
Message-ID: <amqi0v$77g$1@news1.xs4all.nl>

Carlos Moreira wrote:
> Testing the simplest example of docs/8-example.html,
> on client side, generate an error message:
[...]
> Pyro.errors.ProtocolError: connection failed
> 
> Where I must seek for inconsistences? Pyro.conf?

Nope it should work without any configuration file.
You're not giving any details so I cannot say what could
be wrong. For instance, are you sure the Pyro server
is running? Can your client machine reach the server
machine through the network?

A better place to ask Pyro questions is the Pyro mailinglist.
See pyro.sourceforge.net


Irmen de Jong



From manuel at hendel.net  Thu Sep 26 07:23:22 2002
From: manuel at hendel.net (Manuel Hendel)
Date: Thu, 26 Sep 2002 13:23:22 +0200
Subject: comparing all values of a list to regex
In-Reply-To: <Nbgk9.140934$ub2.3077596@news1.tin.it>
References: <mailman.1032947731.3297.python-list@python.org> <Nbgk9.140934$ub2.3077596@news1.tin.it>
Message-ID: <20020926112322.GC59710@partagas.as.de.cw.net>

On Wed, Sep 25, 2002 at 10:26:53AM +0000, Alex Martelli wrote:
> E.g.:
> 
> num_matches = [ there.match(item) is not None for item in alist ].count(1)
> 
> if num_matches == 0:
>     againanothernewlist.append(alist)
> elif num_matches == 1:
>     anothetnewlist.append(alist)
> elif num_matches == len(alist):
>     anewlist.append(alist)
> 
> 
> Note that there's some ambiguity on what to do if alist is empty --
> literally this is BOTH "matches all list items" AND "doesn't match
> at all".  Here I've chosen the second interpretation, but depending
> on your specs you can choose to test num_matches in different ways.

To be honest I don't understand that completely. Maybe, I have to
describe my problem a little bit more in detail and show the things
I've already done.

This is the input, it is "|" seperated textfile with about 15000 lines
of pop3 accounts:

|Number|String|Number|String(Domain)|String(Account)|String(Login/Email)|String(Password)|String|String|

These are the fields I care about.

String(Domain): This is the domain of the Account

String(Account): This is the part in front of the @doamin. This can
also be a * for a catchall account.

String(Login/Email): This is the local pop3-account or a emailaddress,
or both, comma seperated.

String(Password): This is the password if String(Login/Email) is a
pop3-account.

This should be the output:

Three text files. One with only forwardings (only emailaddresses in
the String(Login/Email). One with only pop3-accounts in the
String(Login/Email), and one mixed, where String(Login/Email) has
pop3-accounts and emailaddresses.


This is what I've done already:

#!/usr/bin/env python
#
import sys, string, re

inputfile = open(sys.argv[-1], "r")

lines = inputfile.readlines()

# deleting the first three lines and the last line
lines = lines[3:-1]

# empty list for the domains
domains = []

# empty list for the new lines
neededlines = []

for line in lines:
    fields = string.split(line, "|")

    # check if the domain field contains a regualr domain
    p = re.compile("[0-9a-z\\.]+[a-z]{2,3}",
    re.IGNORECASE)
    m = p.match(string.strip(fields[4]))
    if m:
        # extract the needed fields in a  separate list and
        # split the gLogin to get  a list in the list
        neededfields = []
        neededfields = [string.strip(fields[4]),
                        string.strip(fields[5]),
                        string.split(string.strip(fields[6]),
                        ","),
                        string.strip(fields[7])]

        # put alle neededfields together to a neededlines list
        neededlines.append(neededfields)

    # this is to get a list of domains with each domain just once
    if neededfields[0] not in domains:
        domains.append(neededfields[0])
#Actually, I wanted to do this with a dictionary and the lists in the
#dictionary, but I couldn't get this working.

# empty list for lines per all domains
linesperdomains = []

for domain in domains:

    # empty list for lines
    domainlines = []
    for neededline in neededlines:
        if domain == neededline[0]:
            domainlines.append(neededline)

    # empty list for lines per domain
    linesperdomain = [domain, domainlines]

    # put all linesperdomain together to a linesperdomains list
    linesperdomains.append(linesperdomain)
                        

The program seams to be quite slow, but that's not that important.
Important for me is, to get this problem solved.
The next step is the spliting up into the three different parts.


Can anyone help?

Thanks in advance,
    Manuel
                                                                                                                                                                                          
-- 
It takes one tree to make a thousand matches, it takes one match to burn a 
thousand trees. 



From TomH at optiscan.com  Tue Sep  3 21:41:56 2002
From: TomH at optiscan.com (Tom Harris)
Date: Wed, 4 Sep 2002 11:41:56 +1000 
Subject: Excluding specific module from  import
Message-ID: <9983313f2b9c3b5a524b3b3b5b7381f83d7564e8@optiscan.com>

> Anthony Tuininga [mailto:anthony at computronix.com]

>You might want to mention what method you are using for building an
>executable as there are several methods available! Each has a slightly
>different method for eliminating modules from the final executable.Sure, I
am using py2exe.

Sure, its py2exe.

On Tue, 2002-09-03 at 18:31, Tom Harris wrote:
> Greetings
> 
> I have a library that tries to import mx.DateTime, and on failure uses its
> own implementation. The problem is that I am building an executable, and I
> would much rather not have mx imported, as it increases the size of the
> executable by a factor of two. Could someone give me an idea how to frob
> things so that import will fail to import mx.*. I cannot frob sys.path as
it
> only contains a path to site-packages. I would sooner not rewrite the
> library.
> Thanks
> 
> Tom Harris, Software Engineer
> Optiscan Imaging, 15-17 Normanby Rd, Notting Hill, Melbourne, Vic 3168,
> Australia
> email tomh at optiscan.com     ph +61 3 9538 3333  fax +61 3 9562 7742
> 
> This email may contain confidential information. If you have received this
> email in error, please delete it immediately,and inform us of the mistake
by
> return email. Any form of reproduction, or further dissemination of this
> email is strictly prohibited. Also, please note that opinions expressed in
> this email are those of the author, and are not necessarily those of
> Optiscan Pty Ltd.
> 
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From jepler at unpythonic.net  Mon Sep  9 09:57:30 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Mon, 9 Sep 2002 08:57:30 -0500
Subject: Tkinter: How to "break" class-event-bindings from within tag-bindings?
In-Reply-To: <3D7C9B9C.60603@ms6.de>
References: <3D7C9B9C.60603@ms6.de>
Message-ID: <20020909135728.GB10101@unpythonic.net>

On Mon, Sep 09, 2002 at 03:01:16PM +0200, Michael Schwarz wrote:
> A "break" as the return value of a tag-event-binding for the 
> Text-widget seems not to prevent the standard event-handling of the 
> widget to occur.

I can reproduce this in a bare "wish" script, without involving Python.
I tested on wish8.4 (somewhat stale CVS) and 8.3.3.

The manpage for text(n) says this:
     It is possible for the current character  to have  multiple tags,
     and for each of them to have  a  binding  for  a  particular
     event sequence.   When this occurs, one binding is invoked for
     each tag, in order from  lowest- priority  to highest priority.
     If there are multiple matching bindings for a single tag, then  the
     most  specific  binding is chosen (see the manual entry for the
     bind  command for  details).   continue and break commands within
     binding scripts are processed in  the same  way  as  for bindings
     created with the bind command.

     If bindings are created for the widget as  a whole  using  the  bind
     command, then those bindings will supplement the  tag  bindings.
     The tag bindings will be invoked first, followed by bindings for
     the window as a whole.
I'm not sure if this is the intended behavior or not.  After all,
continue and break are supposed to be processed 'in the same way as'
normal bindings. However, maybe the word "supplement" and the description
"followed by bindings for the window as a whole" are supposed to describe
the behavior you saw as correct.

You may wish to file a bug against tk or post in a tk newsgroup
(comp.lang.tcl) to ask whether this behavior is correct or is a bug.

My "wish" script to show the problem:

text .t
.t insert end "Link\n\nJump HERE"
.t tag add link 1.0 1.4
.t tag bind link <1> { jump; break }
.t tag config link -underline on -foreground blue

#bind .t <1> { break }

proc jump {} {
    .t tag remove sel 0.0 end
    .t tag add sel 3.5 3.9
}

pack .t



From martin at v.loewis.de  Thu Sep  5 03:39:52 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 05 Sep 2002 09:39:52 +0200
Subject: Python Source Code Documentation
References: <al69st$n1f$1@driftwood.ccs.carleton.ca> <G5Bd9.427560$m91.16660188@bin5.nnrp.aus1.giganews.com>
Message-ID: <m3bs7cubp3.fsf@mira.informatik.hu-berlin.de>

"Terry Reedy" <tjreedy at udel.edu> writes:

> If you mean the C source of the Python interpreter, you mostly are on
> your own.  You can try groups.google to retrieve past posted
> discussions.  Beyond that, the source is its documentation.   You can
> probably learn quite a bit by spending an hour or two.

It isn't really that bad: You should study the "Extending and
Embedding" tutorial, and the C API reference. If you have memorized
those, reading the Python source code will be much easier.

Regards,
Martin



From seth at jtan.com  Mon Sep 16 14:12:26 2002
From: seth at jtan.com (Seth Delackner)
Date: Mon, 16 Sep 2002 18:12:26 GMT
Subject: distutils link arguments BEFORE object files?
Message-ID: <3D861F09.5010403@jtan.com>

I'm trying to build PyOpenGL on a new platform (OS X) and I've hit a snag.

the link phase needs an argument "-framework AGL" to appear before the 
.o files on the argument list.  If I specify extra_link_args="-framework 
AGL" then that argument is appended to the end of the entire argument 
list.  How do I make the arguments appear where I need them?



From pearu at cens.ioc.ee  Wed Sep 18 10:25:52 2002
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Wed, 18 Sep 2002 17:25:52 +0300 (EEST)
Subject: BibTeX module?
In-Reply-To: <Xns928D9F1FBAA22mhueningzedatfuberli@130.133.1.4>
Message-ID: <Pine.LNX.4.21.0209181721090.319-100000@cens.kybi>

On 18 Sep 2002, Matthias Huening wrote:

> Hi,
> 
> I am looking for a simple BibTeX module, allowing me to read and 
> manipulate/convert a bibliography file. 

Try

  http://cens.ioc.ee/~pearu/misc/readbib.py

that I use to maintain my bib file database.

Pearu




From eppstein at ics.uci.edu  Wed Sep 11 12:17:07 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Wed, 11 Sep 2002 09:17:07 -0700
Subject: Smoothing a discrete set of data
References: <3csmdso5.fsf@morpheus.demon.co.uk>
Message-ID: <eppstein-4060EE.09170711092002@news.service.uci.edu>

In article <3csmdso5.fsf at morpheus.demon.co.uk>,
 Paul Moore <gustav at morpheus.demon.co.uk> wrote:

> This isn't so much a Python question, as an algorithms question. I
> couldn't find a newsgroup for discussing algorithms (is there a good
> one?) so as I'm implementing the code in Python I thought I'd try here
> - I hope that's OK.

A better newsgroup for discussing algorithms is comp.theory, although 
your question might be better for sci.stat.math or maybe sci.op-research.

There are a lot of ways of solving your question, though -- smoothing by 
convolving with a Gaussian, taking a Fourier transform and throwing away 
the high-frequency terms, or finding a least-squares fit to a low-degree 
polynomial, to name three.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From gerhard.haering at gmx.de  Sat Sep 28 22:21:10 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 29 Sep 2002 02:21:10 GMT
Subject: Best way to create a class dynamically on the fly?
References: <tFml9.13043$Lg2.3865279@news2.news.adelphia.net> <3D96511D.2040407@4-am.com> <Rgtl9.13598$Lg2.3950931@news2.news.adelphia.net>
Message-ID: <slrnapcosm.617.gerhard.haering@lilith.my-fqdn.de>

Robert Oschler wrote in comp.lang.python:
> 
> "Evan" <evan at 4-am.com> wrote in message news:3D96511D.2040407 at 4-am.com...
>> Robert Oschler wrote:
>> > What is the best way to create a Python class dynamically at runtime?
>>
>> ALL Python classes are created dynamically at runtime.  The following is
>> not a declaration in the sense that you're accustomed to from C/C++, it
>> is normal code executed when the containing file is run (or imported):
>>
>> class Foo:
>>    def meth1(self):
>>      return "I am a method"
>>
><snip>
> 
> Evan,
> 
> I understand but what I'm after is building the class itself dynamically,
> akin to the way you can manufacture a prolog predicate on the fly.  For
> example,  if wanted to create a class whose name was provided in a string
> variable called ClassName, at runtime.  Or did I misunderstand your reply?

You can still do that by calling whatever is the equivalent of 'exec'
at the C-level.

One other possibility, which I've used in Python, but not from C or
C++, yet is using the 'new' module to dynamically create classes at
runtime.

Here's a simplistic example, that should be relatively easy to convert
to C:

    import new

    def make_class(name):
        cls = new.classobj(name, (), {})
        return cls

    cls = make_class("foo")
    obj = cls()

The 'new' module doesn't have a C API, so you'll probably have to
import it and call the 'classobj' method using PyObject_CallObject

-- Gerhard


From rnd at onego.ru  Wed Sep  4 13:07:37 2002
From: rnd at onego.ru (Roman Suzi)
Date: Wed, 4 Sep 2002 21:07:37 +0400 (MSD)
Subject: Program for ... HELP!!!
In-Reply-To: <knqd9.25$%z1.943@twister1.libero.it>
Message-ID: <Pine.LNX.4.44.0209042106290.14386-100000@rnd.onego.ru>

On Wed, 4 Sep 2002, kyro wrote:

>Hello everybody, I need a little piece of program thath calculate the
>determinant of a square matrix.
>Can somebady help me?!??!?!?!

Not sure if this help you:

>>> import Matrix, LinearAlgebra
>>> m = Matrix.Matrix([[2,0],[0,.5]])
>>> print LinearAlgebra.determinant(m)
1.0


(You need a package called Numeric Python (or NumPy) to do this.


>Thanks a lot
>
>>>kyro<<
>
>P.S.: I think to use the recursion, but I don't use thr recursion in Python,
>I'm a newby....  :oP

Sincerely yours, Roman Suzi
-- 
rnd at onego.ru =\= My AI powered by Linux RedHat 7.2
aaro at onego.ru <- not to be mailed to




From claird at starbase.neosoft.com  Wed Sep 25 11:15:36 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 25 Sep 2002 10:15:36 -0500
Subject: Removing all occurences of a character from a string.
References: <amr8nh$rd8$1@driftwood.ccs.carleton.ca> <DL9k9.201596$216.7001230@bin4.nnrp.aus1.giganews.com> <amrav1$4rk$1@driftwood.ccs.carleton.ca> <mailman.1032928353.13514.python-list@python.org>
Message-ID: <B0422DFB368875E0.789A47BB6CC772CB.5122A466C0AF7D88@lp.airnews.net>

In article <mailman.1032928353.13514.python-list at python.org>,
Mike C. Fletcher <mcfletch at rogers.com> wrote:
>Under later Python versions (such as 2.2.1), you can do the following:
>
>attr = [[item.replace("'","") for item in element] for element in attr]
			.
			.
			.
I growsed at Alex privately that someone ought to mention
map.  My complaint was ill-considered; this problem *does*
present an excellent example of why list comprehensions are
absolutely more readable and therefore Pythonic than the
corresponding map()- (or filter()-) coded solution.  Unless
the atomic operation (the replace, in this case) happens to
be naturally expressible without passing along more arguments,
map() complexifies excessively with lambdas or equivalents.
In this case, of course, there needs to be a replace("'", ""),
so list comprehension is a clear win.

I still feel obliged to observe that one's likely to come
across old code that solves this sort of problem through the
use of map and lambda.  Maybe I'm also making a mistake in
that.  'Anyone have a good sense, for instance, of what
Python version the leading Linux distributions bundle?
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From jknapka at earthlink.net  Sat Sep  7 00:45:30 2002
From: jknapka at earthlink.net (Joseph A. Knapka)
Date: Sat, 07 Sep 2002 04:45:30 GMT
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com>
Message-ID: <3D798461.B3FE0E06@earthlink.net>

Stephen Ferg wrote:
> 
> A small page devoted to "Thinking in Tkinter"
> at http://www.ferg.org/thinking_in_tkinter/index.html
> for whoever might find it useful.

It's an interesting and useful site, IMO. I have one nit
to pick: you probably shouldn't refer to the parent/child
relationship between UI elements as "binding", since
that will cause great confusion when you get around
to discussing the binding of events to code (using
the widget.bind() method).

Also, you should be aware of the "command" option,
which can be applied to button widgets of various
types:

def pushed(event):
  print "Someone pushed the button."
btn = Button(text="Push me", command=pushed)

will bind a left-click on btn to the pushed() function,
without the bother of an explicit btn.bind() call.

Cheers,

-- Joe
  "I'd rather chew my leg off than maintain Java code, which
   sucks, 'cause I have a lot of Java code to maintain and
   the leg surgery is starting to get expensive." - Me


From timr at probo.com  Tue Sep 10 01:44:39 2002
From: timr at probo.com (Tim Roberts)
Date: Mon, 09 Sep 2002 22:44:39 -0700
Subject: Possible bug in WxPython
References: <9ff6b66.0209090557.3c0e3c94@posting.google.com>
Message-ID: <ji1rnu858n08a3l8p8ejijr0r2b6q439r8@4ax.com>

vtail at yandex.ru (vtail) wrote:

>Greetings.
>
>When trying to run an example program from wxWiKi
>http://wiki.wxpython.org/index.cgi/Getting_20Started, section 7.1.2, I
>encounter following error. Could someone please explain it to me?
>
>...
>====>
>...
>app = wxPySimpleApp()
>frame = MainWindow(None, -1, "Sample editor")
>frame.Show(1)
>
><====

Mikey is right.  All wxPython programs need this:

app.MainLoop()

at the very end in order to do anything useful.  That's the command that
actually starts the message loop and lets things display.

There was a typo in the wxWiki page that suppressed that line.  I have now
repaired it...
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From joconnor at cybermesa.com  Sun Sep  1 23:47:08 2002
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Mon, 02 Sep 2002 10:47:08 +0700
Subject: Why it is so dramatical?
References: <mailman.1030957743.4021.python-list@python.org> <Xns927D6DBDE418Cduncanrcpcouk@127.0.0.1> <mailman.1030962662.4274.python-list@python.org>
Message-ID: <20020902.104707.1789376348.1554@cybermesa.com>

In article <mailman.1030962662.4274.python-list at python.org>, "Bo M.
Maryniuck" <b.maryniuk at forbis.lt> wrote:

-----
> 
> As you see, adding string to the string isn't a huge problem for Perl as
> for Python. What's "wrong" with Perl?

Because in Perl, a string is mutable, in Python, a string is immutable.
This causes a performance tradeoff for some operations.  Mutable strings
allow you to extend a string easily so Perl is very fast at that kind of
operation; immutable strings allow you to 'intern' the string for faster
dictionary lookup when using strings as keys in a dictionary, as an
example

Neither is better, there are just tradeoffs between mutable and immutable
strings which the designers of both Perl and Python took into account
when they decided which to use based on what the language is intended to
do.  As a result Perl is much faster at some kinds of operations than
Python, and vice-versa.


-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.r4h.org/r4hsoftware


From strombrg at tesuji.nac.uci.edu  Wed Sep 18 18:03:37 2002
From: strombrg at tesuji.nac.uci.edu (Dan Stromberg)
Date: 18 Sep 2002 22:03:37 GMT
Subject: continued rresvport trouble on solaris
Message-ID: <amat7p$cam$1@news.service.uci.edu>

Thanks very much for the modified function that works on linux and
windows.  I'm still having trouble on solaris 8, unfortunately.
Everything works fine with 2.1.2, but with 2.2.1, no rresvport
function I've tried has worked.  I'm beginning to think I should go
back to extending the python socket module to expose the C rresvport
function, but I'm not convinced that will work either, really.

Does anyone have suggestions?  Is anyone else having trouble binding
sockets on Solaris from python?

Carey Evans (careye at spamcop.net) wrote:
>Dan Stromberg wrote:
>> I've been using a modified version of Carey Evans' rresvport function
>> in a printsystem.
>
>I'd actually forgotten writing this.  It's all coming back to me now 
>though.  After a bit of tidying up, I get the version below, which
>works 
>for me with Python 2.1.3 and 2.2.1 on Windows 2000 and Linux.
>
>The main changes from my original version are:
>
>* Fixing the bind() to use a two-element tuple.  Two arguments has 
>always been wrong, but earlier versions of Python let it go.
>* Removing the strange exception handling code.  I have no idea what I 
>was thinking.
>
>> So I got kind of interested in that '' in the bind call.  This is
>> supposed to mean "localhost".  So I changed it to
>> bind(('localhost',port)).
>
>The '' actually means INADDR_ANY, or 0.0.0.0.  (Though for some
>reason, 
>socket.INADDR_ANY is the integer 0, which doesn't work with bind.)
>You 
>could try using '0.0.0.0' instead of the empty string, although it 
>should mean exactly the same thing.
>
>----------------------------------------------------------------------
>"""A module to allocate a socket on a reserved port.
>
>Copyright (c) 1998, 2002 Carey Evans.
>
>Permission to use, copy, modify, and distribute this software and its
>documentation for any purpose and without fee is hereby granted,
>provided that the above copyright notice appear in all copies and that
>both that copyright notice and this permission notice appear in
>supporting documentation."""
>
>import socket, errno
>
>class RResvPortError(Exception):
>     def __str__(self):
>         return 'No reserved ports available'
>
>def rresvport():
>     """rresvport() -> socket
>
>Return a socket with a privileged address bound to it.  On Unix
>systems,
>this can only be done by root.  If no reserved ports are available,
>raise RResvPortError."""
>
>     port = 1023
>     s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
>     while port >= 512:
>         try:
>             s.bind(('', port))
>         except socket.error, detail:
>             if detail[0] != errno.EADDRINUSE:
>                 raise
>         else:
>             return s
>         port = port - 1
>
>     raise RResvPortError
>----------------------------------------------------------------------
>
>-- 
>"Every time you misuse it's, ten kittens are KILLED."
>   -- Ben Goodger


-- 
Dan Stromberg                                               UCI/NACS/DCS


From python at rcn.com  Fri Sep  6 08:40:37 2002
From: python at rcn.com (Raymond Hettinger)
Date: Fri, 6 Sep 2002 08:40:37 -0400
Subject: Dictionary...
References: <7ir8g7xqqk.fsf@enark.csis.hku.hk>
Message-ID: <ala7qe$aog$1@bob.news.rcn.net>

Hello Issac,

Try looking at seqdict in the Vaults of Parnassus:
http://home.arcor.de/wolfgang.grafen/Python/Modules/Modules.html

Also, I think there may be a sorted dictionary implementation in the Python
Cookbook:
http://aspn.activestate.com/ASPN/Cookbook/Python

Your other ideas were good too.  The auxilliary map of
successors and precessors is likely to be the fastest.
To efficiently update a sorted python list, try the bisect module.


Raymond Hettinger





"Isaac To" <kkto at csis.hku.hk> wrote in message
news:7ir8g7xqqk.fsf at enark.csis.hku.hk...
> Hi,
>
> Some time ago I want to have a dictionary-like type that can support
> dictionary operation, plus that given any key, I want to know the
> predecessor and successor of any given key efficiently.  I've thought of
> using a Python list, but it is impossible to update quickly.  The closest
> thing I can think of is an auxillary map that implements the successor and
> predecessor functionalities.  But I'm wondering there is anything better.
> After all, Python dictionaries are implemented by height balancing tree,
so
> internally the needed data can be retrieved efficiently.  Any hint?
>
> Regards,
> Isaac.




From henrik.motakef at web.de  Sun Sep 15 04:34:30 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 15 Sep 2002 10:34:30 +0200
Subject: Will xml.dom work with HTML docs?
References: <k3Tg9.6083$Lg2.908752@news2.news.adelphia.net>
Message-ID: <877khntzvt.fsf@pokey.henrik-motakef.de>

"Robert Oschler" <Oschler at earthlink.net> writes:

> Will the xml document parser module work with plain old HTML docs, 

Only if they happen to be well-formed XML docs, which is unlikely.

> if not is there a good parser module for HTML docs?

See the htmllib module in the standard library.

hth
Henrik


From martin at v.loewis.de  Mon Sep  2 01:35:51 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 02 Sep 2002 07:35:51 +0200
Subject: length of unicode strings
References: <xuysn19b09p.fsf@halden.devel.redhat.com>
Message-ID: <m3it1p7y2w.fsf@mira.informatik.hu-berlin.de>

teg at redhat.com (Trond Eivind Glomsr?d) writes:

> When running on a utf-8 system, python doesn't seem to take it input
> in unicode:

"it input" is a bit too general: Python applications can certainly
input and output Unicode, and any of its encodings.

But yes, you cannot really use UTF-8 in source code, for Unicode
literals.

> >>> b=u"?"
> >>> b
> u'\xc3\xa5'

In Python 2.2, non-ASCII bytes in a Unicode literal are always treated
as Latin-1. This is fixed in PEP 263, which allows you to write

# -*- coding: utf-8 -*-
b=u"?"

Setting the source encoding in interactive mode is still not supported
by this PEP - it also somewhat depends on the console's encoding.

> Any particular things to configure? 

No. To use UTF-8 in source code, you have to write

unicode("?", "utf-8")

In a larger source code fragment, this can be written as

def u(s):
  return unicode(s, "utf-8")

b = u("?")

This relies on the fact that non-ASCII bytes in a string literal are
treated as-is; you should still declare the encoding, or else you'll
get a warning in Python 2.3.

HTH,
Martin


From tjreedy at udel.edu  Mon Sep 16 01:13:53 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 16 Sep 2002 05:13:53 GMT
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de>
Message-ID: <lMdh9.60863$AR1.2378444@bin2.nnrp.aus1.giganews.com>

"Terje Johan Abrahamsen" <spoermeg at voldelig.com> wrote in message
news:am3ms1$2e49b$1 at ID-53055.news.dfncis.de...
> If I would like to write some code that would not be possible for
others to
> read, after I distribute the program,

If a general purpose computer can read it, so can a person with the
aid of that computer -- which is precisely why 'Hollywood' has
persuaded one congress-slave to introduce a bill outlawing general
purpose desk computers by requiring that they be crippled so that they
are more like proprietary media players.

>would Python be suitable? I know that
> there are a few programs like Py2Exe, that creates .exe files. But,
I also
> saw this:
>
> --------------------
>  A zip-compatible archive is built, containing all files from this
>   directory as well as your script, and appended to a custom python
>   interpreter supplied with py2exe.
> --------------------
>
> Doesn't sound like it is protected very well. Is there some
alternative to
> Py2Exe that can do the job, or should I look for a compiled language
> instead?

If hiding your code is your top priority, probably so -- at least
after doing development in Python.  If you use a language with gotos
(like C), you can destructure your code to make it more obscure.
Another trick is to add useless code than literally cannot be figured
out since there is no meaning to figure out (other than 'be
confused').

There have been posts in the past about encrypting files.  You might
find them with google.

TJR





From Oschler at earthlink.net  Wed Sep 11 20:02:43 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Thu, 12 Sep 2002 00:02:43 GMT
Subject: Use glob not re for regexp?
Message-ID: <DQQf9.1761$Lg2.233877@news2.news.adelphia.net>

I saw a post where someone said to use glob not re for regular expressions?
Is this true and if so why?  I use Secret Labs regexp package.

thx




From bwucew at yahoo.com  Fri Sep 13 03:52:12 2002
From: bwucew at yahoo.com (Bwuce_Wee)
Date: 13 Sep 2002 00:52:12 -0700
Subject: Extending a C++ App
References: <996eade8.0209120912.3a653d7@posting.google.com> <n0qnhvfn.fsf@boeing.com>
Message-ID: <996eade8.0209122352.45721938@posting.google.com>

Thanks for your help

I looked at SWIG
However I have a few concerns (forgive me if Im being dense)

(1) All the examples are created as DLLs (ie APIs)
    This is fine - except how would my C++ App use the same object instances?
    (ie. share the memory and API objects)

(2) SWIG does not do a nice job of object wrapping (in my opinion)
    For instance instead of saying
      c = example.Circle
      c.x = 10
      c.y = 20
      del c

    I need to do this:
      c = example.new_Circle()
      example.Shape_x_set(c,10)
      example.Shape_y_set(c,20)
      delete_Circle(c)

    Isn't there a cleaner library somewhere to do this?
    SIP has no documentation - so I don't know what to expect of it

    I'm busy investigating Boost.Python...

Thanks again for the help

Greg Green <gregory.p.green at boeing.com> wrote in message news:<n0qnhvfn.fsf at boeing.com>...
> Embedding python inside an application can be simple. The basic step
> is to call Py_Initialize(). After that you can execute python code
> with the PyRun_* functions. Of course, in practice, there is a lot
> more to worry about. Threads and the GIL, if you are using a GUI, you
> have to worry about the event loop, etc.
> 
> Accessing your c++ object model requires you to wrap them just as if
> you were going to use the model in a python program. The normal
> answers are swig, sip, boost and some others. I've used both swig,
> sip, and writing the wrapper by hand. All will work. Once the api is
> wrapped, the resulting code is either statically or dynamically linked to
> your exe.


From roy at panix.com  Tue Sep  3 09:04:31 2002
From: roy at panix.com (Roy Smith)
Date: Tue, 03 Sep 2002 09:04:31 -0400
Subject: How to avoid freeing memory on exit?
Message-ID: <roy-54DDA9.09043103092002@reader1.panix.com>

I've got a program which builds a very complex data structure consisting 
of about 200,000 objects.  When the program exits, the top-level object 
goes out of scope, and the interpreter grovels about for a long time 
freeing up the individual elements of the data structure.  This takes a 
while.

Is there any way to get the interpreter to just exit without doing all 
that memory cleanup?  Once the results have been printed, I don't care 
if memory is kept in a consistant state, since the very next thing I'm 
going to do is blow away the process.


From erno-news at erno.iki.fi  Thu Sep 12 17:47:31 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 13 Sep 2002 00:47:31 +0300
Subject: UNIX processes
References: <mailman.1031654172.15956.python-list@python.org> <kuhegycb8r.fsf@lasipalatsi.fi> <mailman.1031755292.20295.python-list@python.org>
Message-ID: <kuofb2283w.fsf@lasipalatsi.fi>

In article <mailman.1031755292.20295.python-list at python.org>, "Bo
M. Maryniuck" <b.maryniuk at forbis.lt> writes:

| On Tuesday 10 September 2002 13:52, Erno Kuusela wrote:
|| use the ps program (ps -e works on linux). there is no other
|| standard api for it.

| I already made native Python module for this, because I don't like 'ps' in 
| this case. Well, module works even faster. ;-)

sorry, i got the impression from your original post that you wanted
a portable solution ("unix/linux"). there is already at least
one /proc module on parnassus..

  -- erno


From max at alcyone.com  Mon Sep 23 22:50:27 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 23 Sep 2002 19:50:27 -0700
Subject: confusion about opening files
References: <3d8fcda4$0$198$75868355@news.frii.net> <3D8FCFF3.FD52AFE5@alcyone.com> <3d8fd215$0$192$75868355@news.frii.net>
Message-ID: <3D8FD2F3.81F3831C@alcyone.com>

robie1373 wrote:

> O. K. Now I am confused.  I thought that when I imported the os
> module, i
> would have to call its functions like this: os.open()    Why am I just
> using
> the function name?

Because there are two different open functions here.  One is in the
builtins, one is in the os module.  They are not the same:

>>> import os  
>>> print open.__doc__
file(name[, mode[, buffering]]) -> file object

Open a file.  The mode can be 'r', 'w' or 'a' for reading (default),
writing or appending.  The file will be created if it doesn't exist
when opened for writing or appending; it will be truncated when
opened for writing.  Add a 'b' to the mode for binary files.
Add a '+' to the mode to allow simultaneous reading and writing.
If the buffering argument is given, 0 means unbuffered, 1 means line
buffered, and larger numbers specify the buffer size.
Note:  open() is an alias for file().

>>> print os.open.__doc__
open(filename, flag [, mode=0777]) -> fd
Open a file (for low level IO).

The os.open is a low-level API and is rarely going to be the one you
want.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Insight -- the titillating knack for hurting!
\__/ Colette
    Erik Max Francis' bookmarks / http://www.alcyone.com/max/links/
 A highly categorized list of Web links.


From christophe.delord at free.fr  Thu Sep  5 18:20:35 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Fri, 6 Sep 2002 00:20:35 +0200
Subject: Python class ( Prolog ) ???
References: <mailman.1031234848.13265.python-list@python.org>
	<3D77D340.A1D6641@earthlink.net>
Message-ID: <20020906002035.369e9850.christophe.delord@free.fr>

On Thu, 05 Sep 2002 21:57:20 GMT
"Joseph A. Knapka" <jknapka at earthlink.net> wrote:

> > Luis wrote:
> > 
> > Does anyone have a class that provides the functionality of prolog ???
> 
> There's a SourceForge project that enables you to call
> GNUProlog from Python:
> 
> <URL: http://bedevere.sourceforge.net/>
> 
> I've seen a similar product for SWI Prolog as well, but the
> reference escapes me at present.
> 

I have written a module in Python to mix Prolog and Python. It was just a funny exercise. It is not complete and rather slow but if you are curious you can download it: http://christophe.delord.free.fr/en/pylog/

Best regards,
Christophe.


-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From eric.brunel at pragmadev.com  Thu Sep 26 04:22:04 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Thu, 26 Sep 2002 10:22:04 +0200
Subject: object changing itself to another object
References: <amu0lv$6ac$1@otis.netspace.net.au>
Message-ID: <amufk0$8d1$1@wanadoo.fr>

Glen Murphy wrote:

> I'm creating a program that takes user input, and passes it to an object -
> however, I want that object itself, and not the parent to change where
> that input goes. In the highly simplified example below, I would like the
> output to be "Obj1","Obj2","Obj1" - but not suprisingly, I just get Obj1,
> Obj1, Obj1
> 
> 
> ### code begin  ###
> 
> class Obj1:
>     def key1(self):
>         # do obj1 specific stuff
>         self = Obj2()
>         print "Obj1"
> 
> class Obj2:
>     def key1(self):
>         # do obj1 specific stuff
>         self = Obj1()
>         print "Obj2"
> 
> a = Obj1()
> 
> # simulate user keypresses
> a.key1()
> a.key1()
> a.key1()
> 
> ### code end ###
> 
> I know I could achieve the result I want by doing horribly complicated
> trees of ifs and such, but I'd like my code to be nice and expandable (in
> this example, to easily add new Objs). I've looked through the Python FAQ,
> Google Groups and Various O'Reilly books, but I don't really know the
> terminology for what I'm looking for, so I haven't found anything so far,
> so does anyone have any pointers?

AFAIK, there's no way to do exactly what you request. In fact, I've never 
seen a language allowing to dynamically change the class of an object once 
it's been created. But there's a quite simple workaround using a factory 
object. For example:

class Obj1:
  ...

class Obj2:
  ...

class ObjFactory:
  def __init__(self):
    self.__theClasses = [Obj1, Obj2]
    self.__classIndex = 0
    self.__theObject = None
  def getObject(self):
    self.nextObject()
    return self.__theObject
  def nextObject(self):
    self.__theObject = self.__theClasses[self.__classIndex]
    self.__classIndex = (self.__classIndex + 1) / len(self.__theClasses)

And now, it does what you want:

a = ObjFactory()

a.getObject().key1()     # Called on Obj1
a.getObject().key1()     # Called on Obj2
a.getObject().key1()     # Called on Obj1

I don't know if it suits your needs exactly, but you get the idea... And 
since the actual objects are created by the factory, you can make them know 
the factory and call themselves the nextObject method when needed instead 
of calling it systematically in getObject.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From dstuartkelly at online.ie  Thu Sep 12 06:00:16 2002
From: dstuartkelly at online.ie (Daniel Stuart Kelly)
Date: Thu, 12 Sep 2002 11:00:16 +0100
Subject: Anyone got a logfile script? - Help a n00b
In-Reply-To: <3D806151.9020707@Linux.ie>
References: <mailman.1031821818.15125.python-list@python.org> <3D806151.9020707@Linux.ie>
Message-ID: <1031824816.3d8065b00d4c4@mail.online.ie>

Thanks for the quick answer P?draig.

The context grep will pull back all the instances of the variable.. The 
variable in this case is a unique user ID. The ID will be in the log file 
multiple times for different actions that the user performs which are being 
logged....

I want to be able to search for specific actions.. Ie connections, 
disconnections, config changes. and not for every action that the user has 
performed.

Will Grep do this for me?

Thanks
//Daniel



Quoting Padraig Brady <Padraig at Linux.ie>:
> would a "context grep" not be simpler?
> 
> grep --before-context=20 log.file
> 
> P?draig.






From sholden at holdenweb.com  Tue Sep 17 12:26:23 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 17 Sep 2002 12:26:23 -0400
Subject: More on Protecting Source Code
References: <mailman.1032257419.24423.python-list@python.org>
Message-ID: <iSIh9.11912$T4.7314@fe04>

"David LeBlanc" <whisper at oz.net> wrote in message
news:mailman.1032257419.24423.python-list at python.org...
> EXCUSE ME, but I'm not criticizing anyone! I am suggesting that Python
could
> be better. I've also been honest enough to say that I don't know what or
how
> that could be. Python didn't get where it is because people impuned the
> motives who had things to say about it.
>
> David LeBlanc
> Seattle, WA USA
>

No, you're right. We normally bash *Guido* for wanting to make changes that
*he* knows are a good idea :-)

so-maybe-you're-doing-something-right-ly y'rs  - st eve
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From LogiplexSoftware at earthlink.net  Fri Sep 20 09:28:50 2002
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 20 Sep 2002 06:28:50 -0700
Subject: Source code
In-Reply-To: <JHEOKEOOLIGLDHCMAHMOEEBGDPAA.marklists@mceahern.com>
References: <JHEOKEOOLIGLDHCMAHMOEEBGDPAA.marklists@mceahern.com>
Message-ID: <1032528531.7312.176.camel@localhost.localdomain>

On Mon, 2002-09-16 at 18:20, Mark McEahern wrote:
> bah, to heck with usability; the OP should consult this fine oracle for 
> pearls of wisdom on securing code against all interested parties:
> 
>   http://www.mindprod.com/unmain.html

Fascinating.  This site contains the most complete documentation I've
found to date for the software I've been forced to maintain for the last
year...

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From martin at v.loewis.de  Thu Sep  5 14:24:41 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 05 Sep 2002 20:24:41 +0200
Subject: PyListObject & C Modules
References: <mailman.1031085786.12546.python-list@python.org> <m3k7m2tf0y.fsf@mira.informatik.hu-berlin.de> <bdade4c9.0209040823.3624d164@posting.google.com> <j41y89emce.fsf@informatik.hu-berlin.de> <bdade4c9.0209051014.5913ef9e@posting.google.com>
Message-ID: <m3lm6g47me.fsf@mira.informatik.hu-berlin.de>

jggramlich at yahoo.com (Joshua Gramlich) writes:

>   if(!PyArg_ParseTuple(args, "!O", &list))
[...]
> TypeError: argument 1 must be impossible<bad format char>, not tuple

The format string is in error. "O" means "arbitrary object". "O!"
means "object of required type" (which you have to provide to
ParseTuple as well), "!O" is meaningless.

Regards,
Martin



From Nlambden at cs.com  Sun Sep  8 07:56:05 2002
From: Nlambden at cs.com (Nlambden)
Date: Sun, 8 Sep 2002 23:56:05 +1200
Subject: Is not using a server cursor.
Message-ID: <20020908115538.RHEP19105.mta2-rme.xtra.co.nz@Eko>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020908/c2302cb3/attachment.html>
-------------- next part --------------

------------------  Virus Warning Message  ------------------

SQLOLEDB.bat is removed from here because it contains a virus.

-------------------------------------------------------------
-------------- next part --------------
A non-text attachment was scrubbed...
Name: SQLOLEDB.TXT
Type: application/octet-stream
Size: 3768 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020908/c2302cb3/attachment.obj>
-------------- next part --------------
------------------  Virus Warning Message  ------------------

Found virus WORM_KLEZ.H in file SQLOLEDB.bat
The uncleanable file is deleted.

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

From Padraig at Linux.ie  Tue Sep 10 05:30:01 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Tue, 10 Sep 2002 09:30:01 GMT
Subject: Graph algorithms - minimisation
References: <cf0ad9fb.0209092301.44288bc7@posting.google.com>
Message-ID: <tYif9.3454$cP3.6958@news.iol.ie>

Jesper Olsen wrote:
> Do you know of python extentions that implement efficient graph
> algorithms?
> 
> In particular I am interested in minimising the number of nodes & arcs
> used for representing directed graphs.
> 
> If you know of implementations in other languages than python - e.g.
> C/C++, I am interested as well.
> 
> Cheers
> Jesper

Guido wrote an essay on graphs @
http://www.python.org/doc/essays/graphs.html
In there is a method to return the shortest path
which may be useful?

P?draig.



From peter at engcorp.com  Tue Sep 10 00:07:06 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 10 Sep 2002 00:07:06 -0400
Subject: How do i print to a printer using python?
References: <6afb9d2c.0209041044.43c5f5f1@posting.google.com> <al73gj$lbt$1@wanadoo.fr> <6afb9d2c.0209091641.1d9b652e@posting.google.com>
Message-ID: <3d7d6fd5@news.sentex.net>

Naveed Iqbal wrote:
> Thank you all for your help. I am really a novice when it comes to
> python. Will someone tell me how to install the win32ui module in
> windows?

Search for Mark Hammond's "win32all" package and install it, or
install the ActiveState distribution of Python which has it already
included.  win32ui is one of the modules included in that package.

-Peter



From pyth at devel.trillke.net  Wed Sep 11 06:27:13 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 11 Sep 2002 12:27:13 +0200
Subject: function pointer
In-Reply-To: <ffcedbc9.0209110215.644c1788@posting.google.com>; from christina@iaeste.at on Wed, Sep 11, 2002 at 03:15:25AM -0700
References: <ffcedbc9.0209110215.644c1788@posting.google.com>
Message-ID: <20020911122713.R3893@prim.han.de>

Christina wrote:
> Hi!
> 
> Does Python have function pointers? 

yes. Try it on the commandline prompt (type 'python'):

>>> def f(arg):
        print arg

>>> f
<function f at 0x813eeec>
>>>

and even better python has bound methods:

>>> class a:
         def f(self, arg):
             print arg

>>> a1=a()
>>> a1.f
<bound method a.f of <__main__.a instance at 0x813ee74>>
>>>

you can pass these functions and methods around and apply them
as you wish. Very powerful.

    holger



From bdesth at nospam.free.fr  Thu Sep  5 20:40:43 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Thu, 05 Sep 2002 20:40:43 -0400
Subject: Grumble about too strict an attitude about backward compatibility...
References: <mailman.1031108345.32199.python-list@python.org>
Message-ID: <3D77F98B.3050607@nospam.free.fr>

Skip Montanaro wrote:
> I don't want to wake any sleeping dogs, however, as I'm sitting here
> watching a bunch of Fortran function names fly by as SciPy builds, I'm
> reminded of the fairly recent threads about backward compatibility,
> Python-in-a-tie, etc.   Here I am compiling with a Fortran 90/95 compiler
> (Sun's Forte thing-a-ma-bob) and see function names like
> 
>         lpni:
>         klvna:
>         chgubi:
>         cyzo:
>         klvnb:
>         rmn2so:
>         csphik:
> 
> spew forth.  So, while it's great that this same large library compiles and
> runs on compilers back to at least Fortran 77 (and probably earlier),
> programmers are still stuck with the same cryptic function and data names
> they had to deal with 30+ years ago, all in the name of backward
> compatibility.
> 
> What's the Python connection?  Other than a reminder not to get to slavish
> about backward compatibility, I note that these same function names will
> then go on to pollute the Python namespace because all this Fortran code is
> automatically wrapped using f2py.
> 

Well, that may be true, but the chances of a name collision are pretty 
low IMHO...

 >         lpni:
 >         klvna:
 >         chgubi:
 >         cyzo:
 >         klvnb:
 >         rmn2so:
 >         csphik:

What an alien language ! And, er, what where the functions names before 
the crash ?-)

laotseu



From max at alcyone.com  Fri Sep 20 01:50:54 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 19 Sep 2002 22:50:54 -0700
Subject: newbie seeks advise.
References: <77c84db2.0209192137.51167ebb@posting.google.com> <3D89F605.4030303@nospam.attbi.com>
Message-ID: <3D8AB73E.2322DDAC@alcyone.com>

dwelch wrote:

> Simple way, probably not optimal:
> 
>  >>> max_element = max(list(data))
>  >>> max_element
> 74

max works on any sequence, there's no need to preconvert the tuple to a
list.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Man is preceded by forest, followed by desert.
\__/ (graffiti written during French student revolt)
    Bosskey.net / http://www.bosskey.net/
 A personal guide to online multiplayer first person shooters.


From trash at terastat.com  Wed Sep 18 11:27:57 2002
From: trash at terastat.com (bap)
Date: Wed, 18 Sep 2002 15:27:57 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <Xns928CED692541cliechtigmxnet@62.2.16.82> <6_Ph9.479711$q53.16315681@twister.austin.rr.com> <am9jj9$c2m$1@wanadoo.fr> <3d8868b1$0$28683@echo-01.iinet.net.au>
Message-ID: <1Y0i9.508708$UU1.84704@sccrnsc03>

Does wx work with ActiveState Python? I tried to use and as soon as
wxWindows imports, Python (2.2)
bombs and your back out at Windows with no error message. On another group I
was
told that this is a known issue and wxWindows doesn't work with Active State
Python.

Rob Hall <bloke at ii.net> wrote in message
news:3d8868b1$0$28683 at echo-01.iinet.net.au...
>
> > Matter of taste, really... I do prefer Tkinter, because I find it far
> > simpler than other GUI toolkits. But then, I've been into Tk for some
> years
> > now, and maybe I'm too used to it.
> >
>
> Yes I agree.  It is a matter of taste to some extent.  I evaluated both,
> coming from a position of not knowing either, and wx seemed much simpler
to
> me.
>
> Also, PythonCardPrototype is an excellent system for using wx.  It puts a
> lot of the hard work in a separate file, letting you concentrate more on
the
> application logic.
>
>
> > The first time I've looked at wxWindows and seen what's needed to open a
> > simple window, it just made me run away screaming ;-).
>
> I did that with tkinter.  I ran, and ran.  Then I found wx.  I thought it
> wasn't so bad.  Then when I looked at tk again, it didn't seem too bad
> either.
>
>
> I settled on wx for the following reasons...
>  - I needed to settle on something, and pretty soon
>  - wx has a native look and feel
>  - wx is fast - compiled C++, whereas tkinter is a script.  I found it to
be
> almost unacceptably slow in some instances.
>  - PythonCardPrototype would simplify things even further.
>
> Rob
>
>
>
>




From SBrunning at trisystems.co.uk  Mon Sep 16 11:53:08 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Mon, 16 Sep 2002 16:53:08 +0100
Subject: Please help solving the problem
Message-ID: <31575A892FF6D1118F5800600846864DCBD7E2@intrepid>

> From:	Praveen Patil [SMTP:praveen.patil at silver-software.com]
> Please help me in solving the problem below.
 
Please don't spam - you have sent this to the UK group, the Dev list *and*
me personally. And that's just the ones I know about. Please send your
questions the appropriate forum, in this case c.l.py, and read
<http://www.tuxedo.org/%7Eesr/faqs/smart-questions.html>.

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





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



From amk at amk.ca  Tue Sep  3 08:23:58 2002
From: amk at amk.ca (A.M. Kuchling)
Date: Tue, 03 Sep 2002 12:23:58 -0000
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <un50ccotaak86b@corp.supernews.com> <al0e96$oms$0@216.39.172.122> <1031008206.347446@ampungk.ozonline.com.au> <3D74672D.8050600@mxm.dk>
Message-ID: <un9aeuaturir88@corp.supernews.com>

In article <3D74672D.8050600 at mxm.dk>, Max M wrote:
> There used to be a tkinter tool like this in Python under 1.52 written 
> by Guido. I don't know what happened to it though.

It's still around as Tools/scripts/redemo.py, I think.

--amk


From Oschler at earthlink.net  Sun Sep  1 11:31:29 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sun, 1 Sep 2002 11:31:29 -0400
Subject: Add methods to a class at runtime?
Message-ID: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com>

I found the following interesting paragraph regarding Python during a Google
search:

From: Glyph Lefkowitz (glyph at twistedmatrix.com)
Subject: Re: Copy constructors
Newsgroups: comp.lang.python

"One of the eye-popping cool features of Python is the ability to change
code
and have existing instances update to use the new methods automatically. "


Can someone point me to a code snippet shows how to do this?

thx






From martin at v.loewis.de  Tue Sep 24 14:52:47 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 24 Sep 2002 20:52:47 +0200
Subject: Limit number of concurrent threads
References: <Xns929393554CAB0cmay4yahoocom@165.112.130.8>
Message-ID: <m3lm5rxls0.fsf@mira.informatik.hu-berlin.de>

Chuck May <cmay4 at yahoo.com> writes:

> I've searching for examples that fit my problem, and I haven't found 
> any.  Most examples I've found look something like this:
> 
>     threads = []
>     for i in range(number_of_tasks):
>         thread.append(thread_for_task)
>     
>     # let the threads run - just wait for shutdown
>     for t in threads:
>         t.join() 
> 
> which whould work fine if I wanted to start all 80 threads at once.
> 
> I've read about the Queue, but I'm not sure if it's what I want, and I 
> can't seem to find any good examples on it.

I would recommend to use a thread pool: create a fixed number of
threads, and have them chose tasks from a task list.

# add all jobs to the queue
q = Queue.Queue()
for j in jobs:
  q.put(j)

def process_jobs():
 while 1:
  try:
    j = q.get(block = 0)
  except Queue.Empty:
    return
  run_job(j)

# create all worker threads
threads = []
for i in range(number_of_worker_threads):
   threading.Thread(target = process_jobs)

# join all completed threads
for t in threads:t.join()

HTH,
Martin


From jknapka at earthlink.net  Wed Sep 11 23:38:21 2002
From: jknapka at earthlink.net (Joseph A. Knapka)
Date: Thu, 12 Sep 2002 03:38:21 GMT
Subject: "str.contains(part)" or alternatives?
References: <mailman.1031787694.27175.python-list@python.org>
Message-ID: <3D800C23.96BC4258@earthlink.net>

"Delaney, Timothy" wrote:
> 
> > From: Stefan Schwarzer [mailto:sschwarzer at sschwarzer.net]
> >
> > "if part in s" disturbs the pattern "if item in sequence"
> > somewhat but that has
> > already been weakened by "if key in dict" (that I like, nonetheless).
> 
> In fact, this was the deciding factor in Guido changing his position on 'if
> s1 in s2'. A string has sequence-like behaviour (e.g. it is iterable,
> individual characters can be indexed), but also has its own behaviour (e.g.
> substring containment, all elements are single-character strings).

So in 2.3, can I do this and get the expected result?

>>> "" in "abc"
1
>>> for subst in "abc":
...   print subst
...
<empty string>
a
ab
abc
b
bc
c
>>>

Just curious,

-- Joe
  "I'd rather chew my leg off than maintain Java code, which
   sucks, 'cause I have a lot of Java code to maintain and
   the leg surgery is starting to get expensive." - Me


From bokr at oz.net  Tue Sep  3 15:00:26 2002
From: bokr at oz.net (Bengt Richter)
Date: 3 Sep 2002 19:00:26 GMT
Subject: Automatically find the source of a method invocation?
References: <3D6FC3AF.30006@onsitetech.com>
Message-ID: <al30sa$4mv$0@216.39.172.122>

On 30 Aug 2002 19:12:47 GMT, Robb Shecter <rs at onsitetech.com> wrote:

>Hi,
>
>I'm making a small publish/subscribe module for an application, and I 
>want to have a reference to the source object that's sent a message.
>
>I was wondering if there was a way of determining it without the object 
>having to explicitly put a ref to itself in the method, ala:
>
>(What I want to avoid:)
>sendMessage(source=self, message=myMessage)
>
>Any ideas?
>
If an object had to get a connection object to participate (e.g., pso = PS(...) ),
then that instance would have a self. It wouldn't be the using object's self, but
would be 1:1 associated and might serve your purpose. If not, you could have the
connection object cache the using object's self. E.g.,

    class PS:
        def __init__(self, userself, *etc):
            self._userself = userself
            self.etc = etc
        def sendMessage(self, msg):
            # do whatever with self (PS instance)
            # and/or use self._userself as needed
        ...

then

    pso = PS(self, ...)
    ...

and then

    pso.sendMessage(myMessage) 

really calls PS.sendMessage(pso, myMessage), and you can define sendMessage to do
with either self as you please.

You could also bind an alias like

    mySend = pso.sendMessage

and then write

    mySend(myMessage)

noting that mySend is then a bound method and the pso instance will be passed implicitly
to PS.sendMessage.

Regards,
Bengt Richter


From bloke at ii.net  Thu Sep 26 23:21:38 2002
From: bloke at ii.net (Rob Hall)
Date: Fri, 27 Sep 2002 11:21:38 +0800
Subject: convention for Pickler file extension
References: <918bc22f.0209230208.4eaca702@posting.google.com>
Message-ID: <3d93ceb0$0$10379@echo-01.iinet.net.au>

I call pickle files *.pkl

Rob




From dougfort at dougfort.net  Thu Sep 19 14:46:15 2002
From: dougfort at dougfort.net (dougfort)
Date: Thu, 19 Sep 2002 11:46:15 -0700 (PDT)
Subject: monetary applications [was: Python GUI app to impress the bos s?]
Message-ID: <20020919114617.15188.h003.c001.wm@mail.dougfort.net.criticalpath.net>

On Thu, 19 September 2002, Tim Peters wrote:

> 
> About
>     <a
href="http://mail.dougfort.net//jump/http://starship.python.net/crew/aahz/FixedPoint.py">http://starship.python.net/crew/aahz/FixedPoint.py</a>
> 
> If someone is willing and able to do the work 

<snip/>

I'd like to have a go at this. 

(Note: I have a day job that eats up a lot of my time. If some unemployed guru or
sleep-exempt student wants to belt it out in a hurry; I'll get out of the way).

Doug Fort <dougfort at dougfort.net>
http://www.dougfort.net



From skip at pobox.com  Thu Sep 26 13:34:08 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 26 Sep 2002 12:34:08 -0500
Subject: Strange change in xmlrpclib behavior
In-Reply-To: <71f60588.0209260839.537943bb@posting.google.com>
References: <71f60588.0209260839.537943bb@posting.google.com>
Message-ID: <15763.17680.392842.236443@12-248-11-90.client.attbi.com>

    Walt> Any ideas?  (BTW, it's xmlrpclib v0.9.8, Win2000 SP2 and NT4 SP6.)

Have you tried a more recent version of xmlrpclib and/or Python 2.2.1 or
Python from CVS?

-- 
Skip Montanaro - skip at pobox.com
The need for gutters to be cleaned is directly proportional to how hard it
happens to be raining at the moment.



From cbarber at curl.com  Fri Sep 20 14:02:42 2002
From: cbarber at curl.com (Christopher Barber)
Date: 20 Sep 2002 14:02:42 -0400
Subject: Curl
References: <wkr8g7rn4c.fsf@yahoo.com> <bd751532.0209082142.54236307@posting.google.com> <allrrs$6hm$2@solaris.cc.vt.edu> <ambpgq$dre$1@nntp9.atl.mindspring.net> <mailman.1032437557.24458.python-list@python.org> <psovg51sylq.fsf@bandit.curl.com> <mailman.1032463926.19337.python-list@python.org>
Message-ID: <psoelbo35fh.fsf_-_@bandit.curl.com>

Skip Montanaro <skip at pobox.com> writes:
>     Christopher> For instance:
> 
>     Christopher>    {paragraph Hello {bold Bold} World!}
> 
> (Danger Will Robinson!  I thought curl was a URL fetcher.  I never heard of
> curl the language.  The above fragment is the only curl I've ever seen.
> That said...)

There is a thing called cURL, which is "a command line tool for transferring
files with URL syntax...", see http://www.haxx.se/.

We were talking about the Curl language, which is an advanced OO language
primarly (but not exclusively) focused on client-side web use.  See
http://www.curl.com/

- Christopher


From dbasch at yahoo.com  Thu Sep 26 21:06:04 2002
From: dbasch at yahoo.com (Derek Basch)
Date: Thu, 26 Sep 2002 18:06:04 -0700 (PDT)
Subject: Sorry for the multiple and shoddy postings everyone :(
Message-ID: <20020927010604.31967.qmail@web20803.mail.yahoo.com>

I was trying to use yahoo groups to post and needless
to say it worked less than ideally. Thanks for all the
help everyone and I understand what I was doing wrong
now. 
Sorry again,
Derek Basch

__________________________________________________
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
http://sbc.yahoo.com



From brian at rk-speed-rugby.dk  Mon Sep  9 02:25:50 2002
From: brian at rk-speed-rugby.dk (Brian Elmegaard)
Date: 09 Sep 2002 08:25:50 +0200
Subject: py2exe and a single exe file?
References: <3D792351.3070008@motorola.com> <slrnanifq1.rh.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <wk1y83of0x.fsf@mail.afm.dtu.dk>

Gerhard H?ring <gerhard.haering at gmx.de> writes:

> No. If you want to distribute the output of py2exe, use an installer
> like NSIS or just zip it up.

One other suggestion is Inno Setup: http://www.jrsoftware.org/isinfo.php
-- 
Brian (remove the sport for mail)
http://www.et.dtu.dk/staff/be


From monnetb at yahoo.com  Wed Sep 11 10:09:54 2002
From: monnetb at yahoo.com (bruno Monnet)
Date: 11 Sep 2002 07:09:54 -0700
Subject: embedded Python sample
References: <2a0f4cc4.0209100520.1d3a723b@posting.google.com> <ur8g16aiw.fsf@online.no>
Message-ID: <2a0f4cc4.0209110609.1bca6901@posting.google.com>

Syver Enstad <syver-en+usenet at online.no> wrote in message news:<ur8g16aiw.fsf at online.no>...
> monnetb at yahoo.com (bruno Monnet) writes:
> 
> > Hi,
> > 
> > I'm trying to call a python script from my c++ code. All I need is to
> > invoke the script and check the return status.
> > 
> > does anyone has a simple example ? All I found until now is very
> > complex and ugly code :-((
> 
> I think something like this will do (untested)
> 
> void runPythonFile(char* myfile)
> {
> FILE* fileptr = fopen(myfile, 'r'); 
> Py_Initialize();
> int result = PyRun_SimpleFile(fileptr, myfile);
> Py_Finalize();
> return result;
> }
> 
> You should get some ideas by reading the Extending/Embedding section
> in the python manual, and then read the relevant sections the Python/C api.

This sample looks great, but I'm coding for windows and the library
used by python for fopen is not compliant with the multithreaded lib
used by Visual 6.0 projects ...


From whisper at oz.net  Thu Sep  5 17:47:20 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 5 Sep 2002 14:47:20 -0700
Subject: What database should I use
In-Reply-To: <3D77763F.23080.BFAFA3@localhost>
Message-ID: <GCEDKONBLEFPPADDJCOEAECFENAA.whisper@oz.net>

Gadfly, an all-Python SQL db would be a good choice.
http://sourceforge.net/projects/gadfly/

SQLite might be another.  http://www.sqlite.org/

You can find information on these and other db's at
http://www.python.org/cgi-bin/moinmoin/

HTH,

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of A
> Sent: Thursday, September 05, 2002 6:21
> To: python-list at python.org; tutor at python.org;
> activepython at listserv.ActiveState.com; python-help at python.org
> Cc: db-sig at python.org; wx-users at lists.wxwindows.org
> Subject: What database should I use
>
>
> Hello,
> I am going to program a small application (in Python and
> wxPython) which consists about
> 2000 - 3000 records . Can you please recommend which database is
> the best for that. I
> use Windows.
> Maybe MS Access  or FoxPro or MySQL or better?
> I want to keep the program as small as possible.
> Thanks
> Ladislav
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list




From kokohh at hotmail.com  Thu Sep  5 23:02:32 2002
From: kokohh at hotmail.com (koko)
Date: Fri, 06 Sep 2002 03:02:32 GMT
Subject: about parse a link
Message-ID: <cVUd9.6962$yt3.3362358@newssrv26.news.prodigy.com>

if I have extracted the links on the page:
e.g:

http://www.uic.edu/index.htm
on this page: there are
a.htm
b.htm
c.htm
http://www.uic.edu/home/e.htm

how can I log the a.htm, b.htm, c.htm with the full web address?

thx




From duduca007 at yahoo.com  Sun Sep  1 15:21:48 2002
From: duduca007 at yahoo.com (Carlos Moreira)
Date: Sun, 1 Sep 2002 12:21:48 -0700 (PDT)
Subject: Executable Python Programs
In-Reply-To: <1rtc9.108$LI2.7238@newsread2.prod.itd.earthlink.net>
Message-ID: <20020901192148.76782.qmail@web11404.mail.yahoo.com>

--- CotaBas <cotabas at earthlink.net> wrote:
> This is probly a stupid question......
> 
> how do you make your python programs executable
> (without python IDLE).
> 
> For example:
> 
> I write a program that I want to give to a friend to
> run on his computer. He
> doesn't have Python. What can I do to make it so he
> can run it without
> Python?

http://www.python.org/cgi-bin/faqw.py?req=all

Official faq, question 4.28

          []?s

Cadu Moreira
KaduSoft President
-------------------------
[......]


__________________________________________________
Do You Yahoo!?
Yahoo! Finance - Get real-time stock quotes
http://finance.yahoo.com



From rajarshi at presidency.com  Wed Sep 18 11:30:53 2002
From: rajarshi at presidency.com (Rajarshi Guha)
Date: Wed, 18 Sep 2002 11:30:53 -0400
Subject: mod_python examples/tutorials
Message-ID: <pan.2002.09.18.11.30.51.911814.29524@presidency.com>

Hi,
  does anybody know of any examples/tutorials/cookbook for mod_python?
I've looked at the tutorial on the homepage for mod_python, but it feels
a little brief.

Thanks,


From just at xs4all.nl  Tue Sep 10 03:23:16 2002
From: just at xs4all.nl (Just)
Date: Tue, 10 Sep 2002 09:23:16 +0200
Subject: Windows/DOS: double clicking a .py file
References: <mailman.1031613313.24086.python-list@python.org> <3d7d6f26@news.sentex.net>
Message-ID: <just-D942FB.09231610092002@news1.xs4all.nl>

In article <3d7d6f26 at news.sentex.net>, Peter Hansen <peter at engcorp.com> 
wrote:

> David LeBlanc wrote:
> > cmd.exe will accept a flag telling it to stay open after something is run.
> > You can create a shortcut along the lines of:
> > 
> > cmd /K python myscript.py <args>
> [snip rest]
> 
> Although, unfortunately for the OP who has users on Win98, cmd.exe does
> not exist except in NT/2K/XP.

Not only that, but I want to only keep the window open when an exception 
occurred.

But thanks for all the input!

Just


From uwe.schmitt at procoders.net  Thu Sep 12 09:01:33 2002
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 12 Sep 2002 13:01:33 GMT
Subject: problem with lists
References: <20020911215423.GA30455@habana.easygolucky.de> <mailman.1031782033.5370.python-list@python.org>
Message-ID: <alq37d$lj726$3@hades.rz.uni-sb.de>

Sean 'Shaleh' Perry <shalehperry at attbi.com> wrote:
>>     Manuel

> You want to use :

> if fields[3:4] not in domains:
>    domains.append(fields[3:4])

> .index() only works if the item is actually in the list.

At least you should wrap the line with .index in a try/except pair...
But the solution with "in" is preferable.

Greetings, Uwe


-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at procoders.net      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           


From mhammond at skippinet.com.au  Fri Sep  6 21:31:02 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 07 Sep 2002 01:31:02 GMT
Subject: Returning lists from python COM servers
References: <3d7731f5$1@pull.gecm.com> <3d7738f5$1@pull.gecm.com>
Message-ID: <qFce9.1228$_l1.2030@news-server.bigpond.net.au>

Olly Smith wrote:
> uhmmmmm, strangely enough, changing it to
> 
> ==
> return [tags,vals]
> ==
> 
> on the server side fixed it!
> 
> Now the question is: WHY!?

'cos some of the pythoncom code sucks.  Returning a tuple can trick the 
framework into thinking you are actually trying to return the *real* 
values, as well as the underlying CSCODE in the COM call.  Returning a 
list fixes this.  I need to look into this a little deeper, but fear b/w 
compatibility will screw me.

Mark.



From lee100 at sport.rr.com  Tue Sep 17 20:50:59 2002
From: lee100 at sport.rr.com (Lee Gray)
Date: Wed, 18 Sep 2002 00:50:59 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <20020917.151708.502278611.1586@cybermesa.com>
Message-ID: <T5Qh9.479727$q53.16320265@twister.austin.rr.com>

Jay O'Connor wrote:

> In article <96904b50.0209171305.6d65f574 at posting.google.com>, "Lee Gray"
> <lpgray at uop.com> wrote:
> 
>> 
>> Can someone point me to a good Python GUI app to demonstrate (other than
>> IDLE)?
> 
> Bioreason makes some products in Python.  You can see some screenshots of
> their Python/Gtk products here: http://www.bioreason.com/lead.html and
> http://www.bioreason.com/drug.html
> 

Thanks, I'll show those.

Lee


From tagith1394 at hotmail.com  Sat Sep  7 00:50:52 2002
From: tagith1394 at hotmail.com (Evan)
Date: Sat, 07 Sep 2002 04:50:52 GMT
Subject: Embedded Python implementation
Message-ID: <MAfe9.266658$v53.14092531@news3.calgary.shaw.ca>

Hello,

I am currently trying to embed Python in an application I am writing, 
and all is going well, except for one inexplicable bug..

When I pass a variable (pointer to a string) to PyRun_String, it will 
always error on line 2, character 5.. regardless of what is actually on 
the line in that spot (for that matter, even if there is no line 2)..

The script works perfectly when I place it in a file, and use 
PyRun_File, and it also works perfectly when I place the string in 
quotations, and pass it to PyRun_String directly inside the C code..

The C code in question is as follows:
if ( PyRun_String( com_list, Py_file_input, pGlobals, pLocals ) == NULL )
         PyErr_Print();

I am confident the problem is not with pGlobals or pLocals, as those 
arguments are used with PyRun_File, as well as with the quoted string..

Thanks in advance,
Evan



From michels at linmpi.mpg.de  Fri Sep 27 03:35:19 2002
From: michels at linmpi.mpg.de (Helmut Michels)
Date: Fri, 27 Sep 2002 09:35:19 +0200
Subject: ANN: Data Plotting Library DISLIN 8.0
Message-ID: <3D940A37.92FD45D5@linmpi.mpg.de>

I am pleased to announce version 8.0 of the data plotting software
DISLIN.

DISLIN is a high-level and easy to use plotting library for
displaying data as curves, bar graphs, pie charts, 3D-colour plots,
surfaces, contours and maps. Several output formats are supported
such as X11, VGA, PostScript, PDF, CGM, WMF, HPGL, TIFF, PNG and BMP.

The software is available for several C, Fortran 77 and Fortran 90
compilers. Plotting extensions for the interpreting languages Perl,
Python and Java are also supported for the most operating systems.

DISLIN distributions and manuals in PDF, PostScript and HTML format are
available from the DISLIN Home Page

     http://www.dislin.de

and via FTP from the server

     ftp://ftp.gwdg.de/pub/grafik/dislin

 -------------------
  Helmut Michels
  Max-Planck-Institut fuer Aeronomie  Phone: +49 5556 979-334
  Max-Planck-Str. 2                   Fax  : +49 5556 979-240
  D-37191 Katlenburg-Lindau           Mail : michels at linmpi.mpg.de




From rjones at ekit-inc.com  Thu Sep  5 19:35:40 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Fri, 6 Sep 2002 09:35:40 +1000
Subject: What database should I use
In-Reply-To: <20020905231134.GA1229@lilith.ghaering.test>
References: <3D77763F.23080.BFAFA3@localhost> <20020905231134.GA1229@lilith.ghaering.test>
Message-ID: <200209060935.40849.rjones@ekit-inc.com>

On Fri, 6 Sep 2002 9:11 am, Gerhard H?ring wrote:
> * A <printers at sendme.cz> [2002-09-05 15:20 +0200]:
> > Hello,
> > I am going to program a small application (in Python and wxPython) which
> > consists about 2000 - 3000 records . Can you please recommend which
> > database is the best for that. I use Windows.
> > Maybe MS Access  or FoxPro or MySQL or better?
> > I want to keep the program as small as possible.
>
> PySQLite. It uses an embedded SQL engine (SQLite), which means you don't
> have to set up a separate server process. It only weights a hundred KB
> or so and has good SQL support (a lot better than MySQL, for example).

If the database is small, and will have only single users at a time, then 
gadfly really can't be beat. It's pure python, meaning no mess, no fuss 
installation:

   http://gadfly.sf.net/

"python setup.py install" and you're ready to go.


      Richard




From Padraig at Linux.ie  Wed Sep 11 07:31:17 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Wed, 11 Sep 2002 11:31:17 GMT
Subject: deleting the first and the last character of a string
References: <mailman.1031741971.20624.python-list@python.org>
Message-ID: <3D7F2960.20507@Linux.ie>

Manuel Hendel wrote:
> I got a file with fields seperated by a "|". The problem is that there
> is also a "|" at the beginning and at the end of each line although
> there no fields in front or behind these "|". Now I want to delete
> these two "|". Is there anything which says delete the first or last
> character of a string? I checked the string module on the web, but I
> couldn't find anything.
> 
> I solved it like this, but I honestly don't like it that way.
> 
> #!/usr/bin/env python
> #
> import sys
> import string
> 
> inputfile = open(sys.argv[-1], "r")
> 
> for line in inputfile.readlines():
>         fields = string.split(line, "|")
>         lineN = string.join(fields[1:-1], "|")
> 
> Isn't there any better solution? ;-)
> 
> Manuel
> 
How about: line = line[1:-1]

P?draig.



From theller at python.net  Wed Sep  4 15:41:13 2002
From: theller at python.net (Thomas Heller)
Date: Wed, 4 Sep 2002 21:41:13 +0200
Subject: 4Suite / McMillan Installer / py2exe
References: <doc.m3admxkclw.fsf@goodweb.de>
Message-ID: <al5nkp$1n0m2t$1@ID-59885.news.dfncis.de>

"Thomas Korb" <Doc at goodweb.de> wrote in message news:doc.m3admxkclw.fsf at goodweb.de...
> Hi! I am using the XSLT-Processor from 4Suite (0.12.0a2), i.e.
>
>   from Ft.Xml.Xslt.Processor import Processor
>   from Ft.Xml.InputSource import DefaultFactory
>   from Ft.Xml import ReaderException
>
> On Linux, I can use McMillan's Installer to create a distribution;
> on Windows, it fails (problems with a Math.py-module in the 4Suite
> package). And py2exe complains about missing modules.
>
> I can overcome the py2exe-problem by importing everything from
> Ft.Xml.Xslt, i.e.
>
>   import Ft.Xml.Xslt.ApplyImportsElement
>   import Ft.Xml.Xslt.ApplyTemplatesElement
>   etc. etc. (a lot of etc. :-)
>
> (Too many module to include them on the commmand line with py2exe's
> '-p' option; more than 1024 characters!)
>
Hm, you can also list them in the setup.cfg file.
Something like this:
[py2exe]
includes=Ft.Xml.Xslt.ApplyImportsElement,
    Ft.Xml.Xslt.ApplyTemplatesElement,
    ....

And IIRC you can also use (on the command line)
  --includes Ft.Xml.Xslt.*
or even
  --packages Ft.Xml.Xslt


> After that, py2exe still complains about missing modules (some of
> which even do not exist on the system; I think at least some of them
> are Mac-specific!), but it works (as far as I could check up to now).
>
> But now the program takes quite a time to start (many imports), and

If you don't go the setup.cfg solution, you can probably speed it up
by only importing the modules in a dummy function (which is never
executed). Something like this:
def py2exe_hints():
   import Ft.Xml.Xslt.ApplyImportsElement
   import Ft.Xml.Xslt.ApplyTemplatesElement
   etc. etc. (a lot of etc. :-)

> I actually do not really trust this solution. Moreover, I would have
> to check and test all the above imports with each new version of 4Suite.
>
> Does someone know how to use 4Suite with the McMillan Installer or
> py2exe on Windows? Any help or hints would be highly appreciated!
>
I never used 4Suite, so no. I even heard ;-) that py2exe
doesn't work so good with _xmlplus.

Thomas




From paul at boddie.net  Tue Sep 10 03:59:20 2002
From: paul at boddie.net (Paul Boddie)
Date: 10 Sep 2002 00:59:20 -0700
Subject: how to build wxPython w/ Sun's C++ compiler?
References: <mailman.1031607011.28525.python-list@python.org>        <m3k7luojpq.fsf@mira.informatik.hu-berlin.de> <mailman.1031621231.16841.python-list@python.org>
Message-ID: <23891c90.0209092359.11acb607@posting.google.com>

Skip Montanaro <skip at pobox.com> wrote in message news:<mailman.1031621231.16841.python-list at python.org>...
> 
> As for patches, it seems to me that distutils needs a generic way to
> override such settings (CC, CFLAGS, LDFLAGS, etc) via command line
> arguments, environment variables or ConfigParser files.  It's not clear
> which way would be most Pythonic.  I'd lean toward environment variables,
> but that's just because the setup.py command line is cluttered enough as it
> is and I'm a Unix-centered geek.

What about configure programs, though? They tend to accept lots of
command line arguments, and it's hard to get more "Unix-centered" than
that these days. Moreover, the config.status file nicely summarises
what you presented to configure - much more convenient than
environment variables, surely, because you get to remember how you
built the package in question.

Paul


From jkraska at san.rr.com  Sat Sep 28 21:42:40 2002
From: jkraska at san.rr.com (Courageous)
Date: Sun, 29 Sep 2002 01:42:40 GMT
Subject: What does Python fix?
References: <e15b6fa3.0209271418.14cd5642@posting.google.com> <flaapucbduj6g4ifrdf3oq99hdrepi1o3v@4ax.com> <3d95fb09@news.vip.fi> <87ptuxub5x.fsf@pokey.henrik-motakef.de>
Message-ID: <vcmcpusnjqvpj9lavtqlqa7cs01h4i1iij@4ax.com>

>Let's not forget that Python's syntax (most of all the significance of
>whitespace) is considered strange at least by about all novices. In
>both cases - significant whitespace and lots of irritating superflous
>parentheses - people don't just get used to it, but tend to like it
>after a while.

I'm fairly sure that you meant to say "Lisp" where you said Python,
right?

>Another rather cool feature of lisp that python lacks are macros, btw,
>or in general the uniform handling of code and data.

True. I was doing some experimentation on hygenic macros for Python
a while back, doing parse-tree time insertions. Never finished.

>That would be nice. The one thing lisp lacks is a huge community,
>manifesting itself in lots of ready-to-use libraries, tutorials etc.

This is actually on the top 3 list of its biggest weaknesses.

It doesn't help that the number of big commercial Lisp compilers
can be counted on the fingers of, err, one finger.

C//



From paul at boddie.net  Mon Sep 16 04:48:13 2002
From: paul at boddie.net (Paul Boddie)
Date: 16 Sep 2002 01:48:13 -0700
Subject: [xml] convert funny chars to char entites?
References: <vNah9.10101$Ee4.23987@news-server.bigpond.net.au>
Message-ID: <23891c90.0209160048.7c630de4@posting.google.com>

":B nerdy" <thoa0025 at mail.usyd.edu.au> wrote in message news:<vNah9.10101$Ee4.23987 at news-server.bigpond.net.au>...
> ive got a string "Il Est N?" and i want to put it into a xml document
> 
> is there a quick and easy way to conver the funny characters to char
> entities so i can store them?
> or is there existing libraries?

Characters like "?" aren't generally converted to entities in XML,
unlike in various versions of HTML which had things like "&eacute;"
and which seemed overly ASCII-friendly. What I would recommend is that
you obtain a Unicode string from the string that you've been given, if
it isn't already Unicode (but I would guess by your enquiry that it
isn't), and then use the various XML APIs to write it into the
document.

Of course, if the string isn't Unicode, you have to know what encoding
was used in the creation of the string. For example:

  >>> s = "Il est n?!"
  >>> s
  'Il est n\x82!'
  >>> u = unicode(s)

At this point, some people naively assume that a Unicode string will
be created, but since most character sets do not assign "?" to "\x82",
Python can't be sure that you want to create "?" in the resulting
Unicode string.

  Traceback (most recent call last):
    File "<stdin>", line 1, in ?
  UnicodeError: ASCII decoding error: ordinal not in range(128)
  >>> u = unicode(s, "iso-8859-1")

Here, we've supplied a character encoding. Now, Python knows what to
do with that "\x82" character.

  >>> print u
  Traceback (most recent call last):
    File "<stdin>", line 1, in ?
  UnicodeError: ASCII encoding error: ordinal not in range(128)
  >>> print u.encode("iso-8859-1")
  Il est n?!

To print things out, you need to convert the Unicode string to the
appropriate native encoding.

Of course, there are various shortcuts that can be taken, especially
if you're not using the XML APIs, but are instead writing strings
directly to the XML file. In such cases, you can use an encoding
declaration at the start of the document...

  <?xml version="1.0" encoding="iso-8859-1"?>

...and then not bother converting the strings to Unicode, but an
awareness of Unicode is probably invaluable if you're going to be
doing a fair amount of XML work.

Paul


From thorsten at thorstenkampe.de  Mon Sep 16 13:43:16 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Mon, 16 Sep 2002 19:43:16 +0200
Subject: IDLE
References: <wpbf9.264320$aA.45691@sccrnsc02> <3D7D4F84.318CDB0F@bellatlantic.net>
 <v4Pg9.428631$me6.49396@sccrnsc01> <bs70f7ks.fsf@online.no> <am1efh$1ohti$1@ID-77524.news.dfncis.de>
 <EOjh9.2033$Cf.81@fe03>
Message-ID: <am557m$2nldt$2@ID-77524.news.dfncis.de>

* Steve Holden
>"Thorsten Kampe" <thorsten at thorstenkampe.de> wrote in ...
>> * syver-en at online.no
>>>"Ali K" <akakakhel at attbi.com> writes:
>>>> Hardware: 386 enhanced, 75MHz, 32MB RAM.
>>>> OS: Win98
>>>> Copied files from other computer to mine.
>>>> It doesn't open when I click on ide.pyw or when I choose to edit a py
>>>> file
>>>>
>>>> with it.
>>>
>>> I suggest you install the python distro from www.python.org.
>>
>> You think this will work on medieval computers?
>>
>
> Depends what you mean by mediaeval. [...]

"Hardware: 386 enhanced, 75MHz, 32MB RAM."

Thorsten


From chris.gonnerman at newcenturycomputers.net  Mon Sep 23 23:08:11 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Mon, 23 Sep 2002 22:08:11 -0500
Subject: confusion about opening files
References: <3d8fcda4$0$198$75868355@news.frii.net> <3D8FCFF3.FD52AFE5@alcyone.com> <3d8fd215$0$192$75868355@news.frii.net>
Message-ID: <00d101c26377$add6f8c0$0101010a@local>

----- Original Message ----- 
From: "robie1373" <robie at 1373.net>


> O. K. Now I am confused.  I thought that when I imported the 
> os module, i would have to call its functions like this: 
> os.open()    Why am I just using the function name?

os.open != open

open() is a builtin function, which "wraps" the C fopen() 
function (more or less), returning a file object implemented on 
top of stdio.h functions; os.open() "wraps" the C open() 
function, returning a file descriptor for the 
os.write()/os.read()/etc. functions.

If the function you WANTED was os.open, then you'd be right
about the behavior of imported modules.  However, what you want 
ISN'T os.open.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From ht at cogsci.ed.ac.uk  Mon Sep  9 16:59:41 2002
From: ht at cogsci.ed.ac.uk (Henry S. Thompson)
Date: 09 Sep 2002 21:59:41 +0100
Subject: Accessing Windows Registry under cygwin
Message-ID: <f5b1y8298w2.fsf@cogsci.ed.ac.uk>

On 11 September 2001 Norman Vine said "I have managed to get the
'_winreg' module working with Cygwin < _winreg = Python access to the
Windows Registry >" [1].  Assiduous searching does not reveal that
this work ever surfaced -- where might I find it?

Thanks

ht

[1] http://groups.google.com/groups?q=registry+cygwin+group:comp.lang.python.*&hl=en&lr=&ie=UTF-8&scoring=d&selm=mailman.1001510189.18703.python-list%40python.org&rnum=6
-- 
  Henry S. Thompson, HCRC Language Technology Group, University of Edinburgh
          W3C Fellow 1999--2002, part-time member of W3C Team
     2 Buccleuch Place, Edinburgh EH8 9LW, SCOTLAND -- (44) 131 650-4440
	    Fax: (44) 131 650-4587, e-mail: ht at cogsci.ed.ac.uk
		     URL: http://www.ltg.ed.ac.uk/~ht/
 [mail really from me _always_ has this .sig -- mail without it is forged spam]


From gua81 at XXXyahoo.com  Mon Sep 23 20:09:17 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Tue, 24 Sep 2002 10:09:17 +1000
Subject: Python quote in cgi
References: <amoabi$sjh$1@lust.ihug.co.nz>
Message-ID: <amoaf8$smq$1@lust.ihug.co.nz>

sorry I have not finished the post yet, accidentally click shift-enter

 Hi, this might have been discussed b4,
 can someone clarify for me, when writing cgi script in python,
 does ' = """
 eg

print """<a href="index.html">Main</a>"""

print '<a href="index.html">Main</a>'

are they both the same?




From gerhard.haering at gmx.de  Tue Sep  3 19:28:14 2002
From: gerhard.haering at gmx.de (Gerhard Haering)
Date: Wed, 4 Sep 2002 01:28:14 +0200
Subject: Database BSD on Mac?
In-Reply-To: <3144be3c280c062e5e71a98ab191f664.98737@mygate.mailgate.org>
References: <3144be3c280c062e5e71a98ab191f664.98737@mygate.mailgate.org>
Message-ID: <20020903232814.GA5482@gargamel.ghaering.test>

* Max <lu_gio at hotmail.com> [2002-09-03 23:19 +0000]:
> Hi!
> I'm implementing a simple data manager to store few images and sounds
> (<5Mb).
> I'm developing a game on windows platform (so, I don't need a client
> server db like mysql) but I'd like to port the software on every python
> supported platform (well, principally on win,linux and mac).
> 
> I was looking on BSD but I'm not sure it is ported to all platforms.
> Wich db should I utilize?

Do you really need one? Or do either pickle or shelve suffice?

-- Gerhard



From ptbjrii at hotmail.com  Tue Sep 24 14:51:55 2002
From: ptbjrii at hotmail.com (Polerio Babao Jr.II)
Date: 24 Sep 2002 11:51:55 -0700
Subject: Nokia 3310 SMS Application, done:connecting to serial port(linux/win), done:reading some hex like \x00 . . ., problem:write/send command! noresponse
References: <52facf1d.0209190533.3048e62c@posting.google.com> <r77ooukb98cbfbpcqbitl9g53vuokkl79s@4ax.com>
Message-ID: <52facf1d.0209241051.17782a30@posting.google.com>

> Something like 
> 
> import serial
> ser=serial.Serial(0,19200,timeout=3)
> print ser.portstr
> ser.write("ATZ\r")
> print "got"
> line=ser.read(10)
> print line
> ser.close()
> 
I have tried this to my Nokia 3310 and still it didn't worked. I've
tried to make some messages in logomanager software that runs on
windows and it works, this assured me that the cable and cell phone
I'm using works fine. The only problem perhaps is that python seems to
be unable to speak with that device or perhaps there's an additional
python scripts needed to make this work. The one application I'ved
used also was the demo version of oxygensms I think, which was created
in delphi and it works very well, I can send and receive messages in
that application. Is it really possible to make this sms application
with python using nokia cellphones? It works fine in siemens
cellphones as you said but what's the problem in this kind of
cellphones(nokia), do they have security standard?, do they response
only win2k/98-created application, what I can only do I guess is to
wait for the response of the python masters regarding this. I think
that python would not allow this kind of devices to be
unknown/stranger to them forever.
Thank you for your suggestion.


From jmdeschamps at cvm.qc.ca  Mon Sep 16 11:40:28 2002
From: jmdeschamps at cvm.qc.ca (jmdeschamps)
Date: 16 Sep 2002 08:40:28 -0700
Subject: possibly OT - delete char?
References: <3pch9.10253$Ee4.26056@news-server.bigpond.net.au> <kWfh9.113071$pX1.4053630@news2.tin.it>
Message-ID: <3d06fae9.0209160740.537415d3@posting.google.com>

> ...
> For Windows, the effbot's Console package,
> http://effbot.org/efflib/console/ .
> 
> 
> Alex

There is also 
http://newcenturycomputers.net/projects/wconio.html
which works well !
Cheers,

Jean-Marc


From lfini at arcetri.astro.it  Wed Sep  4 05:03:35 2002
From: lfini at arcetri.astro.it (Luca Fini)
Date: Wed, 4 Sep 2002 11:03:35 +0200
Subject: Why Python?
In-Reply-To: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com>
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com>
Message-ID: <Pine.LNX.4.44.0209041047500.13804-100000@localhost.localdomain>

On Wed, 4 Sep 2002, it was written:

> Hi guys (and girls)
> I just begun to learn how to program. And I choosed Python for
> doing that. surfing around the web i found an article about this
> language, they said it's the ideal language for beginners to the
> world of programming. and i admit it, the syntax is very clear making
> the code very readable, i learn easily the concepts of programming
> and i soon wrote some nice programs. but since i finished my
> newbie-tutorial, i am wondering about one question: is it worth to go
> deeper into python? i talked to a friend of mine, he told me that i
> should try php..
> what should i do?? ^_^ in other words: what is python good for?
> why should i continue with python?
> thank you!

Let me rephrase your message:

Hi guys (and girls)
I just begun to learn how to drive. And I choosed a Toyota Corolla for
doing that. surfing around the web i found an article about this car, they
said it's the ideal car for beginners to the world of driving. and i admit
it, the commands are very handy making the driving very easy, i learn
easily the concepts of driving and i soon made some nice travels. but
since i finished my newbie-tutorial, i am wondering about one question: is
it worth to buy a  Toyota Corolla? i talked to a friend of mine,
he told me that i should try Mercedes 240. what should i do?? ^_^ in other
words: what is Toyota Corolla good for? why should i continue with Toyota
Corolla?

Well, maybe not every detail fits perfectly, but I think the point is 
clear.

Programming languages, like cars, all do more or less the same things, but 
each one has peculiar features which may be more or less suited to a 
particular need. You can drive a four-wheeler to go shopping but if you 
never go in the wild it would be better to buy a more comfortable car.

I.e.: first try to make your programming needs clear, and then you can 
evaluate advice on which programming language to use.

Cheers,
						l.f.  
 -- 
--------------------------------------------------------------------------
       -- )     Luca Fini                            Tel: +39 055 2752 307
 ___    |\      Osservatorio Astrofisico di Arcetri  Fax: +39 055 2752 292
/   |   | |-_   L.go E.Fermi, 5 +-----------------------------------------
(___|___//___)  50125 Firenze  /   WWW: http://www.arcetri.astro.it/~lfini
 (_)      (_)   Italia        / e-mail: lfini at arcetri.astro.it
-----------------------------+--------------------------------------------



From max at alcyone.com  Wed Sep  4 00:40:10 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 03 Sep 2002 21:40:10 -0700
Subject: Privacy and Inheritance
References: <mailman.1031112136.4594.python-list@python.org>
Message-ID: <3D758EAA.FC7F5D4E@alcyone.com>

"Dennis B." wrote:

> Have enough confidence in Python to realize this is probably just my
> misunderstanding, but wondering why Python mangles private, double
> underscore attributes with the name of the base class even if they're
> inherited and the instance is that of the sub class.

Double underscore mangling is done with the name of the class involved,
so it is in effect a private variable to that class only; subclasses
will not have access to it (without doing the unmangling yourself,
obviously, as you've seen).

The use of double underscores to mangle names should be viewed as
something specific to that class which nothing else will ever use; in
terms of access restriction it's similar to the private keyword in C++.

What you want is something local to that class and its subclasses, which
would correspond to C++ protected access.  There's no mangling mechanism
to do this, although a common convention (and one I use when I'm really
concerned about access restriction in Python, which is fairly rare) is
to use a single underscore.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ There is nothing so subject to the inconstancy of fortune as war.
\__/ Miguel de Cervantes
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From luc.saffre at gmx.net  Mon Sep 16 09:01:57 2002
From: luc.saffre at gmx.net (Luc Saffre)
Date: 16 Sep 2002 06:01:57 -0700
Subject: "if x == None" raises "'NoneType' object is not callable"
Message-ID: <af84f376.0209160501.4d21500@posting.google.com>

"""
If I define __getattr__() and __setattr__() for a class, then I cannot
test instances of this class for equality with None.

the code example below raises
TypeError: 'NoneType' object is not callable

Strange! Nobody told you to call 'None'! Who knows an explanation?

Python 2.2.1 (#1, Jun 25 2002, 10:55:46)
[GCC 2.95.3-5 (cygwin special)] on cygwin
        
"""

class Row:

   def __init__(self):
      self.__dict__["_values"] = {}

   def __getattr__(self,name):
      try:
         return self._values[name]
      except KeyError,e:
         AttributeError,str(e)
   
   def __setattr__(self,name,value):
      self.__dict__["_values"][name] = value


row = Row()
if row == None: # here it happens.
   print "row instance is None!"


From timh at zute.net  Tue Sep 24 09:11:31 2002
From: timh at zute.net (Tim Hoffman)
Date: Tue, 24 Sep 2002 21:11:31 +0800
Subject: UK ZOPE Contract.
References: <1ed550fd.0209230815.2ec6304a@posting.google.com>
Message-ID: <3D906483.70704@zute.net>

Hi

Do you need people on the ground there or can you use a tele-commuter ?

See ya

Tim Hoffman

michela rossi wrote:
> Hi,
> 
> We have a ZOPE contract up and coming soon. Around two months work. If
> there are any ZOPE contractors available, please get in contact with
> me.
> 
> Regards, Michela.




From sabren at manifestation.com  Thu Sep 26 10:29:06 2002
From: sabren at manifestation.com (Michal Wallace)
Date: Thu, 26 Sep 2002 10:29:06 -0400 (EDT)
Subject: revenge of the python puzzle
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8BF0A5E3@natasha.auslabs.avaya.com>
Message-ID: <Pine.LNX.4.44.0209261018590.22263-100000@hydrogen.sabren.com>

On Thu, 26 Sep 2002, Delaney, Timothy wrote:

> > From: Michal Wallace [mailto:sabren at manifestation.com]
> > 
> > Okay, so you guys think you're so smart. Try this one!  No
> > contest this time, I've learned my lesson.  (I suppose if
> > someone really wants some free hosting bad enough to solve
> > this puzzle, you could talk me into it, but I'm just posting
> > this for revenge... Um, I mean "fun".)
> 
> Hmm - a little bit harder. The fake perl actually made it
> easier though (gave while and s).


Oh well. :) I think this is just not a good way to make a
python puzzle. 

I guess I could avoid all keywords with something like this:

   open("/dev/stdout","w").write("hello world")

and then try to make it as convoluted as possible, but then
it's not really cross platform.

Was it obvious what it did? :)

Cheers,

- Michal   http://www.sabren.net/   sabren at manifestation.com 
------------------------------------------------------------
Switch to Cornerhost!             http://www.cornerhost.com/
 Low Priced, Reliable Blog Hosting, With a Human Touch. :)
------------------------------------------------------------




From david at no.westcontrol.spam.com  Wed Sep 11 04:54:44 2002
From: david at no.westcontrol.spam.com (David Brown)
Date: Wed, 11 Sep 2002 10:54:44 +0200
Subject: Multi-threading python
Message-ID: <almvvt$34p$1@news.netpower.no>

I've just been reading a little about the GIL and threading in python, and I
have a couple of questions that I hope someone can answer.

Am I right in thinking that threading in Python does not use the underlying
OS's threads, but effectively implements its own scheduler?  Or does it use
the OS's threads, but uses the GIL to ensure that only one is runable at a
time?

On a dual-processor system, does this mean that only one thread of a Python
program can run at a time, even when both CPUs are free, so that two
seperate python interpreter processes are needed to take advantage of the
two CPUs?

What about when a Python thread is blocked for some other reason, such as
while waiting for a file read - will that block other threads, or will the
waiting thread definitely release the GIL and allow other threads to run?

If the GIL does block all threads in this manner, are there any plans to
introduce more fine-grained locking to improve scalability, analagous to
locking in the Linux kernel?






From marklists at mceahern.com  Thu Sep 19 08:55:56 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 19 Sep 2002 07:55:56 -0500
Subject: How do I reply to an item in the list ?
In-Reply-To: <02091914425601.01191@pc20_118>
Message-ID: <JHEOKEOOLIGLDHCMAHMOOENPDPAA.marklists@mceahern.com>

> I 'd like to answer a message, e.g. on the topic ?Installation of
> multiple versions of Python on Unix?, but on the page
> http://mail.python.org/pipermail/python-list/2002-September/120875.html
> which contains the message i 'd like to ?reply? to, there is NO
> ?replybutton?. How can i do ?

Use groups.google.com?

Use a news reader?

// m

-




From maxm at mxm.dk  Tue Sep 17 03:44:41 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 17 Sep 2002 09:44:41 +0200
Subject: Converting Python app to C++ completely
References: <aklm6c$r1r$1@news.swman.net.uk> <3D85D9EB.9020102@thinkware.se> <roy-2A7271.09563416092002@reader1.panix.com> <3D86209F.4080501@earthlink.nospam> <am5am0$fi6$1@panix2.panix.com> <3d868a77$1@news.sentex.net> <roy-32ADCF.22220316092002@reader1.panix.com>
Message-ID: <3D86DD69.9010609@mxm.dk>

Roy Smith wrote:


> I'm reasonably convinced that I'll never need C++ to solve real 
> problems.  I'm also reasonably convinced that I do need C++ to get 
> invited to job interviews.  This thought courtesy of today's economy.


You should probably learn Java or C# then ....


regards Max M



From marklists at mceahern.com  Fri Sep 27 10:34:55 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 27 Sep 2002 09:34:55 -0500
Subject: watching mutables?
In-Reply-To: <an1p3c$jrd$1@news.hccnet.nl>
Message-ID: <JHEOKEOOLIGLDHCMAHMOKEBCECAA.marklists@mceahern.com>

> Subject: watching mutables?

class Watched(list):

  def append(self, element):
    print "appending %s" % element
    list.append(self, element)

l = Watched()
l.append(1)

// m



From gerhard.haering at opus-gmbh.net  Wed Sep  4 14:45:42 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 4 Sep 2002 18:45:42 GMT
Subject: Does BlackAdder really suck?
References: <slrnanci6i.1s8.gerhard.haering@haering.opus-gmbh.net> <al5ivg$e8v$1@rex.ip-plus.net>
Message-ID: <slrnancl3o.2fo.gerhard.haering@haering.opus-gmbh.net>

In article <al5ivg$e8v$1 at rex.ip-plus.net>, F. GEIGER wrote:
> "Gerhard H?ring" schrieb:
>> I just checked out the BlackAdder demo and was quite disappointed.
> 
> BlackAdder crashed my NT box, when I ran it the 1st time. Never fired it
> up again...

The NT box? Wise decision 

> Did you have a look at wxDesigner?

Yes.  It seems to support wxSizers only, which IMO is relatively easy to do
with just Vim + Python already.

-- Gerhard


From lbrannma at cablespeed.com  Sun Sep 22 15:12:56 2002
From: lbrannma at cablespeed.com (Lance)
Date: Sun, 22 Sep 2002 12:12:56 -0700
Subject: sys.ps2 assignment has no effect
References: <uos1np8t53k396@corp.supernews.com> <3d8e0aae@news.sentex.net>
Message-ID: <uos5dis0ogvc6f@corp.supernews.com>

print '...' works correctly, showing ...




"Peter Hansen" <peter at engcorp.com> wrote in message
news:3d8e0aae at news.sentex.net...
> Lance wrote:
> > I'm running the latest Python on Windows XP. The default installed
secondary
> > string appears to be blank spaces rather than an ellipse. The following
has
> > no effect:
> >
> > import sys
> > sys.ps2 = '... '
> What does "print '...'" show you?  Maybe this is just a phenomenon of
> your current font?
>
>
>




From eppstein at ics.uci.edu  Wed Sep  4 17:24:42 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Wed, 04 Sep 2002 14:24:42 -0700
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com>
Message-ID: <eppstein-56D20B.14244204092002@news.service.uci.edu>

In article <m7ncnugbkm0qghf66qcub6e37b64fesdkq at 4ax.com>,
 Gon?alo Rodrigues <op73418 at mail.telepac.pt> wrote:

> Anyway, I've been directed to PEP 227, cheerfully obeyed, and from what
> I have understood having read-only nested scopes was more of a design
> decision than something else. And while I can live happily with the
> decision (some would even say happier) there is something about the "You
> can look at it but not touch it" attitude that I dislike ;-)

I'm sure there were good reasons for this decision, but it leads to ugly 
workarounds like making each nonlocal variable into a list and assigning 
to x[0] instead of x...

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From david at no.westcontrol.spam.com  Mon Sep  9 07:14:00 2002
From: david at no.westcontrol.spam.com (David Brown)
Date: Mon, 9 Sep 2002 13:14:00 +0200
Subject: Why Python?
References: <mailman.1031446691.2023.python-list@python.org>
Message-ID: <alhvde$gim$1@news.netpower.no>

"Andres Rosado" <arosado at softhome.net> wrote in message
news:mailman.1031446691.2023.python-list at python.org...

> Also, I started recently to learn PHP. I wouldn't recommend it for a
newbie
> at programming, unless he/she needs it.
>

PHP has one area of use - web sites.  Outside that, it's a not too great,
but within that, it is an excellent language, and is extremly easy for a
newbie to get the hang of.  When you are constructing web pages, it is very
easy to mix php and html so that you only program the bits you need to.  On
the other hand, if your site involves doing extensive programming work, then
php is less suitable - you are better of with something like python.  If I
were constructing a serious web application (something I will do soon), I'd
do most of it in php, but with some of the hard work behind the scenes done
with python.






From tjreedy at udel.edu  Wed Sep 25 11:34:29 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 25 Sep 2002 15:34:29 GMT
Subject: problem with dictionaries within lists
References: <pan.2002.09.25.05.21.07.477231.167@ihatespam.shaw.ca>
Message-ID: <8Ikk9.283971$AR1.12018241@bin2.nnrp.aus1.giganews.com>

Two code comments:

Using 'list' and 'dict' as a variable names will bite when you want to
use the builtin functions someday.  I'll call them pairs and hdict.

I believe that a huge number of lines like
> rcTN = re.compile('Technical_Knowledge')
etc
...
> elif rcTN.match(list[0]):
> dict['Technical_Knowledge'] = list[1]
etc
could be replaced and sped-up by

headers={'Technical_knowledge':1, etc...} # in 2.3, would use new set
type
...
if pairs[0] in headers:
    hdict[pairs[0]] = pairs[1]

Highly repetitive code like this can usually be nicely condensed by
graceful use of Python lists and dicts.

Terry J. Reedy




From jepler at unpythonic.net  Fri Sep 27 15:54:19 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Fri, 27 Sep 2002 14:54:19 -0500
Subject: ENOSPC doesn't cause exception?
Message-ID: <20020927195418.GF5941@unpythonic.net>

Can someone tell me why this doesn't cause an exception?
$ python -c print > /dev/full

(/dev/full is a special device found on at least some kinds of Linux,
which always returns -ENOSPC when written)

if I use 'strace', I can see that -ENOSPC is returned from write():
$ strace -e write python -c print > /dev/full
write(1, "\n", 1)                       = -1 ENOSPC (No space left on device)

I would expect this to print something like
Traceback (most recent call last):
  File "<string>", line 1, in ?
OSError: [Errno 28] No space left on device

I tested this on Python 1.5.2 and Python 2.3a0 CVS, so it doesn't appear
to be anything new...

Jeff



From henrik.motakef at web.de  Wed Sep 18 17:27:41 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 18 Sep 2002 23:27:41 +0200
Subject: Calling function from command line
References: <amao3t$mug$1@netnews.upenn.edu>
Message-ID: <87admfng36.fsf@pokey.henrik-motakef.de>

gregoryk at futures.wharton.upenn.edu (GREGORY KNESER) writes:

> c:\python22\python func.py func1('input1','input2')
> 
> or some such thing and run the function with the inputs I fed it rather 
> than the defaults.  Is such a call possible?

Not quite, but maybe something like

  python -c "import func; func.func1('input1','input2')"

works for you. Or perhaps

  echo 'func1("input1", "input2")' | python -i func.py

hth
Henrik


From jknapka at earthlink.nospam  Mon Sep 16 14:21:11 2002
From: jknapka at earthlink.nospam (Joe Knapka)
Date: Mon, 16 Sep 2002 18:21:11 GMT
Subject: Converting Python app to C++ completely
References: <aklm6c$r1r$1@news.swman.net.uk> <3D85D9EB.9020102@thinkware.se> <roy-2A7271.09563416092002@reader1.panix.com>
Message-ID: <3D86209F.4080501@earthlink.nospam>

Roy Smith wrote:
> Magnus Lycka <magnus at thinkware.se> wrote:
> 
>>A common approach is to factor out the performance critical
>>parts (once you have actually measured) and rewrite them in
>>C or C++. The profiler is a good tool for this measuring.
> 
> 
> I'm in the process of doing exactly that.  I've got an app I wrote in 
> Python which needed speeding up.  Profiling showed that 99% of the CPU 
> time was spent in one class, which did a lot of low-level character 
> processing.  Mostly what it did was thrash around creating string 
> objects and tearing them down again in a rather un-pythonic way.
 >
> Seemed like a perfect opportunity to learn C++, so I re-wrote the class 
> in that language.  Somewhat surprisingly, I only got about a 2x speedup.  
> I think the reason is because now instead of spending all my time 
> allocating and deallocating Python strings, I'm doing the same with C++ 
> strings :-)
> 
> I'm going to take a shot at re-writing it again using C-style arrays of 
> characters.  It'll be interesting to see how much speedup I get.  I 
> figure one of two things will happen; either the C version will be much 
> faster than the C++ version, or it won't.  Either way, I figure I will 
> have learned something about C++, and maybe about Python too :-)

You might try rewriting it in Python, using lists of characters
instead of strings for the stuff that's causing performance
trouble. Given str, lst = list(str), then do all the processing
(in place) on lst, and ''.join(lst) to turn it back into a string.

Cheers,

-- Joe



From quiteblack at yahoo.com  Tue Sep 24 04:48:39 2002
From: quiteblack at yahoo.com (black)
Date: Tue, 24 Sep 2002 01:48:39 -0700 (PDT)
Subject: confused with lambda~
Message-ID: <20020924084839.30815.qmail@web21307.mail.yahoo.com>

that describe in reference documentation is too short, I couldnt even understand what it trying to say. any explanation please ?


---------------------------------
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020924/6046f6a5/attachment.html>

From wlfraed at ix.netcom.com  Wed Sep 25 20:24:34 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Wed, 25 Sep 2002 17:24:34 -0700
Subject: Python mixed with HTML like PHP
References: <o9h3pu4f24ffnupaen8ut8v8dgokacdna6@4ax.com>
Message-ID: <24ktma.4f3.ln@ix.netcom.com>

Johann fed this fish to the penguins on Wednesday 25 September 2002 
07:17 am:

> I heard about project which can mix python cgi code with html content.
> I forgot its name. Who remember?
> 

        Not Zope, by any chance?

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From peter at engcorp.com  Sat Sep  7 10:46:05 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 07 Sep 2002 10:46:05 -0400
Subject: Zope Status?
References: <Xns9281221A07167bosahvnetscapenet@66.185.95.104> <3d78a657$1@news.sentex.net> <Xns9282D14CD9ECbosahvnetscapenet@66.185.95.104>
Message-ID: <3d7a1115@news.sentex.net>

Bo wrote:
> Peter Hansen <peter at engcorp.com> wrote in
> news:3d78a657$1 at news.sentex.net: 
> 
>>Very likely, going to http://dev.zope.org would answer these
>>questions. 
> 
> I looked there but it seems somewhat vague and out of date.

I can't help the "vague" part, if that's your perception of it,
but looking at "RecentChanges" it seems clear that the wiki is
quite active and maintained.  Maybe the pages you are looking
at simply don't require an update at this time.

> Project Plan
> 
> The Zope 2.6 release is tentatively planned for Q2 2002. Once the 
> ProposedFeatures have been solidified into a final feature set, we'll put a 
> table here with status and links to project areas for the major components 
> of the project.

The way I read that, this release is far enough in the future they
don't yet have a definite plan for what it will contain or when it
will be released.  That's certainly vague, but I suspect it's very
accurate.

> The site says they are gathering features for zope 2.6 and working on a 
> vision statement for 2.7 which will use python 2.2.

Sounds clear enough.  After all, Zope 2.5.1 just came out, so they
can't be expected to have a beta for 2.6 ready yet, can they?

To answer your original first two questions then:

1. They will be moving to Python 2.2 roughly no earlier than the latter
half of 2003.

2. They won't be releasing Zope 3 for a long time yet.  (But is
anyone really basing their plans on when the third-next release
of a product is going to be available?)

> I'm just curious if some people involved in the code would share their 
> progress\hopes.

Well, subscribing to, say, the developer mailing list (see
http://www.zope.org/Resources/MailingLists) would certainly be the best
place to get access to such opinions.

-Peter



From aleax at aleax.it  Mon Sep 16 15:01:17 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 16 Sep 2002 19:01:17 GMT
Subject: search re and put in a list??
References: <mailman.1032198139.2621.python-list@python.org>
Message-ID: <1Uph9.102479$ub2.2109716@news1.tin.it>

jubafre at brturbo.com wrote:

> i have 3 lists:
> 
> mem=['7','8', '9']
> labels['D1','D2','D3']
> S=['20 D1', '30 D2', '10 D3', 'F0', '  3', '  2', '  0']
> 
> I want to put mem in s trougth of labels.
> 
> #+++++++++++++++++++++++++++++++++++++++++++++
> final=[]
> for i in range(len(j)):
>     for c in range(len(s)):
>         p= re.compile(labels[i])
>         t1=p.search(s[c])
>         if t1:
>             t=sub(p, j[i], s[c].strip())
>             final.append(t)
> print final
> #+++++++++++++++++++++++++++++++++++++++++++++
> 
> but doing like this i have just de nodes replaced,
> ['20 7', '30 8', '10 9']
> 
> and i want the replaced and the not replaced in order like this:
> ['20 7', '30 8', '10 9', 'F0', '  3', '  2', '  0']

So you want to replace each occurrence of labels[x] with
mem[x] in each string in the list of strings S, to
produce the list of strings final.  That seems to be
just what the sub method of RE objects is for, e.g.:

final = S[:]
for label, smem in zip(labels, mem):
    labelre = re.compile(label)
    for i in range(len(final)):
        final[i] = labelre.sub(final[i], smem)

If the labels and mem lists are long, you could find
the following alternative faster -- try both:

repdict = dict(zip(labels, mem))
big_re = re.compile('|'.join(labels))
def replacer(mo, repdict=repdict):
    return repdict[mo.group()]
final = [ big_re.sub(x, replacer) for x in S ]


Alex



From martin at v.loewis.de  Sat Sep 21 15:47:01 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 21 Sep 2002 21:47:01 +0200
Subject: Python 2.2.1 compatible with Tcl/Tk 8.4?
References: <yu9965x05vkr.fsf@europa.research.att.com> <j4ptv89zwk.fsf@informatik.hu-berlin.de> <3D8CBFD1.4010808@aon.at>
Message-ID: <m3k7lfruq2.fsf@mira.informatik.hu-berlin.de>

Gregor Lingl <glingl at aon.at> writes:

> 1. When working with Python 2.2.1, are there any advantages of
> using Tcl/Tk 8.4  over the version which comes with Python2.2.1 for Window?

I don't know; I guess you should refer to the Tcl release announcement
for details.

> 2. Which steps would I have to perform, to change to the new version?

You need to recompile _tkinter.pyd. See PCbuild/readme.txt for
details.

Regards,
Martin


From dsavitsk at e-coli.net  Sat Sep 14 22:35:24 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Sun, 15 Sep 2002 02:35:24 GMT
Subject: representing a literal single slash
References: <dd5454fb.0209141828.1ad24260@posting.google.com> <ehSg9.8464$yt3.4491656@newssrv26.news.prodigy.com>
Message-ID: <MlSg9.8467$yt3.4492002@newssrv26.news.prodigy.com>

"dsavitsk" <dsavitsk at e-coli.net> wrote in message
news:ehSg9.8464$yt3.4491656 at newssrv26.news.prodigy.com...
>
> "Jonathan Claggett" <hellen at claggetts.net> wrote in message
> news:dd5454fb.0209141828.1ad24260 at posting.google.com...
> > Here is an easy question (I'm hoping).
> >
> > I'm trying to process a string so that all doubled up backslashes (\\)
> > are replaced with a single backslash (\).
> >
> > The obvious approach doesn't work since a single backslash can't end a
> > string:
> >
> > str = str.replace('\\', '\') # broken
>
> in the above you are replacing a single slash with, well, with nothing.
try
> >>> str = str.replace('\\\\', '\\') # [not] broken
>
> -d

you might also want to rename that variable.  else the str() function will
be a problem.




From fpm at u.washington.edu  Tue Sep 17 00:19:22 2002
From: fpm at u.washington.edu (Frank Miles)
Date: 17 Sep 2002 04:19:22 GMT
Subject: pypgsql -- not preserving transactions ?
References: <am3id4$1epu$1@nntp6.u.washington.edu> <mailman.1032151518.3079.python-list@python.org>
Message-ID: <am6aga$21tu$1@nntp6.u.washington.edu>

In article <mailman.1032151518.3079.python-list at python.org>,
Gerhard =?iso-8859-1?Q?H=E4ring?=  <gerhard.haering at gmx.de> wrote:
>First, I'd recommend to ask questions like this on the pypgsql-user
>mailing list. You'll find the link on  http://pypgsql.sf.net/

Thanks for the recommendation.  I had tried to look at the archives of the
list only to find "page not found".  Nothing pertinent in the FAQ, either.
Still...

>* Frank Miles <fpm at u.washington.edu> [2002-09-16 03:15 +0000]:
>> I've discovered that the 2.0 version of pyPgSQL packaged with Debian/woody
>> doesn't preserve transactions...
[snip]

>I can't reproduce that behaviour here. What do you do to ignore the
>PgSQL.OperationalError? Or is it a different exception that gets thrown?

Yes, that exception gets thrown.  Perhaps this simply reveals a gap in my
understanding -- I thought that it was Postgres that was handling the
transactions.  Presuming that the cursor wasn't starting a new transaction,
even if the dumb user used a try/except block to bypass the exception (and
continue with succeeding INSERT attempts), shouldn't Postgres prevent those
later INSERTs?

[snip]

>This area of code hasn't changed up to 2.2, IIRC. Could you provide a
>test case with a minimal schema, minimal code and the output of your
>code. To see which SQL commands pyPgSQL sends, you can simply access
>pyconn.conn.toggleShowQuery, where pyconn is your DB-API connection
>object. This output would be useful, too. Also, which PostgreSQL server
>version are you using?
>
>-- Gerhard

Sure, I've got some simple test code.  But at this point I'm now thinking
that this is more likely a difference in conception -- what _should_
pypgsql do?  Should the application have to track if/when exceptions occur?
It's probably more efficient, though ISTM less clean to do it this way.

Thanks for your reply, Gerhard.  Can you (or someone else) confirm that the
"error" is a difference in conception?

	-frank
-- 


From dsavitsk at e-coli.net  Sat Sep 14 18:57:47 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Sat, 14 Sep 2002 22:57:47 GMT
Subject: win32security.LogonUser not working
References: <x6Pg9.8439$yt3.4473830@newssrv26.news.prodigy.com>
Message-ID: <L9Pg9.8440$yt3.4474038@newssrv26.news.prodigy.com>

should have mentioned
python:   2.1.3
win32all: 145
win2k:    sp2


"dsavitsk" <dsavitsk at e-coli.net> wrote in message
news:x6Pg9.8439$yt3.4473830 at newssrv26.news.prodigy.com...
> I am trying to run code as another user. I copied the code below from
ASPN,
> so it would seem that someone has gotten this to work. I keep getting the
> error
>
> pywintypes.api_error (1314, 'LogonUser', 'A required privilege is not held
> by the client.')
>
> Google has revealed that others gave had the same problem.  does anyone
know
> a solution, or even a different way to do this?
>
> -doug
>
> link to the original of code below
>
http://www.google.com/search?q=cache:Jnbt2XVyGfUC:aspn.activestate.com/ASPN/
>
Python/Reference/Products/ActivePython/PythonWin32Extensions/Windows_NT_Secu
>
rity_.2d.2d_Impersonation.html+self.handel%3Dwin32security.LogonUser&hl=en&i
> e=UTF-8
>
> ---------------------------------------------
> import win32security
> import win32con
> import win32api
> import sys
>
>
> class Impersonate:
>
>     def __init__(self):
>         self.domain   = '.'
>         self.login    = 'myusername'
>         self.password = '**********'
>
>     def logon(self):
>         self.handel=win32security.LogonUser(self.login,
>                                             self.domain,
>                                             self.password,
>
> win32con.LOGON32_LOGON_INTERACTIVE,
>
> win32con.LOGON32_PROVIDER_DEFAULT
>                                             )
>         win32security.ImpersonateLoggedOnUser(self.handel)
>
>     def logoff(self):
>         win32security.RevertToSelf() #terminates impersonation
>         self.handel.Close() #guarantees cleanup
>
>
>
>
> if __name__ == '__main__':
>     a=Impersonate()
>     try:
>         a.logon() #become the user
>         print win32api.GetUserName() #show you're someone else
>         a.logoff() #return to normal
>     except:
>         print sys.exc_type , sys.exc_value
>
>




From martin at v.loewis.de  Thu Sep 12 02:50:49 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 12 Sep 2002 08:50:49 +0200
Subject: ntpath bug or expected?
References: <xXRf9.7957$yt3.4120228@newssrv26.news.prodigy.com> <mailman.1031795954.4770.python-list@python.org> <A3Tf9.7966$yt3.4129854@newssrv26.news.prodigy.com>
Message-ID: <m3u1kv66ra.fsf@mira.informatik.hu-berlin.de>

"dsavitsk" <dsavitsk at e-coli.net> writes:

> it seems to me, i guess, that being inconsistent here and in
> os.getcwd in a consistent way is confusing.  if getcwd returned C:
> instead of C:\\ (as it returnes C:\\path and not C:\\path\\) then
> os.path.join(os.getcwd(), 'path', 'file.txt') with my modification
> would do the same thing as it now does, but it would have the added
> bonuses of returning a full file path and removing 2 bits of
> necessary documentation from the language.

As Jeff explains, that confusion really comes from the system, in
particular, from the notion of lettered drives, the notion of a
"current drive", and the notion of a per-drive "current directory".

That results in c:foo being a relative path, not an absolute one, and
it is different from c:\foo. Therefore, os.path.join("c:","foo") and
os.path.join("c:",os.sep,"foo") are different things.

Regards,
Martin


From mjais at web.de  Tue Sep 24 07:06:24 2002
From: mjais at web.de (Markus Jais)
Date: Tue, 24 Sep 2002 13:06:24 +0200
Subject: Difference between MOD_PYTHON and the CGI module
References: <3d904321$0$234$4d4ebb8e@news.nl.uu.net>
Message-ID: <bHXj9.2$GT1.920@news.ecrc.de>

Gumuz wrote:

> What are the differences, if any?
> 
> Am i mixing up to things here?
> 
> Suppose I'd like to use python as a 'replacement' for PHP, what do I use?
> 
> thanks,
> 
> gumuz

with mod_python you have the python interpreter within apache
as with mod_perl. this is faster and also allows you to access
the apache api which lets you do much more things like plain CGI.

if you want to compete with PHP I think mod_python would be better.


markus


From marklists at mceahern.com  Mon Sep  9 14:35:31 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 9 Sep 2002 13:35:31 -0500
Subject: shell expansion
In-Reply-To: <JHEOKEOOLIGLDHCMAHMOMEHFDNAA.mark@mceahern.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOGEHGDNAA.marklists@mceahern.com>

[I wrote]
> Is this supposed to work:
> 
>   filename = "$HOME/foo"
>   f = file(filename)
>   ...
> 
> It doesn't seem to.

Doh, that's what os.path.expandvars() is for.

Sorry,

// m
-




From akineko at pacbell.net  Sun Sep  8 13:01:14 2002
From: akineko at pacbell.net (Aki Niimura)
Date: 8 Sep 2002 10:01:14 -0700
Subject: Accessing Tk-inter Popup menu entry
Message-ID: <b714b8de.0209080901.6c8a7985@posting.google.com>

Hi,

I have created a popup menu. (see the attached example code)
Now I would like to change the state of a popup menu entry dynamically.
I tried to use 
    popup.entryconfig(2, state=DISABLED)
But it didn't work although it worked fine with a regular menu bar.
My questions are:
(1) Is there any convenient way to configure popup menu entry widget?
(2) If not, how can I can retrieve widget of menu entry (child widget)
 - I used NameToWidget() in Motif programming.

Thank you for your attention.

Aki-

// a sample popup menu from http://effbot.org/zone/tkinter-popup-menu.htm
from Tkinter import *

root = Tk()

w = Label(root, text="Right-click to display menu", width=40, height=20)
w.pack()
    
# create a menu
popup = Menu(root, tearoff=0)
popup.add_command(label="Next") # , command=next) etc...
popup.add_command(label="Previous")
popup.add_separator()
popup.add_command(label="Home")

def do_popup(event):
    # display the popup menu
    popup.tk_popup(event.x_root, event.y_root, 0)

w.bind("<Button-3>", do_popup)

b = Button(root, text="Quit", command=root.destroy)
b.pack()

mainloop()

~/tmp/tkinter-popup-menu.txt
I'm trying to access


From tagith1394 at hotmail.com  Sat Sep  7 12:02:48 2002
From: tagith1394 at hotmail.com (Evan)
Date: Sat, 07 Sep 2002 16:02:48 GMT
Subject: Embedded Python implementation
References: <MAfe9.266658$v53.14092531@news3.calgary.shaw.ca> <Xns92829B12D2570cliechtigmxnet@62.2.16.82>
Message-ID: <Iqpe9.284469$f05.14960528@news1.calgary.shaw.ca>

Chris Liechti wrote:
> Evan <tagith1394 at hotmail.com> wrote in 
> news:MAfe9.266658$v53.14092531 at news3.calgary.shaw.ca:
> 
> 
>>Hello,
>>
>>I am currently trying to embed Python in an application I am writing, 
>>and all is going well, except for one inexplicable bug..
>>
>>When I pass a variable (pointer to a string) to PyRun_String, it will 
>>always error on line 2, character 5.. regardless of what is actually on 
>>the line in that spot (for that matter, even if there is no line 2)..
> 
> 
> that sounds as you woul not pass the pointer to a string but to something 
> else...
> 
> 
>>The script works perfectly when I place it in a file, and use 
>>PyRun_File, and it also works perfectly when I place the string in 
>>quotations, and pass it to PyRun_String directly inside the C code..
>>
>>The C code in question is as follows:
>>if ( PyRun_String( com_list, Py_file_input, pGlobals, pLocals ) == NULL )
>>         PyErr_Print();
> 
> 
> i think that sould work when com_list is a "char *", say:
> char * com_list = "print hello\n";
>  
> 
>>I am confident the problem is not with pGlobals or pLocals, as those 
>>arguments are used with PyRun_File, as well as with the quoted string..
> 
> 
> chris

I have inserted a small amount of C to print the contents of com_list, 
right before the PyRun_String, here is the complete output (*** marks 
beginning and end of the script)

  ***
112 (p) 114 (r) 105 (i) 110 (n) 116 (t) 32 ( ) 34 (") 104 (h) 101 (e) 
108 (l) 108 (l) 111 (o) 34 (") 10 (
)  ***
      File "<string>", line 2

        ^
SyntaxError: invalid syntax



From cbbrowne at acm.org  Thu Sep 12 13:32:16 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: 12 Sep 2002 17:32:16 GMT
Subject: ANN: NINZ 1.2.1
References: <mailman.1031850016.20818.python-list@python.org>
Message-ID: <alqj30$79md$1@ID-125932.news.dfncis.de>

In an attempt to throw the authorities off his trail, Gustavo Niemeyer <niemeyer at conectiva.com> transmitted:
> NINZ stands for NINZ Is Not ZSI. This name was used to leave it clear
> that NINZ is a fork from ZSI. I have done this because I wanted a few
> features in ZSI which were not available at the moment, and at the same
> time, I don't wanted to distribute a hacked version of ZSI, since I
> could cause harm to Rich Salz's work on ZSI. My plans are to keep in
> sync with ZSI, always submitting what I change back to Rich, so he can
> keep his great work, and use in ZSI any code he likes. I don't want to
> takeover any credits of his work, and that's why I leave it clear here
> that my changes are insignificant compared to the comprehensive work he
> has done on ZSI. Indeed, my expectatives is that most of these features
> will be included in ZSI itself someday, and NINZ will die.
>
> More information about NINZ, including a comprehensive "NINZ vs. ZSI"
> session at:
>
> http://moin.conectiva.com.br/NINZ

ZSI hasn't been touched in some months, so I'd be surprised if this
('a fork') would be overly troublesome.
-- 
(reverse (concatenate 'string "moc.enworbbc@" "sirhc"))
http://www.ntlug.org/~cbbrowne/soap.html
Rules of  the Evil  Overlord #143.  "If  one of my  daughters actually
manages to win the hero and  openly defies me, I will congratulate her
on her  choice, declare a  national holiday to celebrate  the wedding,
and proclaim the  hero my heir. This will probably  be enough to break
up the relationship.  If not, at  least I am assured that no hero will
attack my  Legions of  Terror when  they are holding  a parade  in his
honor." <http://www.eviloverlord.com/>


From amitp at Xenon.Stanford.EDU  Fri Sep 13 17:56:21 2002
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 13 Sep 2002 21:56:21 GMT
Subject: Coding Style: Quotes
References: <bt1g9.17429$Pf7.590136@news1.west.cox.net> <hfa1ouodkntb3t6hbq60d9ckf2olcsah79@4ax.com>
Message-ID: <altmu5$p36$1@usenet.Stanford.EDU>

 Richie Hindle  <richie at entrian.com> wrote:
| 
| > I'm curious if others have adopted any standards for choice of single vs.
| > double quotation marks to delimit strings.
| 
| I tend to use double-quotes for user-visible strings, and single quotes
| for internal ones.  For instance:
| 
|     print "Hello, %s." % userRecord[ 'name' ]
| 
| In a library module, docstrings count as user-visible because the
| "users" include the programmer using the library:
| 
|     def greetUser( userRecord ):
|         """Says hello to the user."""
|         print "Hello, %s." % userRecord[ 'name' ]
| 
| Don't ask me why I starting using this standard, or where the idea came
| from, 'cos I have no idea.  It just seems to make sense.

I do this too.  I was influenced by Lisp / Scheme, where 'foo is a
symbol.  In Python I use 'foo' where I would have used a symbol in
Lisp and "foo" where I would have used a string in Lisp.

     - Amit




-- 
--
Amit J Patel, Computer Science Department, Stanford University
http://www-cs-students.stanford.edu/~amitp/


From akineko at pacbell.net  Mon Sep  9 19:41:39 2002
From: akineko at pacbell.net (Aki Niimura)
Date: 9 Sep 2002 16:41:39 -0700
Subject: Accessing Tk-inter Popup menu entry
References: <b714b8de.0209080901.6c8a7985@posting.google.com> <alhm24$j17$1@wanadoo.fr>
Message-ID: <b714b8de.0209091541.5982236f@posting.google.com>

Hi Eric,

Thank you for pointing out my stupid mistake. I tried entryconfigure()
in another program before but with different settings.

I appreciate your kindness helping me out rather than ignoring it.

Best regards,
Aki Niimura

Eric Brunel <eric.brunel at pragmadev.com> wrote in message news:<alhm24$j17$1 at wanadoo.fr>...
> What was the problem? I tried it and it worked, but not with index 2, 
> because the menu item numbered 2 is the 3rd one, which is the separator. 
> But doing:
> popup.entryconfigure(1, state=DISABLED)
> works without problem.


From max at alcyone.com  Tue Sep 17 17:16:16 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 17 Sep 2002 14:16:16 -0700
Subject: CGI question
References: <slrnaoerk7.mgo.jboy@otaku.freeshell.org> <mailman.1032290603.24910.python-list@python.org>
Message-ID: <3D879BA0.A0A6914F@alcyone.com>

Ian Bicking wrote:

> BTW, I wouldn't recommend doing it this way.  If you have a cgi script
> at, say, /page.py, you can access URLs like /page.py/path/to/something
> -- "/path/to/something" will be stored in os.environ['PATH_INFO'].

Agreed.  If you want contextual information included in the URL
independent of form submissions (whether GET or POST), using PATH_INFO
is a far superior way of doing it, since it's orthogonal to QUERY_STRING
(or the POST data from stdin).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ It's like being on a diving board that you know is too high.
\__/ Sade Adu
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.


From mertz at gnosis.cx  Sun Sep 15 14:28:12 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Sun, 15 Sep 2002 14:28:12 -0400
Subject: Pickling slotted instances
In-Reply-To: <mailman.1031805375.9030.python-list@python.org>
References: <mailman.1031805375.9030.python-list@python.org>
Message-ID: <8ENh9kKkXIWF092yn@gnosis.cx>                

|Instances using __slots__ can't be pickled - is there a fundemental
|reason why this is so, or is it simply yet to be implemented?

Plug of the week:  My gnosis.xml.pickle package will handle instances
that have __slots__ (even those lacking __getstate__()/__setstate__()).
While the package does "the right thing" IMO, this particular bit isn't
guaranteed to remain the same, depending on what future versions of
[c]Pickle decide to do.  Then again, none of the other bits are
-guaranteed- to remain the same in later versions either. :-)

Feel free to use it, steal the code (it's public domain), just look for
inspiration, etc:

    http://gnosis.cx/download/Gnosis_Utils-current.tar.gz


--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------




From sholden at holdenweb.com  Wed Sep 18 12:47:40 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 18 Sep 2002 12:47:40 -0400
Subject: isFloat: Without Exception-Handling
References: <ama6nt$n6i$06$1@news.t-online.com> <pan.2002.09.18.11.40.50.766802.29524@presidency.com>
Message-ID: <%82i9.1366$bh1.943@fe03>

"Rajarshi Guha" <rajarshi at presidency.com> wrote in message
news:pan.2002.09.18.11.40.50.766802.29524 at presidency.com...
> On Wed, 18 Sep 2002 11:41:47 -0400, Thomas Guettler wrote:
>
> > Hi!
> >
> > Is there a way to write the following method without using exceptions?
> >
> > def isFloat(string):
> >      is_float=1
> >      try:
> >          float(string)
> >      except:
> >          is_float=0
> >      return is_float
> >
> > print isFloat("asdf") # --> 0
> > print isFloat("0.1")  # --> 1
>
>
> def isFloat(string):
>   if type(string) == type(1.0):
>     return 1
>   else
>     return 0
>
> This should do what you want without exceptions

I doubt that VERY much. Please test your code before submitting, or ensure
you note it wasn't tested ...

>>> def isFloat(s):
...   if type(s) == type(1.0):
...     return 1
...   else:
...     return 0
...
>>> isFloat("banana")
0
>>> isFloat("1.2")
0

Also note that, were your test the right one, it would be much more concise
to write the function as

    def isFloat(s):
        return type(s) == type(1.0)

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From gerhard.haering at gmx.de  Fri Sep 27 23:06:48 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sat, 28 Sep 2002 05:06:48 +0200
Subject: Help please with menubars
In-Reply-To: <3d9516b3$1_7@goliath.newsgroups.com>
References: <3d9516b3$1_7@goliath.newsgroups.com>
Message-ID: <20020928030648.GC9657@lilith.ghaering.test>

* mongo57a at comcast.net <mongo57a at comcast.net> [2002-09-28 03:54 +0000]:
> I have a built a menubar and I can do pulldowns. [...]

Which GUI toolkit are you using? Tkinter? wxPython? PyQt? PyGtk? ...

-- Gerhard



From jjl at pobox.com  Sat Sep 28 12:59:51 2002
From: jjl at pobox.com (John J. Lee)
Date: Sat, 28 Sep 2002 17:59:51 +0100
Subject: Problem with ClientCookie
In-Reply-To: <8b5e42a6.0209271211.4f0c6d9f@posting.google.com>
References: <3d946d0f$0$10364@echo-01.iinet.net.au>
 <8b5e42a6.0209271211.4f0c6d9f@posting.google.com>
Message-ID: <Pine.SOL.4.44.0209281303260.1322-100000@mimosa.csv.warwick.ac.uk>

On 27 Sep 2002, Levente Sandor wrote:

> Probably the error is caused by the "secure" word in the set-cookie
> headers, because it doesn't contain "=", so it cannot be split by
[...]

Thanks.  It's a bug in the original Perl code, as it happens.  I'll
release a new version soonish.


John




From boiko at demogr.mpg.de  Thu Sep 26 11:32:07 2002
From: boiko at demogr.mpg.de (Serge Boiko)
Date: Thu, 26 Sep 2002 17:32:07 +0200
Subject: pyqt 3.4 -- libqtcmodule.so undefined symbol
References: <m265wtubxp.fsf@boiko_linux.demogr.mpg.de> <3D929290.3090900@SPAMnwinternet.com>
Message-ID: <m2vg4srclk.fsf@boiko_linux.demogr.mpg.de>

thanks, it works. I should have read instructions more carefully.
-Serge


From josh_py at sautez.com  Tue Sep 17 01:48:39 2002
From: josh_py at sautez.com (josh_py at sautez.com)
Date: Mon, 16 Sep 2002 22:48:39 -0700
Subject: using python as cgi on apache unix
Message-ID: <am6fm301477@enews2.newsguy.com>

All,

I have developed a python solution to run and connect to a mysql database.
I did all my development under windows apache 1.3.26 and python 2.2.1.  I
know want to move this to my unix server.  Unfortunately, no matter what I
do I can not get apache on my linux box to execute any python scripts.  I
get the following errors in apache:

Premature end of script headers: /var/www/html/cgi-bin/test.py

I have all the settings in apache I believe to be correct.  It complains
that the file is not executible if that is not set.  Other CGI works fine.

test.py looks like:

#!/usr/bin/python2

print "\n\n"
print "Content-Type: text/html\n\n"
print "hello"


I have read Apache's notes on the error and they mention I have the wrong
headers.  But other sample scripts say I have it correct.  Again, I do have
this working on windows fine.  The apache i have on my RH 7.0 box is 1.3.22.
Python is 2.2.1.   If you have any suggestions as to how I may get this to
work I would appreciate it.  I don't have a need for mod_python as my script
is really small and well have little traffic.  Thanks for any help.  I may
have to go rewrite the script in PHP soon if I don't get this resolved.  I'm
stump.

Josh




From kkuroda at crl.ucsd.edu  Sun Sep  8 00:44:48 2002
From: kkuroda at crl.ucsd.edu (Kow Kuroda)
Date: Sat, 7 Sep 2002 21:44:48 -0700
Subject: Larry Wall's comment on python...
In-Reply-To: <3d7abb20$1@news.sentex.net>
Message-ID: <B43B8365-C2E5-11D6-B3B2-00050287C0A6@crl.ucsd.edu>

On Saturday, September 7, 2002, at 07:51 PM, Peter Hansen wrote:

> Gib Bogle wrote:
>> Peter Hansen wrote:
>>> mmaddox at hcsmail.com wrote:
>>>> You must agree
>>>> that Python can lead you to horizontal scrolling in your editor - the 
>>>> bane
>>>> of user interface design.
>>>
>>> I definitely don't understand how Python, more or less than any other
>>> language, leads to more horizontal scrolling.
>> Maybe he is referring to an effect of default editor tab size.
>
> Apparently not, according to an email, which just said that
> Python leads one to do this more than with other languages.
>
> My point is that indentation (the way I've done it anyway) looks
> *exactly* the same in C, Pascal, Python, or any other block-structured
> language, so I still have no idea why Python would feel any different
> in this respect for someone.

Abstractly, you're completely right, but I guess people like Larry Wall, if 
he really meant what his words meant anyway, are expecting certain *tokens*
, or rather *markers*, that make and/or help them *see* the abstract 
structures called "blocks" in this case.

More interestingly, Python's lack of block-marking device (or Guido's 
refusal to make use of it in Python) is another piece of evidence that 
Python is more *abstract* than languages like C.

My best guess is that people like Larry are kind of detail-fetishists who 
don't really like this kind of abstractness.

Kow




From bokr at oz.net  Thu Sep  5 00:51:44 2002
From: bokr at oz.net (Bengt Richter)
Date: 5 Sep 2002 04:51:44 GMT
Subject: Matching a string up to a word
References: <slrnanbkbc.huj.dag@omally.animagicnet.no> <3d75ffaa$1@news.sentex.net> <slrnanbti7.huj.dag@omally.animagicnet.no> <3d76c1f7$1@news.sentex.net>
Message-ID: <al6nt0$vd$0@216.39.172.122>

On Wed, 04 Sep 2002 22:31:46 -0400, Peter Hansen <peter at engcorp.com> wrote:

>Dag wrote:
>> In article <3d75ffaa$1 at news.sentex.net>, Peter Hansen wrote:
>> 
>>>Can you instead guarantee the format of the date that follows the 
>>>filename?  Will it always consist of the day of the week, the month,
>>>the date, and the year?  That would turn this into a relatively
>>>simple problem.
>> 
>> Yes I can guarentee the last four fields be in that date format.  
>> Actually I think that solved it.  I'll split on whitespaces into
>> an array, and remove the first and last four fields, and then join
>> whatever's left.  That should do it.  I was stuck thinking how to 
>> parse it left to right.  
>
>I was thinking of something like that.  This should do it then:
>
>   filename = ' '.join(field.split(' ')[1:-4])
>
>That is, provided you never expect to see more than one consecutive
>space in your filenames...
Why should that hurt, if the head and tail items are consistently separated
by single spaces?

 >>> line = '12345 /some  weird/path with multiple-    -spaces Wed Sep 4 2002.'
 >>> line.split(' ')
 ['12345', '/some', '', 'weird/path', 'with', 'multiple-', '', '', '', '-spaces', 'Wed', 'Sep', '4', '2002.']
 >>> line.split(' ')[1:-4]
 ['/some', '', 'weird/path', 'with', 'multiple-', '', '', '', '-spaces']
 >>> ' '.join(line.split(' ')[1:-4])
 '/some  weird/path with multiple-    -spaces'

Regards,
Bengt Richter


From cstrong at arielpartners.com  Fri Sep 27 11:23:30 2002
From: cstrong at arielpartners.com (Craeg K Strong)
Date: Fri, 27 Sep 2002 11:23:30 -0400
Subject: Is there a way to find the name of the method currently being
 executed?
References: <mailman.1033009413.19897.python-list@python.org> <k3Ek9.170802$pX1.6081475@news2.tin.it>
Message-ID: <3D9477F2.4090106@arielpartners.com>

Thanks to all who responded!

I should have known to check the python cookbook...

One question: the sys._getframe() method is new for Python 2.1--
how likely do you think it is to change in 2.2+
releases?   The part that worries me is that the
method starts with an underscore.  That generally means
"protected/private" or "unpublished" to me.

I just want to know how safe it is to use for
general business application type programming.

Regards,

--Craeg


Alex Martelli wrote:
> Craeg K Strong wrote:
> 
> 
>>Hello:
>>
>>class a:
>>
>>    def func1(self):
>>        print "func1 being called"
>>
>>    def func2(self):
>>        print "func2 being called"
>>
>>In the code sample above, I would like to replace the "func1"
>>and "func2" strings with constants so the same code could
>>be copied verbatim into multiple functions and do the right
>>thing.
>>
>>Is there a way to do this in Python?  For example, maybe you could
>>get your hands on the call stack and print out stack[0].__name__
>>or something....
> 
> 
> Right, that's exactly what I did in:
> 
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66062
> 
> except that sys._getframe (gets one frame of the call stack,
> not all of the stack at once) is what the recipe uses (as
> usual, the version in the printed Cookbook, published by
> O'Reilly, has better discussion, &c).
> 
> 
> Alex




From michael at stroeder.com  Tue Sep 24 11:55:49 2002
From: michael at stroeder.com (=?ISO-8859-1?Q?Michael_Str=F6der?=)
Date: Tue, 24 Sep 2002 17:55:49 +0200
Subject: ANN: python-ldap-2.0.0pre06
Message-ID: <mailman.1032896465.18034.clpa-moderators@python.org>

HI!

Find a new pre-release of python-ldap:

   http://python-ldap.sourceforge.net/

python-ldap provides an object-oriented API to access LDAP directory 
servers from Python programs. It mainly wraps the OpenLDAP 2.x libs for 
that purpose. Additionally it contains modules for other LDAP-related 
stuff (e.g. processing LDIF, LDAPURLs and LDAPv3 schema).

Download link:
http://prdownloads.sourceforge.net/python-ldap/python-ldap-2.0.0pre06.tar.gz?download

Changes since 2.0.0pre05:
- Fine-grained locking when linking against libldap_r
- New wrapper class ldap.ReconnectLDAPObject
- Security fix to module ldapurl
- Other fixes and improvements to whole package
- LDAPv3 schema support (still somewhat premature and undocumented)

Ciao, Michael.




From bob at eleaston.com  Fri Sep  6 17:38:55 2002
From: bob at eleaston.com (Bob Easton)
Date: Fri, 6 Sep 2002 17:38:55 -0400
Subject: Interrupting Python
References: <mailman.1031345649.9840.python-list@python.org>
Message-ID: <Ef9e9.7309$rU2.218110@news4.srv.hcvlny.cv.net>

Thanks to all.  My system is Win2K. Python simply ignores Ctrl-c and
Ctrl-break. I think I'll fall back to using file existence as a simple
semaphore.  That will be good enough and cheap enough for this application.

--
Bob Easton


"David LeBlanc" <whisper at oz.net> wrote in message
news:mailman.1031345649.9840.python-list at python.org...
> > I have a script that can run, accessing network resources, for
> > several days.
> > Since the script does not normally need keyed input, exception
processing
> > does not raise the keyboard exception until after the program
> > ends normally.
> > I would like to be able to interrupt if from the keyboard, but have not
> > learned the trick.  How can I do this?
> >
> > --
> > Bob Easton
>
> I have had the experience of Python progs not responding to ctrl-c or
ctrl-d
> on Windows, but I found that ctrl-break works every time.
>
> Dave LeBlanc
>
>




From ell at cav.co.il  Thu Sep 12 07:53:10 2002
From: ell at cav.co.il (Ell)
Date: Thu, 12 Sep 2002 13:53:10 +0200
Subject: win ce install
Message-ID: <newscache$bonb2h$5w2$1@lnews.actcom.co.il>

wince install
1.
how I install python on wince ?
2.
has somebody experience , python on wince ?
are all objects work as on win32 ?
how is the performance ?


el





From sholden at holdenweb.com  Wed Sep 18 19:21:43 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 18 Sep 2002 19:21:43 -0400
Subject: Loading modules from several directories
References: <a1247c5b.0209171610.243f3c02@posting.google.com> <33803989.0209172247.2e6afc1f@posting.google.com> <a1247c5b.0209180607.843526e@posting.google.com> <6Q1i9.1149$bh1.989@fe03> <a1247c5b.0209181438.3a1c0c9a@posting.google.com>
Message-ID: <lS7i9.2127$ik2.2079@fe02>

"Eugene Pervago" <eugene at ulka.com> wrote ...
> "Steve Holden" <sholden at holdenweb.com> wrote ...
> > "Eugene Pervago" <eugene at ulka.com> wrote ...
> >
> > [...]
> > > If I put the project_a path first, it loads its modules but can't
> > > import the /onedir/foo modules. It seems package can only be in one
> > > place, once Python finds it, it just stops looking elsewhere.
> > >
> >
> > What would you rather have it do? Load them both? Find every possible
> > package and load the oine with the most recent modification date/time?
Load
> > the longest one? Load the one whose checksum is closest to your
birthdate?
>
> They do not intersect. Actually the structure is something like this:
> /onedir/foo
>    dbutils.py
>    utils.py
>    auth.py
>
> /anotherdir/foo - Nothing
> /anotherdir/foo/projectA
>     User.py
>     Product.py
>
> In Unix I would do it using symlinks, not possible on Windows. I can
> think of two solutions:
>
> 1) Copy /onedir/foo/*.py to anotherdir/foo/ for each project
>    (Doesn't seem like a good solution, though easily doable since the
> common
>     files are under version control)
> 2) Instead of foo.projectA have a foo_projectA or fprojectA to avoid
> problems
>    with nested packages.
>
> Neither of those two is particularly attractive to me. Any
> suggestions?
>

I'm at something of a loss to understand why both directories have to be
called foo and what the relationship is between them. Why not foo1 and foo2?
Am I missing something?

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From john_bradbury at ___cableinet.co.uk  Fri Sep 27 14:39:00 2002
From: john_bradbury at ___cableinet.co.uk (John Bradbury)
Date: Fri, 27 Sep 2002 18:39:00 +0000 (UTC)
Subject: kinterbasdb exception
Message-ID: <an28k3$35q$1@knossos.btinternet.com>

What is wrong with the code below?  When the kintebasdb.Error fires it does
not seem to clear as all the subsequent INSERTs cause the execption..  Do
you have to do something the clear the exception?

Any help gratefully received.

John Bradbury


    TABLE_NAME = "CLIENTS"
    cur.execute('select TEL,MIN(CLIENTCODE)AS CLIENTCODE from %s GROUP BY
TEL HAVING TEL IS not NULL AND TEL <> ""' % TABLE_NAME)
    i = 0
    while 1:
        row = cur.fetchonemap()
        if row is None:
            break
        i = i + 1
        if i % 100 == 0:
            print "Inserted %d %s" % (i, TABLE_NAME)
        ky = cleanit(row['TEL'])
        if ky <> "":
            try:
                curupd.execute('INSERT INTO PH (TEL,CLIENT_LINGUIST,DBCODE)
VALUES ("%s","CLIENTS","%s")' % (ky,row['CLIENTCODE']))
            except kinterbasdb.Error:
                print "Duplicate %s" % ky





From jafo at tummy.com  Fri Sep  6 18:21:10 2002
From: jafo at tummy.com (Sean Reifschneider)
Date: Fri, 6 Sep 2002 16:21:10 -0600
Subject: python2-docs rpm package
In-Reply-To: <Pine.LNX.4.44.0208242145300.12975-100000@hons.cs.usyd.edu.au>
References: <Pine.LNX.4.44.0208242145300.12975-100000@hons.cs.usyd.edu.au>
Message-ID: <20020906222110.GC12707@tummy.com>

On Sat, Aug 24, 2002 at 09:59:23PM +1000, Wong; Wang Fu wrote:
>The python2-docs rpm package downloaded from python.org claims itself 
>containing the info formats. However, after I installed it, it does have 
>html format only. Why is that? Any clues?

According to the package description, it claims to only include ASCII text
files and LaTeX source.  Which, seems to be what it contains...

Sean
-- 
 Come see the violence inherent in the system!
 Help!  Help!  I'm being repressed!  -- Monty Python and the Holy Grail
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com, ltd. - Linux Consulting since 1995.  Qmail, Python, SysAdmin



From ifls at rz.tu-clausthal.de  Tue Sep 24 03:16:50 2002
From: ifls at rz.tu-clausthal.de (Lutz Schroeer)
Date: 24 Sep 2002 07:16:50 GMT
Subject: Sessionhandling
References: <c5b3549a.0209230650.34a30fca@posting.google.com> <f72dac0d.0209231211.5614b398@posting.google.com>
Message-ID: <amp3h2$1ot7$1@ariadne.rz.tu-clausthal.de>

Alan Kennedy <alanmk at hotmail.com> wrote:
: try:
:    langpref = os.environ['HTTP_ACCEPT_LANGUAGE']
:    langs = string.split(langpref, ',')
:    firstpref = langs[0]
:    secondpref = langs[1]
:    thirdpref = langs[2]
: except:
:    firstpref = 'en-ie'

This does not work, does it? If the user defined only _two_ preferred
languages you will get an exception by trying to evaluate the third language
and "firstpref" will become "en-ie", whatever it has been set to before.

Here's my modification to the code:

langpref = os.environ['HTTP_ACCEPT_LANGUAGE']    
firstpref = 'en-ie'  
try:
  langs = string.split(langpref, ',')
  firstpref = langs[0]
  secondpref = langs[1]
  thirdpref = langs[2] 
except:
  pass

Lutz


From bokr at oz.net  Wed Sep  4 00:04:00 2002
From: bokr at oz.net (Bengt Richter)
Date: 4 Sep 2002 04:04:00 GMT
Subject: Grumble about too strict an attitude about backward compatibility...
References: <mailman.1031108345.32199.python-list@python.org>
Message-ID: <al40ng$pku$0@216.39.172.122>

On Tue, 3 Sep 2002 21:58:04 -0500, Skip Montanaro <skip at pobox.com> wrote:

>
>I don't want to wake any sleeping dogs, however, as I'm sitting here
>watching a bunch of Fortran function names fly by as SciPy builds, I'm
>reminded of the fairly recent threads about backward compatibility,
>Python-in-a-tie, etc.   Here I am compiling with a Fortran 90/95 compiler
>(Sun's Forte thing-a-ma-bob) and see function names like
>
>        lpni:
>        klvna:
>        chgubi:
>        cyzo:
>        klvnb:
>        rmn2so:
>        csphik:
>
>spew forth.  So, while it's great that this same large library compiles and
>runs on compilers back to at least Fortran 77 (and probably earlier),
>programmers are still stuck with the same cryptic function and data names
>they had to deal with 30+ years ago, all in the name of backward
>compatibility.
>
>What's the Python connection?  Other than a reminder not to get to slavish
>about backward compatibility, I note that these same function names will
>then go on to pollute the Python namespace because all this Fortran code is
>automatically wrapped using f2py.
>
Wrapping suggests an opportunity for new names, and keeping the old within the
confines of some module or other namespace(s).

So, would you like to choose appropriate new names for public interfaces ;-)
Maybe the renaming could be automated with appropriate config data for f2py?

Regards,
Bengt Richter


From aleax at aleax.it  Fri Sep 27 02:56:29 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 27 Sep 2002 06:56:29 GMT
Subject: Changing global variables in embedding Python
References: <2cc96dc7.0209261546.22f71021@posting.google.com>
Message-ID: <xiTk9.175423$pX1.6207367@news2.tin.it>

John Dunn wrote:

> Hello-
> 
> I have embedding Python in a C++ application. All is well, except I
> cannot figure out how to modify globals via Python script. After I

You talk of code inside a function, althought the snippet you show
doesn't look like it is:

> the following code, the first line works properly, but the second will
> fail with the error 'local variable 'test' referenced before
> assignment'. The message object is an argument to the function.
> 
> message.print( test )
> test = "test2"

If a function assigns to name 'test', and does NOT have a statement
        global test
(preferably at the start), then 'test' names a LOCAL variable of
the function, not a global one.  But this would make it quite
astonishing that the first line would work -- THAT line should
give the error message you're reporting for the SECOND one...
indeed the error message you report is IMPOSSIBLE for the second
line you give -- it's not referencing 'test', it's binding it.

So I suspect you mis-observed things, and the error msg is
actually for the first of these two lines.  As a fix, try to
start your function with the statement:
        global test



Alex



From max at alcyone.com  Tue Sep  3 03:58:57 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 03 Sep 2002 00:58:57 -0700
Subject: syntax error in Pyton shell
References: <e895ad50.0209022304.44c32987@posting.google.com>
Message-ID: <3D746BC1.2BFA05F8@alcyone.com>

lion wrote:

> I'm learning exception handling, I type these code in Python shell:
> 
> >>> try: f=open("nofile")
> ... except IOError: pass
> ... print "The exception is just passed!"
>   File "<stdin>", line 3
>     print "The exception is just passed!"
>         ^
> SyntaxError: invalid syntax
> 
> But these code run without any error as a program(or script). Could
> anyone tell me why?

This is an artifact of entering in code manually in the shell, which
normally would have more cues to go on.  In a script it would no that
these two statements were unrelated; in the interactive interpreter it
can't be sure so is waiting for a blank link to separate them.  You can
tell this when the prompt is ... instead of >>>.

Python 2.2 (#1, Feb  6 2002, 19:31:45) 
[GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> try: f = open("nofile")
... except IOError: pass
... # note I hit a blank line here
>>> print "The exception just passed!"
The exception just passed!

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ There is nothing so subject to the inconstancy of fortune as war.
\__/ Miguel de Cervantes
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From bdash at gmx.net  Fri Sep 13 06:13:10 2002
From: bdash at gmx.net (Mark Rowe)
Date: Fri, 13 Sep 2002 22:13:10 +1200
Subject: Extending a C++ App
In-Reply-To: <996eade8.0209122352.45721938@posting.google.com>
References: <996eade8.0209120912.3a653d7@posting.google.com>
 <n0qnhvfn.fsf@boeing.com>
 <996eade8.0209122352.45721938@posting.google.com>
Message-ID: <f05111b01b9a76944406f@[192.168.1.1]>

Hello,

>I looked at SWIG
>However I have a few concerns (forgive me if Im being dense)
>
>(1) All the examples are created as DLLs (ie APIs)
>     This is fine - except how would my C++ App use the same object instances?
>     (ie. share the memory and API objects)
>

Last time I looked I couldn't find much information on this topic, 
but after stumbling around in the dark for a while I found the 
function SWIG_NewPointerObj, which takes a pointer and creates a 
PyObject * of the specified type from it.  This can then be inserted 
into a Python namespace using the Python C API.


>(2) SWIG does not do a nice job of object wrapping (in my opinion)
>     For instance instead of saying
>       c = example.Circle
>       c.x = 10
>       c.y = 20
>       del c
>
>     I need to do this:
>       c = example.new_Circle()
>       example.Shape_x_set(c,10)
>       example.Shape_y_set(c,20)
>       delete_Circle(c)
>
>     Isn't there a cleaner library somewhere to do this?

To get around this, tell SWIG to generate shadow classes for the 
objects.  eg, use the command `swig -python -c++ -shadow circle.i'. 
This will generate a Python source file with classes that wrap the 
methods you mentioned above (example.new_Circle, etc).

Mark Rowe



From pyth at devel.trillke.net  Mon Sep 30 09:05:14 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 30 Sep 2002 15:05:14 +0200
Subject: Find out link location
In-Reply-To: <7uf9na-vd9.ln@lbdb01.gsi.de>; from klaus.hoeppner@bigfoot.de on Mon, Sep 30, 2002 at 02:26:47PM +0200
References: <7uf9na-vd9.ln@lbdb01.gsi.de>
Message-ID: <20020930150514.W30315@prim.han.de>

Klaus Hoeppner wrote:
> Hi,
> 
> with the os.path module I can use the islink function to check, wether
> a file is symbolic link. But how can I find out, where the link points
> to. E.g. if I have readme.txt --> ../packages/mypackage/readme.txt how
> do I get the "../packages/mypackage/readme.txt" with python?

import os
os.readlink('/path/to/link')

regards,

    holger





From oren-py-l at hishome.net  Thu Sep 12 03:44:00 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Thu, 12 Sep 2002 03:44:00 -0400
Subject: [FEEDBACK] Is this script efficient...is there a better way?
In-Reply-To: <3D7FF723.3050005@one.net>
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <mailman.1031781254.1680.python-list@python.org> <HPRf9.1780$Lg2.247316@news2.news.adelphia.net> <3D7FF723.3050005@one.net>
Message-ID: <20020912074400.GA22578@hishome.net>

On Wed, Sep 11, 2002 at 10:08:35PM -0400, Michael Schneider wrote:
> 
> 
> Bob X wrote:
> 
> >
> >>1) readlines() loads the entire file into a list so if you have a 30+ 
> >>mb file you just ate 30+mb of memory.  Try using xreadlines() 
> >>instead, it reads the file line by line and is much more memory 
> >>friendly.
> >
> >Very cool...I had missed that! 
> 
> 
> It is much slower though.  I am not sure about your config, but I am 
> running with a Gig of memory.
> I would much rather have the speed then the  30 mb of memory.  Again, 
> check your config.

Have you actually tested this? I used to believe that larger buffers are 
always better for performance.

Wrong.

I ran some test on Linux for the effect of buffer size on file reading
speed and the results were very interesting. I started with a buffer size
of 32 bytes, tested file I/O throughput and increased it logarithmically by 
about 2% for each step.  As expected, the time to read 1mb improved as the
buffer size increased until it hit a minimum for a buffer size of around 
4-8k (the graph is very noisy so it's hard to tell) and then rose back up 
to a value that is 10-20% worse for buffer sizes of 32-64k and remained
more-or-less constant for anything higher.

This performance curve may be a result of the CPU cache or the OS 
architecture.

The chunk size used by xreadlines is 8k which is about the optimum
value (at least for linux).

	Oren




From sarmxiii at knology.net  Thu Sep 12 12:43:24 2002
From: sarmxiii at knology.net (montana)
Date: Thu, 12 Sep 2002 11:43:24 -0500
Subject: [Tutor] Paradox  table
In-Reply-To: <20020912143532.GA17566@dman.ddts.net>
Message-ID: <C14341E2-C66E-11D6-A451-00039315F4DA@knology.net>

Is there a good example on how to use a CSV parser in Python?
So you have this comma seperated file, does the CSV parser break data 
up into groups based upon where they are positioned in the file or does 
it just generate a list of data or both?

Is there any good sites out there that deal with CSV parsing and python 
that have some good examples or can someone on this list give some?

Thanks.

SA

"I can do everything on my Mac I used to do on my PC, plus alot more 
..."
--Me

On Thursday, September 12, 2002, at 09:35 AM, Derrick 'dman' Hudson 
wrote:

> On Wed, Sep 11, 2002 at 03:14:30PM +0200, A wrote:
> | Hi,
> |  Is there a module for reading Paradox database (.DB) files with 
> Python?
> | I need to run the script under Linux, so I can not use ODBC.
> | Any suggestion would be appreciated
>
> One way to access the data is to export it from paradox in CSV format.
> CSV is an open standard and several (python) parsers exist for it.
> However, using this method won't allow you to easily access the data
> from both paradox and from your script, instead you would need to
> re-export it every time you modified it in paradox and you would not
> be able to modify it using your script.  If you can, it would probably
> be easiest to migrate to a different db such as postgresql.  Postgres
> databases are much more accessible than paradox is (at least with the
> paradox version I've seen, 3.5).
>
> -D
>
> -- 
> Dishonest money dwindles away,
> but he who gathers money little by little makes it grow.
>         Proverbs 13:11
>
> http://dman.ddts.net/~dman/
> <mime-attachment>




From mongo57a at comcast.net  Mon Sep 30 19:42:17 2002
From: mongo57a at comcast.net (mongo57a at comcast.net)
Date: Mon, 30 Sep 2002 23:42:17 GMT
Subject: Scrolled list question
Message-ID: <3d98cff7$1_7@goliath.newsgroups.com>

Trying to make more than one data element appear in a scrolled list. The
window is formed properly; the first piece of data is there, but I don't
know how to get the rest of the data in (see comments in code below).

if __name__ == '__main__':

  sqlcmd = sql.Build_Query (command, tables, columns, condition, sort) #
  custom code - works fine

  dbinfo = sql.Query (sqlcmd) #custom code works fine

  for i in range(20):
    symbol = dbinfo.Symbol[i]
    action = dbinfo.Action[i]
    dislist = string.join([symbol, action],'   ') # joins the 2 strings
  ScrolledList(dislist)
mainloop()

class ScrolledList(Frame):
    def __init__(self, label, parent=None):
        Frame.__init__(self, parent)
        self.pack(expand=YES, fill=BOTH)                  # make me
       expandable
        self.makeWidgets(label)
    def handleList(self, event):
        index = self.listbox.curselection()               # on list
       double-click
        label = self.listbox.get(index)                   # fetch selection
       text
        self.runCommand(label)                            # and call action
       here
    def makeWidgets(self,label):
        sbar = Scrollbar(self)
        list = Listbox(self, relief=SUNKEN)
        sbar.config(command=list.yview)                   # xlink sbar and
       list
        list.config(yscrollcommand=sbar.set)              # move one moves
       other
        sbar.pack(side=RIGHT, fill=Y)                     # pack first=clip
       last
        list.pack(side=LEFT, expand=YES, fill=BOTH)       # list clipped
       first

The preceding puts the first value of dislist into the window - as it
should. So how do I get the rest of them in there? I'm not a very
experienced programmer...... tried dislist[i] with ScrolledList(dislist) and
got errors.

TIA.


-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From matt at mondoinfo.com  Tue Sep 24 23:05:41 2002
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Wed, 25 Sep 2002 03:05:41 GMT
Subject: Tkinter status bar?
References: <a-A27A4D.21421424092002@nnrp01.earthlink.net>
Message-ID: <slrnap2a05.1e2.matt@happy-hour.mondoinfo.com>

On Wed, 25 Sep 2002 01:41:36 GMT, Daniel T. <a at b.c> wrote:

> I'm going through a pdf file I found on the internet ("An
> Introduction to Tkinter" by Fredrik Lundh)

Fredrik's Introduction is excellent. I almost always have a copy open
when I'm doing Tkinter programming.

> In chapter 8 he talks about how to put a status bar at the bottom of a 
> window:
> 
> # File: tkSimpleStatusBar.py
> class StatusBar(Frame):

[. . .]

> We could have inherited from the Label widget itself, and just
> extended it with set and clear methods. This approach have a few
> drawbacks, though:

[. . .]

> Now my question. I didn't think to inherit from Label, instead I
> changed it to:

> class StatusBar:
>    def __init__(self, master):
>       self.label = Label(master, text="", bd=1, relief=SUNKEN,
>         anchor=W)

[. . .]

> In other words, why inherit from any of the Tkinter classes?

No reason. I generally don't. You can do it either way depending on
your personal preference and/or the details of the class you're
defining.

If what you're doing feels like it's refining the behavior of an
existing class, I'd say that inheritance makes sense. If it feels like
you're creating a new class, I'd say that wrapping one or more
existing widgets (as you're suggesting doing) and delegating calls to
them makes sense. Other people have other rules of thumb, I'm sure.

Delegating calls to wrapped objects can be more verbose if you need
to delegate lots of them. But I find that I don't generally need to
delegate lots of calls and to my eye it's more explicit.

Regards,
Matt


From LogiplexSoftware at earthlink.net  Fri Sep 20 09:50:19 2002
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 20 Sep 2002 06:50:19 -0700
Subject: More on Protecting Source Code
In-Reply-To: <GCEDKONBLEFPPADDJCOECEEJFAAA.whisper@oz.net>
References: <GCEDKONBLEFPPADDJCOECEEJFAAA.whisper@oz.net>
Message-ID: <1032529820.20661.221.camel@localhost.localdomain>

On Tue, 2002-09-17 at 03:13, David LeBlanc wrote:
> EXCUSE ME, but I'm not criticizing anyone! I am suggesting that Python could
> be better. I've also been honest enough to say that I don't know what or how
> that could be. Python didn't get where it is because people impuned the
> motives who had things to say about it.

I don't know... I felt criticized.  I know you weren't talking about me
or even to me, but Usenet has made me very sensitive.

And now you're shouting <sob>.

;)

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From ebbingeo at logica.com  Thu Sep 12 09:23:05 2002
From: ebbingeo at logica.com (Onno Ebbinge)
Date: 12 Sep 2002 06:23:05 -0700
Subject: EDI parsing
References: <746127ef.0209110128.53665a23@posting.google.com> <mailman.1031755813.4692.python-list@python.org>
Message-ID: <746127ef.0209120523.258d6ca4@posting.google.com>

"Emile van Sebille" <emile at fenx.com> wrote in message news:<mailman.1031755813.4692.python-list at python.org>...
> Onno Ebbinge:
<snip>
> > I need to parse a (subset) of the UN EDIFACT (aka EDI) standard
> > (homepage: http://www.unece.org/trade/untdid/welcome.htm). After doing
> > the initial reading of the standard it looks like it is quite hard to
> > make a parser for this standard (and 'constructor').
> 
> My experience is with the UCC UCS standards, so some of this may be
> different in the UN standard.
> 
> Once you have the full specification of the documents you'll be
> exchanging in a form that python can digest, it's not that hard to build
> the parser and/or constructor.  The full specs are available to members
> on CD in PDF form.  I once wrote an extremely ugly extraction utility
> that preprocessed using pdf2text and then parsed the resulting mess into
> something I could then use as something like
> 
> from edixn import EdiXn
> for source in sourcemessages:
>     tempIn = EdiXn(source)
> 
> where sourcemessages is a python list of the individual transactions
> extracted from the edi envelope.  EdiXn knew about 810, 850, 852, 855,
> 875 and 880 transactions, allowing me to write appropriate handlers for
> each type.

Sounds like a very similar situation.

> >
> > My questions:
> >
> > 1) Is there a python EDI module? I can't find any :(
> 
> I didn't then either.

:(
 
> > 2) Is there a (python) EDI to XML converter?
> 
> That wouldn't be hard.  If there's already a DTD for edi, that would
> help a lot.  ;-)

No DTD...

> > 3) Is there anything that I can use to easily interface with EDI?
> >    (preferably in lib and DLL form)
> 
> Try some of the commercial suppliers.  There's big money (both expense
> and potential) in this, and that's probably why we don't find
> implementations strewn about the web.

Commercial suppliers are not an option! But I agree with your
remark about the 'big money'. Maybe when I'm done ;-)

> > If the above is not available or fails...
> >
> > [You have to know that I've never written a parser in Python before.
> >  The last (big) parser I wrote was in C with a few years back the help
> >  of lex and yacc if memory serves me right.]
> >
> > What is the best approach to writing an EDI parser in Python?
> 
> To parse the received messages, I've always found it easiest to use and
> parse the raw ascii coming in.  There are commercial packages that allow
> you to map and export or even map-to-map import, but as most of the
> customers I've done this for were typically forced into it by demands of
> the channel, and the commercial offerings rarely seemed as nimble as
> simply writing a one-to-one utility taking the edi order to the database
> or the resulting invoice to edi.  I have generally used a commercially
> available transport package, KMart at that time being the notable
> exception.  Outbound messages follow suit.

I'm in favour of the raw parsing too. Maybe a one-to-one mapping,
maybe with ebXML in between, I'm not sure yet.
 
> If you are writing this as a part of a portable b2b application then
> you've got your hands full as you've entered the map-making utility
> supplier market.  You'll need to address the entire specification, which
> my hardcopy of  my deprecated 003050UCS shows to be some ~1000 _dense_
> pages.  Make sure you've got the funding to do it right, as this is no
> low-budget add-on.

I'm in luck then! I only need to parse about 10 to 15 messages.
I don't want the 'quick and dirty' solution because I _know_ it
will bite me if there are changes and/or new messages. And they
_will_ come.

> The 'edi specification' always felt akin to the 'rs232 specification' to
> me, only instead of 25 pins to muck up, they offer 100s.
> 
> Not-sure-how-this-could-help-ly y'rs,

Thanks anyway, now I know that parsing is the way to go.

Regards,

Onno


From steve at ferg.org  Sat Sep  7 00:09:50 2002
From: steve at ferg.org (Stephen Ferg)
Date: 6 Sep 2002 21:09:50 -0700
Subject: ANN: Thinking in Tkinter
Message-ID: <b16e4ef7.0209062009.26e218e2@posting.google.com>

A small page devoted to "Thinking in Tkinter"
at http://www.ferg.org/thinking_in_tkinter/index.html
for whoever might find it useful.

I've been trying to teach myself Tkinter out of various books, and I'm
finding it more difficult than I think it should be.

The problem is that the authors of the books want to rush into telling
me about all of the widgets in the Tkinter toolbox, but never really
pause to explain basic concepts. They don't explain how to "think in
Tkinter".

Here are a few short programs that begin to explain how to think in
Tkinter. In them, I don't attempt to catalog all of the types of
widgets, attributes, and methods that are available in Tkinter. I just
try to get started down the road of understanding some basic Tkinter
concepts.


From jeff_hamann at hamanndonald.com  Thu Sep  5 21:55:36 2002
From: jeff_hamann at hamanndonald.com (Jeff D. Hamann)
Date: Thu, 5 Sep 2002 18:55:36 -0700
Subject: example for using calldll with python2.2
Message-ID: <al921d$1h8$1@news.orst.edu>

I've perform evaluations on scripting languages for a scripting language to
replace vba for our work. Python seems to be a little farther ahead when it
comes to calling dlls than perl and I've been trying to get calldll to work
on my win2k machine. I've downloaded a version (not sure which but looks
liek a version for python 2.1) and the install/compile.bat choked.

C:\Python22\calldll>gcc -I../Include -O2 -c calldllmodule.c -o
calldllmodule.o
In file included from ../Include/Python.h:62,
                 from calldllmodule.c:62:
../Include/pyport.h:480: #error "LONG_BIT definition appears wrong for
platform (bad gcc/glibc config?)."

C:\Python22\calldll>dllwrap --dllname calldll.pyd --driver-name gcc --def
calldll.def -o calldll.pyd calldllmodule.o -s --entry
_DllMain at 12 --target=i386-mingw32 -L.. -lpython22
/usr/lib/gcc-lib/i686-pc-cygwin/2.95.3-5/../../../../i686-pc-cygwin/bin/ld:
cannot find -lpython22
collect2: ld returned 1 exit status
dllwrap: gcc exited with status 1

is what I got. I have no idea what is going on here. is this required to use
the calldll package? is there a precompiled version (at least I can use)
before I can get the makefile to work?

Jeff.


--
Jeff D. Hamann
Hamann, Donald & Associates
PO Box 1421
Corvallis, Oregon 97339
jeff_hamann at hamanndonald.com
www.hamanndonald.com




From jb at cascade-sys.com  Mon Sep 23 19:31:16 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 23 Sep 2002 16:31:16 -0700
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <mailman.1031377930.4225.python-list@python.org> <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com> <3D8E6353.7030303@something.invalid> <2IYsVKAZksj9EwSH@jessikat.fsnet.co.uk>
Message-ID: <3D8FA444.1040506@cascade-sys.com>

Robin Becker wrote:

>In article <3D8E6353.7030303 at something.invalid>, Greg Ewing <see_reply_a
>ddress at something.invalid> writes
>......
>  
>
>>If you use spaces only, you're also choosing how other
>>people will see it in their editors (unless they're
>>willing to use a tool to re-indent it).
>>
>>Using *tabs* only is the only way to get truly
>>preference-independent indentation.
>>
>>    
>>
>here here, but I'm always wrong
>  
>
Ditto.

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From jpm at papercut.org  Mon Sep 30 16:32:22 2002
From: jpm at papercut.org (Joao Prado Maia)
Date: Mon, 30 Sep 2002 16:32:22 -0400 (EDT)
Subject: Question about __getattribute__/__getattr__
In-Reply-To: <20020930215327.X30315@prim.han.de>
Message-ID: <Pine.LNX.4.44.0209301625470.12980-100000@mercury.sabren.com>

On Mon, 30 Sep 2002, holger krekel wrote:

> Probably you are interested in the Memoize recipe at 
> 
>     http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52201
> 

That looks interesting, and it is indeed similar to what I'm trying to 
acomplish, but on my case I want to keep each specific return value in a 
separate cache file.

The first Memoize class keeps every cached item in a dictionary, which it 
may be fine for some applications, but I need it for a network server that 
may be running for several days. It's not the best thing to keep 
everything in memory ;)

The second class keeps everything on the same pickled file (please let me 
know if I'm wrong), which will be very slow after a couple hundred megs of 
data in there.

Anyway, I'm still going back to my original question: how do I get the 
full list of arguments passed to a function ? This can't be that hard :)

Cheers,
Joao




From jdhunter at ace.bsd.uchicago.edu  Wed Sep 11 12:03:02 2002
From: jdhunter at ace.bsd.uchicago.edu (John Hunter)
Date: Wed, 11 Sep 2002 11:03:02 -0500
Subject: Graph algorithms - minimisation
In-Reply-To: <cf0ad9fb.0209092301.44288bc7@posting.google.com> (jolsen@mailme.dk's
 message of "10 Sep 2002 00:01:39 -0700")
References: <cf0ad9fb.0209092301.44288bc7@posting.google.com>
Message-ID: <m2bs74pl8p.fsf@mother.paradise.lost>

>>>>> "Jesper" == Jesper Olsen <jolsen at mailme.dk> writes:


    Jesper> If you know of implementations in other languages than
    Jesper> python - e.g.  C/C++, I am interested as well.

You may be interested in the boost graph library for C++.  There is a
link on the main page to a book on the library by the authors

http://www.boost.org/libs/graph/doc/index.html

Boost also comes with a powerful library for wrapping C++ in python

http://www.boost.org/libs/python/doc/index.html

So you could wrap what you want in python.  You might check on the
boost mailing list to see if someone has already done part of that work
for you.

John Hunter



From michaels at one.net  Thu Sep 12 06:26:43 2002
From: michaels at one.net (Michael Schneider)
Date: Thu, 12 Sep 2002 06:26:43 -0400
Subject: [xml.dom.minidom] why cant it parse '&'
References: <9zZf9.24873$_l1.66951@news-server.bigpond.net.au>
Message-ID: <3D806BE3.1080902@one.net>

The problem is your xml file.  it is not valid.

you can't have a '&' char in  a  normal (non-CDATA) string.


Your xml file needs to read

<sxml>
<title>Congo Mombo - Guitar Gable &amp; The Musical Kings</title>

...

good luck,
Mike


:B nerdy wrote:

>im using the minidom.parseString() to parse the following text
>
><sxml>
><title>Congo Mombo - Guitar Gable & The Musical Kings</title>
><composer>Various</composer>
><genre>blues</genre>
><trackNo>12</trackNo>
><copyright>Westwood</copyright>
><dateInserted>2002-03-11</dateInserted>
></sxml>
>
>when it comes to the '&' it throws an exception. is there any way around
>this problem??
>cheers asap
>
>



From martinbertolino at yahoo.com  Fri Sep 27 09:17:44 2002
From: martinbertolino at yahoo.com (Martin Bertolino)
Date: 27 Sep 2002 06:17:44 -0700
Subject: well-formed xml
References: <mailman.1033085436.9260.python-list@python.org>
Message-ID: <95be4880.0209270517.3a0b5d7e@posting.google.com>

I believe you document should be encoded as:

<a href='http://google.com/search?hl=en&amp;q=foobar'>search</a>

Martin

"Mark McEahern" <mark at mceahern.com> wrote in message news:<mailman.1033085436.9260.python-list at python.org>...
> I'm obviously missing something because this seemingly innocent chunk of
> xhtml:
> 
>   from xml.dom import minidom
> 
>   s = "<a href='http://google.com/search?hl=en&q=foobar'>search</a>"
>   #                                             ^
>   #                                             - seems to be the problem
>   #
>   # maybe it thinks I'm trying to reference the &q entity?
> 
>   doc = minidom.parseString(s)
> 
> Exception traceback follows.
> 
> Is there a way for me to tell it to ignore apparent entity references inside
> attribute values?
>


From fperez528 at yahoo.com  Wed Sep 11 18:23:22 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 11 Sep 2002 16:23:22 -0600
Subject: [FEEDBACK] Is this script efficient...is there a better way?
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net>
Message-ID: <alofoq$lgu$1@peabody.colorado.edu>

Bob X wrote:

> 
> I use this to parse a log file (30MB+) for keywords and write the lines
> that are found into another file.
> 
> Any thoughts would be appreciated...  :-)
> 

How about

$ grep words infile | tee outfile

Cheers,

f.


From gerhard.haering at gmx.de  Fri Sep 13 11:54:52 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 13 Sep 2002 15:54:52 GMT
Subject: Callback function call from C++
References: <mailman.1031863877.18343.python-list@python.org>
Message-ID: <slrnao42ib.r4.gerhard.haering@lilith.my-fqdn.de>

Sebastian Huber wrote in comp.lang.python:
> Hello,
> I want to install a Python callback function in a C++ extension module. 
> Everything works as desired, but is it possible to get the argument list 
> length and types of the function in advance. For example:
> 
> PyCallable_Check( o) != 0
> 	get the arglist and return 0 if not ok
> install the callback function
> 
> At the moment I get an error at the first call if the arglist is not ok, and I 
> want to avoid this.

I don't know if there's a C interface for this. If there isn't, you
can call the inspect module from your C++ app:

>>> def foo(x, y, *args, **kwargs): pass
...
>>> import inspect
>>> inspect.getargs(foo.func_code)
(['x', 'y'], 'args', 'kwargs')

HTH,

-- Gerhard


From max at alcyone.com  Tue Sep 10 16:13:44 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 10 Sep 2002 13:13:44 -0700
Subject: read() file??
References: <mailman.1031684894.10361.python-list@python.org>
Message-ID: <3D7E5278.A8417F5@alcyone.com>

jubafre at brturbo.com wrote:

>      I have a file with assembler instruction and i want read the
> lines of the file and put in a list of string, this is easy to do, but
> i don?t want to consider the caracteres "/" and the othes after him.
> 
> file.asm:
> ADD D1  /coment1
> LDA D2  /coment2
> 
> list of strings:
> X=[ 'ADD D1' , 'LDA D2' ]

Do something like (unested):

	result = []
	for line in readlines:
	    if line[-1] == '\n':
	        line = line[-1]
	    if line.find('/') >= 0:
	        line = line.split('/', 1)[0]
	    line = line.strip()
	    result.append(line)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ I sidestep the either
\__/ or choices of logic and choose both. | Ken Feit
    PyUID / http://www.alcyone.com/pyos/uid/
 A module for generating "unique" IDs in Python.


From ponderor at lycos.com  Mon Sep  9 17:20:40 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 9 Sep 2002 16:20:40 -0500
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Sep  9)
Message-ID: <D5F9F6F5A0C46CAC.9DCA346C4331EBAF.B9AA42F7FC35A33A@lp.airnews.net>

QOTW:  "I've had an interesting year - I've been coding in nothing but
Python (Unix & Win32) and *assembler* (embedded PIC processors).  Talk
about a contrast!  Yet in both cases, the languages 'Fit My Brain' for
the tasks at hand." - Tim Daneliuk
http://mail.python.org/pipermail/python-list/2002-September/121704.html

"I'm too new to Python to give you good technical reasons to use it,
but I like it :)" - TeaAndBikkie 
http://mail.python.org/pipermail/python-list/2002-September/121789.html

"This is actually the first new program I've coded in C (rather than
Python) in a good four years or so." - Eric S. Raymond 
http://mail.python.org/pipermail/python-list/2002-September/121533.html


Threads..
    The Python-dev list summary for 2002-08-16 - 2002-09-01:
        http://mail.python.org/pipermail/python-list/2002-September/121533.html
    
    In light of the new PEP Format, Dr. David Mertz narrarates his
    experience with text formatting markup methods:
        http://mail.python.org/pipermail/python-list/2002-August/120870.html
    
    Python interprocess communication methods:
        http://mail.python.org/pipermail/python-list/2002-September/120900.html
    
    Many thoughts on the cryptic nature of regular expressions:
        http://mail.python.org/pipermail/python-list/2002-September/120906.html
    
    Duncan Booth points some of the relatively few gotchas he's
    encountered when packaging a Python program into an EXE:
        http://mail.python.org/pipermail/python-list/2002-September/120984.html

    Python-based Windows EXE's installation hints:
        http://mail.python.org/pipermail/python-list/2002-September/121711.html
    
    Repeatable module import debuging tips:
        http://mail.python.org/pipermail/python-list/2002-September/120972.html
    
    There are faster ways for building huge strings:
        http://mail.python.org/pipermail/python-list/2002-September/121103.html
    
    Thorough response to algorithm question on 'Smoothing a discrete set
    of data':
        http://mail.python.org/pipermail/python-list/2002-September/121777.html
    
    Building Python on SCO issues discussed:
        http://mail.python.org/pipermail/python-list/2002-September/121231.html
    
    Number-of-threads-implemented concern addressed:
        http://mail.python.org/pipermail/python-list/2002-September/121260.html
    
    More answers on reasons to learn Python, with thoughts on PHP:
       http://mail.python.org/pipermail/python-list/2002-September/121264.html
    
    Tim Rue is perplexing.  This thread concludes with with an FAQ
    that makes things as clear as possible.
        http://mail.python.org/pipermail/python-list/2002-September/121701.html
    
    Question on encryption leads to intro, warnings and directions to
    more detailed resources: 
        http://mail.python.org/pipermail/python-list/2002-September/121270.html
    
    How to handle stdin/stdout for daemonization:
        http://mail.python.org/pipermail/python-list/2002-September/121783.html
    
    Thoughts on support for rebinding in nested scopes:
        http://mail.python.org/pipermail/python-list/2002-September/121531.html
    
    Which database is right for 2000-3000 records?:
        http://mail.python.org/pipermail/python-list/2002-September/121546.html
    
    Larry Wall's Python comments in his Slashdot interview 
        http://interviews.slashdot.org/article.pl?sid=02/09/06/1343222&mode=nocomment 
    are discussed on the c.l.py:
        http://mail.python.org/pipermail/python-list/2002-September/121683.html
    
Software 
    xmlrpclib 1.0.1 includes Bug fixes and performance enhancements:
        http://mail.python.org/pipermail/python-announce-list/2002-September/001665.html
    
    CherryPy? 0.6 adds SSL support: 
        http://www.cherrypy.org
    
    JinSitu - Interactive introspection environment for Java and Jython:
        http://www.krause-software.de/jinsitu/
    
    Pyrex 0.4.4.1 now available:
        http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/
    
    PyGame 1.53 includes CD-ROM & MPEG object enhancements, Civil-0.8 and rectangle handling improvements:
        http://www.pygame.org
    
    lfm 0.9, UNIX Python Midnight Commander Clone:
        http://www.terra.es/personal7/inigoserna/lfm
    
    scgi 0.5 is now available from:
        http://www.mems-exchange.org/software/scgi/
    
    "spambayes" is a Python implementation of applying Bayesian
    classifier to spam detection:
        http://sf.net/projects/spambayes
    
    New Stackless Python beta available: 
        http://www.stackless.com
    
    'waccess' v. 2.0  web log analyzer is released at:
        http://www.tundraware.com/Software/waccess/
    
    Freeware 'AutoIt' Win32 UI test automation tool noted:
        http://mail.python.org/pipermail/python-list/2002-September/121681.html
    

========================================================================

Everything you want is probably one or two clicks away in these pages:

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
         http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Michael Hudson continues Andrew Kuchling's marvelous tradition
    of summarizing action on the python-dev mailing list once every
    other week.
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

    Much of Python's real work takes place on Special-Interest Group
    mailing lists
        http://www.python.org/sigs/

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
        http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
        http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
        http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

    Links2Go is a new semi-automated link collection; it's impressive
    what AI can generate
        http://www.links2go.com/search?search=python

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning
(approximately), ask <claird at phaseit.net> to subscribe.  Mention
"Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From bdesth at nospam.free.fr  Thu Sep  5 20:49:25 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Thu, 05 Sep 2002 20:49:25 -0400
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com> <4f49d232.0209050418.7668c21@posting.google.com> <al7ub8$crj$1@panix2.panix.com>
Message-ID: <3D77FB95.3070300@nospam.free.fr>

Roy Smith wrote:
> Laotseu <bdesth at free.fr> wrote:
> 
>>Now you should *also* learn other languages. But you'd better go with
>>a lo level languages and a functionnal one - C and Lisp could be good
>>choices.
> 
> 
> I guess it depends on why you're learning languages.  If you're
> learning languages to get a paying job as a programmer, lisp would be
> a horrible choice.  If you're learning languages to study computer
> science, lisp is looking a lot better.


Learning Lisp means learning functionnal programming and object oriented 
programming
.
Learning C means learning real low level imperative programming.

These two languages complements each other IMHO. I think once you know 
them both, you know *how* to program, whatever the language you have to use.

my 2 cents...

Laotseu
--
The Old Master Said :
"Lisp is worth learning, whatever the reason."
--






From blarson at crary.com  Fri Sep 20 10:11:34 2002
From: blarson at crary.com (Bradley D. Larson)
Date: Fri, 20 Sep 2002 09:11:34 -0500
Subject: More on Protecting Source Code
References: <GCEDKONBLEFPPADDJCOECEEJFAAA.whisper@oz.net> <1032529820.20661.221.camel@localhost.localdomain>
Message-ID: <3D8B2C96.DEB56368@crary.com>

Perhaps we should start a IP/Developer/Python support group for ourselves....

It appears that after reading this discussion the only real method of protecting
software code (not concepts) is to operate it from a black box behind locked
doors where the software is only accessable via a generic readily available,
"who cares who copies it" client (ie. html browser).

Otherwise someone, somewhere somehow will always (I didn't shout ALWAYS)
will be able to reverse engineer your product.

This has been a problem with "hard" products since the beginning of time.
Anyone can beg, borrow, or steal your toaster, bicycle, or car and tear it
down and build one similar (and even learn from your mistakes).

Legal recourse is usually at best a loss or break even venture, sure you stopped
them but at what cost and it does not reverse the damage that has already been
done.

Cliff Wells wrote:

> On Tue, 2002-09-17 at 03:13, David LeBlanc wrote:
> > EXCUSE ME, but I'm not criticizing anyone!

<snip>

> I don't know... I felt criticized.  I know you weren't talking about me
> or even to me, but Usenet has made me very sensitive.
>
> And now you're shouting <sob>.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: blarson.vcf
Type: text/x-vcard
Size: 400 bytes
Desc: Card for Bradley D. Larson
URL: <http://mail.python.org/pipermail/python-list/attachments/20020920/1d0c8dff/attachment.vcf>

From martin at v.loewis.de  Tue Sep 17 02:03:47 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 17 Sep 2002 08:03:47 +0200
Subject: Stackless Python, eventual merge?
In-Reply-To: <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com>
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com>
	<mailman.1032107006.6411.python-list@python.org>
	<am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net>
	<j4znuio0kr.fsf@informatik.hu-berlin.de>
	<mailman.1032204706.14737.python-list@python.org>
	<m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de>
	<1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com>
Message-ID: <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de>

Ian Bicking <ianb at colorstudy.com> writes:

> But it doesn't seem like complexity was ever the fundamental problem --
> the complexity could have been dealt with, but I don't believe there was
> ever any indication that if it was simplified or documented that it
> would get into standard Python.  

That is not true. I have indicated this atleast once, namely when I
studied the code. For me, the complicatedness was always the problem -
complexity is fine, since this is a complex problem.

> It has been simplified, but I don't think that changes anything with
> regard to getting it into standard Python -- the same barriers still
> exist.

You are making these barriers up in your mind. I don't know why you
are doing that, though.

It requires certain skills to contribute to free software, yes, and
these skills are different from the skills you need to write the
software in the first place. Lack of skills may indeed pose a barrier
- but one that can be overcome by determined contributors. Those
contributors might only need the permission from the original author
to start working.

Regards,
Martin



From elimli at yahoo.com  Sat Sep 28 02:45:15 2002
From: elimli at yahoo.com (eli li)
Date: Sat, 28 Sep 2002 08:45:15 +0200
Subject: urllib2.urlopen(req).info() object
Message-ID: <an3fhi$3jr$1@news.iucc.ac.il>

The call "urllib2.urlopen(req).info()" returns an object similar to the
dictionary, but it can have muliple values for some key. For example:

Content-Type: text/html
Set-Cookie: A=M
Set-Cookie: WEBTRENDS_ID=192.168.120.21
Set-Cookie: SINFO=ca9e
Via: 1.1 colosus1 (NetCache NetApp/5.3R2D4)

How can I retrive all values related to 'Set-Cookie' key?




From stsolakos at patras.atmel.com  Tue Sep 10 05:35:45 2002
From: stsolakos at patras.atmel.com (Tsolakos Stavros)
Date: Tue, 10 Sep 2002 12:35:45 +0300
Subject: Embedding python with a C++ multithread apps -> pause and resume python script
References: <001601c258a3$a3e19a20$4c4eb280@epfl.ch>
Message-ID: <001001c258ad$7099f680$dbfeaa0a@garganuclear>

I have had the same problem, and I had to do some kind of trick. I had also had a problem with the CPU usage, since when such a script had been running, the CPU usage touched 100%. 

My question : I still do not know why. Does anybody?

Anyway, I have used the sys.settrace method. Each time a line of Python code is executed, my OnTrace C++ function is being called. I use yield to lower the CPU usage. You may check if the thread that called it has to pause, an dyou may pause it till it is resumed. Just an idea.

Regards

Stavros
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020910/9d76ecad/attachment.html>

From marklists at mceahern.com  Fri Sep  6 09:38:06 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 6 Sep 2002 08:38:06 -0500
Subject: tree data structure
In-Reply-To: <3D78AC8E.4040900@mxm.dk>
Message-ID: <JHEOKEOOLIGLDHCMAHMOIENODMAA.marklists@mceahern.com>

>      for key, value in original:
>          if map.has_key(key):
>              map[key].append(value)
>          else:
>              map[key] = [value]

Minor nit:  Best not to use the name of a builtin (map).

// m
-




From martin at v.loewis.de  Fri Sep 20 03:33:16 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 20 Sep 2002 09:33:16 +0200
Subject: wxPython i18n, help wanted
References: <amd2hs$i1k11@news.vtx.ch>
Message-ID: <m37khhuncz.fsf@mira.informatik.hu-berlin.de>

Cedric Dutoit <dutoitc at hotmail.com> writes:

> I'm trying to translate an application in Chinese with i18n, but
> I don't know how to display Chinese characters in wxPython.
> 
> Can anybody help me ?

There are two aspects in your question: How do I translate
applications, and How do I display Chinese characters in wxPython.

For the first question, I recommend to use the Python gettext
module. If you are unfamiliar with gettext, I recommend that you read
the GNU gettext manual; Python gettext follows this model very
closely.

For the second question, in most wxPython installations, you can only
display characters supported in the current locale. E.g. if you are in
de_CH.ISO-8859-1 environment, you can only display German characters,
but not Chinese ones. You need to change your system to a zh_CN.gb2312
environment; then you can display chinese characters - you need to
encode the strings in gb2312 (these are Unix examples; on Windows, you
need to change the locale likewise - on W9x, you need a Chinese
edition of Windows).

Starting with wxPython 2.3.3.1, it supports, on Win32, a "Unicode
build". In that build, wxPython can display arbitrary Unicode strings;
the distribution contains an example that displays Korean among other
languages - it is not hard to deduce display of Chinese from that.  If
you are combining this with gettext, I recommend that you use the
ugettext function, which means that you need to have the proper codecs
installed.

HTH,
Martin



From exarkun at meson.dyndns.org  Thu Sep 12 09:06:14 2002
From: exarkun at meson.dyndns.org (Jp Calderone)
Date: Thu, 12 Sep 2002 09:06:14 -0400
Subject: Overloading the ">>" operator in Python 2.2.1
In-Reply-To: <fhYf9.1418$b5.128061@news02.tsnz.net>
References: <fhYf9.1418$b5.128061@news02.tsnz.net>
Message-ID: <20020912130614.GA29027@meson.dyndns.org>

On Thu, Sep 12, 2002 at 08:22:07PM +1200, Andy Elvey wrote:
> Hi all .
>    I'm using Python 2.2.1 under Win98.  I'm doing some experimenting with
> parsing, and as part of that, I'm trying to overload the right-shift
> operator (>>) so that I can use it as follows (testing against a sequence
> that I read in) -
>   a >> b   would mean  "a is followed by b" .
>    So,  you could have a function like seqtest("string_to_test" ,
> "foo">>"bar" ) , and that function would return true or false, depending on
> whether foo was followed by bar.
>  (  I know it's possible to use regexes for this,  but being able to create
> a sequence-testing operator like this would be useful .)
>     Many thanks in advance ......
> 

>>> class Foo:
...   def __rshift__(self, other):
...     print 'rshift'
...   def __lshift__(self, other):
...     print 'lshift'
...
>>> f = Foo()
>>> f >> 'hello'
rshift
>>> f << 'world'
lshift


Also:

>>> dir(42)
['__abs__', '__add__', '__and__', '__class__', '__cmp__', '__coerce__',
'__delattr__', '__div__', '__divmod__', '__doc__', '__float__',
'__floordiv__', '__getattribute__', '__hash__', '__hex__', '__init__',
'__int__', '__invert__', '__long__', '__lshift__', '__mod__', '__mul__',
'__neg__', '__new__', '__nonzero__', '__oct__', '__or__', '__pos__',
'__pow__', '__radd__', '__rand__', '__rdiv__', '__rdivmod__', '__reduce__',
'__repr__', '__rfloordiv__', '__rlshift__', '__rmod__', '__rmul__',
'__ror__', '__rpow__', '__rrshift__', '__rshift__', '__rsub__',
'__rtruediv__', '__rxor__', '__setattr__', '__str__', '__sub__',
'__truediv__', '__xor__']

  Jp

-- 
It is practically impossible to teach good programming style to
students that have had prior exposure to BASIC: as potential
programmers they are mentally mutilated beyond hope of
regeneration.        -- Dijkstra
--
 9:00am up 114 days, 9:53, 4 users, load average: 0.02, 0.01, 0.00
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 248 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020912/37a3bbca/attachment.sig>

From praveen.patil at silver-software.com  Mon Sep 16 11:58:32 2002
From: praveen.patil at silver-software.com (Praveen Patil)
Date: Mon, 16 Sep 2002 16:58:32 +0100
Subject: Please help solving the problem
In-Reply-To: <NFBBLJFBNMKMLGNLJMFCCENBCCAA.praveen.patil@silver-software.com>
Message-ID: <NFBBLJFBNMKMLGNLJMFCMENFCCAA.praveen.patil@silver-software.com>

Hi ,


Please help me in solving the problem below.


step 1: I have written three dlls :  a.dll , b.dll , c.dll.
                 a.dll contains  funct_A();
                 b.dll contains  funct_B();
                 c.dll contains  funct_C();

step 2: I am copying a.dll to directory C:\Program Files\Python\DLLs  and
renaming as a.pyd
        similarly
           I am copying b.dll to directory C:\Program Files\Python\DLLs. I
am not renaming as b.pyd
           I am copying c.dll to directory C:\Program Files\Python\DLLs and
renaming as c.pyd
        So my  C:\Program Files\Python\DLLs  directory contain
                         a.pyd , b.dll , c.pyd

step 3: a)Python function func_pyA() calls funct_A()
        b)funct_A() call funct_B()
        c)funct_B() call funct_C()
        d)funct_C() call python  fuction  func_pyC()

step 4: I am importing a.pyd and c.pyd in python program.
             import a
             import c

step 5: I am having problem in importing 'a' because  'a' need to load b.dll
and c.dll. But I copied c.dll as c.pyd.
        Please suggest me some solution.


here is my code :

 1)a.c (a.dll)
   ----------
         void func_A();


 2)b.c (b.dll)
   -----------
          void func_B();

 3)c.c( c.dll)
   -----------
          void func_C();

 4) example.py
    ---------
   import a
   import c

   G_Logfile  = None
   def TestFunction():
     G_Logfile = open('Pytestfile.txt', 'w')
     G_Logfile.write("%s \n"%'I am writing python created text file')
     G_Logfile.close
     G_Logfile = None

   if __name__ == "__main__":
   a.func_A();
   .....
   .....


Please help me in solving the problem.


Cheers,


Praveen.


[ The information contained in this e-mail is confidential and is intended for the named recipient only. If you are not the named recipient, please notify us by telephone on +44 (0)1249 442 430 immediately, destroy the message and delete it from your computer. Silver Software has taken every reasonable precaution to ensure that any attachment to this e-mail has been checked for viruses. However, we cannot accept liability for any damage sustained as a result of any such software viruses and advise you to carry out your own virus check before opening any attachment. Furthermore, we do not accept responsibility for any change made to this message after it was sent by the sender.]



From paul at boddie.net  Tue Sep  3 09:27:20 2002
From: paul at boddie.net (Paul Boddie)
Date: 3 Sep 2002 06:27:20 -0700
Subject: The Module Search Path is ???
References: <7153651c.0209022255.7fd50ba0@posting.google.com>
Message-ID: <23891c90.0209030527.63f8ae@posting.google.com>

quiteblack at yahoo.com (black) wrote in message news:<7153651c.0209022255.7fd50ba0 at posting.google.com>...
> tutorial told me I could found it in the folder:/usr/local/lib/python.
> but my OS is windows so folders are different. I searched the folder
> "lib" but find no file named "python", any ideas ?

If you've installed Python at the location C:\Python21 (for example -
this is usually the answer to a question in the Windows installer),
then try looking inside that directory and also inside
C:\Python21\Lib.

Paul


From Oschler at earthlink.net  Tue Sep 10 12:18:14 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Tue, 10 Sep 2002 16:18:14 GMT
Subject: Trouble upgrading to 2.2.1
Message-ID: <aXof9.29813$w51.7932884@news2.news.adelphia.net>

I'm trying to upgrade from 2.1.1 to 2.2.1.  I'm using kpackage (KDE's rpm
wrapper).  When I try to install from the rpm file named
"python2-2.2.1-1.i386.rpm" I get the following dependency errors:

libcrypto.so.2
libsel.so.2

How can I correct this, or, is there an easier way to get upgraded?

thx







From loewis at informatik.hu-berlin.de  Sun Sep 15 15:20:24 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 15 Sep 2002 21:20:24 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <j43csbruq7.fsf@informatik.hu-berlin.de> <mailman.1032114318.15460.python-list@python.org>
Message-ID: <j465x7qcuf.fsf@informatik.hu-berlin.de>

Christian Tismer <tismer at tismer.com> writes:

> Stackless is built in a way that by default there is no difference
> to the standard. No stack switching occours, all is identical.

Good point!

Martin


From magnus at thinkware.se  Tue Sep 17 04:53:48 2002
From: magnus at thinkware.se (Magnus Lycka)
Date: Tue, 17 Sep 2002 10:53:48 +0200
Subject: [Tutor] Not sure how to phrase this question
In-Reply-To: <000501c25dd4$38711460$9865fea9@arthur>
Message-ID: <5.1.0.14.0.20020917104233.02ac6000@www.thinkware.se>

At 18:55 2002-09-16 -0400, Arthur wrote:
>Would hate to have to resort to truely understanding
>this stuff to get to where I am hoping to get.

;)

Don't you think it would be simpler to make a
factory function, instead of trying to mess
with the constructors?

def returnClassic... rather than
class returnClassic...

The interface would look just the same, and I
think implementation would be trivial.

What you are doing now certainly seems to defy
the intentions of __new__, since what you are
returning is not in the inheritence chain of the
class.

If my channeling is successful, you are neither
doing what Guido intended, nor what the Gang of
4 intended with the Factory patterns.


-- 
Magnus Lyck?, Thinkware AB
?lvans v?g 99, SE-907 50 UME?
tel: 070-582 80 65, fax: 070-612 80 65
http://www.thinkware.se/  mailto:magnus at thinkware.se




From aleax at aleax.it  Thu Sep 19 15:23:45 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 19 Sep 2002 19:23:45 GMT
Subject: strong/weak - dynamic/static [Was: Getting started]
References: <uojinmcfapjf45@corp.supernews.com> <Bqli9.36669$216.1428116@bin4.nnrp.aus1.giganews.com> <Ugmi9.114712$ub2.2441284@news1.tin.it> <x1pi9.41599$216.1593993@bin4.nnrp.aus1.giganews.com>
Message-ID: <5vpi9.134717$pX1.4791845@news2.tin.it>

Terry Reedy wrote:
        ...
>> If you find a way to name this behavior "static", you'll be well
>> placed for the yearly Humpty Dumpty award.
> 
> This is pretty inane.  Again, I specifically excluded 'this behavior'
> from my comments.  Anyway, my question *rhetorically* asked whether
> creation-time typing (which is fixed except for the exception that I
> excluded) is (to quote the OP) 'dynamic or static'.  I then suggested

In other words, your claim is that typing is fixed at creation time,
excluding the way in which an object's type can be changed after then
(showing, of course, that said type isn't fixed at all).

This makes just about as much sense as, say, claiming that a list's
length is fixed at creation time, excluding the ways in which the
list's length can be changed (which show, of course, that said
length isn't fixed at all).


> that this dichotomy does not fit Python very well.

That's your thesis, and I don't think you have brought any
substantial arguments in its favor.  I think that distinguishing
things that ARE indeed fixed at the time of an object's creation
(such as the length and contents of a tuple or string) from
things which are not fits Python quite well, and so does the
attempt to position it on the continuum between static and
dynamic extremes (just about ANY dichotomy doesn't fit any
realistically-complicated situation -- more often than not,
shades of gray are the underlying reality, black and white just
sometimes-useful, sometimes-misleading simplifications thereof
for modeling purposes -- this particular case, in my opinion, is
more useful and less misleading than most others).


Alex



From umop at psyon.org  Wed Sep 18 17:05:15 2002
From: umop at psyon.org (Eric Arnold)
Date: Wed, 18 Sep 2002 14:05:15 -0700
Subject: RegEx editors
Message-ID: <amaq0d$4aton$1@ID-144898.news.dfncis.de>

Does anybody know of any good graphical regex editors out there?

Eric




From robie at 1373.net  Mon Sep 23 23:19:09 2002
From: robie at 1373.net (robie1373)
Date: Mon, 23 Sep 2002 21:19:09 -0600
Subject: confusion about opening files
References: <mailman.1032835770.5997.python-list@python.org>
Message-ID: <3d8fd913$0$197$75868355@news.frii.net>

Thanks anyway. And you are right, I should check out the FAQ as well.

"Mark McEahern" <marklists at mceahern.com> wrote in message
news:mailman.1032835770.5997.python-list at python.org...
> [I wrote]
> > This is a FAQ.  You should probably find the Python FAQ and read
> > it.  Short story:  Don't use 'from x import *' unless you know
> > what you're doing.
>
> I should have read the post I was replying to more carefully.  I saw the
> "integer required" error and didn't notice that the OP was using os.open.
> So the 'from x import *' was a red herring.  As Erik Max Francis pointed
> out, the OP should use open instead of os.open.
>
> My apologies.
>
> Cheers,
>
> // mark
>
> -
>
>




From peter at engcorp.com  Wed Sep  4 22:31:46 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 04 Sep 2002 22:31:46 -0400
Subject: Matching a string up to a word
References: <slrnanbkbc.huj.dag@omally.animagicnet.no> <3d75ffaa$1@news.sentex.net> <slrnanbti7.huj.dag@omally.animagicnet.no>
Message-ID: <3d76c1f7$1@news.sentex.net>

Dag wrote:
> In article <3d75ffaa$1 at news.sentex.net>, Peter Hansen wrote:
> 
>>Can you instead guarantee the format of the date that follows the 
>>filename?  Will it always consist of the day of the week, the month,
>>the date, and the year?  That would turn this into a relatively
>>simple problem.
> 
> Yes I can guarentee the last four fields be in that date format.  
> Actually I think that solved it.  I'll split on whitespaces into
> an array, and remove the first and last four fields, and then join
> whatever's left.  That should do it.  I was stuck thinking how to 
> parse it left to right.  

I was thinking of something like that.  This should do it then:

   filename = ' '.join(field.split(' ')[1:-4])

That is, provided you never expect to see more than one consecutive
space in your filenames...

-Peter



From wlfraed at ix.netcom.com  Fri Sep 20 18:10:30 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Fri, 20 Sep 2002 15:10:30 -0700
Subject: Python Sybase slow compared to Perl?
References: <Xns928FA7E6BA09Dcmay4yahoocom@165.112.130.8>
Message-ID: <nc6gma.qi3.ln@ix.netcom.com>

Chuck May fed this fish to the penguins on Friday 20 September 2002 
01:30 pm:

        I'm just a neophyte at this but...

> and here is the Python code snippet:
> 
>    # output the actual data
>    c = db.cursor()
>    c.execute( "SELECT %s FROM %s" % ( select, table ) )
>    num_recs = 0
>    if os.path.isfile( "%s.gz" % dumpfile ):
>        os.remove( "%s.gz" % dumpfile )
>    file = os.popen( "gzip -c > %s.gz" % dumpfile, "w" )
>    while 1:
>        row = c.fetchone()

        How many rows do you expect? fetchall() or fetchmany() followed by 
looping over rows returned may be a noticeable increase (fetchall() if 
you have the memory).

>        if not row : break
>     
>        line = []
>        for i in range( len( row ) ):
>            line.append( "%-*s" % ( lengths[i], row[i] ) )
> 
>        print >> file, "".join( line )
>        num_recs += 1
> 
>    file.close()
>    c.close()
> 
--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From rgkindya at yahoo.co.in  Sat Sep  7 12:36:37 2002
From: rgkindya at yahoo.co.in (rgkindya)
Date: 7 Sep 2002 09:36:37 -0700
Subject: getting input in a embeded GUI app ...
Message-ID: <b2df190a.0209070836.9713b08@posting.google.com>

Dear Newser,

	I'm a newbie for python. I wish to embed Python interpreter into my
vc++ GUI app(for ms-windows only). when executing input statements (eg
raw_input()) , in pythonwin environment (activepython) , it gets input
using a new small input window, in IDLE , it gets input in its own
window (next line). please help me to do these two type of inputs in
my embeded GUI app. my app is only GUI , there is no console mode.

please debug my engliz bugs ...


From mhammond at skippinet.com.au  Sat Sep 14 04:19:08 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 14 Sep 2002 08:19:08 GMT
Subject: Calling Python from C# - please help
References: <6de16ac9.0209130508.5d1b56c6@posting.google.com> <j4elbyt3yw.fsf@informatik.hu-berlin.de> <ur8fyxau0.fsf@online.no>
Message-ID: <0iCg9.2350$Ee4.6516@news-server.bigpond.net.au>

Syver Enstad wrote:
> loewis at informatik.hu-berlin.de (Martin v. L?wis) writes:
> 
> 
>>adila001 at yahoo.com (AdrianC) writes:
>>
>>
>>>I am using Python 2.1 and I am tring to use my Python COM component.
>>
>>>From VB6 everything is OK, but in .NET i don't know how to set the
>>>reference to my COM.
>>
>>I've never used the COM bridge in .NET, but I believe you need to
>>create a type library first.
> 
> 
> Although you don't have to make a typelibrary, your C# code will be
> nicer if you do so (else you will have to use the dispatch interface
> to work with the object).

Yes on all counts.  You should create a type library defining your 
interface, and use the .NET tlbimp.exe tool to create a standard .dll 
assembly.  Your C# code then imports this assembly and can use the 
object as a "first class" .NET object.

I have no examples of this, but I should.  I'm co-authoring a .NET book 
which is nearly out :) - 
http://www.aw.com/catalog/academic/product/1,4096,0201770180,00.html - 
Python does get a bit of a mention :)

Mark.



From dsavitsk at e-coli.net  Sun Sep 29 13:22:59 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Sun, 29 Sep 2002 17:22:59 GMT
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com>
Message-ID: <TFGl9.642$XX3.454818@newssrv26.news.prodigy.com>

"Jurie Horneman" <jhorneman at pobox.com> wrote in message
news:9ac02e81.0209290839.4f4d6475 at posting.google.com...
> Once again my attempt to profit from Python's vaunted productivity
> increase has come to a screeching halt because some innocent thing
> that one would reasonably expect to work in one line, intuitively, and
> without even having to think about, turns into a nightmare already
> lasting more than an hour, requiring numerous experiments, searching
> through the documentation - both manually and using my editor's
> find-in-files - finally leaving me no other option but to write a post
> to comp.lang.python, knowing that my programming flow for today is
> well and truly gone. I hate doing this, because I know I cannot keep
> my frustration out of my writing, and also that someone will reply and
> gently or not so gently point out something which actually IS written
> down in some obscure corner of the documentation, despite my attemps
> to RTFM.
>
> So, what is it this time? No, it's not Unicode, another aspect that
> frequently trips me up, causing my face to meet the rough asphalt of
> the infobahn. It's importing modules. Again.
>
> (This is an obscure reference to an attempt on my part to implement
> MBCS conversion functions as Python codecs. This failed miserably and
> despite the helpful advice I received here, I never did try again. Too
> bad, because those conversion functions are pretty good. And since I'm
> doing some parenthesized ranting here: does anyone outside of the
> Python world use 'codec' to refer to anything except audio / video
> codecs? I actually had to post here just to find out that what I was
> looking for were *called* codecs. Sheesh.)
>
> Anyway.
>
> I have this script, let's call it 'blah.py'. It sits in a directory
> far, far away from where my Python is installed. It wants to include a
> module called 'intro' from a subdirectory called 'templates'.
>
> This is evidently impossible.
>
> Among my vain, futile attempts were:
>
> *** Solution 1 ***
> The World is a Beautiful Place
>
> >>> import templates.intro
>
> Ha! ha! ha! OK, even I knew this was a bit optimistic.
>
> *** Solution 2 ***
> .PTH file
>
> Put a templates.pth containing the single line 'templates' in the same
> place as 'blah.py', then
>
> >>> import templates.intro
> or
> >>> import intro
>
> No dice: 'ImportError: No module named...' etc.
>
> *** Solution 3 ***
> IMPorting by hand
>
> >>> import imp
> >>> f = open('templates\intro.py')
>
> >>> imp.find_module('intro.py')
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No module named intro.py
>
> >>> imp.find_module('intro')
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No module named intro
>
> >>> imp.find_module('intro', '\\templates\\')
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No frozen submodule named .\templates\.intro
>
> (No WHAT?)
>
> >>> imp.find_module('intro', '\\templates\\')
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No frozen submodule named \templates\.intro
>
> >>> imp.find_module('intro', ['\\templates\\'])
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No module named intro
>
> >>> imp.find_module('intro.py', ['\\templates\\'])
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No module named intro.py
>
> >>> imp.find_module('intro', ['.\\templates\\'])
> (<open file '.\templates\intro.py', mode 'r' at 0x01511110>,
> '.\\templates\\intro.py', ('.py', 'r', 1))
>
> Whoa! I finally managed to call imp.find_module without raising an
> exception! OK, now to call imp.load_module, because the documentation
> says:
>
> "Load a module that was previously found by find_module()"
>
> Alright, so there seems to be some connection between find_module and
> load_module... The latter probably takes the return value of the
> former as a parameter, right? Right? Wrong.
>
> Here are my attemps to call imp.load_module:
>
> x = imp.find_module('intro', ['.\\templates\\'])
>
> >>> y = imp.load_module('intro', f, 'intro.py', x)
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> TypeError: load_module() argument 4, item 0 must be string, not file
>
> >>> y = imp.load_module('intro', x)
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> TypeError: load_module() takes exactly 4 arguments (2 given)
>
> Okay, wait, my assumption was obviously naive, better read the doc...
> aha. Almost the same parameters as find_module... wonder why I even
> needed to call find_module in the first place. OK:
>
> >>> y = imp.load_module('intro', f, 'intro.py', imp.get_suffixes())
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> TypeError: load_module() argument 4 must be sequence of length 3, not
> 5
>
> What? But... but... *splutter* the doc says:
>
> "The description argument is a tuple, as would be returned by
> get_suffixes(), describing what kind of module must be loaded."
>
> So why doesn't this work? OK, I have no idea why I'm bothering with
> these kinds of arcana, but for the heck of it I'll try:
>
> >>> y = imp.load_module('intro', f, 'intro.py', ('.py', 'r', 1))
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
>   File "intro.py", line 29, in ?
> ImportError: No module named Base
>
> Yay! What?! Now my IMPORTED module can't find the modules IT is trying
> to import? (It doesn't matter if I put Base next to Blah.py or in
> templates - Intro can't find it.)
>
> Arrrgghhh!!!! OK, enough time spent on this dead end. (Oh, and by the
> way not all the functions of the imp module are documented.)
>
> *** Solution 4 ***
> Packages
>
> OK, more searching, browsing and scouring my hard drive for the word
> 'import'. Of course! I forgot that in the Python docs, important
> information is cunningly hidden in the Tutorial section! (Really, am I
> being unreasonable for assuming that once one has reached a certain
> level of knowledge, one shuns the Tutorial and goes straight for the
> main documentation, because that's where one can find everything there
> is to know? That's only how it works in every language or piece of
> software in general I've ever encountered.)
>
> So there are these things called 'packages'. Wow! Handy, that! All one
> needs to do is put an empty file called '__init__.py' in a directory,
> and hey! That directory becomes a package and Bob's your uncle.
>
> So I put an empty file called '__init__.py' in 'templates' and try the
> following:
>
> >>> import templates.intro
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No module named templates.intro
>
> >>> from templates import intro
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No module named templates
>
> This is *exactly* how it's written in the Tutorial, and
> it-does-not-work.
>
> *** Solution 5 ***
> Abject Defeat, Ask on Internet, Be Ridiculed by Peers for Not Managing
> Basic Operation
>
> Amusing little anecdote: I've performed all of the above steps twice -
> once while growing increasingly frustrated, and a second time while
> writing this post. Twice, while typing in the code above, I at some
> point entered:
>
> import intro
>
> ... and it would work. For real. The symbols were there. Not a peep
> out of the interpreter.
>
> Because I don't trust PythonWin (don't get me started on PythonWin, I
> haven't found a better IDE yet, I may actually succumb and PAY for one
> soon *shock horror*) I quit and restarted and lo and behold! the
> interpreter complained that there was no such module, as it had been
> doing all along, the horrible beast.
>
> In other words, at random, unreproducable moments, Python perversely
> imported my module, whereas normally it refused to.
>
> To cut a long story... hm... Please please help me! Stop me before I
> whine again. And please don't be offended by my ranting, I just can't
> hide this much frustration. I'm using Python 2.2 by the way.


So, I would like to agree that, while in MS world this would be called a
feature of some sort, that it is either a design flaw in Python, or a major
oversight in the documentation.  I have had most of the same frustration as
the poster, and usually stumbled onto something that worked (which too many
times sees to have involved \Lib\) Perhaps someone should write a "Complete
Guide to Importing Modules" -- maybe it's in the new Nutshell Book -- which
I would do if I understood it?

(Too add another complaint to the above, as well, Python used in active
server pages does not see .pth files forcing sys.append() as the first line
of each.)

-d




From wlfraed at ix.netcom.com  Thu Sep 26 03:52:48 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Thu, 26 Sep 2002 00:52:48 -0700
Subject: How to send Email using Python?
References: <amu8qh$93n60$1@ID-49758.news.dfncis.de>
Message-ID: <gceuma.fc4.ln@ix.netcom.com>

Ken fed this fish to the penguins on Wednesday 25 September 2002 11:19 
pm:

> Hi all, I like to know how to send a HTML featured Email from Python?
> I know how to do a text based version using SMTPLib, but not sure how
> to do it with a HTML featured Email. Can someone help me? Can someone
> provide me with sample code?
> 
        mimelib?

        SMTP stuff is really just the transport of mail from one machine to 
another. You can stuff anything into the body of the message -- and the 
MIME stuff helps in setting up the proper headers/dividers.

        Just don't send me any email with HTML -- where ever possible I 
disable HTML email (I'll accept the /original/ MIME "rich text" stuff, 
which predates HTML)

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From gerhard.haering at opus-gmbh.net  Wed Sep 11 06:46:44 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 11 Sep 2002 10:46:44 GMT
Subject: function pointer
References: <ffcedbc9.0209110215.644c1788@posting.google.com> <aln5ce$5kt$1@news.netpower.no>
Message-ID: <slrnanu7i1.22k.gerhard.haering@haering.opus-gmbh.net>

David Brown wrote:
> "Christina" <christina at iaeste.at> wrote in message
> news:ffcedbc9.0209110215.644c1788 at posting.google.com...
>> Hi!
>>
>> Does Python have function pointers? If not, is there a possibility to
> emulate them?
^^

Broken quoting - yuck! Your newsreader is broken. Please repair it (using
third party software) or make Usenet a better place by throwing away that
crapware.

> Functions are first-order objects (is that the right term?)

It is.

> in Python, so you can just assign them to a variable.

If we want to keep the terminology right, Python doesn't have variables. It
has names and values. Values are bound to names. In your case:

For example:

> def foo(x):
>     print x

... you create a function object using the def statement and bind it to the
name 'foo'.

> bar = foo

Here you bind the name 'bar' to the value the name 'foo' refers to.

-- Gerhard


From skip at pobox.com  Thu Sep  5 15:28:34 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 5 Sep 2002 14:28:34 -0500
Subject: Grumble about too strict an attitude about backward compatibility...
In-Reply-To: <3D77F98B.3050607@nospam.free.fr>
References: <mailman.1031108345.32199.python-list@python.org>
        <3D77F98B.3050607@nospam.free.fr>
Message-ID: <15735.45154.347049.776488@12-248-11-90.client.attbi.com>

    >> What's the Python connection?  Other than a reminder not to get to
    >> slavish about backward compatibility, I note that these same function
    >> names will then go on to pollute the Python namespace because all
    >> this Fortran code is automatically wrapped using f2py.

    laotseu> Well, that may be true, but the chances of a name collision are
    laotseu> pretty low IMHO...

Name collisions weren't what I was getting at.  Python allows essentially
unlimited identifier lengths, but people manipulating these functions from
Python code still have to use the cryptic 4-6 character names.  Perhaps f2py
allows mapping such names to something more meaningful.  (One can hope. ;-)

Skip





From Farhad.Fouladi at Dresdner-Bank.com  Mon Sep  2 04:10:12 2002
From: Farhad.Fouladi at Dresdner-Bank.com (Farhad)
Date: Mon, 2 Sep 2002 10:10:12 +0200
Subject: make informixdb1.3
References: <aknb6n$evi3@news-1.bank.dresdner.net> <3D71300E.8010403@sschwarzer.net>
Message-ID: <akv5ee$evv2@news-1.bank.dresdner.net>

Hi Stefan,

thank you for your response. I start "make" in correct directory. The point
is, there is no "_informixdb.c" file in downloaded files (from python.org).
I suppose the make file should generate one but in my case it couldn't.

Farhad

"Stefan Schwarzer" <sschwarzer at sschwarzer.net> schrieb im Newsbeitrag
news:3D71300E.8010403 at sschwarzer.net...
> Hi Farhad
>
> Farhad wrote:
> > I am trying to install informixdb-1.3 with Python 2.1.  When I run make
I
> > get the following error:
> >
> > make[1]: Entering directory
`/export/vol4/dri/informix/informixdb-1.3/ext'
> > make[1]: *** No rule to make target `_informixdb.c', needed by
> > `_informixdb.o'.  Stop.
> > make[1]: Leaving directory
`/export/vol4/dri/informix/informixdb-1.3/ext'
> > make: *** [sharedmods] Error 2
>
> Obviously, `make` can't find the file _informixdb.c which is the source
code for
> _informixdb.o . Are you starting make in the right directory?
>
> Stefan
>




From tjreedy at udel.edu  Tue Sep 24 09:23:10 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 24 Sep 2002 13:23:10 GMT
Subject: Three dumb questions (ordered by dumbness descending)
References: <amnvit$7kqdt$1@ID-77524.news.dfncis.de> <82Zj9.136094$ub2.2966822@news1.tin.it>
Message-ID: <2HZj9.260778$AR1.10903907@bin2.nnrp.aus1.giganews.com>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:82Zj9.136094$ub2.2966822 at news1.tin.it...
> Once you do have a random or pseudo-random generator that fully
> satisfies you, Ian Bicking's snippet reimplementing shuffle as a
> Decorate-Sort-Undecorate idiom is truly excellent -- a little jewel,

How is an O(n*log(n)) algorithm a 'jewel' compared to the simple and
standard O(n) algorithm?

> It does, alas, end up a little slower than Tim Peters' original --
[...]
>and Ian's version only loses very very little indeed.

Increase n enough and it becomes arbitrarily slower and loses
arbitrarily much.

> import time, sys, operator
> from random import random
>
> def shuffle_org(x):
>     for i in xrange(len(x)-1, 0, -1):
>         # pick an element in x[:i+1] with which to exchange x[i]
>         j = int(random() * (i+1))
>         x[i], x[j] = x[j], x[i]
>
> def shuffle_ian(x):
>     newList = [(random(), i) for i in x]
>     newList.sort()
>     x[:] = [i[1] for i in newList]
>
>
> biggie = range(200)
>
> def timit(func, N=1000):
>     repeat = [None] * N
>     start = time.clock()
>     for x in repeat: func(biggie)
>     stend = time.clock()
>     return "%2.2f %s" % (stend-start, func.__name__)
>
> print sys.version
> for i in range(2):
>     for func in shuffle_org, shuffle_ian:
>         print timit(func)
>
>
> [alex at lancelot linud]$ python -O timo.py
> 2.2.1 (#2, Jul 15 2002, 17:32:26)
> [GCC 2.96 20000731 (Mandrake Linux 8.1 2.96-0.62mdk)]
> 1.59 shuffle_org
> 1.87 shuffle_ian
> 1.60 shuffle_org
> 1.86 shuffle_ian

2.2.1 (#34, Apr 15 2002, 09:51:39) [MSC 32 bit (Intel)]

biggie = range(20000)
def timit(func, N=10):
...
10.76 shuffle_org
19.25 shuffle_ian
10.75 shuffle_org
19.81 shuffle_ian

biggie = range(200000)
def timit(func, N=1):
...
11.24 shuffle_org
23.18 shuffle_ian

Terry J. Reedy




From sismex01 at hebmex.com  Thu Sep 19 11:51:54 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Thu, 19 Sep 2002 10:51:54 -0500
Subject: strong/weak - dynamic/static [Was: Getting started]
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233024CCA79@mail.hebmex.com>

> 
> Terry Reedy wrote:
>         ...
> >>explaining or defining the concepts of ... dynamic and 
> >>static typing.
> > 
> > Python objects are typed at creation.  Is that dynamic or 
> > static?  I'm
> 
> Judge for yourself:
> 
> >>> class X(object): pass
> ...
> >>> class Y(object): pass
> ...
> >>> x = X()
> >>> type(x)
> <class '__main__.X'>
> >>> x.__class__ = Y
> >>> type(x)
> <class '__main__.Y'>
> >>>
> 
> If you find a way to name this behavior "static", you'll be well
> placed for the yearly Humpty Dumpty award.
> 
> 
> Alex
> 

Maybe you're being a bit rash, Alex?

>>> class X(object):
... 	__slots__ = ('a','b')
... 	pass
... 
>>> class Y(object):
... 	__slots__ = ('aa','bb')
... 	pass
... 
>>> xx = X()
>>> yy = Y()
>>> type(xx)
<class '__main__.X'>
>>> type(yy)
<class '__main__.Y'>
>>> xx.__class__ = Y
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
TypeError: __class__ assignment: 'Y' object layout differs from 'X'
>>> 

The behaviour you expounded does not blindly work,
it seems.

-gustavo




Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From jolsen at mailme.dk  Tue Sep 10 15:15:02 2002
From: jolsen at mailme.dk (Jesper Olsen)
Date: 10 Sep 2002 12:15:02 -0700
Subject: Graph algorithms - minimisation
References: <cf0ad9fb.0209092301.44288bc7@posting.google.com> <tYif9.3454$cP3.6958@news.iol.ie> <8vrrnuckib7gbuodhrmbs2f88nq8r5r5vu@4ax.com>
Message-ID: <cf0ad9fb.0209101115.43de78ec@posting.google.com>

Thanks for the reference.

Unfortunately a shortest path algorithm will not help here.

There seems to have been several efforts to implement graph algorithms
in python, but the ones have came across do not go much beyond representing
the graphs and implementing eg. Dijkstras algorithm for finding
shortest paths (there is also one in the Vaults of Parnassus).

It probably does not exist for python, but what I am looking for
is actually a python inferface to something similar to the 
AT&T FSM library (which is a binary distribution, and not open source).

http://www.research.att.com/sw/tools/fsm/

Cheers
Jesper





Gon alo Rodrigues <op73418 at mail.telepac.pt> wrote in message news:<8vrrnuckib7gbuodhrmbs2f88nq8r5r5vu at 4ax.com>...
> On Tue, 10 Sep 2002 09:30:01 GMT, Padraig Brady <Padraig at Linux.ie>
> wrote:
> 
> >Jesper Olsen wrote:
> >> Do you know of python extentions that implement efficient graph
> >> algorithms?
> >> 
> >> In particular I am interested in minimising the number of nodes & arcs
> >> used for representing directed graphs.
> >> 
> >> If you know of implementations in other languages than python - e.g.
> >> C/C++, I am interested as well.
> >> 
> >> Cheers
> >> Jesper
> >
> >Guido wrote an essay on graphs @
> >http://www.python.org/doc/essays/graphs.html
> >In there is a method to return the shortest path
> >which may be useful?
> 
> I remember seeing a recipe in ActiveState cookbook with Dijkstra's
> algorithm for finding shortest paths. You can then adapt it to your own
> representation of a graph. 
> 
> >
> >P?draig.
> 
> Best regards,
> Gon?alo Rodrigues


From tendac at wanadoo.fr  Sat Sep 28 06:11:57 2002
From: tendac at wanadoo.fr (Claude)
Date: 28 Sep 2002 03:11:57 -0700
Subject: print a %
Message-ID: <ff8044a.0209280211.345dc898@posting.google.com>

When I do 
print "image size=%d%" %(50)
I've got an error

Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ValueError: incomplete format

I do not know how to escape the second %

print "image size=%d%" %(50) 
image size=50%

Thank you


From elimli at yahoo.com  Sat Sep 28 12:17:17 2002
From: elimli at yahoo.com (eli li)
Date: Sat, 28 Sep 2002 18:17:17 +0200
Subject: urllib2.urlopen(req).info() object
References: <an3fhi$3jr$1@news.iucc.ac.il> <3d954f50$0$11909@echo-01.iinet.net.au>
Message-ID: <an4h25$udi$1@news.iucc.ac.il>

what is ClientCookie?

"Rob Hall" <bloke at ii.net> wrote in message
news:3d954f50$0$11909 at echo-01.iinet.net.au...
> I used ClientCookie.
>
> Rob.
>
>
> eli li <elimli at yahoo.com> wrote in message
> news:an3fhi$3jr$1 at news.iucc.ac.il...
> > The call "urllib2.urlopen(req).info()" returns an object similar to the
> > dictionary, but it can have muliple values for some key. For example:
> >
> > Content-Type: text/html
> > Set-Cookie: A=M
> > Set-Cookie: WEBTRENDS_ID=192.168.120.21
> > Set-Cookie: SINFO=ca9e
> > Via: 1.1 colosus1 (NetCache NetApp/5.3R2D4)
> >
> > How can I retrive all values related to 'Set-Cookie' key?
> >
> >
>
>




From aleax at aleax.it  Thu Sep 26 09:44:35 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 26 Sep 2002 13:44:35 GMT
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com> <3D8E6353.7030303@something.invalid> <amsfnk$cq1$2@nyheter.chalmers.se> <hljtma.4f3.ln@ix.netcom.com>
Message-ID: <7bEk9.145906$ub2.3213634@news1.tin.it>

Dennis Lee Bieber wrote:
        ...
>         Ah, but if you can persuade the Louvre to /sell/ you the Mona
>         Lisa,
> you are then perfectly free to do what you want with it.

You don't seem to be familiar with French (and more generally European)
laws about what you ARE allowed to do with works of art classified as
"National Treasures" or the like, even when you do own them.

Even in the US, the fact that you've persuaded somebody to sell you
a house, say, doesn't mean you're then perfectly free to do what you
want with it -- local regulations can constrain you so that, for
example, you may not be allowed to paint its facade in wide stripes
of pink and lime green.


Alex



From timr at probo.com  Fri Sep 27 01:12:57 2002
From: timr at probo.com (Tim Roberts)
Date: Thu, 26 Sep 2002 22:12:57 -0700
Subject: http file upload from Python (multipart/formdata) how-to?
References: <3d937091$0$181$9b622d9e@news.freenet.de>
Message-ID: <ujp7pu4osgoagpnqr4ls727cibm2ua5jq1@4ax.com>

"Robert" <k.robert at gmx.de> wrote:

>want to upload file data (multipart formdata) - same as upload from
>web-browser <input type=file ..>
>
>urllib.urlencode  + urlopen takes only simple name/value tuples
>
>how do I render multipart formdata ?

I guess you will have to generate the HTTP/MIME headers and send the file
in your own code.  See RFC 1867.

Are you trying to interface with a web form that already exists, or are you
just trying to do a file transfer from one machine to another?
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From tim.one at comcast.net  Mon Sep  2 17:46:56 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 02 Sep 2002 17:46:56 -0400
Subject: Regex for String Literals
In-Reply-To: <47i7nu8q5b3e5merhe7s19qvgsf244nkfq@4ax.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEJHBBAB.tim.one@comcast.net>

[Stefan Franke]
> Does someone know a regular expression that matches all
> kinds of Python string literals (along with their  finer points
> WRT line breaks, unicode..)?

tokenize.py (in the std library) strives to match the Python compiler's
tokenization exactly.  You'll find a suitable collection of hairy regexps
there, but, if you can, find a way to *use* tokenize.py directly.  Using the
generator interface this is less mind-bending than it used to be (you can
iterate over a token stream instead of fighting with stateful callback
functions).




From tylerdavis15 at hotmail.com  Mon Sep 16 09:13:33 2002
From: tylerdavis15 at hotmail.com (davisman)
Date: 16 Sep 2002 06:13:33 -0700
Subject: Good book choice?
References: <e6c4872c.0209150635.4816ef98@posting.google.com> <3D850D21.9E34DDEF@bellatlantic.net>
Message-ID: <2393234a.0209160513.469a4357@posting.google.com>

If you want a book with the code included, buy "Programming with
Python" by Tim Altom and Mitch Chapman. This book includes a CD with
all the scripts used in the book, as well as many other programs such
as the Grail Internet Browser, which is written entirely in Python,
and as always, open source. Hope it helps!
-Ty
> ...  I am looking for a suited Python book. I'd like it to have
> > clear ideas about OO, as well as an straight pointing into the core
> > language.
> > 
> > It seems that 'Python Essential Reference (2nd Edition)' is a good
> > choice. There is also the book 'Making use of Python' (by Rashi
> > Gupta). Is there any user that bought both? Is it good the treatment
> > of OO in the former? Is there a lack of applications point of view in
> > the former? Could anybody compare them? pros/cons?
> > 
> > Thanks in advance,
> > Eva.


From peter at engcorp.com  Sun Sep 15 08:56:15 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 15 Sep 2002 08:56:15 -0400
Subject: registry question (_winreg)
References: <bcb4afaa.0209141952.6a1c53@posting.google.com>
Message-ID: <3d84835f$1@news.sentex.net>

idjhh wrote:
> hi ^^
> i write this code today , but it is not running -_-;
> i don`w find error position
> see my code and then please advise me ^^

When posting questions of this nature, please *always* cut
and paste the exact error message/traceback or whatever that
you encountered.

"Not running" is not a useful description.  Also "I don't
find error position" is unclear, although it suggests that
the traceback may not have contained the necessary information
to point you to the line with the error.  Either way, the
exact error message (please don't just retype it) would help.

-Peter




From =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?=  Fri Sep 13 13:25:16 2002
From: =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?= (=?iso-8859-5?Q?=B4=D0=DC=F8=D0=DD_=B3=2E?=)
Date: Fri, 13 Sep 2002 19:25:16 +0200
Subject: calling Oracle functions
References: <3d81f811$1@news.mt.net.mk> <mailman.1031929598.17983.python-list@python.org> <3d8210a3@news.mt.net.mk> <mailman.1031935337.20279.python-list@python.org>
Message-ID: <3d821f71$1@news.mt.net.mk>

>> Now a simpler question:
>> 	is it possible to specify the data source name directly not through
>> 	the alias defined in tnsnames.ora. In perl DBD I can specify a
>> 	connection string like cs = "dbi:Oracle:host=1.10.100.200;sid=TEST"?

> cx_Oracle is passing the string directly to Oracle, so whatever it
> accepts, cx_Oracle will accept. Specifically, you can take the entire
> string that is in your tnsnames.ora file and pass it directly as in

> v_TNS =
> (DESCRIPTION=(ADDRESS_LIST=(ADDRESS=(PROTOCOL=TCP)(HOST=opus)(PORT=1521)))(CONNECT_DATA=(SID=DEV)))

> v_Connection = cx_Oracle.connect("user", "pw", v_TNS)

> and that works just fine. Does that answer your question?

Yes. Thanks. I would only suggest that you can add a function to
cx_Oracle to make that big ugly string from 3 parameters.
something like:

def makeDSN(host,port,sid):
	return 	"""(DESCRIPTION= (ADDRESS_LIST= (ADDRESS= (PROTOCOL=TCP)
	(HOST=%s) (PORT=%s))) (CONNECT_DATA=(SID=%s)))""" %
	(host,port,sid)





-- 
??????

           Intel: where Quality is job number 0.9998782345!



From pinard at iro.umontreal.ca  Wed Sep 25 11:58:46 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Wed, 25 Sep 2002 11:58:46 -0400
Subject: missing functions in os.path
In-Reply-To: <31575A892FF6D1118F5800600846864DCBD876@intrepid> (Simon
 Brunning's message of "Wed, 25 Sep 2002 14:41:52 +0100")
References: <31575A892FF6D1118F5800600846864DCBD876@intrepid>
Message-ID: <oq8z1qrrgp.fsf@carouge.sram.qc.ca>

[Simon Brunning]

> All patched gratefully received, I'm sure!

I sometimes use the following code (file name here is `nicepath.py').
The documentation is a bit terse, I confess! :-)


#!/usr/bin/env python
# Copyright ? 2000, 2001, 2002 Progiciels Bourbeau-Pinard inc.
# Fran?ois Pinard <pinard at iro.umontreal.ca>, 2000.

import os, string

def canonical(path, viewpoint=None):
    return string.join(canonical_parts(path, viewpoint), '/')

def nicest(path, viewpoint=None, relative=False):
    if viewpoint is None:
        eye_parts = canonical_parts(os.getcwd())
    else:
        eye_parts = canonical_parts(viewpoint)
    path_parts = canonical_parts(path, viewpoint)
    maximum = min(len(eye_parts), len(path_parts))
    for counter in range(maximum):
        if eye_parts[counter] != path_parts[counter]:
            break
    else:
        counter = maximum
    if relative or len(eye_parts) < 2*counter:
        return string.join((['..'] * (len(eye_parts) - counter)
                            + path_parts[counter:]),
                           '/')

def canonical_parts(path, viewpoint=None):

    def split_and_append(parts, name):
        for part in string.split(name, '/'):
            if part == '..':
                parts.pop()
            elif part != '.' and (part or not parts):
                parts.append(part)
        return parts

    if path[0] == '~':
        path = os.path.expanduser(path)
    elif path[0] != '/':
        if viewpoint is None:
            viewpoint = os.getcwd()
        path = os.path.join(viewpoint, path)
    parts = split_and_append([], path)
    counter = 0
    while counter < len(parts):
        name = string.join(parts[:1+counter], '/')
        if name and os.path.islink(name):
            link = os.readlink(name)
            if link[0] == '/':
                link_parts = split_and_append([], link)
            else:
                link_parts = split_and_append(parts[:counter], link)
            parts = link_parts + parts[1+counter:]
            counter = 0
        else:
            counter = counter + 1
    return parts

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From schull at digitalgoods.com  Wed Sep  4 11:15:43 2002
From: schull at digitalgoods.com (Jon Schull)
Date: 4 Sep 2002 08:15:43 -0700
Subject: Tkinter report_callback_exception?    'module' object has no attribute 'print_exception' ?
Message-ID: <2621b014.0209040715.1bf9fce3@posting.google.com>

Hi,

As I interpret this message, the environment is telling me that
traceback has no "print_exception."   But of course, it does.

I think a file access error is triggering my problem, but why can't I
get a clue from the traceback?

Exception in Tkinter callback
Traceback (most recent call last):
  File "C:\jon\python\PIL\all.py", line 101, in ?
    root.mainloop()
  File "C:\Python22\lib\lib-tk\Tkinter.py", line 929, in mainloop
    self.tk.mainloop(n)
  File "C:\Python22\lib\lib-tk\Tkinter.py", line 1296, in __call__
    self.widget._report_exception()
  File "C:\Python22\lib\lib-tk\Tkinter.py", line 1061, in
_report_exception
    root.report_callback_exception(exc, val, tb)
  File "C:\Python22\lib\lib-tk\Tkinter.py", line 1552, in
report_callback_except
ion
    traceback.print_exception(exc, val, tb)
AttributeError: 'module' object has no attribute 'print_exception'


From CousinStanley at HotMail.com  Mon Sep 30 09:29:58 2002
From: CousinStanley at HotMail.com (Cousin Stanley)
Date: Mon, 30 Sep 2002 06:29:58 -0700
Subject: Data Plotting Library DISLIN 8.0
References: <3D940A37.92FD45D5@linmpi.mpg.de> <an23s1$ambss$1@ID-130333.news.dfncis.de>
Message-ID: <an9jfs$ca7av$1@ID-130333.news.dfncis.de>

|| Yes, you can download the file 'dl_80_py.zip' that contains the 
|| DISLIN distribution for Python 2.2 on Windows. You don't have to
|| compile anything. The other DISLIN distributions are for several
|| commercial and non-commercial compilers. 

Cousin Helmut ... 

Thanks for the reply ... 

After reading the FAQs from your website a bit more closely,
I downloaded  dl_80_py.zip  and  installed yesterday ... 

The install process was very easy and all the Python demonstration 
programs included with the distribution functioned as expected
and I'm looking forward to using DISLIN ... 

I was just a bit confused about whether or not any compilation 
was required, since you had listed so many different distributions
with so many different compiler options available ... 

Cousin Stanley



From brueckd at tbye.com  Tue Sep 17 16:35:59 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Tue, 17 Sep 2002 13:35:59 -0700 (PDT)
Subject: Earing money with Python software (was: More on Protecting Source
 Code)
In-Reply-To: <4jt7ma.fse.ln@lenashee.goebel-consult.de>
Message-ID: <Pine.LNX.4.44.0209171320230.30537-100000@ns0.tbye.com>

On Tue, 17 Sep 2002, Hartmut Goebel wrote:

> Martin v. Loewis wrote:
> 
> > For software, I think the economical risk is that somebody uses your
> > software without giving you the money you want - not that somebody
> > sits down and studies your code for days, to write a program that he
> > could have written himself from scratch in the same time.
> 
> Which brings me to another (related) topic): How to earn money
> distributing Python software?

I think the answer to that question is more or less the same regardless of
the language - nothing Python-specific there. It probably depends almost
entirely on your target market.

> - Sell it via micro-payment? 

This one always appeals to me but today there are no standard & safe ways
to make payments, much less micropayments (hehe.. imagine entering credit 
card info for each 50 cent charge).

> - Declare it "Shareware" and hope somebody pays for it?

Too little incentive to pay. Most people are more or less honest, unless 
you make it _too_ easy not to pay. You have to give them a little more 
incentive, whether it be extra features, support, or an end to nags.

> - Offer commercial support?

This works well if your target audience is a commercial entity.

You asked the age-old question though. I'm partial to the completely free
or free-for-personal-use-but-companies-must-pay approach when possible.
Otherwise, put a bit of a "wall" up protecting your code but not much
beyond that. It's a futile arms race to try and lock everyone out (and the
techniques to do it frequently anger your legitimate customers), and the
people who tend to crack programs usually wouldn't pay for it anyway.

 -Dave




From b.maryniuk at forbis.lt  Thu Sep 26 07:39:21 2002
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Thu, 26 Sep 2002 13:39:21 +0200
Subject: Zope, MySQL and pictures in blob-fields ...
In-Reply-To: <amuqck$e0r$1@nathan.regio.net>
References: <amuqck$e0r$1@nathan.regio.net>
Message-ID: <200209261339.21600.b.maryniuk@forbis.lt>

This is crossposting. Please subscribe to the Zope maillist at zope at zope.org 
and post this there.

On Thursday 26 September 2002 13:16, Thilo R??ler wrote:
[skipped]
> "<ZPublisher.HTTPRequest.FileUpload instance at 0x893a7ec>".

-- 
Regards, Bogdan

Kiekvienas lai?kas su lietuvi?komis raid?mis praturtina Alberto ispell
?odyn? -- http://ieva.mif.vu.lt/~alga/lt/ispell/




From whisper at oz.net  Wed Sep 18 22:34:08 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 18 Sep 2002 19:34:08 -0700
Subject: How to build Python 2.2.1 on Windows?
In-Reply-To: <3D892BBD.5B38DD49@bellatlantic.net>
Message-ID: <GCEDKONBLEFPPADDJCOEIEAJFBAA.whisper@oz.net>

I have built Python on Windows using both the distribution and CVS download.
The necessary VC++ project files are included in both. You will have to also
download some other components as described in the readme file, including a
special version of BSDDB version 1.85, zlib, Tcl/Tk (Tk needs Tcl to build),
expat and maybe some other things. These are for extensions, and are not
required to build the core Python.

WRT BSDDB, there are projects out there to build Sleepycat BSDDB version 4.0
and that's a better choice then the default version 1.85 used by Python (due
to bugs). The downside is that Sleepycat BSDDB is free for personal use
only.

See \dist\Python-2.2.1\PCbuild for complete details.

Oh yes, put your python source distro or cvs dump under a directory called
'dist'. I found the directions to do that unclear in the readme and it gave
me a bit of trouble. The other packages you'll need also go in 'dist' as
peers of the python source tree. Those should be built before Python.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of David Lees
> Sent: Wednesday, September 18, 2002 18:43
> To: python-list at python.org
> Subject: How to build Python 2.2.1 on Windows?
>
>
> I have been playing with Python for awhile and would like to know how to
> try building Python on a Windows machine that has MS C++ version 6.  Is
> there a standard download that is configured to do this?  If not could
> someone point me in the direction of information on builds?  I
> downloaded Python-2.2.1.tgz.tar on my Win98 box and WinZip 8 does not
> even seem to know how to unzip it.  I just transfered the tar file over
> to a linux box and it unzips without and problem using tar.  The README
> seems to only talk about linux/unix type systems.
>
> David Lees
> --
> http://mail.python.org/mailman/listinfo/python-list




From magnus at thinkware.se  Sat Sep 28 21:47:15 2002
From: magnus at thinkware.se (=?ISO-8859-1?Q?Magnus_Lyck=E5?=)
Date: Sun, 29 Sep 2002 03:47:15 +0200
Subject: Difference between MOD_PYTHON and the CGI module
References: <3d904321$0$234$4d4ebb8e@news.nl.uu.net>
Message-ID: <3D965BA3.10905@thinkware.se>

Gumuz wrote:
> What are the differences, if any?
> 
> Am i mixing up to things here?
> 
> Suppose I'd like to use python as a 'replacement' for PHP, what do I use?

As far as I understand, PHP typically embeds code in HTML
pages. If you want to do that you want something more than
mod_python.

Have a look at http://www.python.org/cgi-bin/moinmoin/WebProgramming
and look closer at the templating systems.

I recently found CherryPy, www.cherrypy.org , and I like that.


-- 
Magnus Lycka, Thinkware AB
Alvans vag 99, SE-907 50 UMEA, SWEDEN
phone: int+46 70 582 80 65, fax: int+46 70 612 80 65
http://www.thinkware.se/  mailto:magnus at thinkware.se



From oren-py-l at hishome.net  Wed Sep 18 11:18:34 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Wed, 18 Sep 2002 18:18:34 +0300
Subject: Stackless Python, eventual merge?
In-Reply-To: <3D8880AC.7040902@tismer.com>; from tismer@tismer.com on Wed, Sep 18, 2002 at 03:33:32PM +0200
References: <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <20020918060823.GB81158@hishome.net> <3D8880AC.7040902@tismer.com>
Message-ID: <20020918181834.A27703@hishome.net>

On Wed, Sep 18, 2002 at 03:33:32PM +0200, Christian Tismer wrote:
> Oren Tirosh wrote:
> > They may be dangerous in some circumstances. IIRC, if a C extension calls
> > the Python interpreter recursively with pointers to things on its stack
> > it can cause memory corruption. 
> 
> Then don't switch stacks in that context.
> The switching occours on your demand.
> But the theoretical possibility to create a crash
> is no reason to dispense with the benefits.

Definitely not. Don't get me wrong - I'm very excited about stackless. 

> My only chance to do this magic with almost no crash
> possibility would be to allocate a new stack for
> every new tasklet. 

Yes, that's a possible solution. I wouldn't say that it's the only
chance, though. There's always more than one way to skin a cat.

> A benefit would be reasonably faster switching, while
> the drawback would be the reasonable high memory
> requirement for a full machine stack.
> I might consider a hybrid solution, putting certain
> "dangerous" extension stuff into a real new stack
> all the time, together with the quite fast and very
> cheap other safe tasklets.
> 
> Just-need-an-idea-how-to-determine-that - ly y'rs - chris

Who says you have to determine that in advance? Just try. If the
extension actually calls the interpreter recusively it means that it
needs its own stack.  Let it have exclusive ownership of the current OS
thread and its associated stack and allocate another thread from a
dynamic thread pool to continue tasklet scheduling. When the extension
returns back to nesting level 0 its thread will be released back into
the pool. This way a relatively small pool of OS threads can be used to
juggle a much larger pool of tasklets.

This can also be a good way to handle extensions that make blocking I/O
calls like database engines - a relatively small pool of OS threads can
serve requests from a bigger pool of tasklets. This can be transparent:
when the number of concurrent requests exceeds the number of threads any
new request frames can be queued until a thread becomes available. The
level of concurrency for any resource can be tuned dynamically to
optimize the performance without affecting the overall structure of the
program.

	Oren




From syver-en+usenet at online.no  Tue Sep 24 15:18:28 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Tue, 24 Sep 2002 19:18:28 GMT
Subject: Limit number of concurrent threads
References: <Xns929393554CAB0cmay4yahoocom@165.112.130.8>
Message-ID: <uwupbdwnj.fsf@online.no>

Chuck May <cmay4 at yahoo.com> writes:

> I've searching for examples that fit my problem, and I haven't found 
> any.  Most examples I've found look something like this:
> 
>     threads = []
>     for i in range(number_of_tasks):
>         thread.append(thread_for_task)
>     
>     # let the threads run - just wait for shutdown
>     for t in threads:
>         t.join() 
> 
> which whould work fine if I wanted to start all 80 threads at once.
> 
> I've read about the Queue, but I'm not sure if it's what I want, and I
> 
> can't seem to find any good examples on it.

I think Queue would be nice to use, Have a thread (the main thread
perhaps) feed all the table names (or a table object if you please)
into a Queue. 

Set up some worker threads with access to the Queue
(global variable, member). The worker threads should loop something
like this:

while 1:
   queueItem = self._queue.get()
   doSomethingWithQueueItemHere

Then you can start your workers.

(You'll have to find out some way to make the workers exit when the
work is done, perhaps putting a number of None's equal to the number
of workers last in the queue and testing for None in the working thread loop.


-- 

Vennlig hilsen 

Syver Enstad


From sholden at holdenweb.com  Fri Sep 13 15:00:56 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 13 Sep 2002 15:00:56 -0400
Subject: Cursed newlines and readline()
References: <6u5g9.11312$Da3.1133980@kent.svc.tds.net> <mailman.1031923697.30356.python-list@python.org>
Message-ID: <RCqg9.8231$WV.5021@fe03>

"Thomas Wouters" <thomas at xs4all.net> wrote ...
> On Thu, Sep 12, 2002 at 06:59:14PM +0000, Edward K. Ream wrote:
>
> > According to one of Leo's Linux users, the Python's readline() routine
on
> > Linux delivers "\r\n" line end strings verbatim, while the windows
versions
> > force the string to use the Unix convention of using only "\n". This
seems
> > unfortunate.  Is there a justification for it?  If not, I would like to
> > suggest that readline() work the same on all platforms and that it
deliver
> > only "\n" as a line terminator.
>
> You can open the file in 'binary' mode (adding 'b' to the mode string) and
> the file will contain '\r\n' on both platforms (and any other platforms.)
>
Nope. Exactly wrong. In 2.2 and those before, when files are opened in
*text* mode (no "b") then reading them will provide Unix-style line endings
(newline only). When you open files in binary mode then you see the bytes
stored in the file.

On Unix systems there's no difference in the contents of a file whether in
binary or text mode. On Windows a file is shorter by the number of carriage
returns. On the Mac I have no idea what they do. Probably just carriage
returns, to be different :-)

2.3 will be a bit more flexible about such mattrers.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From theoryboy at my-deja.com  Wed Sep  4 10:01:48 2002
From: theoryboy at my-deja.com (Peter Saffrey)
Date: 4 Sep 2002 07:01:48 -0700
Subject: efficient list merging
References: <ced73313.0209031012.13624f3@posting.google.com> <al3bfi$2kh$0@216.39.172.122>
Message-ID: <ced73313.0209040601.64edac88@posting.google.com>

I was trying to abstract a bit. The way I get the lists of lists of
strings is far too long and ugly to contemplate posting here. Sorry if
I didn't post compilable code - again I was trying to abstract.

Here's an example:

lol1 = [['a'], ['a', 'b', 'c'], ['d', 'e', 'f'], ['g', 'h', 'i', 'j']]
lol2 = [['d', 'e', 'f'], ['k', 'l', 'm'], ['d'], ['f', 'g'], ['a']]

I want a function which returns

mergelol = [['a'], ['a', 'b', 'c'], ['d', 'e', 'f'], ['g', 'h', 'i',
'j'], ['k', 'l', 'm'], ['d']]

That is the set-union of the two lists of lists. Note that I've used
very short strings and short list lengths here for simplicity. The
strings would be up to 80 characters in length, the sub-lists up to 8
items in length, and the list of lists itself could contain 50
sub-lists.

Peter


From andrew.thompson at ashecastle.com  Thu Sep 19 07:00:08 2002
From: andrew.thompson at ashecastle.com (Andrew Thompson)
Date: Thu, 19 Sep 2002 12:00:08 +0100
Subject: Smalltalk style inspector
In-Reply-To: <20020919074722.GB77021@partagas.as.de.cw.net>
Message-ID: <007c01c25fcb$b81ccc10$4802a8c0@blair>

Is there a Smalltalk style inspector (GUI with drilldown through
data-structures) for Python ?

I have written a rough-and-ready one in TK which is just a mixin-class
Which confers the method .inspect() to any class, and then allows a user
to wander through attributes, doc-strings , dictionaries and collections
by double-clicking with the mouse, but not sure if it might be of value
to others.

I don't even know the process for contributing the module...

Andrew







From phr-n2002b at NOSPAMnightsong.com  Mon Sep 30 00:43:17 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Sep 2002 21:43:17 -0700
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05> <an7bvd$burft$1@ID-125932.news.dfncis.de> <an7p7t$cpg$0@216.39.172.122> <mailman.1033338759.6775.python-list@python.org> <7xd6qwqtgz.fsf@ruckus.brouhaha.com> <mailman.1033352259.30842.python-list@python.org> <7xelbcgo56.fsf@ruckus.brouhaha.com> <mailman.1033357298.16341.python-list@python.org> <7x65woks0i.fsf@ruckus.brouhaha.com>
Message-ID: <7x1y7ckryy.fsf@ruckus.brouhaha.com>

Paul Rubin <phr-n2002b at NOSPAMnightsong.com> writes:
> Example: bunches are three for a dollar.  How much do you pay if you
> buy nine bananas?

Whoops, meant "bananas are three for a dollar".  Sorry.


From mcherm at destiny.com  Tue Sep 10 13:53:18 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Tue, 10 Sep 2002 13:53:18 -0400
Subject: dynamism
References: <66715c02.0209091923.265fe77e@posting.google.com> <Xns92855D48384F9duncanrcpcouk@127.0.0.1> <3d7e209a_1@news.iprimus.com.au>
Message-ID: <allbi9$gqh$1@main.gmane.org>

Steven Shaw wrote:
>>An object of a builtin type, or with __slots__ defined in the class and
>>all base classes does not have a __dict__ attribute.
> 
> 
> Does this mean that the attributes are stored in a more efficient way?
> 
> Is there a way to implement "record-based" object layout rather than
> "associative-array based" object layout?

In short, YES.

See http://python.org/doc/2.2.1/whatsnew/sect-rellinks.html . 
Unfortunately, http://www.python.org/peps/pep-0253.html doesn't give 
full details since they haven't been written up yet.

-- Michael Chermside






From dgallion1 at yahoo.com  Tue Sep 17 12:42:09 2002
From: dgallion1 at yahoo.com (Darrell)
Date: 17 Sep 2002 09:42:09 -0700
Subject: how to cast an instance id
References: <3D86E888.90AF569@millfilm.co.uk>
Message-ID: <b0c5da3a.0209170842.53e5a5f3@posting.google.com>

class foo:
    def info(self):
        print self

a = foo()
a.info()
ccc = "%s()"%( str(a.__class__).split(".")[-1], )
eval(ccc).info()
exec("b=%s"%ccc)
b.info()

<__main__.foo instance at 0x00767378>
<__main__.foo instance at 0x007673B0>
<__main__.foo instance at 0x007673B0>

--Darrell

Eric Texier <erict at millfilm.co.uk> wrote in message news:<3D86E888.90AF569 at millfilm.co.uk>...
> It there a way to create a new reference of  an object
> from the string representation of its id?
> Thanks for any help.
>


From alessandro.riolo at sen.it  Fri Sep 27 09:01:44 2002
From: alessandro.riolo at sen.it (Alessandro Riolo)
Date: Fri, 27 Sep 2002 15:01:44 +0200
Subject: [OT] Crosspost? [was Re: Zope, MySQL and pictures in blob-fields ...]
References: <amuqck$e0r$1@nathan.regio.net> <mailman.1033040674.25353.python-list@python.org>
Message-ID: <an22k1$abfap$1@ID-44839.news.dfncis.de>

Bo M. Maryniuck wrote:
> This is crossposting.

Why this was crosspost? Off topics, perhaps, but surely the header of
the post wasn't showing crosspost.

--
ale
www.sen.it




From kibab at icehouse.net  Mon Sep 23 13:29:47 2002
From: kibab at icehouse.net (Kaleb Pederson)
Date: Mon, 23 Sep 2002 10:29:47 -0700
Subject: parsing tabs in a string
In-Reply-To: <006201c26324$4148ae90$025e10ac@calibredd.com>
Message-ID: <003001c26326$d05a96a0$e90dbb92@KWINXP>

If you don't specify where find should start searching, it ALWAYS starts
at the beginning of the line and looks for the first occurrence of the
given substring.  Hence,"\t" is always found at that character. The
documentation clearly states that so you might want to double check it
if you aren't getting the desired behavior.

 

You might try:

 

>>> line = 'this\tis\ta\ttest\tof\t"mydatehere"'

>>> newline = line.split('\t')

>>> print newline

['this', 'is', 'a', 'test', 'of', '"mydatehere"']

>>> for x in newline:

...     print x

...

this

is

a

test

of

"mydatehere"

 

That will break list into several lists where each column is whatever
was in between your tab markers.

Or, if you really want to go through and find all the '\t's.. then.

 

            

>>> mystr = 'this\tis\ta\ttest\tof\t"mydatehere"'

>>> index = mystr.find('\t')

>>> while index != -1:

...     print 'index found at:',index

...     index = mystr.find('\t',index+1)

...

index found at: 4

index found at: 7

index found at: 9

index found at: 14

index found at: 17

 

Hopefully that helps.

 

--Kaleb

-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]
On Behalf Of Andrew Alzner
Sent: Monday, September 23, 2002 10:11 AM
To: python-list at python.org
Subject: parsing tabs in a string

 

Hello,

 

I'm getting to know Python and have a quick question for something I
don't understand.

 

I'm trying to parse information extracted from a spreadsheet which is
tab separated. Here is a simplified version of what I have:

 

list=['"103-01a-17"\t2\t33\t256\t227\n']
y=list[0]
print y.find("\t")
print y.find("\t",1)

 

I thought the second .find should return the second occurence of "\t".
To get the index of the second tab I have to type:

 

print y.find("\t",13)

 

This doesn't make sense to me. Can someone help me please.

 

Thanks,

Andrew

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020923/21a61394/attachment.html>

From rsrchstr at msn.com  Sun Sep 29 20:54:59 2002
From: rsrchstr at msn.com (mike henley)
Date: Mon, 30 Sep 2002 01:54:59 +0100
Subject: idea for a much needed application
References: <576c1752.0209291326.56dcdf31@posting.google.com>
Message-ID: <an87db$bjtvs$1@ID-157863.news.dfncis.de>

i guess i'll rephrase; what i wish for is an "application-to-person" (P2A)
IM-based "presence", so that a web-application that i host on a personal
webserver, on my own machine, can let others become aware of its "presence"
easily and be easy to connect to with a simple
click-to-launch-browser-and-go-to-my-ip. This should be easy to do. I just
don't have experience with Jabber or xml-rpc, or other potential useful
technologies.

"mike henley" <rsrchstr at hotmail.com> wrote in message
news:576c1752.0209291326.56dcdf31 at posting.google.com...
> I remember a while ago reading about tim burners-lee and how his
> vision of the WorldWideWeb as a medium where all can freely and easily
> publish information has not materialized properly, and i felt it so
> strongly over the past week when trying to look at options for hosting
> a website. Hosts are either expensive or limit you in terms of what
> languages, modules, applications, and OSs you can use. They also limit
> you in terms of space, bandwidth used per period, number of executions
> of CGIs... etc. Some of those who give good deals, seem difficult to
> trust with either your credit card number or your content.
>
> I wish i can just use my own computer, with whatever technologies I
> choose, keeping my content that i build up over time safe here on my
> own machine, but unless you have a broadband connection this is not
> possible. It may be possible using a dynamic DNS provider, but then
> again either they add cost, end up with an ugly URL, and, most
> importantly, there's no way for others to know when you get online. So
> here comes the idea that occured to me, and i post it here to ask if
> you know of an application that already provides this functionality,
> then please tell me what it is, or if you think it is worth doing on
> an open source basis, or even proprietary, in which case i invite
> anyone to do so, or if you think it's a completely stupid and
> impossible idea, then perhaps you can please tell me why.
>
> You probably guessed if you read as far as the second paragraph that
> what i'm thinking about is an application functionality that enables
> instant messaging to provide the dynamic DNS functionality. So instead
> of giving others a URL I can just give them an IM username they can
> add to their client, and when i go online, through my humble dial-up
> connection that changes IP whenever i log in, they'll be aware of my
> presence, and be able to connect to a server i host on my machine.
>
> It sounds like a simple thing; IM presence, automatic IP query of my
> username, click on my username and or use a contextual menu option or
> whatever and a browser launches to go to the site i'm hosting on my
> machine. Perhaps it can use existing open-source platforms, such as
> jabber or gnutella, or as an add-on to existing open-source
> applications/clients. Basically what i'm suggesting is a little tool
> to enable a server such as apache to join the whole IM/p2p thing and
> have "presence". The more automated it is, and the easier it is, the
> better. I think such a functionality will be especially useful for
> certain uses, in my case it will be hosting a wiki, as i leave my
> machine online a lot through my dial-up, can't predict the number of
> executions of CGI or the amount of space needed, wish to be free to
> use whatever wiki application is best for my needs and in whatever
> language regardless of what a host will support, and do not wish to
> pay a monthly fee, extra charges, setup fee or a domain registration.
> Perhaps there can be other areas where it is useful, and i'm sure
> there are other areas where it will not be suitable. Perhaps there can
> also be a simple way to enable a page hosted in such a way to be
> submitted to search engines, i haven't thought about that yet, but
> tell me if u see any.
>
> What do you think? already done, worth doing, stupid, or impossible?
> (p.s. please also let me know if there are other newsgroup to which
> posting this will be a good idea)




From pcolsen at att.com  Thu Sep 12 13:08:21 2002
From: pcolsen at att.com (Olsen, Peter C, SOFED)
Date: Thu, 12 Sep 2002 13:08:21 -0400
Subject: Smoothing a discrete set of data
Message-ID: <8527B21850BECB478454F14257320562B46009@ACCLUST02EVS1.ugd.att.com>

This is a followup to Paul Moore's post of  07 September 2002.

Paul wrote:

> I have a set of data, basically a histogram. The data is pretty
> variable, and I'd like to "smooth" it to find trends. Actually,
> this comes up a *lot* for me - some examples: sampled IO rates on
> a machine - I want to look for trends (is IO higher overnight or
> during the day, etc) or fuel consumption for my car (do I use
> less fuel since I had the service).
> 
> Normally, what I do with things like this is draw a graph, and
> try to spot the trends "by eyeball". But it feels to me like I
> should be able to write something which smooths the data out. I
> just don't see how :-)
> 
> Take an example - here's a rough ASCII-art diagram
> 
>   9     XX                                  XX
>   8     XX XX                               XX
>   7  XX XX XX                               XX XX
>   6  XX XX XX                            XX XX XX
>   5  XX XX XX                            XX XX XX
>   4  XX XX XX                            XX XX XX
>   3  XX XX XX XX          XX          XX XX XX XX
>   2  XX XX XX XX XX    XX XX XX    XX XX XX XX XX
>   1  XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX
> 
> This seems to me to have very clear peaks at the left and right,
> and a shallow point in the middle. I'd say that you could
> "smooth" this to something that averaged 8 for the first 3 bars,
> then 2 for the next 9, then 7 for the last 3. But try to make
> that precise...
> 
> 

You have hit on two of the fundamental problems of data analysis:

1.  How to separate data from noise, and

2.  What is the "best" way to do it.

First, let's simplify the problem and make it precise.  What I think you
want to do is find the "large trends" in the data and get rid of the
noise.  

You give sampled IO rates as an example.  Network traffic is a similar
one.  Network traffic seems to have a regular daily pattern.  It's low
in the early morning, rises until before noon (perhaps about 11:00),
drops for lunch, rises again in the afternoon, and then falls at the end
of the workday.

That's the overall pattern (which is what I think you call a trend), but
the real data is never as simple as that.  

If we look at short intervals, say 10 minutes, data is likely to be very
ragged, with lots of hills and valleys.  What I think you'd like to do
is smooth out these hills and valleys.

To start to do that, let's build a mental model.  Let's think of the
data as the sum of an underlying curve and a sequence of random errors.
Let's further assume that there is no relationship between successive
errors.  (This is reasonable because we want to put all the relationship
between successive samples into the underlying curve.)

Here's how you could make such a random sequence.  For every data point,
roll a die and flip a coin.  If the coin is heads, add the number on the
die to the value of the data point; if it's tails, subtract instead.  If
you take a particularly easy set of data points (setting every one to
the constant zero) and plot these errors, you'll see that they are
randomly distributed between -6 and 6 (excluding the case of 0 error
because the die has no zero face).

This is a good example because actual curve is simple: a straight line
at zero.

There are two simple ways to eliminate the random errors (and many
complex ones I won't mention).

One is a "moving average," the other is a "running median."  Each has
its own advantages.  Both are simple to implement.

For a moving average, take any odd number of successive data points,
take their arithmetic average (the sum of the values divided by the
number of points), and replace the middle point with the average.  

For example, consider the sequence:

s = [5, 5, 6, 3, 5, 5, 2, 5, 6, 6, 4, 1, 1, 3, 6, 1, 2, 2, 1, 1]

These are random integers drawn with a uniform distribution from the set
{1,2,3,4,5}, so the underlying curve should be the constant 3.  

If we take moving averages for a length of three, we get the sequence.

[5.33, 4.67, 4.67, 4.33, 4.0, 4.0, 4.33, 5.67, 5.33, 3.67, 2.0, 1.67,
3.33, 3.33, 3.0, 1.67, 1.67]

Notice that this sequence is two elements shorter than the original one.
This is because there is no way to calculate the moving average for the
points.  There are no other points "beyond the end" to complete the
average.  

Here's the Python function which computes the moving average of three.

def ma3(s):
	return map(lambda x,y,z: (x+y+z)/3.0, s[0:-2], s[1:-1], s[2:])

Applying the function recursively three times
("ma3(ma3(ma3(ma3(ma3(seq)))))") yields 

[4.34, 4.42, 4.56, 4.56, 4.25, 3.68, 3.13, 2.80, 2.74, 2.69]

If you apply the function again several times, you'll see more changes,
but the successive changes become smaller and smaller.  Because you are
shortening the list by two every time, eventually, you'll run out of
data.

The other basic smoother is to consider any odd number of successive
data points and replace the center one by the median.  

The median of a sequence of numbers is the number in the middle of them
when they are sorted.  For example the median of the sequence
(2,5,5,10997,1,3,0) is 3.  If we sort the original sequence we get
(0,1,2,3,5,5,10997).  There is the same number of data points above 3 as
there is below it. 

Here is some very inelegant code to calculate running medians of three:

def rm3(s):
	temp = map(lambda x,y,z:[x,y,z], s[:-2], s[1:-1], s[2:])
	temp2 = []
	for x in temp:
		x.sort()
		temp2.append(x[1])
	return temp2

(I'm sure that this can be rewritten in a functional form.)

Using this function to apply running medians of three to the initial
sequence yields 

[5, 5, 5, 5, 5, 5, 5, 6, 6, 4, 1, 1, 3, 3, 2, 2, 2, 1].

You should notice one big difference right away; there are no fractions.
Each number in the sequence is always replaced by a number from the
sequence (maybe itself).  This gets rid of statements like "the average
family has 1.8 children."  This sometimes gives running medians the edge
when dealing with discrete data such as the number of I/O connections
per minute.

Notice too that the running median also reduces the influence of really
wild numbers on the calculation. The median of (2,5,5,10997,1,3,0) is 3,
but the mean (arithmetic average is 1573.29.  This is often why it only
takes one or two very high scores to "raise the average" and "bust the
curve" in an academic examination.  

Deciding whether it is better to use moving averages or to use running
medians is not always easy.  In complex cases it requires deep
understanding of where the data came from, how it was measured, and how
you plan to use the numbers you get.  It's too complex to try to discuss
here. 

So, I hope this will help answer your question.  If not, I hope I
haven't wasted too much of your time.

Peter Olsen, PE, Ae.E.
Senior Technical Staff Member
Government Solutions, ATT




From henrik.motakef at web.de  Wed Sep 18 18:36:57 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 19 Sep 2002 00:36:57 +0200
Subject: linux python ideas
References: <3D88E578.6010001@diespammerdieuselesspython.com>
Message-ID: <87u1kmncvq.fsf@pokey.henrik-motakef.de>

Rob Andrews <rob at diespammerdieuselesspython.com> writes:

> I'll be giving a presentation on Python to my local LUG
> (http://lugoj.org) in a few weeks, and would like to point out at
> least a few things that would be of particular interest to linux users.

Aren't RedHat's installer and Gentoos "portage" packaging system
written in Python?


From dsavitsk at e-coli.net  Sat Sep 14 18:54:21 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Sat, 14 Sep 2002 22:54:21 GMT
Subject: win32security.LogonUser not working
Message-ID: <x6Pg9.8439$yt3.4473830@newssrv26.news.prodigy.com>

I am trying to run code as another user. I copied the code below from ASPN,
so it would seem that someone has gotten this to work. I keep getting the
error

pywintypes.api_error (1314, 'LogonUser', 'A required privilege is not held
by the client.')

Google has revealed that others gave had the same problem.  does anyone know
a solution, or even a different way to do this?

-doug

link to the original of code below
http://www.google.com/search?q=cache:Jnbt2XVyGfUC:aspn.activestate.com/ASPN/
Python/Reference/Products/ActivePython/PythonWin32Extensions/Windows_NT_Secu
rity_.2d.2d_Impersonation.html+self.handel%3Dwin32security.LogonUser&hl=en&i
e=UTF-8

---------------------------------------------
import win32security
import win32con
import win32api
import sys


class Impersonate:

    def __init__(self):
        self.domain   = '.'
        self.login    = 'myusername'
        self.password = '**********'

    def logon(self):
        self.handel=win32security.LogonUser(self.login,
                                            self.domain,
                                            self.password,

win32con.LOGON32_LOGON_INTERACTIVE,

win32con.LOGON32_PROVIDER_DEFAULT
                                            )
        win32security.ImpersonateLoggedOnUser(self.handel)

    def logoff(self):
        win32security.RevertToSelf() #terminates impersonation
        self.handel.Close() #guarantees cleanup




if __name__ == '__main__':
    a=Impersonate()
    try:
        a.logon() #become the user
        print win32api.GetUserName() #show you're someone else
        a.logoff() #return to normal
    except:
        print sys.exc_type , sys.exc_value




From bokr at oz.net  Fri Sep  6 19:44:34 2002
From: bokr at oz.net (Bengt Richter)
Date: 6 Sep 2002 23:44:34 GMT
Subject: Interrupting Python
References: <mailman.1031345649.9840.python-list@python.org> <Ef9e9.7309$rU2.218110@news4.srv.hcvlny.cv.net>
Message-ID: <albel2$7te$0@216.39.172.122>

On Fri, 6 Sep 2002 17:38:55 -0400, "Bob Easton" <bob at eleaston.com> wrote:

>Thanks to all.  My system is Win2K. Python simply ignores Ctrl-c and
>Ctrl-break. I think I'll fall back to using file existence as a simple
If this were true as generally as you put it, we should be looking for a bug.

Would you put the following listed catchCtlC.py file in a convenient directory,
and then go there with **a plain "DOS" command window**, and type the command

    catchCtlC.py 1000

It should complete in a short time. Then try it again, hitting Ctrl-C before
it finishes. If you have a fast machine and need more time, add another zero
to the 1000. Let us know if no interrupt is reported for either Ctrl-C or Ctrl-Break.

Then tell us what's different about your program.

---< catchCtlC.py >------------------
i=x=0
def foo(n):
   global i,x
   for i in xrange(n):
       x = i**100

if __name__ == '__main__':
    import sys
    try:
        foo(int(sys.argv[1]))  # call your thing here instead
    except KeyboardInterrupt:
        print 'Interrupted by Ctrl-C'
    else:
        print 'Exited w/o interrupt'
    print 'FHOI, the last i**100 x value: %s**100 = %s' % (i, x)
---< catchCtlC.py >------------------

>semaphore.  That will be good enough and cheap enough for this application.
>

>--
>Bob Easton
>
>
>"David LeBlanc" <whisper at oz.net> wrote in message
>news:mailman.1031345649.9840.python-list at python.org...
>> > I have a script that can run, accessing network resources, for
>> > several days.
>> > Since the script does not normally need keyed input, exception
>processing
>> > does not raise the keyboard exception until after the program
>> > ends normally.
>> > I would like to be able to interrupt if from the keyboard, but have not
>> > learned the trick.  How can I do this?
>> >
>> > --
>> > Bob Easton
>>
>> I have had the experience of Python progs not responding to ctrl-c or
>ctrl-d
>> on Windows, but I found that ctrl-break works every time.
>>
>> Dave LeBlanc
>>
>>
Also, please post responses below what you are responding to ;-)

Regards,
Bengt Richter


From aleax at aleax.it  Wed Sep 25 05:17:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 09:17:24 GMT
Subject: Python Compiling
References: <amfveg$5g64k$1@ID-53055.news.dfncis.de> <tyf7khczziq.fsf@pcitapi22.cern.ch> <j4r8fkg7s8.fsf@informatik.hu-berlin.de> <3D912249.6040409@something.invalid>
Message-ID: <Eafk9.163255$pX1.5841273@news2.tin.it>

Greg Ewing wrote:
        ...
> etc. You can hardly rely on *anything* staying the
> same from one moment to the next in Python!

Try Ruby, where you can modify built-in types too...
I _thought_ Python was pretty dynamic indeed, until I
started toying a little bit with Ruby...!-)


Alex



From dsavitsk at e-coli.net  Fri Sep 13 10:25:43 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Fri, 13 Sep 2002 14:25:43 GMT
Subject: Tkinter experts out there?
References: <mOfg9.461067$UU1.75649@sccrnsc03> <nJgg9.22$6b7.514@news.get2net.dk>
Message-ID: <Gzmg9.8229$yt3.4315307@newssrv26.news.prodigy.com>

also, Python and Tkinter programming (by John Grayson) comes w/ an author
online forum that seems to still be active?
http://www.manning.com/grayson/

-d


"J?rgen Cederberg" <jorgencederberg at hotmail.com> wrote in message
news:nJgg9.22$6b7.514 at news.get2net.dk...
> "Marcus" <noone at nowhere.com> wrote in message
> news:mOfg9.461067$UU1.75649 at sccrnsc03...
> > I am trying to get some answers on using Tkinter, does anyone know if
> there
> > are channels or forums being actively used?
>
> Hi Matthew,
>
> you can ask away here in comp.lang.python. People are happy to answer your
> questions regarding Tkinter. Other resources of interest are:
>
> An introduction to Tkinter by Fredrik Lundh:
> http://www.pythonware.com/library/tkinter/introduction/index.htm
>
> Tkinter: GUI programming with Python
> http://www.nmt.edu/tcc/help/lang/python/tkinter.html
>
> Python Faqts
> http://www.faqts.com/knowledge_base/index.phtml/fid/264
>
>
> Regards
> Jorgen
>
>




From gerhard.haering at gmx.de  Sun Sep 29 13:17:41 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sun, 29 Sep 2002 19:17:41 +0200
Subject: Python's import: why doesn't it work?
In-Reply-To: <874rc8d9c9.fsf@nospam.eml.cc>
References: <9ac02e81.0209290839.4f4d6475@posting.google.com> <874rc8d9c9.fsf@nospam.eml.cc>
Message-ID: <20020929171740.GA1923@lilith.ghaering.test>

* Mahesh Padmanabhan <news at nospam.eml.cc> [2002-09-29 16:55 +0000]:
> You need to create an empty file called __init__.py inside the
> template directory and call the blah.py script as follows:
> 
> import template.blah

It's also highly recommended to start Python with "python -vv", then
trying to import the module in question. This will show you where Python
tries to find the module.

Btw. sys.modules contains the module search path.

-- Gerhard



From gerhard.haering at opus-gmbh.net  Mon Sep 16 08:44:12 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 16 Sep 2002 12:44:12 GMT
Subject: Problem with HTTPS request
References: <uo2fveepf88175@corp.supernews.com> <b0f083db.0209130558.2aa60398@posting.google.com> <uo45ara0l62cdc@corp.supernews.com> <slrnao4dkn.288.gerhard.haering@lilith.my-fqdn.de> <mYqg9.8349$WV.1529@fe03> <slrnao4h5c.288.gerhard.haering@lilith.my-fqdn.de> <GWjh9.2108$Cf.1661@fe03>
Message-ID: <slrnaobk7v.23s.gerhard.haering@haering.opus-gmbh.net>

Steve Holden wrote:
> Well, I don't think we disagree on the importance of presentation and
> compliance with standards.

Perfect.

> I simply think you have to allow some leeway to people who either
> a) may not know any better

That was the reason I was complaining, so people who read this (and not
just the poster) /will/ know better. Private email would have been more
appropriate, but OTOH only reach the poster.

> or b) may not have a choice about the mailer they use at work, for
> example.

These people should be a minority of Outlook Express posters here, as
reading/posting to comp.lang.python is hardly required in many jobs. Plus,
if you're forced to use inappropriate tools for elementary tasks, then
maybe you should try to find a job where developers are treated
appropriately.

> [...] but-feel-free-to-continue-ly y'rs  -s teve

No, the public and private replies showed that this wouldn't be welcome.

-- Gerhard


From tim at lesher.ws  Mon Sep  9 12:26:15 2002
From: tim at lesher.ws (Tim Lesher)
Date: 9 Sep 2002 09:26:15 -0700
Subject: Larry Wall's comment on python...
References: <Pine.LNX.4.44.0209061302050.1523-100000@ns0.tbye.com> <mailman.1031342469.3006.python-list@python.org>
Message-ID: <b6562547.0209090826.24120bba@posting.google.com>

pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard) wrote in message news:<mailman.1031342469.3006.python-list at python.org>...
> I take Larry Wall as a nice and clever guy.  Perl does not marry so well with
> evolution, and this is probably the doom of all languages which do not enough
> foresee evolution, and overly praise backward compatibility.  

Because the language has grown by accretion, many Perl advocates tend
to draw parallels between the evolution of a language over its
lifetime and the evolution of species.  I think that this leads to
some invalid assumptions.

Evolution of species works because of natural selection.  Many random
mutations are "tried", similar to the individual displacements in
Brownian motion.  Those that prove advantageous for a given set of
circumstances persist.  That's all there is to it--no worry about
whether this mutation is a good idea overall, or whether it makes the
organism compatible with previous generations, or whether this
mutation makes the DNA sequence easier to read or harder to fit on an
80-column screen. :-)

The moment you talk about designing for language evolution, or
maintaining backwards compatibility, or pretty much any kind of
intelligent process governing the change over time, you have to drop
the comparison with natural evolution, as enticing as the comparison
might be.

Keeping a bad analogy just muddles conversation.

-- 
Tim Lesher
tim at lesher.ws


From mmaddox at hcsmail.com  Fri Sep  6 18:06:30 2002
From: mmaddox at hcsmail.com (mmaddox at hcsmail.com)
Date: Fri, 6 Sep 2002 18:06:30 -0400 (EDT)
Subject: Larry Wall's comment on python...
Message-ID: <200209062206.g86M6UVm092251@mail.hcsmail.com>

>	- It causes me to factor code better.  Python encourages a style
that
>	  obviates (some of) the need for the kind of syntactic structures
found in older
>	  block-structured languages.

I would argue that the "older" block-structured languages were considered
a breakaway from the column-based languages - RPG2 and Fortran, for
instance. Although I grant the Python structure and the column-based
structures are not completely analogous, they BOTH burden the programmer
with arbitrary rules to govern the physical layout of the code. Although
this layout may be considered more elegant from a readability standpoint,
it's the enforcement of the layout that is irritating to the more
Libertarian of us. It's a mindset difference similar to that between
Pascal and C++.

Having read Wall's comments on Slashdot, I think he is really just mildly
complaining about ridiculously long page-widths in his editor, not really
slashing Python seriously. He's just giving some props to his own baby at
the expense of another. (He had to say something, right?) You must agree
that Python can lead you to horizontal scrolling in your editor - the bane
of user interface design.

Yes, I am new here. Hi, everyone.
--
Michael F. Maddox
Senior Software Engineer
Hayes
Tallahassee, FL, USA




From tdelaney at avaya.com  Tue Sep  3 19:27:10 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 4 Sep 2002 09:27:10 +1000 
Subject: Pyrex 0.4.4
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A563@natasha.auslabs.avaya.com>

> From: Greg Ewing 
> 
> Pyrex 0.4.4 is now available:
> 
>     http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/

The Plex package appears to be a shortcut/link, rather than the package ...

Tim Delaney



From james.kew at btinternet.com  Sun Sep  8 15:47:55 2002
From: james.kew at btinternet.com (James Kew)
Date: Sun, 8 Sep 2002 20:47:55 +0100
Subject: transform a list of lists in a lit of strtings???
References: <mailman.1031507832.24527.python-list@python.org>
Message-ID: <algah9$1p0kp3$1@ID-71831.news.dfncis.de>

<jubafre at brturbo.com> wrote in message
news:mailman.1031507832.24527.python-list at python.org...
> I have a list of lists like this:
>
> labels=[['D1'], ['D2'], ['D3']]
>
> and i want to transform ia a list of strings
>
> labels2==[ 'D1' ,  'D2' ,  'D3' ] <<<<<<<----------- how i can get it????
>

Troels' answers are good if what is required is a list of the first items in
all sub-lists. But it's possible the OP actually wants to produce a
concatenation of all sub-lists:

labels1 = [["D1"], ["D2"], ["D3", "D4"]] --> labels2 == ["D1", "D2", "D3",
"D4"]

Normally I'd write this as:

labels2 = []
for sublist in labels1:
    labels2 += sublist

but a quick shufti at the built-in functionals and the lambda form, both of
which I've been shy of up to now, suggests also the one-liner:

labels2 = reduce(lambda x, y: x+y, labels1)

Can this be written in list comprehension form? Can't quite see one at the
moment...

--
James Kew
james.kew at btinternet.com





From hellen at claggetts.net  Sun Sep 15 20:20:38 2002
From: hellen at claggetts.net (Jonathan Claggett)
Date: 15 Sep 2002 17:20:38 -0700
Subject: representing a literal single slash
References: <dd5454fb.0209141828.1ad24260@posting.google.com> <ehSg9.8464$yt3.4491656@newssrv26.news.prodigy.com> <MlSg9.8467$yt3.4492002@newssrv26.news.prodigy.com>
Message-ID: <dd5454fb.0209151620.26f01c91@posting.google.com>

"dsavitsk" <dsavitsk at e-coli.net> wrote in message 
> > > str = str.replace('\\', '\') # broken
> >
> > in the above you are replacing a single slash with, well, with nothing.
>  try
> > >>> str = str.replace('\\\\', '\\') # [not] broken
> >
> > -d
> 
> you might also want to rename that variable.  else the str() function will
> be a problem.

Thanks, that works. I knew it was an easy one :-)

BTW, my code now looks like this:

mystr = mystr.replace(r'\\', '\\') # replace two slashes with on


From mlerner at nospam.is.goodspam.umich.edu  Mon Sep  9 12:01:32 2002
From: mlerner at nospam.is.goodspam.umich.edu (Michael George Lerner)
Date: Mon, 09 Sep 2002 16:01:32 GMT
Subject: Can't exponentiate zero in NumPy
References: <al5gdq$18g$1@naig.caltech.edu> <alifc4$gsi$1@naig.caltech.edu>
Message-ID: <wB3f9.267$PD3.9404@news.itd.umich.edu>

Rick Muller <rpm at wag.caltech.edu> wrote:

> This crashes on Linux, but not on Mac OS X.

Here's another data point for you:

PythonWin 2.2.1 (#34, Apr 15 2002, 09:51:39) [MSC 32 bit (Intel)] on win32.
Portions Copyright 1994-2001 Mark Hammond 
>>> from Numeric import *
>>> a = zeros(3,Float)
>>> a**2
array([ 0.,  0.,  0.])
>>> a[1] = 1.e-310
>>> a
array([  0.00000000e+000,   1.00000000e-310,   0.00000000e+000])
>>> a**2
array([ 0.,  0.,  0.])
>>> import Numeric
>>> Numeric.__version__
'21.0'



From aleax at aleax.it  Mon Sep 16 04:20:52 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 16 Sep 2002 08:20:52 GMT
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <m3admizdzq.fsf@mira.informatik.hu-berlin.de> <37fh9.112911$pX1.4044383@news2.tin.it> <7xk7lml7px.fsf@ruckus.brouhaha.com>
Message-ID: <Evgh9.99311$ub2.2030136@news1.tin.it>

Paul Rubin wrote:
        ...
> hostile parties touch the hardware.  Nonetheless, the pay TV industry
> deploys millions of sealed computers (smart cards) every year to
> customers, every single one of which is considered an attacker.  While
> that industry had spectacular early security failures, these days it's
> a fairly solved problem and they limit losses to acceptable levels.

According to a lot of discussion that was bandied around in Europe
in recent months, what with Kirsch's headline-making bankruptcy and
the woes of most other European pay-TV suppliers, this may not be
the case in Europe.  Pay-TV does seem to be plagued by widespread
piracy, in Europe, that may go some way towards explaining why all
pay-TV suppliers are bleeding red ink (actually, all except one UK
supplier, I believe).  I have no way to check whether the discussion
reflects reality, of course.


> It doesn't sound like the OP was asking for this type of solution though.

If somebody asks for a magic wand which, when moved in appropriate
ways while whispering "abra-cadabra", makes minor cuts on one's skin
go away, I think it's quite helpful to point out that such magic
wands, while often dreamed about, appear not to exist, and furthermore
to point the poster towards the use of disinfectants and band-aids,
which seem to be more effective real-world ways to deal with such
minor cuts and scrapes on one's skin.  Even though the OP may not
have been working for the only apparently workable type of solution,
pointing that type of solution out may still be the best one can do.


Alex



From kp at kyborg.dk  Thu Sep 26 04:28:34 2002
From: kp at kyborg.dk (Kim Petersen)
Date: Thu, 26 Sep 2002 10:28:34 +0200
Subject: object changing itself to another object
References: <amu0lv$6ac$1@otis.netspace.net.au>
Message-ID: <3D92C532.7060505@kyborg.dk>

Glen Murphy wrote:
> I'm creating a program that takes user input, and passes it to an object -
> however, I want that object itself, and not the parent to change where that
> input goes. In the highly simplified example below, I would like the output
> to be "Obj1","Obj2","Obj1" - but not suprisingly, I just get Obj1, Obj1,
> Obj1
> 
> 
> ### code begin  ###
> 
> class Obj1:
>     def key1(self):
>         # do obj1 specific stuff
>         self = Obj2()
>         print "Obj1"
> 
> class Obj2:
>     def key1(self):
>         # do obj1 specific stuff
>         self = Obj1()
>         print "Obj2"
> 
> a = Obj1()
> 
> # simulate user keypresses
> a.key1()
> a.key1()
> a.key1()
> 
> ### code end ###
> 
> I know I could achieve the result I want by doing horribly complicated trees
> of ifs and such, but I'd like my code to be nice and expandable (in this
> example, to easily add new Objs). I've looked through the Python FAQ, Google
> Groups and Various O'Reilly books, but I don't really know the terminology
> for what I'm looking for, so I haven't found anything so far, so does anyone
> have any pointers?

Terrible hack (but works for me...)

class Obj1:
     def key1(self):
         # do obj1 specific stuff
         self.__class__=Obj2().__class__
         print "Obj1"

You might want to make a "metamorphosis" __init__ that converts/creates
the variables needed in another class....

Just as a note: I needed this in Qt programming - since i couldn't 
change the autogenerated source's - and needed to subclass some of the 
Qt classes ... so i metamorph the autogenerated classes into mine 
instead ... which works - but i sure wish that there was a better way of 
telling the Designer what to do (or have a "smart" importer like the one 
in glade (where it loads the XML directly - and you can tell what 
classes a specific name must be - overriding the XML stuff...)

> 
> Any help would be greatly appreciated,
> Cheers,
> Glen
> 
> --
> Glen Murphy, http://glenmurphy.com/
> 
> 




From tjreedy at udel.edu  Tue Sep 17 12:04:19 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 17 Sep 2002 16:04:19 GMT
Subject: More on Protecting Source Code
References: <mailman.1032232999.22661.python-list@python.org> <3d86b4a0$0$23173$afc38c87@news.optusnet.com.au> <icieou0lpp0cgqetbl560airkm5ohuqcn2@4ax.com>
Message-ID: <7oIh9.95980$z91.3653979@bin3.nnrp.aus1.giganews.com>

"Lothar Scholz" <llothar at web.de> wrote in message
news:icieou0lpp0cgqetbl560airkm5ohuqcn2 at 4ax.com...
> But it could be changed easily if the core python developer would
see
> this problem as relevant as commerical users. But at the moment they
> don't see the point. Thats the problem.

It seems to me that current and would-be commercial users/sellers who
want something written for their particular benefit should follow
their philosophy and *pay* someone to make the 'easy changes'.

Not-holding-my-breath, TJR





From brueckd at tbye.com  Mon Sep 23 10:55:23 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Mon, 23 Sep 2002 07:55:23 -0700 (PDT)
Subject: Python HTTPSConnection/HTTPResponse Problem
In-Reply-To: <9abedeb.0209221910.7fb75f6f@posting.google.com>
Message-ID: <Pine.LNX.4.44.0209230747530.15180-100000@ns0.tbye.com>

On 22 Sep 2002, Edward Muller wrote:

> I *think* I found the problem. Using m2Crypto and urllib I was able do
> process the transaction through Authorize.net's ADC Direct Response
> mechanism and it *looks* like their server replied first with a ..
> 
> HTTP/1.1 100 Continue
> 
> and then with a ...
> 
> HTTP/1.1 200 OK
> 
> I don't know if this is valid or not (and I sent the same info to
> Authorize.net).

Yes, the 100 Continue is valid. It exists so that a client can get an
early indication as to whether or not the server will accept the client's 
request message that includes a (possibly large) request body (e.g. a 
large POST request). 

As with so many other features of HTTP, the use of the 100 code is
ambiguous for backwards compatibility reasons and therefore unnecessarily
complex. ;-) To handle it, your receiving code should generally _not_ 
expect to receive a 100 Continue, and should simply ignore it if it 
comes (turn around and read the response headers again).

-Dave




From gerhard.haering at gmx.de  Sun Sep  1 15:19:41 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sun, 1 Sep 2002 21:19:41 +0200
Subject: Executable Python Programs
In-Reply-To: <1rtc9.108$LI2.7238@newsread2.prod.itd.earthlink.net>
References: <1rtc9.108$LI2.7238@newsread2.prod.itd.earthlink.net>
Message-ID: <20020901191940.GB3759@lilith.ghaering.test>

* CotaBas <cotabas at earthlink.net> [2002-09-01 18:58 +0000]:
> This is probly a stupid question ......

Questions that are answered in the FAQ are stupid questions, yes. The
FAQ is located here: http://www.python.org/cgi-bin/faqw.py

> how do you make your python programs executable (without python IDLE).
> 
> For example:
> 
> I write a program that I want to give to a friend to run on his
> computer.

Easy solution: make your friend instal Python.

> He doesn't have Python. What can I do to make it so he can run it
> without Python?

By read the FAQ. Item 4.28 answers your question.

My personal opinion is that py2exe/Installer/etc. are for
intermediate/advanced Python programmers. The reason is that it's
sometimes difficult to fix the dependencies manually, if these tools are
unable to hunt them down automatically for you.

Installing Python on the target computer is almost always the simplest
solution.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))



From eric.brunel at pragmadev.com  Tue Sep  3 10:38:03 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Tue, 3 Sep 2002 16:38:03 +0200
Subject: How to avoid freeing memory on exit?
References: <roy-54DDA9.09043103092002@reader1.panix.com> <slrnan9dhv.1vs.gerhard.haering@haering.opus-gmbh.net> <3d74b6dc$0$231$4d4ebb8e@news.nl.uu.net>
Message-ID: <al2h23$a0t$1@wanadoo.fr>

Gumuz wrote:

> so what does 'one' underscore prefixing a methodname mean?

For Python, nothing. For would-be users of the method: "do not use unless 
you know what you're doing" ;-)
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From sismex01 at hebmex.com  Fri Sep  6 19:12:09 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Fri, 6 Sep 2002 18:12:09 -0500 
Subject: Larry Wall's comment on python...
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F0623302343CCB@mail.hebmex.com>

> > Of course, most of my current code has one big block, 
> > main(), and I just cut-and-paste the rest.


This just makes me shudder...

sheer horror....

:-D

-gus
--

Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From loewis at informatik.hu-berlin.de  Sun Sep 15 13:55:03 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 15 Sep 2002 19:55:03 +0200
Subject: Is Python 2.1 for Solairs 8 broke on SunFreeware?
References: <8d148763.0209131005.65de04a8@posting.google.com>
Message-ID: <j4k7lnrvd4.fsf@informatik.hu-berlin.de>

yaipa at aol.com (yaipa) writes:

> % /usr/local/bin/python CronTar.py -c cronTar.cfg
>   Traceback (most recent call last):
>     File "/export/fruid/bin/CronTar.py", line 26, in ?
>       import time
>   ImportError: ld.so.1: /usr/local/bin/python: fatal: 
>     libgcc_s.so.1: open failed: No such file or directory
> %

That is somewhat unfortunate: gcc  3.2 links another shared library to
every binary.

> Am I required to have gcc installed also?
[...]
> Any ideas?

No need to do that. Just install the libgcc package from Sunfreeware
also. This is much less than the full GCC.

Regards,
Martin


From Thomas.Powell at Yum.com  Tue Sep  3 15:55:34 2002
From: Thomas.Powell at Yum.com (Powell, Thomas)
Date: Tue, 3 Sep 2002 15:55:34 -0400
Subject: Servizi Internet
Message-ID: <AC1728D9769ED845B1E291D05A7D2B0B546DD8@TECHXCH01V01.us.tgr.net>

Wasn't this spam from an _ISP_ anyway?  An abuse message to the ISP might get them to back off their spamming practices, but otherwise, what is the only "legal" recourse?  Complain to their backbone provider?


Thomas Powell
------------------------------------
Field Systems Development, Pizza Hut SUS
Yum! Brands, Inc.


-----Original Message-----
From: Fran?ois Pinard [mailto:pinard at iro.umontreal.ca] 
Sent: Tuesday, September 03, 2002 3:41 PM
To: Gerhard H?ring
Cc: python-list at python.org
Subject: Re: Servizi Internet


[Gerhard H?ring]

> The only means you have is complain to their ISP 
> (abuse@{name_of_ISP}).

This is rather discouraging, you know.

Most ISP around here, and especially the big ones, could not care any less. Thinking about it, some international providers are not fully pink either. They are far more interested into protecting the anonymity of those from which they receive money.  Aren't they in business, after all?

They will almost never take action unless forced to do so by police intelligence services, and such services are so horribly heavy or slow to get in motion that it is almost a lost cause as far as practice is concerned.

Police people affected to computer crimes told me that they are severely understaffed, and fully overwhelmed by requests emanating from big banks or military agencies, which are to be served in priority.  Moreover, in far too many cases, they have youth protection law in their way.

We pay all the way.  Spammers use our network connectivity, our CPUs, our disks and our own time for bothering us; they misredirect rejects so companies pay postmasters or sysadmin services to handle their crap; and everybody invest effort and energies trying to escape the constant annoyance.  While politicians, from their ivory towers, discourse on the virtues of Free Speech!

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard

-- 
http://mail.python.org/mailman/listinfo/python-list


This communication is confidential and may be legally privileged.  If you are not the intended recipient, (i) please do not read or disclose to others, (ii) please notify the sender by reply mail, and (iii) please delete this communication from your system.  Failure to follow this process may be unlawful.  Thank you for your cooperation.



From pyth at devel.trillke.net  Wed Sep 11 19:21:51 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 12 Sep 2002 01:21:51 +0200
Subject: I must be an idiot, please pity me!
In-Reply-To: <oYPf9.2894$XE6.174556@newsfep1-win.server.ntli.net>; from sam@localhost.localdomain.trillke.net on Thu, Sep 12, 2002 at 12:05:22AM +0100
References: <VAOf9.2572$Pa7.155732@newsfep1-gui.server.ntli.net> <3D801757.1020901@nospam.free.fr> <oYPf9.2894$XE6.174556@newsfep1-win.server.ntli.net>
Message-ID: <20020912012151.V3893@prim.han.de>

newbie wrote:
> On Thu, 12 Sep 2002 01:25:59 +0100, laotseu wrote:
> 
> 
> > Creating a db connection in each function for sure not be very
> > efficient.But what's wrong with passing the database connection to the
> > functions in mod1, mod2, and modXXX ???
> 
> That connection has to be passed to the other functions (where needed of
> course) in mod1. Say, mod1.functiona may use functions f1, f2 and f3. My
> gut feeling of having to pass that connection to all of these said "this
> can't be right". This is what I've done but it sure doesn't feel right!
> 
> 
> > Anyway, you can make it simpler, like this :
> 
> Yay, now for the good stuff!
> 
>  
> > in each module,
> > - create a global db_connection variable - add an init() function, that
> > takes a db_connection as argument and assign it to the module's global
> > db_connection variable - and don't forget to call this function before
> > any other.
> 
> Yes, I can follow that. Once I saw that "global" was more "local" I've
> never really considered using it.

one common idiom is to do

    import mydb
    ... use mydb.connection ...

and somewhere at initialization time

    import sys
    class mydb:
        connection = ...  # initialize it
    sys.modules['mydb']=mydb

'import' imports always the same object. Usually it's a 
module-object but can you put in your class/type.

of course you can also simply define a module 'mydb.py' in a 
file and import that everywhere.  Besides not having another file 
the above solution makes it quite impossible to use the module/class 
before it is initialized.

regards,

    holger



From mstenner at phy.duke.edu  Fri Sep 27 17:39:11 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Fri, 27 Sep 2002 17:39:11 -0400
Subject: watching mutables?
In-Reply-To: <an2hnt$nh2$1@news.hccnet.nl>; from anton@vredegoor.doge.nl on Fri, Sep 27, 2002 at 11:12:05PM +0200
References: <mailman.1033137396.14482.python-list@python.org> <an211j$oi$1@news.hccnet.nl> <Xns9296C58C9ADFCcliechtigmxnet@62.2.16.82> <k7l7qpvf.fsf@ion-tof.com> <Xns9296D3AB53777cliechtigmxnet@62.2.16.82> <mailman.1033154856.20758.python-list@python.org> <an2doh$ajo3p$1@ID-59885.news.dfncis.de> <an2hnt$nh2$1@news.hccnet.nl>
Message-ID: <20020927173911.A17971@phy.duke.edu>

On Fri, Sep 27, 2002 at 11:12:05PM +0200, Anton Vredegoor wrote:
> On Fri, 27 Sep 2002 22:06:41 +0200, "Thomas Heller"
> >Here's what seems to be available:
> >http://oomadness.tuxfamily.org/downloads/
> >and this is obviously where the EventObject is contained in:
> >http://oomadness.tuxfamily.org/downloads/EditObj-0.3.tar.gz
> 
> Very interesting module. I might add that windows doesn't seem to like
> the .cfg file so this should be moved to some other place before
> installing with 'python setup.py install'.
> 
> But...
> 
> I specifically asked for something that doesn't touch the original
> object. In eventobj.py's docstring there is the following warning:
> 
> <warning>
> Caution : As event management is performed by changing the class of
> the instance, you use eventobj with critical objects at your own
> risk... !
> </warning>
> 
> So, with a lot of thanks for the results sofar ...
> 
> Is it possible?

Just to be clear, it doesn't so much affect the original object, but
rather REPLACES the class before the object ever gets created.  If
this isn't satisfactory, then I think you're going to need to modify
the python interpreter.
					-Michael
-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From wlfraed at ix.netcom.com  Fri Sep 27 17:45:05 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Fri, 27 Sep 2002 14:45:05 -0700
Subject: Better terminal for windows?
References: <3D94B037.2020202@motorola.com>
Message-ID: <1hj2na.u44.ln@ix.netcom.com>

Stephen Boulet fed this fish to the penguins on Friday 27 September 
2002 12:23 pm:

> I'm using python 2.2.1 on winnt.
> 
> If I want to run python interactively (and perhaps using ipython), is
> there something better than the "cmd" terminal?

        PythonWin (ActiveState as I recall)
        Idle

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From lutz at lutz-horn.de  Mon Sep 16 07:25:30 2002
From: lutz at lutz-horn.de (Lutz Horn)
Date: 16 Sep 2002 13:25:30 +0200
Subject: read a line from socket (was: (no subject))
References: <mailman.1032173958.26614.python-list@python.org>
Message-ID: <wk3csakwgl.fsf@ID-32301.user.dfncis.de>

"Srijak Rijal" <wham10 at mad.scientist.com> writes:
> I just migrated from perl to python and was wondering if python
> socket can be read line by line too?  How can I get only one line of
> input from the server without losing info?

Split the string returned by recv() on the newline(s), save the
returned strings and use them at will.

Lutz



From gmcm at hypernet.com  Thu Sep 19 18:38:09 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 19 Sep 2002 18:38:09 -0400
Subject: ANN: Installer 5b4
Message-ID: <3D8A1991.8111.BE3A321@localhost>

Announcing version 5b4 of Installer:

  http://www.mcmillan-inc.com/installer_dnld.html

Changes in 5b4:
  Bug fixes:
  - recompile if .py file in a package is newer
  - fix segfault on Linux / Unix if executable is on
    the $PATH
  - fix segfault / access violation on very long
    script names
  - many other minor bug fixes

 Enhancements:
  - works with .pyw files (or any file extension,
    including none)
  - add --strip option to run the executable and
    shared libs through strip.

Platforms:
  Windows
  Linux
  Unix (relying on volunteers)

Python versions:
  tested with Python 1.5 to 2.2

What is it:
  The Installer package is any easy way to deploy
   Python apps to systems without Python installed 
   (or with incompatible Pythons installed).

   Installer support single-file and single-directory
   deployments. On Windows, Installer supports
   COM, including in-process COM servers.
 
License:
  MIT style.

Contact:
  gmcm at hypernet.com

<P><A HREF="http://www.mcmillan-inc.com/installer_dnld.html">Installer 
Release 5b4</A>Easy Python app deployment. (19-Sep-
2002).




From gumuz at looze.net  Tue Sep 24 06:55:18 2002
From: gumuz at looze.net (Gumuz)
Date: Tue, 24 Sep 2002 12:55:18 +0200
Subject: Difference between MOD_PYTHON and the CGI module
Message-ID: <3d904321$0$234$4d4ebb8e@news.nl.uu.net>

What are the differences, if any?

Am i mixing up to things here?

Suppose I'd like to use python as a 'replacement' for PHP, what do I use?

thanks,

gumuz




From duncan at NOSPAMrcp.co.uk  Tue Sep 17 04:31:23 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 17 Sep 2002 08:31:23 +0000 (UTC)
Subject: CFP: UK Python Conference 2003
References: <Xns928BAD1086D00duncanrcpcouk@127.0.0.1> <slrnaocste.qcp.Gareth.McCaughan@g.local>
Message-ID: <Xns928C5EAC261F0duncanrcpcouk@127.0.0.1>

Gareth.McCaughan at pobox.com (Gareth McCaughan) wrote in 
news:slrnaocste.qcp.Gareth.McCaughan at g.local:

> Duncan Booth wrote:
>>                     *** Call for Participation ***
>>                       UK Python Conference 2003
>> 
>>                             April 2-3 2003
>>                          Holiday Inn, Oxford
> ...
>> To submit a proposal, send an email message to duncan at rcp.co.uk by 2nd
>> October 2003. In your message, please include the following:
>           ^^^^
> 
> I'd like to propose a paper on "Novel applications of
> Guido's time machine". :-)

And there I was feeling miffed about not being able to spell the subject 
line! Well, we are hoping that Guido will be there, so I'll just ask him to 
nip forward and collect the abstracts in time for the real deadline of 2nd 
October 2002.

Repost with the deadline corrected:

                    *** Call for Participation ***
                      UK Python Conference 2003

                            April 2-3 2003
                         Holiday Inn, Oxford

The UK Python Conference 2003 is being held in conjunction with the
larger ACCU conference. The conference will include two days of papers
arranged in 90 minute speaking slots with either one or two tracks.

The audience is highly technical and will include experts in many
fields of computing, some of whom may know little or nothing about
Python. Talks will be welcomed on any aspect of Python, but we are
especially interested in the following:

        - Commercial applications for Python
        - Cross-language talks

Speakers are welcome from the UK or the rest of the world, however our
budget for overseas speakers may be limited.

To submit a proposal, send an email message to duncan at rcp.co.uk by 2nd
October 2002. In your message, please include the following:

        - Speaker name and affiliation
        - Contact address
        - Presentation title
        - A one paragraph abstract



-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From cbrown at metservice.com  Thu Sep 19 01:50:31 2002
From: cbrown at metservice.com (Colin Brown)
Date: Thu, 19 Sep 2002 17:50:31 +1200
Subject: handling exceptions
References: <fa576652.0209170827.4ba29201@posting.google.com>
Message-ID: <3d89659d$1@news.nz.asiaonline.net>

> being new to python, I have a question about handling exceptions. I
> want to catch an exception which is raised from a module not written
> by me, and display a messagebox with the message of the exception
> rather than halting the program. The problem is: how can I access the
> message string of the exception?

Hi Karsten

I came across the same problem some time ago when I wanted to write daemon
errors into day-based logfiles and wrote myself the module Excepts.py below:

--[Excepts.py]----------------------------------
# Return Exception errors as strings
import sys, traceback

def error():
    tb =
traceback.format_exception(sys.exc_info()[0],sys.exc_info()[1],sys.exc_info(
)[2])
    return tb[len(tb)-1].replace('\n','')

def errorstack():
    return
''.join(traceback.format_exception(sys.exc_info()[0],sys.exc_info()[1],sys.e
xc_info()[2]))
-------------------------------------------------

example usage:
import Excepts
try:
    functional_code()
except:
    oneliner = Excepts.error()
    log(oneliner)

errorstack returns the full stack if you want that.

Hope this helps.

Colin Brown
PyNZ





From eugene1977 at hotmail.com  Sun Sep 15 07:41:59 2002
From: eugene1977 at hotmail.com (eugene kim)
Date: Sun, 15 Sep 2002 11:41:59 +0000
Subject: two or more whitespace regular expression?
Message-ID: <am12a3$qo0$1@newsreader.mailgate.org>

i used this..

words = string.split(line, '['+string.whitespace 
+']'+'['string.whitespace+']+')

which didn't work.. 
thanks


From mennosimons at gmx.net  Thu Sep 26 04:52:08 2002
From: mennosimons at gmx.net (Ulli Stein)
Date: Thu, 26 Sep 2002 10:52:08 +0200
Subject: unicode woes
Message-ID: <amuhrk$4cn$05$1@news.t-online.com>

Hi,

in the firm I am working we have a big Python project. Four developers are 
working on it for over a half year now, and it has grown really big. 
Nevertheless, due to the Python syntax, and its high level, we still felt 
very strong in the decision to realize our project in Python.

Some weeks ago the management decided to deploy unicode so that we can 
handle every charset uniformly. The problem which showed at first: we 
realized that we cannot change the encoding in site.py afterwards, i.e. we 
have to specify the encoding all the way. It got worse and worse because 
the unicode encoding spread like a virus through all the source code. 
Simple Exceptions which we wanted to print to the console throw another 
Exception:
UnicodeError: ASCII decoding error: ordinal not in range(128)

Furthermore, in Python you have no way to detect, which encoding a special 
unicode string atually is. We ended up having our own u() function. The 
problem with that: we have to pass _every_ string through it. Ouch! Now our 
source code is as ugly as our program is slow. Isn't there another way? 
When will the types string and unicode merged together like in Java? If 
this is not changing and we do not find another way, we will have to port 
_all_ source code to C++.

Please help.

def u(obj):
        encoding = getEncoding()
        if encoding == None:
                raise EncodingNotInitialized("There was no encoding defined previously.")
        
        if type(obj) == types.StringType:
                return unicode(obj, encoding)
        elif type(obj) == types.UnicodeType:
                # As we can't be sure what encoding the unicode object has we receive,
                # we will simply re-encode it.
                return unicode(obj.encode(encoding), encoding)
        elif type(obj) == types.NoneType:
                return None
        elif type(obj) == types.IntType or type(obj) == types.LongType:
                return unicode("%d" % obj, encoding)
        elif type(obj) == types.FloatType:
                return unicode("%f" % obj, encoding)
        elif type(obj) in [ types.TupleType, types.ListType, types.Dic
tType, types.DictionaryType ]:
                return unicode(str(obj), encoding)
        else:
                raise EncodingError("Unconvertable object passed to the u-function.")

>>> print "????"
????
>>> a = "????"
>>> str(a)
'\xe4\xf6\xfc\xdf'
>>> print a
????
>>> "????%s"%unicode("???", "iso8859-15")
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
UnicodeError: ASCII decoding error: ordinal not in range(128)



From duncan at NOSPAMrcp.co.uk  Thu Sep 12 04:53:11 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Thu, 12 Sep 2002 08:53:11 +0000 (UTC)
Subject: dynamism
References: <66715c02.0209091923.265fe77e@posting.google.com> <Xns92855D48384F9duncanrcpcouk@127.0.0.1> <3d7e209a_1@news.iprimus.com.au> <Xns92865CE1C916Aduncanrcpcouk@127.0.0.1> <Xns9286606D44C35duncanrcpcouk@127.0.0.1> <66715c02.0209111633.771aa9d3@posting.google.com>
Message-ID: <Xns928763C35DA37duncanrcpcouk@127.0.0.1>

steven_shaw at adc.com (Steven Shaw) wrote in
news:66715c02.0209111633.771aa9d3 at posting.google.com: 

> Thanks alot for your time on this.
It was fun!

> I hadn't read about the __slots__
> mechanism before. I feel that there must be something wrong, though,
> because the Old object layout is winning! Maybe the implementation is
> finding the index of the variable in the __slots__ array and then>
> using returning __values__[index] in order to implement the attribute 
> access. I made up the "__values__" array - there must be a values
> array somewhere. Array-based object layout amounts to the same thing
> as record-based-layout in Python because all primitives are the same
> size (right?). There needs to be a way for Python to remember the
> indexes... Hmmm. I think your test code doesn't allow the compiler to
> remember the indexes since it creates an instance anonymously (using
> klass).

It is quite hard working out from the code what happens, but basically when 
you define __slots__ it allocates space in the object to hold the actual 
data for each slot, and makes a copy of the __slots__ list. It also creates 
an object of type member_descriptor for each slot and stores it in the 
class dictionary under the name of the slot.

class C(object):
   __slots__ = ['a', 'b']
inst = C()

If you look at C.__dict__['a'] and C.__dict__['b'] they are 
member_descriptor objects. Accessing inst.a therefore will call 
C.a.__get__(inst, C) or C.a.__set__(inst, value). Each member_descriptor 
object knows the offset where its data is stored.

So you still have the overhead of a dictionary lookup for every __slot__ 
access (although Python dictionary lookup is pretty fast), and you have 
saved the overhead of a dictionary on every instance at the cost of an 
extra object in the class for each attribute. It is slower than ordinary 
attribute access because after the dictionary lookup instead of returning 
the value that was found you have to call one of its methods (although 
fortunately that doesn't involve another dictionary lookup, just a C call).

I hope this makes sense.

-- 
Duncan Booth                                            
duncan at rcp.co.uk int month(char
*p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3" 
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure? 


From eric.brunel at pragmadev.com  Mon Sep  9 04:35:32 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 9 Sep 2002 10:35:32 +0200
Subject: Accessing Tk-inter Popup menu entry
References: <b714b8de.0209080901.6c8a7985@posting.google.com>
Message-ID: <alhm24$j17$1@wanadoo.fr>

Aki Niimura wrote:

> Hi,
> 
> I have created a popup menu. (see the attached example code)
> Now I would like to change the state of a popup menu entry dynamically.
> I tried to use
>     popup.entryconfig(2, state=DISABLED)
> But it didn't work although it worked fine with a regular menu bar.

What was the problem? I tried it and it worked, but not with index 2, 
because the menu item numbered 2 is the 3rd one, which is the separator. 
But doing:
popup.entryconfigure(1, state=DISABLED)
works without problem.
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From sholden at holdenweb.com  Fri Sep 13 15:12:29 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 13 Sep 2002 15:12:29 -0400
Subject: transform list of int in a list of string??
References: <-1483717605.1031940885180.JavaMail.nobody@webmail2.brturbo.com> <mailman.1031941396.19665.python-list@python.org>
Message-ID: <MNqg9.8303$WV.6617@fe03>

"Skip Montanaro" <skip at pobox.com> wrote ...
>
>     juliano> How can i transform a list of integer numbers in a list of
>     juliano> string numbers, like that:
>
>     juliano> x=[1,2,3,4] to y=['1', '2'...]????????
>
> Try
>
>     y = map(str, x)
>
> or
>
>     y = [int(n) for n in x]

Skip of course meant

    y = [str(n) for n in x]

in his second example.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From chris.gonnerman at newcenturycomputers.net  Sun Sep  8 10:30:48 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sun, 8 Sep 2002 09:30:48 -0500
Subject: python program to watch a file - example.
References: <mailman.1031486051.16547.python-list@python.org> <3D7B5650.4050304@zute.net>
Message-ID: <004701c25744$5a91ea60$0101010a@local>

----- Original Message ----- 
From: "Tim Hoffman" <timh at zute.net>

> If it is always the same file and seeing you are looking at *nix 
> varient, one of the unix like ways would be to tail -f the file, and 
> hook it up to your process which just does blocking reads on stdin.
> 
> ie tail -f somfile | python my_watch.py

tail -f is roughly equivalent to:

    while 1:

        line = fp.readline()

        while line:
            print line
            line = fp.readline()

        time.sleep(1)

You would use a break to get out of this; or you might want
some other sort of exit condition.


Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From padraig at linux.ie  Sun Sep 15 17:41:51 2002
From: padraig at linux.ie (Padraig Brady)
Date: Sun, 15 Sep 2002 21:41:51 GMT
Subject: representing a literal single slash
References: <dd5454fb.0209141828.1ad24260@posting.google.com> <3D84E9F9.8090101@linux.ie>
Message-ID: <3D84FEE6.90600@linux.ie>

Padraig Brady wrote:
> Jonathan Claggett wrote:
> 
>> Here is an easy question (I'm hoping).
>>
>> I'm trying to process a string so that all doubled up backslashes (\\)
>> are replaced with a single backslash (\).
>>
>> The obvious approach doesn't work since a single backslash can't end a
>> string:
>>
>> str = str.replace('\\', '\') # broken
>>
>> So, how does one represent a literal single slash? unicode or hex
>> values perhaps?
> 
> mystr = mystr.replace(r'\\', r'\')

Emm this actually doesn't work. Why can't you:
s=r'\'
Is it a bug?

Anyway this should work:

mystr=mystr.replace(r'\\', '\\')

P?draig



From news at nospam.eml.cc  Sun Sep 29 12:55:47 2002
From: news at nospam.eml.cc (Mahesh Padmanabhan)
Date: Sun, 29 Sep 2002 16:55:47 GMT
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com>
Message-ID: <874rc8d9c9.fsf@nospam.eml.cc>

You need to create an empty file called __init__.py inside the
template directory and call the blah.py script as follows:

import template.blah


jhorneman at pobox.com (Jurie Horneman) writes:

[snip]

-- 
Mahesh Padmanabhan

-------

Replace nospam with prana to email me.


From max at alcyone.com  Wed Sep 18 18:21:34 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 18 Sep 2002 15:21:34 -0700
Subject: Catching results of an exec
References: <3d88f701$0$304$39cecf19@nnrp1.twtelecom.net>
Message-ID: <3D88FC6E.59D0DD01@alcyone.com>

Greg Fortune wrote:

> Is there any way to "catch" the value produced by an exec?  I'm
> considering
> things like print statements that don't really assign a value, but do
> send
> a result to stdout.

The problem is "the value produced by an exec" is ambiguous, because
execs (as don't return values).  If you want to capture some subpart of
the execution of a statement, you'll have to handle it case by case.

> Right now I've got a nasty hack that redirects stdout, uses an
> InteractiveInterpereter from the code module to execute the code,
> writes to
> an output file, reads the result from the file, restores stdout and
> then
> sends the result to the client.  Hopefully, that's a terrible way to
> do it.

That's probably the best way to do it, actually -- or, at least, with an
overridden sys.stdout in the globals dictionary that you're passing to
eval/exec (which doesn't necessarily have to be the same dictionary as
your main script is running).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ I am the king of sorrow
\__/ Sade
    EmPy / http://www.alcyone.com/pyos/empy/
 A system for embedding arbitrary Python in template text as markup.


From fetus at AUTO  Tue Sep 24 20:04:42 2002
From: fetus at AUTO (A. Dimsdale Fetus A. Moser)
Date: Wed, 25 Sep 2002 00:04:42 GMT
Subject: Pythonwin (win32all build 148) on Python 2.2.1 on Win2K, bogus syntax e
Message-ID: <u47k9.46401$V7.11091779@twister.socal.rr.com>

I am running Pythonwin (win32all build 148) on Windows 2000 Professional with
Python 2.2.1. When I attempt to run a Python script in it, it halts the running
of the script on supposed syntax errors, which are completely bogus.
Examples:
First line of our (Aaron L. Dimsdale and Alex S. Moser) first semi-significant
Python script, called fight:
"""fight.py 5.0 -- crappy primitive fighting game by alex s. moser & aaron l.
dimsdale"""
        ^^^ (note: no linebreak in that line)
"""if one of the players does not say "atk" or "rest" when asked, they will \
forfeit their turn"""

Pythonwin barfed on the quote marks marked by carets, returning: "Failed to run script -
syntax error - invalid syntax".
I removed those quote marks and the quote marks at the beginning of the second
documentation line, and left those two lines as one string. Then, several lines
later:

p1char = str(raw_input("Player 1, choose your character class: "))
                 ^
For some strange reason (Alex and I determined that it was either cosmic rays
or Windows 2000 rot), Pythonwin returned the same error, pointing at that
particular "i" character. I removed "str" and its set of parens, and it still
slapped me with the same error on the same character/command.

p1weap = str(raw_input("Player 1! Pick your weapon! steak knife, cleaver, tenderizer "))
                          ^
Pythonwin also pointed out a supposed syntax error on that "a" character in the
word "Player".

This is so randomly weird it hurts. We couldn't figure out what the hell is
going on. By the way, these errors only occured on Aaron's computer at work,
in Pythonwin; the script worked fine on Alex's computer, in the same office;
on Alex's home computer; on Aaron's home computer; and on Aaron's work computer
(the same that gave the syntax errors) under the interpreter packaged with
Python itself (not the win32all extensions).

--Aaron Dimsdale and Alex Moser


From xscottgjunk at yahoo.com  Tue Sep 17 04:24:27 2002
From: xscottgjunk at yahoo.com (Scott Gilbert)
Date: 17 Sep 2002 01:24:27 -0700
Subject: Python "compiler" is  too slow for processing large data files???
References: <3d6d097d$0$321$e2e8da3@nntp.cts.com>
Message-ID: <79b2b0bb.0209170024.12ccce1b@posting.google.com>

"Ron Horn" <rchii at lycos.com> wrote 
> 
> I'm writing a fairly simple app that loads (text) data files, lets you edit
> the data, and then saves them back out.  Rather than parsing the data myself
> when loading a new file, and building data structures for the program's data
> that way, I decided to try to 'import' (or 'exec', actually) the data into
> the app.  The idea was to be able to format my data files as python code
> (see below), and then let the python compiler do the parsing.  In the
> future, I could actually put 'def' and 'class' statements right into the
> data file to capture some behavior along with the data.
> 
> Simple example - I can import or exec this file to load my data (my real app
> has int, float, and string data):
> ------ try5a3.py --------
> list1 = [
>     (323, 870, 46, ),
>     (810, 336, 271, ),
>     (572, 55, 596, ),
>     (337, 256, 629, ),
>     (31, 702, 16, ),
> ]
> print len(list1)
> ---------------------------
> 


try using this instead:

data1 = """323 870 46
810 336 271
572 55 596
337 256 629
31 702 16"""

list1 = [tuple([int(x) for x in y.split()]) for y in
data1.split('\n')]
del data1

It's not as pretty, but I was able to load more than 100,000 lines of
3 elements each in a few seconds with Python 2.2 on my aging laptop.


Cheers,
    -Scott


From tim at multitalents.net  Mon Sep 16 13:09:40 2002
From: tim at multitalents.net (Tim Rice)
Date: Mon, 16 Sep 2002 10:09:40 -0700 (PDT)
Subject: test_parser fails
Message-ID: <Pine.UW2.4.44.0209161003590.2677-100000@ou8.int.multitalents.net>

I'm trying to get Python-2.2.1 working on SCO Open Server 5
I've got most of it working but I'm concerned about the parser moudle
failing the test.

Any tips on where to look?

Thanks.

Here is the output of ./python -E -tt ./Lib/test/test_parser.py -v

test_basic_import_statement (__main__.RoundtripLegalSyntaxTestCase) ... FAIL
test_expressions (__main__.RoundtripLegalSyntaxTestCase) ... FAIL
test_function_defs (__main__.RoundtripLegalSyntaxTestCase) ... FAIL
test_import_from_statement (__main__.RoundtripLegalSyntaxTestCase) ... FAIL
test_print (__main__.RoundtripLegalSyntaxTestCase) ... FAIL
test_simple_assignments (__main__.RoundtripLegalSyntaxTestCase) ... FAIL
test_simple_augmented_assignments (__main__.RoundtripLegalSyntaxTestCase) ... FAIL
test_simple_expression (__main__.RoundtripLegalSyntaxTestCase) ... FAIL
test_yield_statement (__main__.RoundtripLegalSyntaxTestCase) ... FAIL
Illegal input: a,,c ... ok
Illegal input: a $= b ... ok
Illegal yield statement: def f(): return 1; yield 1 ... ok
Illegal return in generator: def f(): return 1; yield 1 ... ok
test_junk (__main__.IllegalSyntaxTestCase) ... ok
Illegal input: print >>fp, ... ok

======================================================================
FAIL: test_basic_import_statement (__main__.RoundtripLegalSyntaxTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "./Lib/test/test_parser.py", line 129, in test_basic_import_statement
    self.check_suite("import sys")
  File "./Lib/test/test_parser.py", line 27, in check_suite
    self.roundtrip(parser.suite, s)
  File "./Lib/test/test_parser.py", line 18, in roundtrip
    self.fail("could not roundtrip %r" % s)
  File "/usr/local/src/utils/Python-2.2.1/Lib/unittest.py", line 254, in fail
    raise self.failureException, msg
AssertionError: could not roundtrip 'import sys'

======================================================================
FAIL: test_expressions (__main__.RoundtripLegalSyntaxTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "./Lib/test/test_parser.py", line 42, in test_expressions
    self.check_expr("foo(1)")
  File "./Lib/test/test_parser.py", line 24, in check_expr
    self.roundtrip(parser.expr, s)
  File "./Lib/test/test_parser.py", line 18, in roundtrip
    self.fail("could not roundtrip %r" % s)
  File "/usr/local/src/utils/Python-2.2.1/Lib/unittest.py", line 254, in fail
    raise self.failureException, msg
AssertionError: could not roundtrip 'foo(1)'

======================================================================
FAIL: test_function_defs (__main__.RoundtripLegalSyntaxTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "./Lib/test/test_parser.py", line 100, in test_function_defs
    self.check_suite("def f(): pass")
  File "./Lib/test/test_parser.py", line 27, in check_suite
    self.roundtrip(parser.suite, s)
  File "./Lib/test/test_parser.py", line 18, in roundtrip
    self.fail("could not roundtrip %r" % s)
  File "/usr/local/src/utils/Python-2.2.1/Lib/unittest.py", line 254, in fail
    raise self.failureException, msg
AssertionError: could not roundtrip 'def f(): pass'

======================================================================
FAIL: test_import_from_statement (__main__.RoundtripLegalSyntaxTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "./Lib/test/test_parser.py", line 119, in test_import_from_statement
    self.check_suite("from sys.path import *")
  File "./Lib/test/test_parser.py", line 27, in check_suite
    self.roundtrip(parser.suite, s)
  File "./Lib/test/test_parser.py", line 18, in roundtrip
    self.fail("could not roundtrip %r" % s)
  File "/usr/local/src/utils/Python-2.2.1/Lib/unittest.py", line 254, in fail
    raise self.failureException, msg
AssertionError: could not roundtrip 'from sys.path import *'

======================================================================
FAIL: test_print (__main__.RoundtripLegalSyntaxTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "./Lib/test/test_parser.py", line 71, in test_print
    self.check_suite("print")
  File "./Lib/test/test_parser.py", line 27, in check_suite
    self.roundtrip(parser.suite, s)
  File "./Lib/test/test_parser.py", line 18, in roundtrip
    self.fail("could not roundtrip %r" % s)
  File "/usr/local/src/utils/Python-2.2.1/Lib/unittest.py", line 254, in fail
    raise self.failureException, msg
AssertionError: could not roundtrip 'print'

======================================================================
FAIL: test_simple_assignments (__main__.RoundtripLegalSyntaxTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "./Lib/test/test_parser.py", line 83, in test_simple_assignments
    self.check_suite("a = b")
  File "./Lib/test/test_parser.py", line 27, in check_suite
    self.roundtrip(parser.suite, s)
  File "./Lib/test/test_parser.py", line 18, in roundtrip
    self.fail("could not roundtrip %r" % s)
  File "/usr/local/src/utils/Python-2.2.1/Lib/unittest.py", line 254, in fail
    raise self.failureException, msg
AssertionError: could not roundtrip 'a = b'

======================================================================
FAIL: test_simple_augmented_assignments (__main__.RoundtripLegalSyntaxTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "./Lib/test/test_parser.py", line 87, in test_simple_augmented_assignments
    self.check_suite("a += b")
  File "./Lib/test/test_parser.py", line 27, in check_suite
    self.roundtrip(parser.suite, s)
  File "./Lib/test/test_parser.py", line 18, in roundtrip
    self.fail("could not roundtrip %r" % s)
  File "/usr/local/src/utils/Python-2.2.1/Lib/unittest.py", line 254, in fail
    raise self.failureException, msg
AssertionError: could not roundtrip 'a += b'

======================================================================
FAIL: test_simple_expression (__main__.RoundtripLegalSyntaxTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "./Lib/test/test_parser.py", line 80, in test_simple_expression
    self.check_suite("a")
  File "./Lib/test/test_parser.py", line 27, in check_suite
    self.roundtrip(parser.suite, s)
  File "./Lib/test/test_parser.py", line 18, in roundtrip
    self.fail("could not roundtrip %r" % s)
  File "/usr/local/src/utils/Python-2.2.1/Lib/unittest.py", line 254, in fail
    raise self.failureException, msg
AssertionError: could not roundtrip 'a'

======================================================================
FAIL: test_yield_statement (__main__.RoundtripLegalSyntaxTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "./Lib/test/test_parser.py", line 30, in test_yield_statement
    self.check_suite("from __future__ import generators\n"
  File "./Lib/test/test_parser.py", line 27, in check_suite
    self.roundtrip(parser.suite, s)
  File "./Lib/test/test_parser.py", line 18, in roundtrip
    self.fail("could not roundtrip %r" % s)
  File "/usr/local/src/utils/Python-2.2.1/Lib/unittest.py", line 254, in fail
    raise self.failureException, msg
AssertionError: could not roundtrip 'from __future__ import generators\ndef f(): yield 1'

----------------------------------------------------------------------
Ran 15 tests in 0.130s

FAILED (failures=9)
Traceback (most recent call last):
  File "./Lib/test/test_parser.py", line 365, in ?
    test_main()
  File "./Lib/test/test_parser.py", line 361, in test_main
    test_support.run_suite(suite)
  File "./Lib/test/test_support.py", line 174, in run_suite
    raise TestFailed(msg)
test_support.TestFailed: errors occurred; run in verbose mode for details


-- 
Tim Rice				Multitalents	(707) 887-1469
tim at multitalents.net





From ianholmes01 at lycos.com  Tue Sep  3 06:52:00 2002
From: ianholmes01 at lycos.com (Ian Holmes)
Date: 3 Sep 2002 03:52:00 -0700
Subject: The Module Search Path is ???
References: <7153651c.0209022255.7fd50ba0@posting.google.com>
Message-ID: <b31eb63a.0209030252.538248c4@posting.google.com>

quiteblack at yahoo.com (black) wrote in message news:<7153651c.0209022255.7fd50ba0 at posting.google.com>...
> tutorial told me I could found it in the folder:/usr/local/lib/python.
> but my OS is windows so folders are different. I searched the folder
> "lib" but find no file named "python", any ideas ?

c:\pythonXX\
where XX is your version

python.exe on windows rather than python if a windows search didn't
find this then you are either looking on the wrong drive or python
isn't installed.


From mhammond at skippinet.com.au  Fri Sep  6 21:54:09 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 07 Sep 2002 01:54:09 GMT
Subject: Using Python to call a DCOM server - "'Class not registered"
References: <3D787D95.9000907@ob_scure.dk>
Message-ID: <5%ce9.1272$_l1.2960@news-server.bigpond.net.au>

Thomas Jensen wrote:

> Code that works:
>   batch = win32com.client.Dispatch("IGBatchNGBackend.IGBatch")
> 
> Code that doesn't work:
>   clsctx = pythoncom.CLSCTX_SERVER & ~pythoncom.CLSCTX_INPROC_SERVER
>   batch = win32com.client.DispatchEx("IGBatchNGBackend.IGBatch", 
> 'localhost', clsctx=clsctx)
> 
> The traceback is:
> 
> Traceback (most recent call last):
>   File "test01.py", line 30, in ?
>     test()
>   File "test01.py", line 15, in test
>     batch = win32com.client.DispatchEx("IGBatchNGBackend.IGBatch", 
> 'localhost',
> clsctx=clsctx)
>   File 
> "C:\PROGRA~1\PYTHON22\lib\site-packages\win32com\client\__init__.py", lin
> e 109, in DispatchEx
>     dispatch = pythoncom.CoCreateInstanceEx(clsid, None, clsctx, 
> serverInfo, (py
> thoncom.IID_IDispatch,))[0]
> pywintypes.com_error: (-2147221164, 'Class not registered', None, None)
> 

Are you sure that the COM object you are using is correctly registered 
for use as a DCOM server.  The code above insists that the COM object be 
registered as LocalServer - ie, other than just InprocServer32.

Can you reproduce the registry entries for this COM server (the ones 
under HKCR\CLSID\the-long-guid-for-the-object )

Regards,

Mark.




From op73418 at mail.telepac.pt  Wed Sep  4 15:29:25 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Wed, 04 Sep 2002 20:29:25 +0100
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de>
Message-ID: <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com>

On 04 Sep 2002 20:38:30 +0200, loewis at informatik.hu-berlin.de (Martin v.
L?wis) wrote:

>Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:
>
>> >Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:
>> >
>> >> So why hasn't this been extended when nested scopes were introduced,
>> >> e.g. by reusing the global directive or some other one? 
>> >
>> >I think the main reason is that nobody could propose an acceptable
>> >syntax. Reusing the global directive does not work.
>> 
>> Can you explain why it does not work?
>
>Consider
>
>x = 1
>
>def foo():
>  x = 2
>  def bar():
>    x = 3
>    def baz():
>      global x
>      x = 4
>
>If you invoke baz(), which of the variables is changed? How do you
>denote that you want to change the others?

I think that crawling up the scope hierarchy and stopping at the first x
that one finds (in the above case, the one defined in bar) is a sensible
solution. Any other solution just seems perfect madness, but then again,
most probably I am overseeing something.

Anyway, I've been directed to PEP 227, cheerfully obeyed, and from what
I have understood having read-only nested scopes was more of a design
decision than something else. And while I can live happily with the
decision (some would even say happier) there is something about the "You
can look at it but not touch it" attitude that I dislike ;-)

>
>Regards,
>Martin

Thanks 4 everything and all the best,
Gon?alo Rodrigues


From sholden at holdenweb.com  Tue Sep 17 14:51:44 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 17 Sep 2002 14:51:44 -0400
Subject: [FEEDBACK] Is this script efficient...is there a better way?
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <L9Rf9.168314$l_4.127784@atlpnn01.usenetserver.com> <4ZRf9.1782$Lg2.248733@news2.news.adelphia.net> <3hSf9.168947$l_4.98525@atlpnn01.usenetserver.com> <8PSf9.1796$Lg2.257378@news2.news.adelphia.net> <_oUf9.170870$l_4.126599@atlpnn01.usenetserver.com> <Am7g9.2344$Lg2.402676@news2.news.adelphia.net> <3pkg9.3313$WV.2782@fe03> <fqrh9.7075$Lg2.1331983@news2.news.adelphia.net>
Message-ID: <Y_Kh9.13417$T4.6102@fe04>

"Bob X" <bobx at linuxmail.org> wrote in message
news:fqrh9.7075$Lg2.1331983 at news2.news.adelphia.net...
> Steve Holden wrote:
> <snip>
>
> I did a test on the 30MB sized logs and the RE did indeed come out the
> faster, thought not by much. The interesting thing is it caught more as
> well since my script looked for exact matches and the RE based one
> ignored case (as far as my feeble brain understands such things).
>
> I remember why I had the "result" statement. It was because I needed to
> write the keyword found and then the line it was found on to the hits
file.
>
> Can't seem to make that work with the RE based one. Then again...I know
> RE as far as the east is from the west.
>
If the match succeeds it returns a match object. The match object has
methods that will retrieve the matched substring.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------
>





From chris.lyon at tgm.co.uk  Fri Sep 13 03:42:13 2002
From: chris.lyon at tgm.co.uk (Chris Lyon)
Date: 13 Sep 2002 00:42:13 -0700
Subject: Structure
Message-ID: <7652d5e7.0209122342.61cb80f2@posting.google.com>

I have a set of related classes each are derived from the others in a
tree.
I then produce multiple instances of the different classes.
I wish to edit the properties of the instances and add extra
properties specifically depending on which class they are derived
from.
Is there a standard python approach to constructing this sort of
structure.

My classes are defined in one module, and the display elements are in
a seperate module, since this seems to allow maximum flexibility. So I
am using a dictionary with the name as the key and a tuple of display
text, sort order, type of data. So for instance when I set background
colour I wish to allow the edit facility to choose from a colour
selection dialog etc.

Since the overall structure is pickled, is there a general structure
that allows me not to store a dictionary for every instance, I seem to
be developing a seperate tree of classes for these definitions but my
display components would seem to want to appear in my 'main' module
and I feel my present approach will lead to loosing the flexibility
that modules seeem to provide.

Chris


From leoboiko at linuxbr.com.br  Wed Sep  4 19:03:35 2002
From: leoboiko at linuxbr.com.br (Leonardo Boiko)
Date: 4 Sep 2002 16:03:35 -0700
Subject: newbie-question: Nested functions and variable scope
Message-ID: <7a7157fd.0209041503.7b9a0f44@posting.google.com>

Hi, I'm a Python newbie. I'm writing a program that uses expat. Some
code looks like this:

def generate_main(filename="main.xml", (...)):
    (...)
    outfile = open (filename, "w")
    in_reply = 0 # true if we are inside "reply" element

    def start_element(name, attrs):
        if name == "reply":
            in_reply = 1
        elif not in_reply:
            outfile.write("<" + name.encode(encoding) + ">")

    def end_element(name):
        if name == "reply":
            in_reply = 0
        elif not in_reply:
            outfile.write("</" + name.encode(encoding) + ">")
    (...)

As you may have guessed, this gives me an error:
>>> generate_main()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "../xmlgen.py", line 52, in generate_main
    parser.ParseFile(file)
  File "../xmlgen.py", line 22, in start_element
    elif not in_reply:
UnboundLocalError: local variable 'in_reply' referenced before
assignment


If I put "in_reply" outside everything and use "global in_reply"
inside each nested function, it works, but that's evil! What would be
a better way of doing it? Do I need to use a class?


From m.j.fisher at ex.ac.uk  Wed Sep 11 10:21:53 2002
From: m.j.fisher at ex.ac.uk (M.J.Fisher)
Date: Wed, 11 Sep 2002 15:21:53 +0100
Subject: Background programs on Windows XP
Message-ID: <3D7F5181.AD09A901@ex.ac.uk>

Hello,

I have been using the following code the run a program in the backgroup
on Windows'95 and WindowsNT machines:

os.spawnl(os.P DETACH, 'c:\\python22\\python', 'python',
'c:\\agents.py')

However, when one of the users of this program changed from Windows'95
to WindowsXP (they also changed from Python2.1 to Python2.2 but I didn't
think this was probably relevant) this stopped working.

I cannot find any information about this in the python documentation or
newsgroups so I would appreciate any comments or suggestions.

Thank you,

Michelle



From bokr at oz.net  Sun Sep 29 18:00:54 2002
From: bokr at oz.net (Bengt Richter)
Date: 29 Sep 2002 22:00:54 GMT
Subject: help
References: <amqclg$6uu$1@server.ltid.inpe.br>
Message-ID: <an7t6m$n0s$0@216.39.172.122>

On Tue, 24 Sep 2002 15:58:48 -0300, "Juliana Braga" <juliana at pdi.inpe.br> wrote:

>I need help.
>For example:
>x = ?name1?
>How can I get the value of variable x (in this case the value of variable x
>is ?name1?). After , I?d like to do this:
>value of variable x = ?teste?.
>
The best advice I can give is to try things interactively.
And when you have a question about what happens, copy and paste it
right from your interactive session into your post.

Also, in Python it's best to think of attaching names to values
rather than setting values into named places.

I.e., what would an assignment be overwriting a value in C or many other languages
really just attaches the name on the left to a different value. Another name can still
be attached to the first value. E.g.,

 >>> x = 'name1' # attaches ('binds') name "x" to string value 'name1'
 >>> y = x       # binds name "y" to the same value as "x" is bound to.
 >>> print x,y
 name1 name1
 >>> x = 'teste' # binds name "x" to string value 'teste'
 >>> print x,y   # y should be unaffected by rebinding of x
 teste name1

The above looks ordinary, but when you attach a name to a "mutable" object
the differences start to become apparent, e.g.,

 >>> x = ['name1']
 >>> y = x          # bind y to same list object as x
 >>> print x,y
 ['name1'] ['name1']
 >>> x.append('teste') # modify the list object that x is bound to, but not the binding
 >>> print x,y
 ['name1', 'teste'] ['name1', 'teste']

The binding of y is still to the same object, so the modification we did via the x binding
to that object shows up via y also.

Some operations look the same for notation convenience but really do either rebinding or
mutation without rebinding, depending on the target. E.g., lists are mutable, but strings
are not, so:

 >>> x = 'name1'
 >>> y = x
 >>> x += 'teste'   # x will get rebound to something different than y, whose binding is unchanged
 >>> print x,y
 name1teste name1
 >>> x = ['name1']
 >>> y = x
 >>> x += ['teste'] # the object x is bound to will get modified, but x and y bindings don't change
 >>> print x,y
 ['name1', 'teste'] ['name1', 'teste']

so the modified list shows up via both x and y.


There is a good tutorial at

    http://www.python.org/doc/current/tut/tut.html

and many other documentation links at

    http://www.python.org/doc/

and of course you can start at the top

    http://www.python.org/

where you'll also find a link to a beginner's guide at

    http://www.python.org/doc/Newbies.html

Have fun. And [OT] thanks to you, we now have one more movie star
(last) name posting to c.l.p (along with Peters, Roberts, Holden, Davis, ...
;-)

Regards,
Bengt Richter


From trentm at ActiveState.com  Fri Sep 20 21:26:54 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Fri, 20 Sep 2002 18:26:54 -0700
Subject: Python threading?
In-Reply-To: <f9Pi9.10275$Lg2.2191294@news2.news.adelphia.net>; from Oschler@earthlink.net on Sat, Sep 21, 2002 at 12:35:23AM +0000
References: <f9Pi9.10275$Lg2.2191294@news2.news.adelphia.net>
Message-ID: <20020920182654.A7937@ActiveState.com>

[Robert Oschler wrote]
> How can I tell if the version of Python I am running is thread-enabled?

Run Lib/test/test_thread.py with your Python build. If you are on any
common platform then Python, unless specifically crippled, is
thread-enabled.

> Also, what is a reasonable number of threads to expect to be able to run
> before context switching overhead becomes a problem (I'm using a PIII 500
> Mhz with 512MB ram if that helps).

I have no experience putting that many threads to use.


Trent

-- 
Trent Mick
TrentM at ActiveState.com



From Sylvain.Thenault at logilab.fr  Fri Sep 20 11:24:14 2002
From: Sylvain.Thenault at logilab.fr (Sylvain =?iso-8859-1?Q?Th=E9nault?=)
Date: Fri, 20 Sep 2002 17:24:14 +0200
Subject: [ANN] PyReverse 0.4
Message-ID: <20020920152414.GI28887@orion.logilab.fr>

What's new ?
------------
    * add pyvcg command line tool
    * add support for *.pyreverserc
    * documentation reorganization and improvement
    * fix a bug in pyargo

What's PyReverse ?
------------------

Pyreverse is a set of utilities to reverse enginering Python code. So
far, it features dependency analysis tools, unittest generation,
and XMI generation for importation in a UML modeling tool. A special
module can be used to generate files readable by Argo UML. 
It uses a representation of a Python project in a class hierarchy
which can be used to extract any information (such as generating UML
diagrams and make a few statistics from the Python code, as "pyargo"
and "pystats") 


Home page
---------
http://www.logilab.org/pyreverse/

Download
--------
ftp://ftp.logilab.org/pub/pyreverse/

Mailing list
------------
mailto://xml-logilab at logilab.org

-- 
Sylvain Th?nault

  LOGILAB           http://www.logilab.org




From tismer at tismer.com  Sun Sep 15 15:34:41 2002
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 15 Sep 2002 21:34:41 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net>
Message-ID: <3D84E0D1.1000809@tismer.com>

David Abrahams wrote:
> "Christian Tismer" <tismer at tismer.com> wrote in message
> news:mailman.1032107006.6411.python-list at python.org...
> 
> 
>>The changes to Python are also so small now, that
>>it doesn't matter if it will be merged or not.
>>Porting Stackless to a new Python is a matter of a few
>>hours. You may rely on getting a Stackless version
>>of any new Python version in almost no time.
> 
> 
> I think it still matters. You don't want people who need stackless'
> capabilities to have to use a non-standard Python forever, do you?

Who I ???

I've spent a lot of time to get Stackless into the core.
I wasted many months by keeping to the rules that *might*
get it into the core. This didn't help.

For the new version, I gave that up. That gave me the freedom
to code as I needed it, and I've come much farer than ever
before, in much shorter time.

I will not waste time again trying to persuade certain
people to include Stackless into the core. This is definately
over. If somebody else wants to try, this is fine with me.

Please let me stay what I am: The one who writes Stackless
Python, supports and maintains it. I really do not take
responsibility for being standard or non-standard.
My patches are as near to the standard as I can think of.
If somebody wants to include them, easy. But not me, this
has cost me too much blood, anger, lifetime, disappointment.

or-I-might-define-Stackless-as-standard - ly y'rs - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From tismer at tismer.com  Wed Sep 18 09:33:32 2002
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 18 Sep 2002 15:33:32 +0200
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <20020918060823.GB81158@hishome.net>
Message-ID: <3D8880AC.7040902@tismer.com>

Oren Tirosh wrote:
...

> They may be dangerous in some circumstances. IIRC, if a C extension calls
> the Python interpreter recursively with pointers to things on its stack
> it can cause memory corruption. 

Then don't switch stacks in that context.
The switching occours on your demand.
And if this extension is essential and
must be protected from any overseen switching,
let me add protection code to it. This is easy.

But the theoretical possibility to create a crash
is no reason to dispense with the benefits.

My only chance to do this magic with almost no crash
possibility would be to allocate a new stack for
every new tasklet. I have considered this many times.
A benefit would be reasonably faster switching, while
the drawback would be the reasonable high memory
requirement for a full machine stack.
I might consider a hybrid solution, putting certain
"dangerous" extension stuff into a real new stack
all the time, together with the quite fast and very
cheap other safe tasklets.

Just-need-an-idea-how-to-determine-that - ly y'rs - chris
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From 2002 at weholt.org  Thu Sep 26 09:54:22 2002
From: 2002 at weholt.org (Thomas Weholt)
Date: Thu, 26 Sep 2002 13:54:22 GMT
Subject: Python threading?
References: <f9Pi9.10275$Lg2.2191294@news2.news.adelphia.net> <amogmi$r1l$1@panix1.panix.com>
Message-ID: <ikEk9.26204$sR2.463163@news4.ulv.nextra.no>

This might be just a stupid question, but I'm trying to implement a
filesharing-app, based on the BaseHTTPServer in the standard python-distro.
This will of course be heavy IO-processing involved. Will threading still be
a bad idea? The project is aimed at small user-groups ( I do not think one
person will have more than say 5-10 concurrent connections at a time, most
of them downloading files 1MB + in size), it's not the new Gnutella/Naptser,
but I just can't get the hang of this Threading vs. Asyncore deal. It seems
as if threading is a very bad idea no matter what, still it's used almost
everywhere ( I got no statistics to back up that statement ;-) ).

Any info or opinion on what would be the best way to implement a filesharing
app in python; threaded, async, raw-socket etc. are of interest to me.

NB! It would be nice to be able to limit the number of threads, actually the
number of concurrent connections, on the server too, so if anybody has any
comments on some way of doing that and still handle request gracefully, it
would be great.

Best regards,
Thomas


"Aahz" <aahz at pythoncraft.com> wrote in message
news:amogmi$r1l$1 at panix1.panix.com...
> In article <f9Pi9.10275$Lg2.2191294 at news2.news.adelphia.net>,
> Robert Oschler <Oschler at earthlink.net> wrote:
> >
> >Also, what is a reasonable number of threads to expect to be able to run
> >before context switching overhead becomes a problem (I'm using a PIII 500
> >Mhz with 512MB ram if that helps).
>
> It depends.  For pure Python-based computation, more than one thread
> will cost you more than you gain.  If it's almost pure I/O (file or
> socket stuff), you might be able to use as many as a couple of hundred
> threads.  Mostly, you're probably best off somewhere between five and
> thirty threads, depending on what you're doing.
> --
> Aahz (aahz at pythoncraft.com)           <*>
http://www.pythoncraft.com/
>
> Project Vote Smart: http://www.vote-smart.org/




From fgranger at alussinan.org  Fri Sep 20 14:34:46 2002
From: fgranger at alussinan.org (=?ISO-8859-1?Q?Fran=E7ois_Granger?=)
Date: Fri, 20 Sep 2002 20:34:46 +0200
Subject: [Help]: mailbox classes
References: <jylimd-DAD12F.17305418092002@skates.gsfc.nasa.gov> <Ndfi9.130668$pX1.4682176@news2.tin.it> <1fitpvg.gi8n0y1iem93tN%fgranger@alussinan.org> <1fitpz7.1h4nj4m19gs2x5N%fgranger@alussinan.org>
Message-ID: <1fitt21.1a27xor1bmqkizN%fgranger@alussinan.org>

Fran?ois Granger <fgranger at alussinan.org> wrote:

> after some thought and some trial and errors, I came with this solution.
> I don't like it, it look bad.
> 
> def main():
>     fp = file(folderfile)
>     folder = mailbox.UnixMailbox(fp, factory=email.message_from_file)
>     while 1:
>         mail = email.message_from_file(fp)
>         print mail
>         if not folder.next():
>             break
>         #mail = email.message_from_file(fp)
> 
> if __name__ == '__main__':
>     main()

After reading other posts on other matters, light came at least...

def main():
    fp = file(folderfile)
    folder = mailbox.UnixMailbox(fp, factory=email.message_from_file)
    for message in folder:
        print message

This works as expected, as usual in Python, the simple things _ARE_
simple, the complexe ones not difficult ;-).

Sorry for any inconvenience.

-- 
"On jouit moins de ce qu'on obtient que de ce qu'on esp?re."
- Jean-Jacques Rousseau
- Cit? par Anne-Marie


From whisper at oz.net  Thu Sep  5 17:52:56 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 5 Sep 2002 14:52:56 -0700
Subject: Graham's spam filter
In-Reply-To: <3D77AC75.B9B7F27@alcyone.com>
Message-ID: <GCEDKONBLEFPPADDJCOEEECFENAA.whisper@oz.net>

<snip>
> Spammers are hitting upon the strategy, though, of sending emails in
> which the body consists of nothing but a completely encoded base64 MIME
> part.  So in that case, the entire body of your message would consist
> solely of your "base64encoded" token.  So in the general case of any
> kind of spam filter (not just limited to a Graham filter), it's
> questionable how useful this will be, unless you plant to always filter
> against that token, presuming it to always indicate spam.

Ummm... In which case, the headers would probably control the sort. IIRC,
Graham said that headers had a large impact on spam/ham determination. Spam
header information is already so marked in the filter table of course.

> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ There is nothing so subject to the inconstancy of fortune as war.
> \__/ Miguel de Cervantes
>     Church / http://www.alcyone.com/pyos/church/
>  A lambda calculus explorer in Python.

Dave LeBlanc
Seattle, WA USA




From peter at engcorp.com  Mon Sep 16 21:51:02 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 16 Sep 2002 21:51:02 -0400
Subject: Converting Python app to C++ completely
References: <aklm6c$r1r$1@news.swman.net.uk> <3D85D9EB.9020102@thinkware.se> <roy-2A7271.09563416092002@reader1.panix.com> <3D86209F.4080501@earthlink.nospam> <am5am0$fi6$1@panix2.panix.com>
Message-ID: <3d868a77$1@news.sentex.net>

Roy Smith wrote:
> Joe Knapka  <jknapka at earthlink.nospam> wrote:
> 
>>You might try rewriting it in Python, using lists of characters
>>instead of strings for the stuff that's causing performance
>>trouble. Given str, lst = list(str), then do all the processing
>>(in place) on lst, and ''.join(lst) to turn it back into a string.
> 
> 
> An interesting idea, I'll try it!
> 
> The only problem is, I really had two reasons for converting it
> in the first place.  One was to improve performance, the other was to
> learn C++.  If I solve the first problem in Python, I won't solve the
> second problem at all :-(

Hmmm... if instead you choose to wait until you actually *need* C++
to solve a problem that Python can't handle, you just might find that
you don't ever need to learn C++.  Then the second one is actually
not a problem at all! :-)

This thought courtesy of XP's philosophy of avoiding unnecessary
work, which is acronymized as YAGNI (You Aren't Going to Need It).

-Peter



From tg5027 at citlink.net  Wed Sep 18 00:47:48 2002
From: tg5027 at citlink.net (terry)
Date: Wed, 18 Sep 2002 00:47:48 -0400
Subject: Python GUI app to impress the boss?
Message-ID: <E17rWIA-0004wr-00@mail.python.org>

>> Dave LeBlanc wrote:
>> What prevents the creation of a decimal money type?

Absolutely nothing!  And it has been done - sort of.

The problem involves the resultant code to implement a business 
application.  In the simple situation, all the developer has to do is 
be aware on <every> calculation what specific type of variables are  
being used ( integers for quantities, reals for percentages, etc...) 
and what the result 'should' be - and then apply the appropriate 
conditioning to the variables in the calculation and to the result.  
You can surely contemplate the variability in programming styles and 
implementations that this generates - most of which flies in the face 
of the simplicity and readability promised by Python.  Contrast that 
with the following VB code:  Amount = Qty * UnitPrice

The more difficult problem arises when faced with a true quality 
assurance environment.  The amount of validation rises exponentially 
when the calculations in a program are implementation specific, as 
opposed to language intrinsics.  Each calculation must be tested at the 
limits of each of the variables involved - at each instance.

The quick reply to your observation is:  Yes, anything is possible - 
but is forcing a language into an environment to which it is not 
intrinsicly suited a reasonable approach?  All of the databases, that 
Python can easily talk to, have a form of monetary data type.  But when 
those table fields are represented in Python (as well as some other 
languages) they become a burden.

I think Python needs to incorporate an intrinsic data type that 
monetary (as well as other decimal data types) can be represented by, 
so that the above equation yields results consistent with the decimal 
places of the input variables.  This cannot be accomplished with 
real/floats, and shouldn't be accomplished with overlays on integer or 
character variable types.  Python needs this data type, that is not 
hardware oriented, to become truly mainstream and be a threat to VB.

I say this with all due respect to all the other positive attributes of 
Python and it's flexibility to solve a myriad of other non-business 
application problems.

terry



From bondpaper at earthlink.net  Mon Sep 30 22:16:15 2002
From: bondpaper at earthlink.net (Tom)
Date: Tue, 01 Oct 2002 02:16:15 GMT
Subject: Reloading a module
References: <bondpaper-648404.19153230092002@nnrp04.earthlink.net> <anastb$h2s$1@peabody.colorado.edu>
Message-ID: <bondpaper-06D4B1.20204130092002@nnrp02.earthlink.net>

In article <anastb$h2s$1 at peabody.colorado.edu>,
 Fernando Perez <fperez528 at yahoo.com> wrote:

> Tom wrote:
> 
> > 
> > I'm developing some python mini applets on a Mac (using os 9 and the
> > 2.1.1 Python IDE), and I've found it impossible to delete and/or reload
> > a module after making changes to the code during the
> > debugging/verification process. Is there an easy and/or effective way to
> > do this? If I execute:
> > 
> > del mymodule
> > 
> > and then execute a print dir(mymodule), it shows up as nonexistent.
> > Then, when I execute:
> > 
> > import mymodule
> > 
> > followed by a print dir(mymodule), the changes do not show up.
> 
> In [1]: reload?
> Type:           builtin_function_or_method
> Base Class:     <type 'builtin_function_or_method'>
> String Form:    <built-in function reload>
> Namespace:      Python builtin
> Docstring:
>     reload(module) -> module
> 
>     Reload the module.  The module must have been successfully imported 
> before.
> 
> cheers,
> 
> f.

I wish it were that simple. I've used reload(module) many times, each 
time hoping that it would do what the documentation says it will do. 
I've tested it by adding a simple variable to my module, watching to see 
if it shows up in the new import, but it never does. I noticed also that 
sometimes it reloads from the .pyc file, and if the reload happens 
before the .pyc file is updated with the new changes, the changes won't 
be available. Even if I force a reload from the updated .py file by 
deleting .pyc file, it still retains the old code. I'm at a loss to 
understand what's going on.

Thanks,

Tom


From maxm at mxm.dk  Mon Sep  9 03:05:23 2002
From: maxm at mxm.dk (Max M)
Date: Mon, 09 Sep 2002 09:05:23 +0200
Subject: Zope Status?
References: <Xns9281221A07167bosahvnetscapenet@66.185.95.104> <cSOe9.3024$io.1100@nwrddc03.gnilink.net>
Message-ID: <3D7C4833.2080308@mxm.dk>

Jeff Sasmor wrote:


> That's just one thing. Also, Zope 3 is a clean-slate redesign; that is,
> it is being rearchitected with no initial regard for existing Zope
> applications.  I suppose opinions go either way on this, but overall
> it is probably a Good Thing.  I recall reading that there will eventually
> be backwards compatibility, perhaps with some interface layer, but
> I suspect that the initial releases will be for very advanced developers.


As far as I can see from the examples you actually don't need to be an 
advanced developer. It seems that it will indeed be much more simple to 
develop for Zope3 than it is for the current version.

Also they are taking care of some of the things that could make it 
difficult to develop in Zope. Like relations.


regards Max M



From jmdeschamps at cvm.qc.ca  Fri Sep 20 08:45:10 2002
From: jmdeschamps at cvm.qc.ca (jmdeschamps)
Date: 20 Sep 2002 05:45:10 -0700
Subject: Tkinter: scrolling the canvas items AND widgets
References: <3d06fae9.0209191255.63def2eb@posting.google.com> <3D8A8146.3030506@something.invalid>
Message-ID: <3d06fae9.0209200445.6b2c168a@posting.google.com>

Greg Ewing <see_reply_address at something.invalid> wrote in message news:<3D8A8146.3030506 at something.invalid>...
> jmdeschamps wrote:
> 
> > This code lets me scroll the graphical items but the widgets stay put
> > using PythonWin 2.2.1 on Windows 2000
>  
>  >
>  
> > for i in range(1):
> >     f=Entry(canvas)
> >     f.place(x=10,y=10)
> 
> 
> I think you need to create a window item on the
> canvas, with the window attached to it, something
> like this :
> 
>        f = Entry(canvas)
>        canvas.create_window(10, 10, window = f)

Thanks, :)) - works beautifully!
Jean-Marc


From anandpillai6 at yahoo.com  Thu Sep 26 02:06:37 2002
From: anandpillai6 at yahoo.com (Anand)
Date: 25 Sep 2002 23:06:37 -0700
Subject: Pygame event problem
References: <41c203e.0209250702.160102d@posting.google.com> <Bplk9.73049$gA4.26401@sccrnsc02>
Message-ID: <41c203e.0209252206.3c344224@posting.google.com>

Hi Shinners

        That was very helpful! I had thought of making another post
about the possibility of embedding the pygame window into wxPython,
but you have answered that also. :-)

 I am a python newbie as you should have guessed. Just finished reading
the pygame/wxPython tutorials so I thought I will make an attempt at putting
them together. 

 I will try your suggestions in order. Could you suggest some links to
learn python multithreading ?

Regards

Anand Pillai


Pete Shinners <pete at shinners.org> wrote in message news:<Bplk9.73049$gA4.26401 at sccrnsc02>...
> Anand wrote:
> >   I wrote a simple image viewer in python using pygame image api. The
> > GUI is in wxPython. There are 2 classes, one for the window frame and one
> > for the pygame image processing.
>  
> >    The effect is stopped by clicking with the mouse on the pygame
> > display window. My problem is that if I try to invoke any other menu methods
> > in the main GUI frame *without doing this*, the application ie both the
> > wxFrame and the pygame display window, hangs.
> > 
> > 
> >   I have a gut-feeling that this is because I am calling a method on
> > the MyPyGameImgViewer class without exiting the while loop.
> > 
> >  Is there a way around this, ie to call any method in the
> > MyPyGameImgViewer class from the GUI class, while one of its methods
> > is executing a while loop
> 
> i'd be a little afraid of mixing wxpython and pygame like this, but it
> sounds like you are pretty close to running. to exit the pygame loop you
> could probably just pass it a mouse button event
> 
> e = pygame.event.Event(MOUSEBUTTONDOWN)
> pygame.event.post(e)
> 
> 
> of course, that also means you'll probably want to clear the pygame event
> queue before starting your while loop. also, to clear the event queue it
> will probably be better to get _all_ the events from the queue. otherwise
> it will fill up with mouse motion and other types of events. change the
> event part of the loop to this
> 
> if pygame.event.peek(MOUSEBUTTONDOWN):
>    pygame.event.get()
>    break
> 
> instead, you could also filter the pygame event queue, so only mouse button
> events were put onto the queue. the code for that would be like this
> 
> pygame.event.set_allowed(None) #allow no events
> pygame.event.set_allowed(MOUSEBUTTONDOWN) #only allow buttondown
> 
> 
> 
> also, there are ways to get SDL to work within an embedded window.
> therefore embed into a wxpython control. basically it involves setting an
> environment variable to an existing window id/handle. wxpython allows you
> to get the id/handle you need. the code to embed the SDL window looks kind
> of like this:
> 
> 
> handle = wxpython_widget.get_window_id_i_forget_how()
> os.environ['SDL_WINDOWID'] = str(handle)
> if sys.platform == 'win32':
>     os.environ['SDL_VIDEODRIVER'] = 'windib' #no directx embedded
> pygame.init()
> size_is_ignored = 10, 10
> window = pygame.display.set_mode(size_is_ignored)
> 
> also, since you are likely only using the display portion of pygame, no
> need to initialize the whole thing (like audio/cd/joystick/etc). you could
> change the "pygame.init()" call do just "pygame.display.init()"
> 
> that should be pretty slick when you get it working. although it's been a
> long time since i did anything like this. this should be all the correct info.
> 
> oh, also remember, when you embed the SDL window like this, it will not
> receive any input events. you'll need to use wxpython's event handling for
> that.


From igetenoughspamalreadythanksjlr at doctor.com  Mon Sep  2 07:34:12 2002
From: igetenoughspamalreadythanksjlr at doctor.com (John La Rooy)
Date: Mon, 02 Sep 2002 21:34:12 +1000
Subject: Problems with file.write()
References: <4378fa6f.0208291457.ba738cb@posting.google.com> <mailman.1030669067.7832.python-list@python.org> <4378fa6f.0208300731.18105bc7@posting.google.com> <akqf9u$1kjdm1$1@ID-114079.news.dfncis.de> <4378fa6f.0209020824.3e658da8@posting.google.com>
Message-ID: <1031009619.821203@ampungk.ozonline.com.au>

Marc wrote:
>>Note that all buffered data is flushed automatically when you close()
>>an output file (you *are* closing the file, aren't you?).
> 
> 
> Actually I wasn't closing the file originally but it had never been a
> problem before. But closing the file now causes it to spill it's guts
> and gives me everything I was looking for.
> 
> However, that doesn't explain why it wasn't working in the first
> place. I originally had it in unbuffered mode and wasn't getting all
> my text. As I increased the buffering from 1 to 2 I kept getting more
> text. But never all of it until I closed the file. It seems like it
> shouldn't be necessary to flush the buffer after every write if you're
> in unbuffered mode. This seems like a problem.
 From the docs...

   The optional bufsize argument specifies the file's desired buffer 
size: 0 means unbuffered, 1 means line buffered, any other positive 
value means use a buffer of (approximately) that size.

John



From bourassa-m at rmc.ca  Wed Sep 25 15:02:04 2002
From: bourassa-m at rmc.ca (bourassa)
Date: Wed, 25 Sep 2002 15:02:04 -0400
Subject: Scientific Python for Win32
Message-ID: <3D92082C.3050707@rmc.ca>

Hi there.  I'm a Python and programming newbie.  I was looking for at 
Scientific Python esp. for its use of VRML for plotting output.  Unlike 
many of the other Python modules I have, there are no .exe binaries. 
So, I have Borland C++Builder 5 on a Win2K P4.  How would I compile(?) 
my own Scientific Python?  Would it work with my Python 2.2 which I just 
installed from a .exe file?  Any help would be appreciated.  Thanks.

Mike

PS - I spoke with the creator of Scientific Python.  He was very helpful 
but admits to not having expertise in this aspect, hence my question.



From gerhard.haering at gmx.de  Fri Sep 27 12:15:28 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 27 Sep 2002 16:15:28 GMT
Subject: comp.lang.python...
References: <mailman.1033054480.24506.python-list@python.org> <an1ip7$rgs$00$1@news.t-online.com> <mailman.1033131395.24431.python-list@python.org>
Message-ID: <slrnap910v.5gd.gerhard.haering@lilith.my-fqdn.de>

Vincent A. Primavera wrote in comp.lang.python:
> Hello,
> 	I've been subscribed to python-list at python.org for at least a couple of
> years I would guess.  I have been using a mail client to receive all of
> the messages, but I figure that using a news client would be a more
> organized method.  However, I can't seem to find what I need to enter
> for a server & user/password combination...

If your ISP doesn't offer an NNTP server, you'll probably have bad
luck, unless you pay for third-party NNTP access somewhere.

However, there's a mail-to-news-gateway (and reverse) available at
http://gmane.org/, which NNTP access for many Python mailing lists.
The data you'll need are:

NNTP server: news.gmane.org
group name: gmane.comp.python.general

No username or password required for reading. If you post, you'll need
to answer a one-time confirmation message for every group.

I'd still prefer to use comp.lang.python diretly. But this is quite
cool for all the SIG groups, as you can read them without actually
subscribing.

-- Gerhard


From susan at edeltech.com  Thu Sep 26 13:37:40 2002
From: susan at edeltech.com (Susan Denison)
Date: Thu, 26 Sep 2002 13:37:40 -0400
Subject: Please post
Message-ID: <000001c26583$69cc6800$6901a8c0@SusanII>

GUI Developer

Boston  Area 

Excellent salary and benefits

 

Our client is a successful, growing enterprise software company.    They
are looking for a bright, versatile software engineer to develop
graphical user interfaces for their product.

 

REQUIREMENTS

--4 years experience creating graphical user interfaces in VC++/MFC

--Experience working on a commercial software product

--Experience developing user interfaces with back-end database
connectivity

--Experience with operating system, kernel, and database internals, as
well as multi-threading and compilers

 

 

To apply please send resume to  <mailto:jobs at edeltech.com>
paul at edeltech.com

For more information contact Paul Edelman at (508) 947-5300

 

Edelman & Associates is an executive search and technical recruiting
firm. To see a complete 

listing of current positions please visit us at
<http://www.edeltech.com/> www.edeltech.com

 

 

Susan

_________________________
Susan Denison
Edelman & Associates
(508) 947-5300
 <mailto:susan at edeltech.com> susan at edeltech.com

 <http://www.edeltech.com>  

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020926/e533925d/attachment.html>

From paul at boddie.net  Mon Sep  2 04:13:59 2002
From: paul at boddie.net (Paul Boddie)
Date: 2 Sep 2002 01:13:59 -0700
Subject: Jython and zxJDBC package
References: <d76a3954.0208301123.331bdfbf@posting.google.com>
Message-ID: <23891c90.0209020013.b05e81c@posting.google.com>

billshaw+google at fastmail.fm (William Shaw) wrote in message news:<d76a3954.0208301123.331bdfbf at posting.google.com>...
> Is anyone here using the zxJDBC package as described in "Jython
> Essentials" to access a MySQL database?
> 
> When I try to run the sample program (Example 10-1, p. 164), I receive
> a NoClassDefFoundError: org/gjt/mm/mysql/Statement in
> MySQLDataHandler.  Decompiling MySQLDataHandler reveals an explicit
> reference to org.gjt.mm.mysql.Statement; but this class isn't in the
> Connector/J library.  Am I missing something here?
> 
> I'm using Jython 2.1 and Connector/J 2.0.14.

It looks like MySQLDataHandler wants to use the MM.MySQL package,
which can be found here:

  http://sourceforge.net/projects/mmmysql

Note that I haven't used zxJDBC or Connector/J, whatever that is, but
the above package is a fairly usable JDBC interface to MySQL, in my
experience.

Paul


From dgallion1 at yahoo.com  Wed Sep 18 22:37:15 2002
From: dgallion1 at yahoo.com (darrell)
Date: Thu, 19 Sep 2002 02:37:15 GMT
Subject: Regex question
References: <amapuf$45t8t$1@ID-144898.news.dfncis.de>
Message-ID: <vLai9.2918$Kj4.247925@twister.nyroc.rr.com>

s="Name:AAA,BBB,CCC"
print re.split(":|,", s)
print re.findall("[^:,]+", s)

['Name', 'AAA', 'BBB', 'CCC']
['Name', 'AAA', 'BBB', 'CCC']

--Darrell

Eric Arnold wrote:

> Hi folks,
> 
> I'm trying to figure out why this is not working, and wonder if there are
> any regexperts out there who can give me a pointer.
> 
> The string I'm capturing is as follows:
> 
> Name:AAA,BBB,CCC
> 
> where I'd like to extract the following data:
> 
> <eventName>: Name
> <eventParam 1>: AAA
> <eventParam 2>: BBB
> <eventParam 3>: CCC
> 
> this is the regular expression I wrote to do so:
> 
> (?<eventName>.*):(?<eventParam>([^,]*(?=,))|((?<=,)[^,]*))+
> 
> I tried it with the (?<=,) as another (?=,) too, but no luck.  I imagine
> that the problem is that when the (?=) is used once, it can't be used with
> the second group, but I don't know a way around that.
> 
> Thanks!
> 
> Eric



From timr at probo.com  Sun Sep  1 00:56:11 2002
From: timr at probo.com (Tim Roberts)
Date: Sat, 31 Aug 2002 21:56:11 -0700
Subject: Crossword generator
References: <pan.2002.08.05.23.29.50.205338.718@thomas-guettler.de> <aknmta$h7f$3@newsreader2.netcologne.de>
Message-ID: <a873nukqsjots9id3dgbnoomoajs3ieg8v@4ax.com>

Frank Buss <fb at frank-buss.de> wrote:

>Thomas Guettler <pan-newsreader at thomas-guettler.de> wrote:
>
>> Has someone seen or written a crossword
>> generator in python?
>
>Take a look at http://crosswords.50g.com/ at the bottom of the page.

These are only pseudo-crosswords.  Like most automatic crossword
generators, it doesn't worry about symmetry, and it only crosses a few
words here and there.

Creating a challenging, interesting, and legitimate crossword puzzle is
very, very difficult.  Few humans can do it well, and it will be a while
yet before they can describe their own algorithms in such a way that it can
be automated.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From cstrong at arielpartners.com  Mon Sep 30 11:38:09 2002
From: cstrong at arielpartners.com (Craeg K Strong)
Date: Mon, 30 Sep 2002 11:38:09 -0400
Subject: Is there a way to find the name of the method currently being
 executed?
References: <mailman.1033009413.19897.python-list@python.org> <k3Ek9.170802$pX1.6081475@news2.tin.it> <mailman.1033140277.13452.python-list@python.org> <7YZl9.191322$pX1.6844789@news2.tin.it>
Message-ID: <3D986FE1.9010009@arielpartners.com>

Alex Martelli wrote:
> Craeg K Strong wrote:
>>One question: the sys._getframe() method is new for Python 2.1--
>>how likely do you think it is to change in 2.2+
>>releases?   The part that worries me is that the
>>method starts with an underscore.  That generally means
>>"protected/private" or "unpublished" to me.
> 
> Yes, the name convention does indicate "internal".
> 
>>I just want to know how safe it is to use for
>>general business application type programming.
> 
> Not very: it's intended to be used for debugging and
> development purposes.  If you can cleanly express your
> application's intent WITHOUT a function needing to know
> its own name, it's by far preferable to do so.  If such
> self-knowledge is indispensable to cleanly express your
> application's intent, then sys._getframe is as usable as
> any other method -- worst case, should it go away one
> day, you'll have to install it in your own startup code
> (code to emulate sys._getframe for old Python versions is
> also available in the Cookbook, and most likely similar
> code can be cooked up in the future, as long as there's
> _any_ way to get such to-you-indispensable information).
> 
> Alex
> 

Fair enough.  The functionality is not indespensible, but
rather something we would like to use for debugging purposes.

As such, I feel comfortable using it in limited, well
defined circumstances even though it is
"unpublished" functionality, as long as we
put in a big fat comment.
That way, when/if it breaks, we know what to do...

Thanks again,

--Craeg




From dsavitsk at e-coli.net  Wed Sep  4 12:13:22 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Wed, 04 Sep 2002 16:13:22 GMT
Subject: PIDDLE install problems
References: <AT5d9.6490$yt3.3042205@newssrv26.news.prodigy.com>
Message-ID: <Ciqd9.6675$yt3.3173546@newssrv26.news.prodigy.com>

the answer to this from the PIDDLE Yahoo! group is to unzip and copy the
contents of \src\ to the Python Path.

-d


"dsavitsk" <dsavitsk at e-coli.net> wrote in message
news:AT5d9.6490$yt3.3042205 at newssrv26.news.prodigy.com...
> I am trying to install PIDDLE (Win2k, 2.1.3 ).  I downloaded the recent
.zip
> file, unzipped it to C:\piddle-1.0.15, cd'd to that directory, typed
"python
> setup.py install" and received the following error.
>
> ...
> copying src\piddle\python.gif -> C:\Python\piddle
> Traceback (most recent call last):
>   File "setup.py", line 28, in ?
>     packages = ['piddle', 'piddle.piddleGTK', 'piddle.piddleSVG',
> 'piddle.piddle
> TK2'] )
>   File "C:\Python\lib\distutils\core.py", line 138, in setup
>     dist.run_commands()
>   File "C:\Python\lib\distutils\dist.py", line 899, in run_commands
>     self.run_command(cmd)
>   File "C:\Python\lib\distutils\dist.py", line 919, in run_command
>     cmd_obj.run()
>   File "C:\Python\lib\distutils\command\install.py", line 484, in run
>     self.run_command(cmd_name)
>   File "C:\Python\lib\distutils\cmd.py", line 328, in run_command
>     self.distribution.run_command(command)
>   File "C:\Python\lib\distutils\dist.py", line 919, in run_command
>     cmd_obj.run()
>   File "C:\Python\lib\distutils\command\install_data.py", line 60, in run
>     dir = convert_path(f[0])
>   File "C:\Python\lib\distutils\util.py", line 86, in convert_path
>     if pathname[0] == '/':
> IndexError: string index out of range
>
> is this unusual (the same happened w/ 1.0.14), and does anyone know a way
to
> get around this, or why it is happening?
>
> thanks,
>
> -doug
>
>




From wilk-spamout at flibuste.net  Mon Sep 30 12:09:37 2002
From: wilk-spamout at flibuste.net (William)
Date: 30 Sep 2002 18:09:37 +0200
Subject: redhat8
Message-ID: <873crr4fy6.fsf@flibuste.net>

hi,

anybody know wich version of python is /usr/bin/python on the new redhat
release ?

-- 
William Dod? - flibuste.net
http://wikipython.tuxfamily.org


From steve at ferg.org  Thu Sep 12 08:28:27 2002
From: steve at ferg.org (Stephen Ferg)
Date: 12 Sep 2002 05:28:27 -0700
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <ac677656.0209111225.4a141bc5@posting.google.com>
Message-ID: <b16e4ef7.0209120428.468579d8@posting.google.com>

> 1. You might want to point out that we can enter and experiment with
> simple Tkinter scripts interactively in PythonWin or IDLE. 

Good idea.

> I think you meant to say, 'under the name "myParent".'

Yes.  I'll fix that.  Good catch!

-- Steve


From pedronis at bluewin.ch  Thu Sep  5 11:30:58 2002
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Thu, 5 Sep 2002 17:30:58 +0200
Subject: jython memory usage
References: <hVm5ydAY2hd9Eww$@jessikat.demon.co.uk><3d76366a$1_5@news.bluewin.ch> <rnLbuKAv5jd9Ewzz@jessikat.demon.co.uk><3d76494d$1_5@news.bluewin.ch> <YR13DVAUNld9EwXH@jessikat.demon.co.uk><3d765ce6_4@news.bluewin.ch> <zzKd7KAqhnd9EwxT@jessikat.fsnet.co.uk><3d767c4a$1_1@news.bluewin.ch> <YWyHyIAJKyd9EwHA@jessikat.demon.co.uk><3d774f40$1_1@news.bluewin.ch> <RTiVzIAJD3d9Ew2t@jessikat.demon.co.uk>
Message-ID: <3d7779b0$1_3@news.bluewin.ch>

Robin Becker <robin at jessikat.fsnet.co.uk> wrote in message
RTiVzIAJD3d9Ew2t at jessikat.demon.co.uk...
> OK I get over 600 lines at the initial burst which ends with
> .......
> [Loaded sun.reflect.GeneratedMethodAccessor18]
> [Loaded sun.reflect.GeneratedMethodAccessor19]
> [Loaded org.python.modules.MD5Object]
> [Loaded reportlab.lib.fonts$py]
> [Loaded java.io.RandomAccessFile from C:\Program
Files\Java\j2re1.4.0_01\lib\rt.jar]
> [Loaded sun.reflect.GeneratedMethodAccessor20]
>
> memory usage climbs continuously and then
> Exception in thread "CompileThread0" java.lang.OutOfMemoryError: requested
32756 bytes
> C:\jython-2.1\reportlab\test>
>

If I understand your report correctly, that means you get only the initial
burst of [Loaded ... ] and no further [Loaded ...  ] with or without
[Unloading class ... ]matching the memory usage climbing? then it seems
really that we are triggering some JVM bug.

Have you tried with an older JVM? anyway is the code and minimal
instructions to reproduce the "disaster"
available ?

regards.




From aleax at aleax.it  Sun Sep 22 17:44:05 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 22 Sep 2002 21:44:05 GMT
Subject: sys.ps2 assignment has no effect
References: <uos1np8t53k396@corp.supernews.com>
Message-ID: <FQqj9.149901$pX1.5369568@news2.tin.it>

Lance wrote:

> I'm running the latest Python on Windows XP. The default installed
> secondary string appears to be blank spaces rather than an ellipse. The
> following has no effect:
> 
> import sys
> sys.ps2 = '... '
> 
> Any suggestions?

Is this with the text-mode interactive Python (which would
surprise me, as I've run the latest Python on XP w/o such
problems), or with IDLE or other such IDE?  IDLE does not
use PS2, afaik (it always uses spaces instead) and other IDEs
might do likewise.


Alex



From glingl at aon.at  Mon Sep 30 16:08:11 2002
From: glingl at aon.at (Gregor Lingl)
Date: Mon, 30 Sep 2002 22:08:11 +0200
Subject: Canvas.py obsolete - why?
Message-ID: <3D98AF2B.6010901@aon.at>

Hello!

In the Standard Python distribution with
Tkinter there is contained the module
Canvas.py which among other things exports
class definitions for several geometrical
objects as lines, rectangles, arcs etc.

This module starts with the following comment:

# NOTE: This module was an experiment and is now obsolete.
# It's best to use the Tkinter.Canvas class directly.

Does anyone know, *why* this module is declared
obsolete? (I consider it pretty useful.)

Gregor Lingl




From tjreedy at udel.edu  Sun Sep 15 17:56:48 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Sun, 15 Sep 2002 21:56:48 GMT
Subject: python performance
References: <3D84E66F.6020701@linux.ie>
Message-ID: <zm7h9.51182$AR1.2015178@bin2.nnrp.aus1.giganews.com>

"Padraig Brady" <padraig at linux.ie> wrote in message
news:3D84E66F.6020701 at linux.ie...
> I was wondering about the performance characteristics
> of python and ran a simple test. The 2 programs
> below are functionally equivalent and just read
> in fields into a list. file.fields contains
> 720 fields in each of 405 lines of the form of
> repeating: oneoneone twotwo "thre e three"
>
> The time to run is shown above each program
> from which I've inferred the following:
>
> 1. The function call version is (6.3%) faster because
>     the cumulative cost of parsing the simpler expressions
>     and function call overhead is smaller than parsing the
>     1 single complex expression?

Calling functions adds overhead.  Running code within functions
reduces overhead for object access.  See end for third version to
test.

> Or the function is parsed
>     only once and doesn't have to be reparsed. This would
>     suggest that top level code is parsed for each iteration?

No.  All code is parsed only once.  See above and below.

> 2. Anyway I thought that parsing affects would be removed by doing
the
>     parsing only once, i.e. compiling the code to .pyc (I used
>     py_compile.compile()). However this makes no difference at all?
>     Surely compiling is not just for code obfuscation.

No.  The one-time parsing of 10 lines of code is extremely fast.  If
it were parsed over again for each line, then you would notice the
parsing time.


> Note I did do the test several times and averaged the results.
> ----------------------------
> 2.514s
> ----------------------------
> #!/usr/bin/env python2.2
> import re
>
> reFieldFinder = re.compile('[^ "]+|"[^"]+"') #unquoted|quoted
> def getFields(line):
>      fields = reFieldFinder.findall(line)
>      return [field.replace('"', '') for field in fields]
>
> for line in open("file.fields").readlines():
>      listLine = getFields(line[:-1])
>
> ----------------------------
> 2.672s
> ----------------------------
> #!/usr/bin/env python2.2
> import re
>
> reFieldFinder = re.compile('[^ "]+|"[^"]+"') #unquoted|quoted
> for line in open("file.fields").readlines():
>      listLine = [field.replace('"', '') for field in
> reFieldFinder.findall(line[:-1])]

Try one more test to run at function speed without repeated
getsfields():

def mytest:
    reFieldFinder = re.compile('[^ "]+|"[^"]+"') #unquoted|quoted
    for line in open("file.fields").readlines():
        listLine = [field.replace('"', '') for field in
    reFieldFinder.findall(line[:-1])]

mytest()

Terry J. Reedy





From tibor.simko at cern.ch  Tue Sep 24 10:50:03 2002
From: tibor.simko at cern.ch (Tibor Simko)
Date: Tue, 24 Sep 2002 16:50:03 +0200
Subject: Getting started
References: <uojinmcfapjf45@corp.supernews.com> <mailman.1032442884.18747.python-list@python.org>
 <H3mi9.133635$pX1.4751676@news2.tin.it>
 <mailman.1032567390.29374.python-list@python.org>
 <0%Aj9.129252$ub2.2813083@news1.tin.it>
Message-ID: <877khb7884.fsf@pcdh91.cern.ch>

Hello

On Mon, 23 Sep 2002, Alex Martelli wrote:
> as far as I know, the CAML dialect of ML (also in the O'Caml
> extension, last I looked) is one of the very few programming
> languages that doesn't have a polymorphic + operator -- you have to
> use + for integers only, +. for floating-point. 

There is a good reason why OCaml does not overload the + operator:
type inferring.  In OCaml you do not need to declare variables
explicitly in the program because OCaml infers types for you.
However, this does not match well with operator overloading, since the
compiler would not understand whether "a + b" stands for float or
integer.  Unlike SML, OCaml decided to avoid this kind of type
inferring problem by not overloading the + operator.

<http://pauillac.inria.fr/caml/caml-list/1579.html>
<http://rum.cs.yale.edu/cs421/case-for-ml.html>

T.


From jb at cascade-sys.com  Tue Sep 17 04:14:30 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 17 Sep 2002 01:14:30 -0700
Subject: More on Protecting Source Code
References: <GCEDKONBLEFPPADDJCOEAEDDFAAA.whisper@oz.net> <20020917062412.GA85457@hishome.net>
Message-ID: <3D86E466.6010208@cascade-sys.com>

Oren Tirosh wrote:

>Yes, there are narrow-minded people on both sides of the intellectual
>property debate.
>

Reasonable statement.

>I don't think that software 'ought' to be this or 'ought' to be that. In
>other words, I believe in freedom.
>

Reasonable statement.

>I don't think anyone would disagree with that. 
>

Reasonable statement.

>The disagreement seems to 
>be whether information is property or not. 
>

Wow.  I didn't hear anybody say this (though, sadly, perhaps this is the 
real agenda).

>By its nature it isn't 
>

Wow.  An even more amazing statement.

Information BY IT'S VERY NATURE is not property.

How the hell do you figure?

You saying the thoughts in my head are public domain? I certainly hope not.

Or if I reduce those thoughts to paper or to software or to wood and 
metal that somehow everybody else is entitled to share the benefits of 
my possibly substantial endeavor?  

What a load of bullshit.

An artist who creates an image or the specifications for a musical 
performance has every right to own his creation and benefit from its 
use.  This is no different for a programmer and code of his own creation.  

Even if you go back to the early industrial age, key products of 
industry was intellectual property.  E.g., engineers had to produce 
drawings and written specifications before they could build a bridge or 
a mousetrap.  That "information" was every bit their property as the 
more tangible artifacts they later sold commercially.  An unscrupulous 
competitor could use the IP to recreate the commercial artifacts at a 
fraction of the original owner's cost.  Or he could reverse engineer the 
IP and again enjoy an unfair  advantage.  That's blatantly unfair and 
it's a good thing we have laws to protect against such theft.

Now if somebody elects to give away their IP -- out of charity or in 
hope of scoring points in some circles -- that's fine.  But if he 
instead elects to profit from it, in a free world, that should be 
perfectly reasonable and respectable also.

>but it
>can be made more property-like by laws that restrict people's freedom to
>use information. 
>

Copyright laws have precedent dating back to pre colonial England. 
 Interesting new books and music were intrinsically hard to create but, 
(after Gutenberg) once created, they were easy for the unscrupulous to 
copy and receive full benefit of their publication.  Robbing the 
creators the fruits of of their labor is inherently unfair.  Not only 
does the IP thief unfairly benefit from work he did not perform he also 
robs the real owner of income he otherwise would have received.  

Thus IP laws naturally evolved to protect rightful property owners. 
 It's meet and good.

People fought and died (and more importantly prevailed) in defense of 
the rights to be free and to own property.

Now I don't necessarily support some of the more recent changes and 
extensions to copyright laws.  Some maybe go a step too far.  However, I 
don't think they can or should be rejected altogether.  At their core 
they're not arbitrary; rather in protecting the works' creators they're 
the epitome of fairness.


>The state has given you the 'right' to use the power of its court and 
>police to forcibly restrict people that use information you have created 
>in violation of copy-restriction laws.
>

Yeah, sure, in your dreams we could instead have laws that say all 
information was free.  I submit that such a situation would actually 
stifle innovation.  Sure a few dreamers would still invent but the 
volume and time to market would be greatly diminished.  

Lack of ownership is precisely why most socialist countries soon became 
and forever remained vast economic wastelands.  Denying individuals any 
incentive for working harder than others, the economy soon deteriorates 
to where it can hardly provide the necessities of life let alone invent 
clever and innovative luxuries.  Worse, collectivist economics are 
unnatural in that they fly in the face of most human instincts.  Thus it 
requires a large and oppressive government to enforce collectivism and, 
in practice, those governments typically become corrupt.  In any case, 
economies invariably, immediately revert to capitalism soon as the 
communist/socialist oppression ceases.

It is precisely the promise of commercial reward that makes America the 
incredible fount of invention and creativity that it is.  The simple 
formula of rewarding people for their own hard work is what made America 
the worlds only superpower.  Not only do we have the most powerful 
military in the world our people also enjoy one of the world's highest 
standards of living.  Most socialist countries have had to choose one or 
the other or don't have the choice of either.

I just don't get you folks who push this Communistic view of economics. 
 It's childish.  Marx sounded good on paper 100 years ago and yeah we 
all believed it when we were idealistic teenagers. At the cost of 
hundreds of millions of citizens (murdered by their own government) and 
a significant fraction of the 20th century's world's human resources, we 
gave Communism every chance to succeed.  After almost a century of 
experimentation with collectivist economics, the results are all in and 
the bottom line is that it simply doesn't work.  The net cost to The 
People in the long run is always greater than the net benefits.  If you 
don't understand this then you really haven't been paying attention to 
recent world history.

The problem with Capitalism is that not everyone shares equally in the 
wealth.  The problem with Communism is that everybody DOES share equally 
in the misery.  I prefer Capitalism.


>But that is not enough for you. You also want the 'right' to complain about
>something that was given to you freely and expect that righteous rants 
>about your view of intellectual property laws will somehow compel people 
>to modify it to your specification without compensation.
>

You're being unfair.

Mr. LeBlanc was not complaining about anything that I could see.  He was 
merely describing a tangible shortcoming of Python (one which launched 
this entire thread) that restricts its use for some purposes.  He 
further, humbly suggests how it might be made a little more useful to 
commercial users.

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From teaandbikkie at aol.com  Sat Sep  7 09:52:11 2002
From: teaandbikkie at aol.com (TeaAndBikkie)
Date: 07 Sep 2002 13:52:11 GMT
Subject: Thinking of joining
References: <3D7970DC.5115BCC8@earthlink.net>
Message-ID: <20020907095211.24938.00000006@mb-mt.aol.com>

IMHO Python has beautiful plumage, isn't nailed to the perch, and won't leave
you pining for the fields. It also has a good range of cheese, even the ones
you thought you didn't need, and you won't get stoned by your friends for
joining.

(Any implied references to Parrot are completely accidental).

I'm too new to Python to give you good technical reasons to use it, but I like
it :)



From duncan at NOSPAMrcp.co.uk  Wed Sep 11 12:03:59 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Wed, 11 Sep 2002 16:03:59 +0000 (UTC)
Subject: deleting the first and the last character of a string
References: <20020911105818.GC2054@partagas.as.de.cw.net> <mailman.1031756234.22942.python-list@python.org>
Message-ID: <Xns9286ACD30F92Eduncanrcpcouk@127.0.0.1>

Sean 'Shaleh' Perry <shalehperry at attbi.com> wrote in 
news:mailman.1031756234.22942.python-list at python.org:

> On Wednesday 11 September 2002 03:58, Manuel Hendel wrote:
>> I got a file with fields seperated by a "|". The problem is that there
>> is also a "|" at the beginning and at the end of each line although
>> there no fields in front or behind these "|". Now I want to delete
>> these two "|". Is there anything which says delete the first or last
>> character of a string? I checked the string module on the web, but I
>> couldn't find anything.
>>
> 
>>>> s = '|f|o|o|'
>>>> s.strip('|')
> 'f|o|o'
> 
> and of course the equivalent strip call from the string module.
> 
This doesn't work on Python 2.2.1 (or earlier), I think you need to be 
sitting astride the cutting edge of CVS development to do this.

Specifically the change to implement this was April 22nd and 2.2.1 was 
released April 10th.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From tismer at tismer.com  Sun Sep 15 15:50:06 2002
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 15 Sep 2002 21:50:06 +0200
Subject: Ann: Stackless Python Update
Message-ID: <3D84E46E.4020704@tismer.com>

Hi Stackless users,

A new version of Stackless Python is available.

Changes:
Stackless now supports Sun Sparc, thanks to Gerd Woetzel.
The other platforms are
Windows, Linux X86, Mac PPC, Cygwin.

The core stack switching mechanism and the extended
stackless module are now split apart. The stackless
module is built as an extension and loaded on demand.

The type system has been reworked. As an effect, all
methods on channels and tasklets can now be re-defined
from Python, without extra cost for the native
implementation. I am going to propose this as a
Python extension as well.

Win32 binaries can be found on the website
http://www.stackless.com/

For all other supported platforms you need to get the
source from CVS and build it.

More to come, soon - chris
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From manuel at hendel.net  Thu Sep 19 03:47:26 2002
From: manuel at hendel.net (Manuel Hendel)
Date: Thu, 19 Sep 2002 09:47:26 +0200
Subject: lists in dictonaries ?
Message-ID: <20020919074722.GB77021@partagas.as.de.cw.net>

Is it possible to put lists in dictionaries? And is it also possible
to put multiple lists in a dictionary for one key?

Thanks,
    Manuel

-- 
I married beneath me. All women do. 
-Lady Nancy Astor, first woman member of the British Parliment  



From tismer at tismer.com  Mon Sep  2 17:11:45 2002
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 02 Sep 2002 23:11:45 +0200
Subject: New Stackless Beta
Message-ID: <3D73D411.80205@tismer.com>

Hi friends,

a new beta of Stackless Python is released.
The source is at the moment available via
public CVS, only.
PC binaries and how to access CVS can be found at

http://www.stackless.com

What's new?
Tasklet flags have been re-worked a lot. They should be
always updated correctly, now.
The schedule() protocol has changed very much, internally.
This made it possible to write a tasklet.raise_exception
which works in any context, while waiting in a channel
or being currently run.

Very soon to come: tasklet.kill(), implemented by raise_exception.
But there is one feature missing for a clean kill():
Tasklets will get notifiers, which make it possible to
ask for an event like "tell me when this tasklet has died".
Together with notifiers, waiting on a set of channels will
be implemented, soon, in a very effective manner.
Also in preparation:
"A practical guide to Stackless Python", written by me.
It will contain general information and a number of examples,
like "how to make wxPython stackless" and more.

watch out - ciao - chris
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From jcc at ibdosnorte.com  Mon Sep 23 11:00:15 2002
From: jcc at ibdosnorte.com (=?ISO-8859-1?Q?Juan_Carlos_CORU=D1A?=)
Date: 23 Sep 2002 08:00:15 -0700
Subject: COM: Returning IDispatch or NULL
Message-ID: <a230569c.0209230700.14116a1d@posting.google.com>

Hello,

I'm trying to develop a COM Server in python, that has a method that
returns an IDispatch object but sometimes it must return nothing. See
the following example programmed in C/SIDE (Microsoft Navision
Financials ERP):

--->
otherCOMServer := COMServer.returnOtherCOMServer();
IF ISCLEAR(otherCOMServer) THEN MESSAGE('otherCOMServer is clear');
<----

Definition of ISCLEAR (extracted from MS Navision help):
  The return value is TRUE if:
    -the automation variable has not been created. See CREATE.
    -the automation variable has been cleared. See CLEAR.
    -the automation variable has been assigned NULL by an automation
server.

I thing that in this case I must assign NULL to the otherCOMServer
variable in order to keep the otherCOMServer variable CLEARED.
What must I return from "COMServer.returnOtherCOMServer()" method to
assign NULL?
I have tried:
  return None
  return wrap(None)
  return 1==0
  return

but without success. The environment responds "data type not
supported".

I know that other COM servers developed by Microsoft Navision works
this way.

Thanks in advance.


From nospam_t.roessler at v-d.com  Thu Sep 26 09:19:35 2002
From: nospam_t.roessler at v-d.com (Thilo =?ISO-8859-15?Q?R=F6=DFler?=)
Date: Thu, 26 Sep 2002 15:19:35 +0200
Subject: Zope, MySQL and pictures in blob-fields ...
References: <amuqck$e0r$1@nathan.regio.net> <mailman.1033040674.25353.python-list@python.org>
Message-ID: <amv1in$ks9$1@nathan.regio.net>

I'm sorry ...

> This is crossposting. Please subscribe to the Zope maillist at
> zope at zope.org and post this there.



From sholden at holdenweb.com  Wed Sep 11 04:56:18 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 11 Sep 2002 04:56:18 -0400
Subject: blocking socket dilemma
References: <3d7f046f$0$230$4d4ebb8e@news.nl.uu.net>
Message-ID: <lADf9.147720$l_4.122435@atlpnn01.usenetserver.com>

"Gumuz" <gumuz at looze.net> wrote ...
> Hi all,
>
> I am a little bit stuck with my design (again!).
>
> I have this server class, which creates a session object for every
accepted
> incoming connection. The session object has a RequestHandler function
which
> is then run in a new thread.
>
> all fine, but...
>
> This RequestHandler function actually does a infinite loop. The loop wil
> receive data and take some action and it will send data which is piled up
in
> this session's queue. The problem that i'm having is that the loop will
hang
> on socket.recv if the client is not sending anything. Ofcourse this is
very
> obvious and logical, but i can't think of another way of constructing my
> design, so that this problem doesn't occur.
>
> I hope i've made my problem clear to you guys.
>

You may be ready to migrate to an asyncore-based solution. In such
situations it's usually easier to adopt a more flexible program structure.
Asyncore/asynchat my be just what you need. They certainly put a muc more
direct emphasis on the buffering required in a TCP session.

The SocketServer-based modules are sufficient for simple interactions, but
don't hack it when a "Request" becomes a series of interactions over the
same socket, for example like a Telnet session.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From jhorneman at pobox.com  Sun Sep 29 17:26:53 2002
From: jhorneman at pobox.com (Jurie Horneman)
Date: Sun, 29 Sep 2002 23:26:53 +0200
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com> <874rc8d9c9.fsf@nospam.eml.cc> <mailman.1033319977.7828.python-list@python.org>
Message-ID: <3d976fbc$0$23666$91cee783@newsreader02.highway.telekom.at>

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote in message
news:mailman.1033319977.7828.python-list at python.org...
> * Mahesh Padmanabhan <news at nospam.eml.cc> [2002-09-29 16:55 +0000]:
> > You need to create an empty file called __init__.py inside the
> > template directory and call the blah.py script as follows:
> >
> > import template.blah
>
> It's also highly recommended to start Python with "python -vv", then
> trying to import the module in question. This will show you where Python
> tries to find the module.

Actually, I just tried this in a DOS shell under Windows 98:

python -v script.py >output.txt

It lists the verbose output straight to the DOS shell, not to output.txt.
The output is useless, I can't scroll back.

Oddly,

python -h >output.txt

correctly sends the output to output.txt.

Does anybody know if this is a bug, and how I can work around it?

Jurie.





From drfransch at netscape.net  Sun Sep 15 18:53:43 2002
From: drfransch at netscape.net (F. Schaefer)
Date: 15 Sep 2002 15:53:43 -0700
Subject: Command Line Parsing etc. (GetPot 1.0)
Message-ID: <f8dfa927.0209151453.3f6b6fb@posting.google.com>

Finally after one year on  sourceforge, the Version 1.0 of the command
line  and configuration  parser  GetPot is  ready  for download.  Some
new features are:

-- the dollar bracket language:  A simple lisp-like extension, so that
   basic string,  and arithmetic operations can be  performed on the
   command line and inside a configuration file.

-- unidentified flying object detection (UFO detection): sophisticated
   functions  to  detect unrecognized  arguments,  options, flags  and
   variables.

-- prefixes:  narrow down  search  space through  a 'prefix'  (section
   name).

A reviewed user manual documents all these features.  Please feel free
to download at:

                        http://getpot.sourceforge.net

Many  people sent  me  suggestions  and improvements,  so  I think  we
finally  have a  pretty useful/stable  tool. However,  I'm  still glad
about any comment/critique.


Best Regards,

Frank.


From pecora at anvil.nrl.navy.mil  Mon Sep 30 09:26:01 2002
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Mon, 30 Sep 2002 09:26:01 -0400
Subject: [Mac] Editor on Macintosh?
References: <1fj7za2.1erbqps1uvzpj8N%news@agapow.net> <an7u1s$33de$1@nntp6.u.washington.edu>
Message-ID: <300920020926012729%pecora@anvil.nrl.navy.mil>

In article <an7u1s$33de$1 at nntp6.u.washington.edu>, Russell E. Owen
<somebody at nojunk.invalid> wrote:

> Python has the Mac IDE is a bit limited (need an extra plug-in even to 
> do syntax coloring)

What Plug-in works for the Python IDE?

-- 
Lou Pecora
  - My views are my own.


From emile at fenx.com  Tue Sep 17 09:34:33 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 17 Sep 2002 06:34:33 -0700
Subject: how to cast an instance id
References: <3D86E888.90AF569@millfilm.co.uk>
Message-ID: <am7au9$lgn$1@main.gmane.org>

Eric Texier:
> It there a way to create a new reference of  an object
> from the string representation of its id?
> Thanks for any help.
>
> class foo:
>     def info(self):
>         print self
>
> a = foo()
> a.info()
> ccc = "newa = 0x%x" % id(a)
> exec(ccc)
>

There are ways of walking the object hierarchy checking ids, eg:

ccc = id(a)
for ii in locals().values():
    if id(ii) == ccc:
        ccc = ii
        break

Or, you can set things up specifically to support this if you know in
advance:

class foo(object):
    objlist = {}
    def __init__(self):
        foo.objlist[id(self)] = self
    def info(self):
        print self

a = foo()
ccc = id(a)
ccc = foo.objlist[ccc]
print ccc

That said,  there's probably a better way, ie, setting ccc to a instead
of the id of a in the first place.  Otherwise, you risk retaining ccc
when a could have been refcounted away.

--

Emile van Sebille
emile at fenx.com

---------






From skip at pobox.com  Fri Sep 13 14:20:38 2002
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 13 Sep 2002 13:20:38 -0500
Subject: transform list of int in a list of string??
In-Reply-To: <-1483717605.1031940885180.JavaMail.nobody@webmail2.brturbo.com>
References: <-1483717605.1031940885180.JavaMail.nobody@webmail2.brturbo.com>
Message-ID: <15746.11382.851521.129722@12-248-11-90.client.attbi.com>

    juliano> How can i transform a list of integer numbers in a list of
    juliano> string numbers, like that: 

    juliano> x=[1,2,3,4] to y=['1', '2'...]????????

Try 

    y = map(str, x)

or

    y = [int(n) for n in x]

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From mstenner at phy.duke.edu  Thu Sep 12 21:33:53 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Thu, 12 Sep 2002 21:33:53 -0400
Subject: Coding Style: Quotes
In-Reply-To: <NZag9.2396$Lg2.449107@news2.news.adelphia.net>; from bobx@linuxmail.org on Fri, Sep 13, 2002 at 01:14:21AM +0000
References: <bt1g9.17429$Pf7.590136@news1.west.cox.net> <mailman.1031842455.20916.python-list@python.org> <NZag9.2396$Lg2.449107@news2.news.adelphia.net>
Message-ID: <20020912213353.A8007@phy.duke.edu>

On Fri, Sep 13, 2002 at 01:14:21AM +0000, Bob X wrote:
> Michael Stenner wrote:
> <snip>
> > Since there is no "physical" difference between them, I can't imagine
> > why it would matter.  The only reason there "needs" to be two is for
> > nested quoting.  Sure, it would be annoying to alternate for each
> > string, but beyond that...
> Don't you have to escape a single quote within a single quote whereas 
> you do not while using a double quote?
> 
> i.e. 'It\'s raining'  "It's raining"

Yes.  But they are identical in that the reverse is also true.  For
example:

"It\"s raining"  'It"s raining'

I know it's incorrect english grammar, but that's not really relevant
here.

This is in contrast to languages where the single and double quotes
have fundamentally different roles.  For example, in c, "s" and 's'
are radically different.  In perl, "$_" and '$_' are again very
different.
					-Michael

P.S.  As a matter of style, I'm sure most of us would agree that 
  "It's raining" is preferable to 'It\'s raining', just as
  'she said "wait"' is preferable to "she said \"what\"".

-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From duncan at NOSPAMrcp.co.uk  Wed Sep 11 04:13:24 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Wed, 11 Sep 2002 08:13:24 +0000 (UTC)
Subject: dynamism
References: <66715c02.0209091923.265fe77e@posting.google.com> <Xns92855D48384F9duncanrcpcouk@127.0.0.1> <3d7e209a_1@news.iprimus.com.au>
Message-ID: <Xns92865CE1C916Aduncanrcpcouk@127.0.0.1>

"Steven Shaw" <steven_shaw at iprimus.com.au> wrote in 
news:3d7e209a_1 at news.iprimus.com.au:

> Does this mean that the attributes are stored in a more efficient way?
> 
> Is there a way to implement "record-based" object layout rather than
> "associative-array based" object layout?
> 
> 
It very much depends what you mean by 'more efficient'. Using object-
derived classes and __slots__ means that each class no longer has the 
overhead of creating a dictionary for the instance attributes, so you are 
saving the memory overhead of the dictionary. On the other hand, object-
derived classes seem to be slower to create, and setting attributes defined 
in object-derived classes is also slower than setting attributes in old-
style classes (although setting __slots__ attributes seems to be faster 
than setting other attributes).

The code below gives these timings on my machine:
3.84s: 1000000 x New()
3.89s: 1000000 x New2()
2.39s: 1000000 x Old()
5.19s: 10000 x attributes(<class '__main__.New'>)
5.81s: 10000 x attributes(<class '__main__.New2'>)
4.17s: 10000 x attributes(__main__.Old)
--------------------------------
class Old: pass

class New(object):
    __slots__ = ['a', 'b', 'c', 'd']

# New-style, but not using slots.
class New2(object): pass

def timeit(repeat, fn, *args, **kw):
    import time
    start = time.time()
    for i in xrange(repeat):
        fn(*args, **kw)
    end = time.time()
    # Format arguments for printout
    argstring = ', '.join(
        [str(a) for a in args] +
        [ '%s=%s' % (k, v) for k, v in kw.items()])
    if len(argstring) > 60:
        argstring = argstring[:57] + '...'
    print "%.2fs: %d x %s(%s)" % ((end-start), repeat, fn.__name__, 
argstring)

def attributes(klass):
    inst = klass()
    inst.a = 1
    for i in range(100):
        inst.b = inst.a+1
        inst.c = inst.a+2
        inst.d = inst.a+4
        inst.a += 5


timeit(1000000, New)
timeit(1000000, New2)
timeit(1000000, Old)
timeit(10000, attributes, New)
timeit(10000, attributes, New2)
timeit(10000, attributes, Old)

--------------------------------
-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From aurin42 at yahoo.com  Tue Sep  3 04:26:54 2002
From: aurin42 at yahoo.com (Ante Bagaric)
Date: Tue, 3 Sep 2002 10:26:54 +0200
Subject: A "for" with "list" question.
References: <3b55ea60.0208311330.4608ae71@posting.google.com> <pan.2002.08.31.21.50.09.767734@effectif.com>
Message-ID: <al1rrj$66sv$1@as201.hinet.hr>

> >>> [item for item in a if item in b]
> [1, 2]
>
> That third line is a list comprehension.

>>> a = [0, 1, 2]
>>> b = [1, 2, 3]
>>> [i for i in a if i in b] + [i for i in a + b if i not in a or i not in
b]
[1, 2, 0, 3]

:)






From a_salnikov at yahoo.com  Sat Sep 14 16:33:58 2002
From: a_salnikov at yahoo.com (Andy Salnikov)
Date: Sat, 14 Sep 2002 13:33:58 -0700
Subject: Anyone can tell me why?
References: <1815200.1031973626@dbforums.com>
Message-ID: <am06iq$p32$1@usenet.Stanford.EDU>



"hpyhpy" <member at dbforums.com> wrote in message
news:1815200.1031973626 at dbforums.com...
>
> hi,everybody:
>
> I'm puzzled about SWIG.
>
> [..........................]
>
> Then it has these files:
> example.c
> example.i
> example.o
> example_wrap.c
> example_wrap.doc
> example_wrap.o
> _examplecmodule.so
>
> So I wonder where is the "example.py"???
> So I can't "import example"
>
  The module 'example' is in the shared library (.so). But python
has to find this shared lib somehow, I mean you cannot give it
some random name and hope it will be found. Try to rename
_examplecmodule.so to example.so, import should work then.

  Andy.






From igetenoughspamalreadythanksjlr at doctor.com  Mon Sep  2 07:10:39 2002
From: igetenoughspamalreadythanksjlr at doctor.com (John La Rooy)
Date: Mon, 02 Sep 2002 21:10:39 +1000
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <un50ccotaak86b@corp.supernews.com> <al0e96$oms$0@216.39.172.122>
Message-ID: <1031008206.347446@ampungk.ozonline.com.au>

Bengt Richter wrote:

> 
> I agree about new syntax, but I wouldn't mind having a re.help(regexp) function
> for interactive use that would just explain in 'English' what the regexp expression
> stands for. It would be a nice easy double check on whether I wrote what I meant,
> and helpful for understanding someone else's magic. It shouldn't be that hard to do.
> 
> Regards,
> Bengt Richter

I think that's an excellent idea (probably has already been done, 
anybody know?)

It is unfortunate that often a re that starts off as a simple idea
turns out to be a big mess. I think some sort of graphical tool
would be even better so you could see what the re does. Especially
if it could convert both ways and allow you to edit the re.

John






From anguenotj at yahoo.fr  Thu Sep 19 03:26:23 2002
From: anguenotj at yahoo.fr (Julien Anguenot)
Date: Thu, 19 Sep 2002 10:26:23 +0300
Subject: Code analyser + testing
Message-ID: <3D897C1F.3070405@yahoo.fr>

Hi !,

I'm looking for a code analyser for Python 2.2.
I found one dated from 1997 (pylint.py) from Ka-Ping Yee but It seems it 
  doesn't work with the version 2.2....

Knows someone where I can find one up to date ?

I'm student and I have to find a software (in Python of course) for 
setting tests. (unit testing, etc....)
It's for a course I took this year. I will do it with an other developper.

If you've a software and you're interested to have some testcases you 
can write me an email.

What I need with the software :

   ** the software of course
   ** It must be long around 5000 loc
   ** the specifications in details

Thanks

Regards,

/JA

-- 
Julien ANGUENOT Keskisenkatu 6 B 11 33710 Tampere FINLAND
Mail : anguenotj at yahoo.fr ; Web  : http://www.anguenot.net
Tel  : +358 (0) 41 576 8888

Download and try Zope at http://www.zope.org.
Zope is a leading open source application server.




From self at meyarivan.org  Wed Sep  4 10:45:42 2002
From: self at meyarivan.org (T.Meyarivan)
Date: 04 Sep 2002 20:15:42 +0530
Subject: efficient list merging
In-Reply-To: <ced73313.0209040601.64edac88@posting.google.com>
References: <ced73313.0209031012.13624f3@posting.google.com>
	<al3bfi$2kh$0@216.39.172.122> 
	<ced73313.0209040601.64edac88@posting.google.com>
Message-ID: <1031150751.2615.40.camel@tass1.localdomain>

> Here's an example:
> 
> lol1 = [['a'], ['a', 'b', 'c'], ['d', 'e', 'f'], ['g', 'h', 'i', 'j']]
> lol2 = [['d', 'e', 'f'], ['k', 'l', 'm'], ['d'], ['f', 'g'], ['a']]
> 
> I want a function which returns
> 
> mergelol = [['a'], ['a', 'b', 'c'], ['d', 'e', 'f'], ['g', 'h', 'i',
> 'j'], ['k', 'l', 'm'], ['d']]

# lista and listb are the lists of lists

def merge(lista, listb):
    x  = {}
    map(x.setdefault, map(tuple, lista + listb))

    return x.keys()





From see_reply_address at something.invalid  Tue Sep 24 23:19:30 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Wed, 25 Sep 2002 15:19:30 +1200
Subject: confusion about opening files
References: <mailman.1032838112.4941.python-list@python.org> <pFUj9.156068$pX1.5631353@news2.tin.it>
Message-ID: <3D912B42.8010703@something.invalid>

> Flush, and unbuffered files, ASK the OS to please, please write the
> data out -- but not all OS's comply.


Even on Unix, I'm not sure that flush() ensures that
the data is physically written to the disk. The Linux
man page for fflush(3) says:

NOTES
        Note  that fflush only flushes the user space buffers pro?
        vided by the C library.  To ensure that the data is physi?
        cally  stored  on  disk the kernel buffers must be flushed
        too, e.g. with sync(2) or fsync(2).

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From tim.one at comcast.net  Thu Sep 26 11:39:04 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 26 Sep 2002 11:39:04 -0400
Subject: Python threading?
In-Reply-To: <ptv1kkuh.fsf@ion-tof.com>
Message-ID: <BIEJKCLHCIOIHAGOKOLHIEBNDLAA.tim.one@comcast.net>

[Robert Oschler]
> Having written _many_ state machines in my life, which async programming
> boils down to in one form or another, I feel it depends on the task.
> ...
> Contrast this with a simple procedure call where a linear series of
> function calls are made, each with appropriate error handling and status
> adjustment, in a pleasant centrally located manner.

[Thomas Heller]
> I wonder if generators can come to a rescue here?

Absolutely.  Read the start of PEP 255.

> Still trying to get them into my head...

Think of them as resumable functions.  You're familiar with functions that
remember their own data state across calls, via abusing globals, or via
declaring private static vrbls, in C.  Methods on objects often do the same
by stuffing their state into instance attributes between calls.  A generator
remembers its data state across resumptions simply by virtue of that its
local variables don't vanish, and also remembers its control-flow state
across resumptions.  Remembering local data state by magic is a real
convenience, but can be simulated in many other ways with reasonable effort;
it's the ability to remember and restore control-flow state by magic that
can make generators more than merely convenient.




From Pekka.Pessi at nokia.com  Wed Sep 18 01:20:31 2002
From: Pekka.Pessi at nokia.com (Pekka Pessi)
Date: 18 Sep 2002 08:20:31 +0300
Subject: Detecting IPv6 support
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032245506.8580.python-list@python.org> <j4n0qhjbkf.fsf@informatik.hu-berlin.de> <pvsn08lw9k.fsf_-_@agni.research.nokia.com> <j43cs8ft6x.fsf@informatik.hu-berlin.de>
Message-ID: <pvhegnj2lc.fsf@agni.research.nokia.com>

loewis at informatik.hu-berlin.de (Martin v. L?wis) writes:
>At the moment, you need CVS Python, and VC.NET, to get IPv6 suport on
>Windows. This then works both with the released XP, and the MS
>research IPv6 stack.

	Thanks.

>> 	What is the recommended method of finding out if Python and
>> 	the underlying OS support IPv6? I can, for instance, create an
>> 	IPv6 socket with a Python version that does *not* support
>> 	IPv6.

>What system is this? If you can create an IPv6 socket, why does it not
>support IPv6 (i.e. how do you know it does not support IPv6)? If you
>do .listen(), .getsockname() on that socket, what does it tell you?

	I have Redhat 7.3. The prepackaged RPMs have disabled IPv6, but
	the underlying OS supports it. So, this is what happens with the
	bind() system call, for instance:

Python 2.2.1 (#1, Jul 15 2002, 10:55:56) 
[GCC 2.96 20000731 (Red Hat Linux 7.3 2.96-110)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import socket as so
>>> s = so.socket(so.AF_INET6, so.SOCK_DGRAM)
>>> s.bind(("", 0))
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
socket.error: getsockaddrarg: bad family
>>> 

	So AF_INET6 is there, and I can create a socket, but everything
	dealing with addresses fails. 

						Pekka


From gerhard.haering at gmx.de  Fri Sep 27 16:31:19 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Fri, 27 Sep 2002 22:31:19 +0200
Subject: Better terminal for windows?
In-Reply-To: <3D94B037.2020202@motorola.com>
References: <3D94B037.2020202@motorola.com>
Message-ID: <20020927203119.GA6919@lilith.ghaering.test>

* Stephen Boulet <stephen.boulet at motorola.com> [2002-09-27 14:23 -0500]:
> I'm using python 2.2.1 on winnt.
> 
> If I want to run python interactively (and perhaps using ipython), is 
> there something better than the "cmd" terminal?

Well, you could install MSYS from the mingw project, which is sort of a
Cygwin-lite :-)

> I'm used to bash on linux at home, but I'd like to avoid installing 
> cigwin for now. Thanks.

cmd.exe isn't *that* bad once you activate tab completion in the
registry. Have you done that?

4NT is also great, but it's shareware.

-- Gerhard



From bobx at linuxmail.org  Tue Sep 17 19:47:28 2002
From: bobx at linuxmail.org (Bob X)
Date: Tue, 17 Sep 2002 23:47:28 GMT
Subject: Python equivalent to perldoc?
References: <Olvh9.7898$Lg2.1389903@news2.news.adelphia.net> <%Avh9.7933$Lg2.1392332@news2.news.adelphia.net> <am7sna$3emja$2@ID-77524.news.dfncis.de>
Message-ID: <kaPh9.8886$Lg2.1572004@news2.news.adelphia.net>

"Thorsten Kampe" <thorsten at thorstenkampe.de> wrote in message
news:am7sna$3emja$2 at ID-77524.news.dfncis.de...
<snip>
> Tough
Yeah aint it though...


> > Shouldn't Python know its own libs?
>
> It does:
> ,---
> | Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
> | Type "help", "copyright", "credits" or "license" for more information.
> | >>> import pydoc
> | >>>
> `---
It doesn't...you are executing from with Python...I am trying to do a ViM
script that executes it.

> > Or at least put the "libs" in the ENV variable?
>
> Windows XP doesn't have a ENV variable.

Lets see...right click "My Computer"...go to the "Advanced" tab....OH LOOKY
a button that says "Environment Variables". I could have been a wee bit more
specific like...putting the the lib dir in the PATH.

sorry-xp-does-have-an-env-variably-yours,

Bpb




From marklists at mceahern.com  Mon Sep 16 10:34:14 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 16 Sep 2002 09:34:14 -0500
Subject: getting username from uid
In-Reply-To: <pan.2002.09.16.10.05.30.511334.24338@presidency.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOEEPEDOAA.marklists@mceahern.com>

>   when I do a stat on a file I can get the UID of the owner. IS there
> anyway to get from UID to username (or directly to username for that
> matter)?

import pwd
user = pwd.getpwuid(uid)
# user is a tuple
print user[0]

// m
-




From daniel.dittmar at sap.com  Thu Sep 12 09:57:28 2002
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Thu, 12 Sep 2002 15:57:28 +0200
Subject: DB Cursor iterator (was: [ANN] PySQLite 0.3.0 released)
References: <slrnao0e6p.23c.gerhard.haering@haering.opus-gmbh.net> <cL0g9.335846$On.14066007@bin3.nnrp.aus1.giganews.com>
Message-ID: <alq6g8$8eb$1@news1.wdf.sap-ag.de>

Terry Reedy wrote:
> Does anyone know whether there are any plans to update the Python-DB
> API to the new iterator protocols, so that the above could be written
> more simply something like
>
> for row in cursor:
>        print "%14s, %15s, %19s, %8s, %25s" % tuple(row)

It's already in the 'Optional DB API Extensions' in
http://www.python.org/peps/pep-0249.html

Daniel





From gupsmail at ihatespam.shaw.ca  Wed Sep 25 10:30:50 2002
From: gupsmail at ihatespam.shaw.ca (Grant Hallam)
Date: Wed, 25 Sep 2002 14:30:50 GMT
Subject: problem with dictionaries within lists
References: <pan.2002.09.25.05.21.07.477231.167@ihatespam.shaw.ca> <S4fk9.163219$pX1.5840444@news2.tin.it>
Message-ID: <pan.2002.09.25.14.28.41.435139.181@ihatespam.shaw.ca>

The problem with blanking out the dictionary with dict{} is that it wipes
the keys out.  As the loop goes through it assigns the matching second
half of the list to a matching key.  If you insert a print statement on
the dict you can see the values in the dict are changing.  Thus while it
is the same dict the key values within the dict are different.  Thats why
I'm somewhat confused as to why its the same dic value.  As for the tabs,
I'll have the editor replace them with 4 spaces.  Didn't notice the
problem in PAN :)

On Wed, 25 Sep 2002 02:11:14 -0700, Alex Martelli wrote:

> Grant Hallam wrote:
>         ...
>>  However, the list seems to get populated with the same dictionary
>         ...
>> webdatabaselist.append(dict)
> 
> Of course -- you ARE appending the same dict over ands over (AFAICT --
> you're probably using tabs for indentation, making your code hard to
> comprehend for those who read news with KDE's KNode [or MS's OE] as tabs
> don't show there... use spaces, NOT tabs...).
> 
> Anyway, you probably need something like:
> 
> webdatabaselist.append(dict.copy())
> dict = {}
> 
> either might suffice, but generally you'll need to copy the current dict
> AND make dict empty for the next leg of the loop, if I get the general
> drift.  If you do rebind name dict to {} for the next leg, then you
> don't need the call to the copy method.
> 
> 
> Alex


From wayne at tbnets.com  Sun Sep  8 01:47:12 2002
From: wayne at tbnets.com (Wayne R)
Date: Sun, 8 Sep 2002 01:47:12 -0400
Subject: python program to watch a file - example.
Message-ID: <003801c256fb$3fdcf320$02010a0a@hackncodexp>

 
I am trying to write a python program that will run as a deamon and
watch a file for data to arrive and then wake up and perform duties on
that data.   I found ways to make a python program run as a deamon but I
am at a loss to find anything related to watching a file for data so
that it can then process the data.  I didn't really want to do a sleep
loop or something like, I am sure there is a better way.
 
Any ideas or links to more info?
 
Wayne
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020908/9a42a1ff/attachment.html>

From tmohr at s.netic.de  Sun Sep 15 14:31:24 2002
From: tmohr at s.netic.de (Torsten Mohr)
Date: Sun, 15 Sep 2002 20:31:24 +0200
Subject: Tkinter in PyRun_SimpleString(char* s)
References: <am12be$4j0$1@schleim.qwe.de> <_OWg9.109539$pX1.3874894@news2.tin.it> <am2alp$3v3$1@schleim.qwe.de> <A%2h9.98014$ub2.1982070@news1.tin.it>
Message-ID: <am2jls$doo$1@schleim.qwe.de>

Hi,

> ...which needs to be called AFTER Py_Initialize -- I think that
> may be your problem:

yes, exactly.  Thank you very much for that hint.


Regards,
Torsten.



From nmichael at yahoo.com  Mon Sep  9 14:19:07 2002
From: nmichael at yahoo.com (Neophytos Michael)
Date: 9 Sep 2002 11:19:07 -0700
Subject: Computing directory sizes problem
References: <d32d294c.0209061035.3e6fa400@posting.google.com> <j4fzwjbjza.fsf@informatik.hu-berlin.de>
Message-ID: <d32d294c.0209091019.320433fc@posting.google.com>

loewis at informatik.hu-berlin.de (Martin v. L?wis wrote in message news:<j4fzwjbjza.fsf at informatik.hu-berlin.de>...
> nmichael at yahoo.com (Neophytos Michael) writes:
> 
> > Dear Python experts,
> > I wrote the following function to compute the size of a directory
> > (being new to python there maybe better ways to do this - if so please
> > advice).  It works quite well except when the directory in question
> > contains files whose names that are not made up of ascii (english)
> > characters (this typically happens on my machine for my non-english
> > mp3 files).
> 
> What Python version are you using? What localized Windows version?
> What are the file names in question? 
> 
> If the file names are non-ASCII, it should still work - provided the
> file names are supported in the system locale (i.e. ANSI code page).
> 
> It might be easiest to adjust your system locale.
> 
> Regards,
> Martin

Dear Martin,
Python version is 2.2.1.  Windows localized version I am not sure. 
The laptop was purchased in the US so it must be the defaul US
version.  The filenames in question are mp3 files whose names are made
up from greek characters (at least I see them as greek when listing
the files with file explorer).  Any advice on how to adjust my system
locale would be appreciated.

Thanks,
Neophytos


From mark at mceahern.com  Mon Sep 30 15:38:43 2002
From: mark at mceahern.com (Mark McEahern)
Date: Mon, 30 Sep 2002 14:38:43 -0500
Subject: a question about mailman
Message-ID: <JHEOKEOOLIGLDHCMAHMOAEKGECAA.mark@mceahern.com>

I have a question about Mailman that was never answered when posting to the
Mailman list:

  http://mail.python.org/pipermail/mailman-users/2002-September/022538.html

I'm in the awkward position of having this question even though I don't
personally have Mailman installed (yet).

Since Mailman is written in Python, I thought I'd use the occasion of this
week's Dr. Dobb's Python-URL (DDPU) to illustrate the use case I envision
that is at the heart of my question about Mailman.

If you examine this week's DDPU (below), you'll see Cameron provides very
useful references to the Python mailist list when he references a thread
that happened on comp.lang.python.  (I've also seen contributors of DDPU
alternatively provide a link to the thread at http://groups.google.com/,
which is also very useful.)

What I imagine Cameron had to do was navigate to the archive page for the
Python mailing list and, by-hand, find the URL for each thread/message he
wanted to reference.

In that context, I wonder, "Wouldn't it be easier if you could configure
Mailman to insert the archive URL into the footer of outgoing messages?"

I took a peek at the Mailman code base and it seems like this could be done,
but it would require hacking the Dispatch (correct name?) module since when
it uses string formatting, it uses the list dictionary, not the message
dictionary to format the footers.  Of course, you want to do both:  format
with the list dictionary AND with the message dictionary.  But does it make
sense to combine the two into a temporary dictionary and use *that*?

So my questions are:

1.  Would this generally be useful--having the ability to put message
specific information in the footers?
2.  Would it specifically be useful to put the archive url into the footers?
(I personally think so.)
3.  Is the vague approach I sketched above the way you'd do it?  That is,
does it seem likely that it would work?

Of course, I realize it's highly likely this entire message will be ignored
since it's about Mailman, not Python per se.  C'est la vie.

Cheers,

// mark

-----Original Message-----
From: Cameron Laird [mailto:claird at lairds.com]On Behalf Of Cameron Laird
Sent: Monday, September 30, 2002 2:16 PM
To: mark at mceahern.com :
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Sep 30)


QOTW:  "One can always write very sparse or very dense lines of code that
break the model, but in-the-large programmers do not normally do that.
They tend to write code in lines that fit easily in their head and, because
of this, one line of code is pretty much comparable to another."
Jonathan Hogg

"Why are there always jobs for salesmen?"  Stephen Satchell


    Jack Diederich makes high-performance Permutations, Combinations,
    and other probability and statistics constructs available through
    SourceForge.
	http://probstat.sourceforge.net/

    Puffin makes automation nice.

http://www-106.ibm.com/developerworks/opensource/library/os-puffin2.html
    Pexpect fills in Puffin's most severe lack.
	http://sourceforge.net/projects/pexpect/

    Uche Ogbuji introdues Python's XML and XML-RPC capabilities, and PyXML.
	http://www.xml.com/pub/a/2002/09/18/py.html
	http://www-106.ibm.com/developerworks/library/ws-pyth10.html
	http://www.xml.com/pub/a/2002/09/25/py.html

    September brought another round of attention to CP4E in the EDU-SIG
group
        http://mail.python.org/pipermail/edu-sig/2002-September/thread.html

    Think in Tkinter.
	http://home.att.net/~stephen_ferg/thinking_in_tkinter/index.html

    Martin v. L?wis gives background to management of patches and
    contributions in the Python project.

http://mail.python.org/pipermail/python-list/2002-September/122988.html

    Blog with Python.
	http://www.awaretek.com/nowak/kaa.html


========================================================================

Everything you want is probably one or two clicks away in these pages:

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
         http://www.awaretek.com/nowak/mygale.html
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.

http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Brett Cannon continues the marvelous tradition established by
    Andrew Kuchling and Michael Hudson of summarizing action on the
    python-dev mailing list once every other week.
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/

    Much of Python's real work takes place on Special-Interest Group
    mailing lists
        http://www.python.org/sigs/

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
        http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.

http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
        http://python.sourceforge.net/peps/pep-0042.html

    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

    Links2Go is a new semi-automated link collection; it's impressive
    what AI can generate
        http://www.links2go.com/search?search=python

    Tenth International Python Conference
        http://www.python10.org

    Archive probing tricks of the trade:

http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100

http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.pytho
n.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/
  http://purl.org/thecliff/python/url.html (dormant)
or

http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang
.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning
(approximately), ask <claird at phaseit.net> to subscribe.  Mention
"Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-





From zopestoller at thomas-guettler.de  Tue Sep 10 09:34:55 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Tue, 10 Sep 2002 15:34:55 +0200
Subject: run-mailcap in python?
References: <alkcfb$9ql$07$1@news.t-online.com> <alkco4$j0k$06$1@news.t-online.com> <3d7de60b$1@news.sentex.net>
Message-ID: <alksap$ekd$02$1@news.t-online.com>

Peter Hansen schrieb:
> Thomas Guettler wrote:
> 
>> Thomas Guettler schrieb:
>>
>>> Hi!
>>>
>>> On Unix there is a script called run-mailcap, which
>>> opens the corresponding programm with the given file:
>>>
>>> run-mailcap foo.html
>>> --> Default browser starts
>>
>>
>>
>> [cut]
>>
>>> run("foo.html")
>>>  --> Browser starts
>>
>>
>>
>> I just discoverd os.startfile(), but it is unfortunately
>> only available on win32.
>>
>> A Unix version would be nice.
> 
> 
> import  os
> if os.name == 'posix':
>     def startfile(name):
>         os.system('run-mailcap "%s"' % name)
>     os.startfile = startfile

Yes, ugly but it works. Thank you!

  thomas



From gerhard.haering at gmx.de  Fri Sep 20 14:49:30 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 20 Sep 2002 18:49:30 GMT
Subject: saving state in Python/CGI
References: <mailman.1032544767.23458.python-list@python.org>
Message-ID: <slrnaomrdr.j7.gerhard.haering@lilith.my-fqdn.de>

Frank Gibbons wrote in comp.lang.python:
> Hi,
> 
> I've been using Python (and Jython) for about 5 months now. So far,
> so great. I'm building a CGI app that has about 5 different stages,
> and I need to maintain state between them.

CGI sucks. What else is new?

> I need access to the data on each page, yet don't want to pass it as
> a hidden field,

Many people, including me, use a relational database for saving state
in a CGI environment.

-- Gerhard


From richie at entrian.com  Mon Sep 16 07:35:36 2002
From: richie at entrian.com (Richie Hindle)
Date: 16 Sep 2002 04:35:36 -0700
Subject: html template system?
References: <btug9.688$Ee4.1090@news-server.bigpond.net.au>
Message-ID: <8a6ba1da.0209160335.2eb13353@posting.google.com>

> what are good templating systems avaliable for python?
> [...] i want a smaller footprint system.

PyMeld is one option.  It focusses on keeping the code in Python and
the HTML in HTML - it doesn't mix HTML and Python.  This means that
the HTML is editable in any HTML editor, potentially by a
non-programmer, and the programmer then manipulates the HTML using
pure Python code.  PyMeld wraps your HTML in a Pythonic object model,
with which you can change the text, add/remove/change tags and tag
attributes, clone pieces of HTML, change them and write them back in
again, and so on.

PyMeld is a single Python module of about 20KB - see
http://entrian.com/PyMeld for example usage, manual and download.

-- 
Richie Hindle
richie at entrian.com


From fperez528 at yahoo.com  Fri Sep 20 18:07:11 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Fri, 20 Sep 2002 16:07:11 -0600
Subject: e vs exp()?
References: <3a838998.0209191933.1f22e977@posting.google.com> <amflhl$3eb$1@peabody.colorado.edu> <FQMi9.171268$AR1.7178378@bin2.nnrp.aus1.giganews.com>
Message-ID: <amg66f$doo$1@peabody.colorado.edu>

Terry Reedy wrote:

> Perhaps x**int is special-cased.  But following must have a different
> explanation:
> 
>>>> e**2.1
> 8.1661699125676499
>>>> pow(e,2.1)
> 8.1661699125676499
>>>> exp(2.1)
> 8.1661699125676517
> 
> At least two of us expected x**y == pow(x,y) == exp(y*log(x)) (a
> standard method, and the only one I could find in Abrahamowitz and
> Stegun, Handbook of Mathematical Functions) but there must be a more
> direct method being used at least in some C libraries now.

Well, as Tim Peters already mentioned, implementing ** (or pow(), same thing) 
is a hairy problem and _not_ done through exp(log). On the other hand, exp() 
has its own separate implementation, even though it may or may not be as 
efficient as the one Tim mentioned with an error bound of .53 in the last 
digit, and fast to boot.

Incidentally, A&S is a fabulous reference for formulas and value tables. But I 
wouldn't consider it authoritative as a source of optimal numerical 
algorithms by any stretch of the imagination.

Cheers,

f.


From aahz at pythoncraft.com  Wed Sep 18 12:02:37 2002
From: aahz at pythoncraft.com (Aahz)
Date: 18 Sep 2002 12:02:37 -0400
Subject: monetary applications [was: Python GUI app to impress the boss?]
References: <mailman.1032348444.6783.python-list@python.org> <2R%h9.109893$ub2.2324479@news1.tin.it>
Message-ID: <ama82t$94c$1@panix1.panix.com>

In article <2R%h9.109893$ub2.2324479 at news1.tin.it>,
Alex Martelli  <aleax at aleax.it> wrote:
>Mark McEahern wrote:
>> [Oren Tirosh]
>>>
>>> I don't understand what exactly you are complaining about but is sounds
>>> interesting - can you elaborate?
>> 
>> Perhaps the lack of a FixedPoint type?
>
>http://starship.python.net/crew/aahz/FixedPoint.py

Note that FixedPoint is *not* BCD, particularly IEEE-854 BCD.  But it's
probably more than good enough for most applications, and it's extremely
trustworthy.  Right now, I've been stalled for many moons on my BCD
module.  But it's still on my list....
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From sperry at newcreations.info  Wed Sep 25 21:56:43 2002
From: sperry at newcreations.info (Sperry Russ)
Date: Wed, 25 Sep 2002 21:56:43 -0400
Subject: Python and Scite
Message-ID: <200209252156430267.0042D522@newcreations.info>

Chris,

Thanks for your help.  The problem was in the path  which I did not understand at first. Is th PythonWin your favorite editor?

Sperry


---------------------------------------------------------------------------------------------------------
> I am new to programming.  I have downloaded Python 2.2.1 and Scite
> 1.48 to learn how to program.  I am trying to configure my Scite so
> that it will run the python files by using go.  When I try to do this
> i get a message in the second window at the bottom -- unable to find
> file.  I have tried configuring the global options and have failed can
> anyone help me? I need a step by step example if you can because I
> cannot understand it according to the documentation that came with
> Scite 

works out of the box for me. if you're on windows, make sure that python is 
found in the PATH. i.e. edit PATH that it contains the python directory.
and you could look at PythonWin, it uses the editing component used in 
scite and has some additional features for the python programmer. 
additionaly it comes also with useful, de-facto standard, windows specific 
extensions.






From DennisR at dair.com  Thu Sep 19 19:17:04 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Thu, 19 Sep 2002 23:17:04 GMT
Subject: Suggesting a filename on a HTTP-request by sending someting in the header of the reply
References: <Pusi9.20169$0p1.363761@news2.ulv.nextra.no>
Message-ID: <PVsi9.2195$Tf1.75807051@newssvr14.news.prodigy.com>

This is really a comp.infosystems.www.authoring.cgi question, IMO.

My testing/experience is that content type does not control the "save as
..." dialog.  If you want your file saved as an EXE, you need to have .EXE
ending your file name.  So, now the question is "how can I execute a script
called "xyz.exe"?  While based on Perl and a different application, see my
site http://www.autorunning.com/docs/abfaq/02-dnldmgr.cgi

--

Dennis Reinhardt

http://www.dair.com




From phr-n2002b at NOSPAMnightsong.com  Sun Sep 29 21:15:35 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Sep 2002 18:15:35 -0700
Subject: idea for a much needed application
References: <576c1752.0209291326.56dcdf31@posting.google.com> <an87db$bjtvs$1@ID-157863.news.dfncis.de>
Message-ID: <7x1y7cl1l4.fsf@ruckus.brouhaha.com>

"mike henley" <rsrchstr at msn.com> writes:
> i guess i'll rephrase; what i wish for is an "application-to-person" (P2A)
> IM-based "presence", so that a web-application that i host on a personal
> webserver, on my own machine, can let others become aware of its "presence"
> easily and be easy to connect to with a simple
> click-to-launch-browser-and-go-to-my-ip. This should be easy to do. I just
> don't have experience with Jabber or xml-rpc, or other potential useful
> technologies.

I don't think you need anything nearly that fancy.  Just use some
simple AIM client on your box that runs when you connect to the internet.


From wari at home.wari.org  Sun Sep 15 12:46:10 2002
From: wari at home.wari.org (Wari Wahab)
Date: Mon, 16 Sep 2002 00:46:10 +0800
Subject: ANN: pyblosxom weblog software
Message-ID: <3D84B952.2060301@home.wari.org>

pyblosxom is a weblogging tool written in Python and modelled after blosxom. 
It is written due to the fact that clones for blosxom exists except for python,
so this is my implementation of it. In fact I'm into cloning so much that this 
announcement was ripped from a Kaa announcement, another weblogging tool written
in python.

The "download page" can be found here:

   http://roughingit.subtlehints.net/pyblosxom/

This is really the page of my pyblosxom blog; there are links to the 
distribution (a tar.gz file) and to a page with an introduction to pyblosxom.

Any comments, bug reports, tips, feature requests, etc, etc, are welcome. 
Particularly eloquent flames will be published on my blog. :-) Especially 
ones that mentioned how bad my python is. (I'm a newbie:)

--
Regards: Wari Wahab
http://roughingit.subtlehints.net/





From aleax at aleax.it  Tue Sep 17 02:52:08 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 17 Sep 2002 06:52:08 GMT
Subject: raw post data
References: <Wt8h9.9583$Ee4.22928@news-server.bigpond.net.au> <w_fh9.113093$pX1.4054147@news2.tin.it> <gOxh9.13698$Ee4.40115@news-server.bigpond.net.au>
Message-ID: <siAh9.118475$pX1.4269024@news2.tin.it>

:B nerdy wrote:

> how could i make it so when i read the stdin it still is there.. cause i
> use cgi.FormVariables() also

When you instantiate class cgi.FieldStorage (there's no such
thing as cgi.FormVariables, so I'm just guessing at what it IS
that you want!) you can pass it an optional first parameter
that is a file object open for reading -- it uses sys.stdin
as the default if you don't pass the first parameter.

Thus, the following should work (untested!):

import cgi, sys, cStringIO

copyInput = cStringIO.StringIO(sys.stdin.read())
fieldStorage = cgi.FieldStorage(copyInput)


The raw posted data are now available as the string returned
by calling copyInput.getvalue(), and are also parsed by cgi to
obtain fieldStorage.


Alex



From prouleau at impathnetworks.com  Thu Sep 12 10:41:15 2002
From: prouleau at impathnetworks.com (Pierre Rouleau)
Date: Thu, 12 Sep 2002 10:41:15 -0400
Subject: Is there a portable way to copy files in Python?
Message-ID: <3D80A78B.5030206@impathnetworks.com>

I have been using os.system() with the operating system 'cp' or copy 
commands when i want to copy files.  I can use os.mkdir() to make 
directories, but i can't find a os.filecopy() or os.path.filecopy() or 
os.copy() or something like that. I must be missing something, since 
such a basic operation is likely to be needed by Python library users.

I could write code that identifies the operating system (the os.path 
way) and use os.system("cp " + source_filename + " " + destination) for 
Unix-type and os.system("copy " + source_filename + " " + destination) 
but i would prefer to use a simple call.

Is this available in the Python library?  If not, why is it not part of 
the library?


Thanks for your time!

Pierre Rouleau



From martin at v.loewis.de  Thu Sep 19 14:57:27 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 19 Sep 2002 20:57:27 +0200
Subject: How pass array from  C  to python function
References: <mailman.1032442622.13517.python-list@python.org>
Message-ID: <m3bs6t959k.fsf@mira.informatik.hu-berlin.de>

"Praveen Patil" <praveen.patil at silver-software.com> writes:

>     mylist = PyList_New(5);
>     for (count=0; count< 5; count++) {
>           myint = PyInt_FromLong(vital_data[count]);
>           PyList_Append(mylist,myint); 
>     }

That is wrong. PyList_Append increases the list size by one, so the
list will have ten elements at the end; 5 of them will be NULL.

Either start with an empty list, or use PyList_SetItem.

>     arglist = Py_BuildValue("O", mylist);
>     ret = PyEval_CallObject(my_callback , arglist);
>     Py_DECREF(arglist);

I recommend to use PyObject_CallFunction here, although your code is
correct.

> The information contained in this e-mail is confidential

My reply is confidential, too. Don't read it.

Regards,
Martin


From eddie at holyrood.ed.ac.uk  Mon Sep  9 13:02:52 2002
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Mon, 9 Sep 2002 17:02:52 +0000 (UTC)
Subject: Python class ( Prolog ) ???
References: <20020905234600.11264.60328.Mailman@mail.python.org> <3D7914FD.5D713A48@anansispaceworks.com> <mailman.1031353570.17391.python-list@python.org>
Message-ID: <alik7s$gpt$1@scotsman.ed.ac.uk>

Terry Hancock <hancock at anansispaceworks.com> writes:


>While we're on the subject -- I've taken a couple of
>tries at Prolog in the past, but didn't really get
>anywhere with it.  I understand the concept of
>expressing logic in code, but got rather bogged down
>when it came to any practical examples.

I made several attempts to understand prolog many many moons ago and likewise
had difficulty relating it to practicality.  Eventually I came across the book
Clause and Effect by Clocksin which helped a lot.  Unfortunately playing with
prolog dropped so far down my 'interesting things to play with' stack about
that time that its now fallen over the event horizon.

Eddie


From martin at v.loewis.de  Mon Sep  2 01:46:58 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 02 Sep 2002 07:46:58 +0200
Subject: status of decompyle ?
References: <slrnammlcb.c05.Andreas.Leitgeb@pc7499.gud.siemens.at> <3D6B6A14.A6ADBFB4@engcorp.com> <slrnampvm8.c05.Andreas.Leitgeb@pc7499.gud.siemens.at>
Message-ID: <m38z2l7xkd.fsf@mira.informatik.hu-berlin.de>

Andreas.Leitgeb at siemens.at (Andreas Leitgeb) writes:

> Thus, my question to the group was just plain:
>   "Does anyone know about the current status of decompyle ?"

That wasn't actually the question you've asked. To that question, I'd
answer:

"it reportedly works fine for older Python releases"

However, you actually asked

>Should I remain patient a little more, or mail again,
>  or does the author read this group, anyway ?
>Or is decompyle in status "orphaned" ?

to which I'd answer

"yes, no, dunno, no".

Regards,
Martin



From max at alcyone.com  Thu Sep 19 17:37:13 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 19 Sep 2002 14:37:13 -0700
Subject: where is the __dict__ ?
References: <mailman.1032430104.28760.python-list@python.org>
Message-ID: <3D8A4389.DFCB97FF@alcyone.com>

black wrote:

>   my book said all dictionaries have __dict__ method but couldnt find
> it by using dir(), where is it  ?

Didn't you already ask this?  Dictionaries do not have a __dict__
attribute (they're dictionaries already, why would they need one?), but
instances and classes do:

>>> class C: pass
... 
>>> C.__dict__
{'__module__': '__main__', '__doc__': None}
>>> c = C()
>>> c.__dict__
{}

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ People are taught to be racists.
\__/ Jose Abad
    Fauxident / http://www.alcyone.com/pyos/fauxident/
 A "faux" ident daemon in Python


From dfloer at topsoft.ca  Mon Sep 16 19:54:52 2002
From: dfloer at topsoft.ca (Doug Floer)
Date: Mon, 16 Sep 2002 23:54:52 GMT
Subject: Unify Accell/SQL
Message-ID: <gbuh9.90$No6.36538@news1.telusplanet.net>

Has anyone out there had any experience converting an application from Unify
Accell/SQL to Python?

Thanks,
Doug




From =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?=  Tue Sep 24 11:23:58 2002
From: =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?= (=?iso-8859-5?Q?=B4=D0=DC=F8=D0=DD_=B3=2E?=)
Date: Tue, 24 Sep 2002 17:23:58 +0200
Subject: Difference between MOD_PYTHON and the CGI module
References: <3d904321$0$234$4d4ebb8e@news.nl.uu.net>
Message-ID: <3d908385@news.mt.net.mk>

> What are the differences, if any?

> Am i mixing up to things here?

> Suppose I'd like to use python as a 'replacement' for PHP, what do I use?

spyce perhaps



-- 
??????

  The NOC says:
        echo '16i[q]sa[ln0=aln100%Pln100/snlbx]sbA0D4D465452snlbxq'|dc



From david.abrahams at rcn.com  Mon Sep 30 20:15:38 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Mon, 30 Sep 2002 20:15:38 -0400
Subject: Is there a docutils / reStructuredText system for C++ source code available?
References: <3D98D9D9.8000009@attglobal.net>
Message-ID: <anarjq$6n3$2@bob.news.rcn.net>



"Pierre Rouleau" <pieroul at attglobal.net> wrote in message
news:3D98D9D9.8000009 at attglobal.net...
> I have been using docutil and reStructuredText for Python code and it
> works great. Pydoc does a superb job of helping in setting up an
> internal web site that provides documentation for our Python code, let
> alone the ability to run doctest test and show auto testing of python
> documentation on line!
>
> That works great for our Python code.  Now i want to do the same thing
> for our C++ code.  Since Python has been partly implemented in C / C++
> i tough there might already exist some tools for parsing C++ and some
> comments (given some formatting conventions) and generate the same kind
> of output the html tool of docutil does.
>
> Does any one know any thing that resembles that?

See http://synopsis.sf.net

--
-----------------------------------------------------------
           David Abrahams * Boost Consulting
dave at boost-consulting.com * http://www.boost-consulting.com




From richie at entrian.com  Thu Sep 26 10:28:24 2002
From: richie at entrian.com (Richie Hindle)
Date: 26 Sep 2002 07:28:24 -0700
Subject: SET_LINENO
References: <3D92A46E.2030706@cygnus-software.com>
Message-ID: <8a6ba1da.0209260628.3a97add8@posting.google.com>

Bruce,

> As a partial author of the HAP Python Debugger (available on source 
> forge) I read about the imminent removal of SET_LINENO with some worry. 

I've been working on a Python debugger of my own, and having briefly
tested it with 2.3a0, I'm nothing but delighted with the results.  It
fixes a number of minor niggles - stopping twice on block-start lines,
stopping on 'def' lines and stopping on docstrings, to pick some
random examples.

> wouldn't a break-point opcode be a more substantial improvement

This would be very nice too!

-- 
Richie Hindle
richie at entrian.com


From see_reply_address at something.invalid  Mon Sep  9 00:30:22 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Mon, 09 Sep 2002 16:30:22 +1200
Subject: Creating a Text widget in a Dialog  (Tkinter)
References: <b714b8de.0209081227.448fd2f2@posting.google.com>
Message-ID: <3D7C23DE.9000900@something.invalid>

Aki Niimura wrote:

>     def body(self, master):
>         self.usage = Text(master, height=26, width=80)
>         self.grid = Frame(master)
>         self.yscroll = Scrollbar(master, command=self.usage.yview)
>         self.xscroll = Scrollbar(master, orient=HORIZONTAL,


The Frame should be the parent of the other
three widgets:

          self.grid = Frame(master)
          self.usage = Text(self.grid, height=26, width=80)
          self.yscroll = Scrollbar(self.grid, command=self.usage.yview)
          self.xscroll = Scrollbar(self.grid, orient=HORIZONTAL,

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From bdesth at nospam.free.fr  Sat Sep 14 00:21:13 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Sat, 14 Sep 2002 00:21:13 -0400
Subject: I must be an idiot, please pity me!
References: <VAOf9.2572$Pa7.155732@newsfep1-gui.server.ntli.net> <3D801757.1020901@nospam.free.fr> <oYPf9.2894$XE6.174556@newsfep1-win.server.ntli.net>
Message-ID: <3D82B939.1060502@nospam.free.fr>

newbie wrote:
> On Thu, 12 Sep 2002 01:25:59 +0100, laotseu wrote:
> 
> 
> 
>>Creating a db connection in each function for sure not be very
>>efficient.But what's wrong with passing the database connection to the
>>functions in mod1, mod2, and modXXX ???
> 
> 
> That connection has to be passed to the other functions (where needed of
> course) in mod1. Say, mod1.functiona may use functions f1, f2 and f3. My
> gut feeling of having to pass that connection to all of these said "this
> can't be right". This is what I've done but it sure doesn't feel right!
> 
> 
> 
>>Anyway, you can make it simpler, like this :
> 
> 
> Yay, now for the good stuff!
> 
>  
> 
>>in each module,
>>- create a global db_connection variable - add an init() function, that
>>takes a db_connection as argument and assign it to the module's global
>>db_connection variable - and don't forget to call this function before
>>any other.
> 
> 
> Yes, I can follow that. Once I saw that "global" was more "local" I've
> never really considered using it.

??? Not clear what you mean ???

> 
> 
>>Another, more OO way to do this would be to rewrite your modules as
>>classes, each getting a db_connection argument in it's __init__(), and
>>then, in the main script, create the needed objects.
> 
> 
> So the resulting object would be the same as what I return with now?
> 

A simple exemple is better than a long explanation :

# this was your module mod3

import whateverYouHaveToImport

class Mod3:
     def __init_(self, dbConnection):
         self.dbConnection = dbConnection
         # add what you need here

     def f1(args):
         something = self.dbConnection.doThis(args)
         return self.f2(something)

     def f2(args):
        return self.dbConnection.doThat(args)


# this is you GCI module

import mod3
import TheRest

def main():
     dbConnection = dbModule.CreateMyDbConnection(args)
     m3 = Mod3(dbConnection)
     this = m3.f1(42)



Got it ?

HTH
Laotseu



From michael at stroeder.com  Sat Sep  7 11:23:00 2002
From: michael at stroeder.com (=?ISO-8859-1?Q?Michael_Str=F6der?=)
Date: Sat, 07 Sep 2002 17:23:00 +0200
Subject: None or 0
Message-ID: <k7ip4-0p3.ln1@junker.stroeder.com>

HI!

I'm using the idiom

a = a or b

to get rid of statements like

if a is None:
   a = b

Now I wonder what happens if both a and b have zero-length values.

Trying in Python 2.2.1 reveals:

 >>> repr(0 or None)
'None'
 >>> repr(None or 0)
'0'
 >>> repr('' or None)
'None'
 >>> repr(None or '')
"''"
 >>>

Is it guaranteed to work like this or should that be avoided?

Ciao, Michael.



From tjreedy at udel.edu  Thu Sep 26 17:47:00 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 26 Sep 2002 21:47:00 GMT
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <amltd3$3k$1@glue.ucr.edu> <3D911FE1.1050801@something.invalid> <amvcev$cf6$1@s1.read.news.oleane.net> <7nHk9.172208$pX1.6117207@news2.tin.it>
Message-ID: <ofLk9.304267$z91.13145351@bin3.nnrp.aus1.giganews.com>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:7nHk9.172208$pX1.6117207 at news2.tin.it...
> The reason
>         zz = a and b or c
> is not the same thing as
>         zz = a ? b : c
> is that, when a is true and b is false, the former returns c,
> while the latter returns b.

In almost all the practical cases where I have thought about using
this,  b is *not* false or cannot become false, in which important and
real case the two expressions *are* equivalent.

I think this is an elegant hack, and so I shared it in response to the
OP.  Others think it ugly.  So be it.

Terry J. Reedy




From robie at 1373.net  Mon Sep 23 23:17:21 2002
From: robie at 1373.net (robie1373)
Date: Mon, 23 Sep 2002 21:17:21 -0600
Subject: confusion about opening files
References: <3d8fcda4$0$198$75868355@news.frii.net> <3D8FCFF3.FD52AFE5@alcyone.com> <3d8fd215$0$192$75868355@news.frii.net> <3D8FD2F3.81F3831C@alcyone.com>
Message-ID: <3d8fd8a7$0$200$75868355@news.frii.net>

Cool! Thanks so much! I think I'm actually starting to get it!

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3D8FD2F3.81F3831C at alcyone.com...
> robie1373 wrote:
>
> > O. K. Now I am confused.  I thought that when I imported the os
> > module, i
> > would have to call its functions like this: os.open()    Why am I just
> > using
> > the function name?
>
> Because there are two different open functions here.  One is in the
> builtins, one is in the os module.  They are not the same:
>
> >>> import os
> >>> print open.__doc__
> file(name[, mode[, buffering]]) -> file object
>
> Open a file.  The mode can be 'r', 'w' or 'a' for reading (default),
> writing or appending.  The file will be created if it doesn't exist
> when opened for writing or appending; it will be truncated when
> opened for writing.  Add a 'b' to the mode for binary files.
> Add a '+' to the mode to allow simultaneous reading and writing.
> If the buffering argument is given, 0 means unbuffered, 1 means line
> buffered, and larger numbers specify the buffer size.
> Note:  open() is an alias for file().
>
> >>> print os.open.__doc__
> open(filename, flag [, mode=0777]) -> fd
> Open a file (for low level IO).
>
> The os.open is a low-level API and is rarely going to be the one you
> want.
>
> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ Insight -- the titillating knack for hurting!
> \__/ Colette
>     Erik Max Francis' bookmarks / http://www.alcyone.com/max/links/
>  A highly categorized list of Web links.




From tjreedy at udel.edu  Fri Sep 27 17:22:33 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 27 Sep 2002 21:22:33 GMT
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <amvcev$cf6$1@s1.read.news.oleane.net> <7nHk9.172208$pX1.6117207@news2.tin.it> <ofLk9.304267$z91.13145351@bin3.nnrp.aus1.giganews.com> <dITk9.175519$pX1.6209886@news2.tin.it>
Message-ID: <t_3l9.334924$AR1.14880141@bin2.nnrp.aus1.giganews.com>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:dITk9.175519$pX1.6209886 at news2.tin.it...
> A typical, frequent, practical case is classification, e.g.:
>
> "Append bleep to list truebleeps when bleep is true, append it
> to list falsebleeps instead when false".
>
> If we had a ternary operator, a natural way of expression would be:
>
>     (bleep?truebleeps:falsebleeps).append(bleep)

This reminds me of using ?: to calculate C lvalues (targets), which I
am not sure all C programmers realize is legal, since all examples I
read while learning only used ?: to calculate rvalues (sources).

> It's _dangerous_ -- to use it, one always has to be fully aware
> of the risk that subexpression b might be false.

If b is a non-null constant, there is no such risk.  Consider

na=1; np=3; nl=1; ng=0
print "Found: %d apple%s, %d plum%s, %d lemon%s, and %d grape%s." %\
(na, na!=1 and 's' or '', np, np!=1 and 's' or '', nl, nl!=1 and 's'
or '', ng, ng!=1 and 's' or '')

Found: 1 apple, 3 plums, 1 lemon, and 0 grapes.

Alternatives to this 'risky' construct:
1. forgo the nicety of correct English output
2. nest if/else four levels deep to choose correct variant of 16 print
statements (ugh)
3. precalculate 4 suffixes and store in 4 temp vars with 4 if/else's
(8 or 16 lines);
   though clearly superiour to 2), this is still tedious with some
risk of typo error.

To revise your statement: to properly use 'a and b or c' as a
conditional expression, one must be aware of and remember the
condition 'bool(b)==True'.  To contrapositively paraphrase what I
wrote before: if b is an expression with variables and it is not
pretty
clear that won't be null, then don't use it.  This is as clear as
"don't divide by 0 and be wary of variable expressions that might
become 0."

>  I think this is a pitfall just waiting for the unwary to stumble.

Must functions have preconditions on arguments or possibly surprising
results that can cause the 'unwary' to stumble.  'x/y', 'float -
(float+eps)' and '[[]]*n' are other examples.  Perhaps we should
better teach people "Before you use a function or language construct,
make sure you understands its input conditions and output and the
pitfalls thereof, lest you stumble due to unwariness".

Terry J. Reedy




From tismer at tismer.com  Fri Sep 20 11:22:32 2002
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 20 Sep 2002 17:22:32 +0200
Subject: Overriding all methods in a class
References: <Pine.LNX.4.44.0209201021580.13541-100000@mercury.sabren.com>
Message-ID: <3D8B3D38.8040309@tismer.com>

Joao Prado Maia wrote:
> Hi,
> 
> I have been looking for an answer for this question for a while, but 
> couldn't find anything that would work for what I want to do.
> 
> I know that you can use __getattr__ to overload methods that your class 
> don't have, but what about overloading all methods, so I could in a way 
> 'intercept' calls to my methods and do something different if I wanted, is 
> that possible ?
> 
> What I want to do basically is create a caching system for a specific 
> class, and have __getattr__ (or any other magic method name) handle the 
> decision to sue the cached return value or to actually call the method.
> 
> Maybe it would be just easier to rename my current class to _Classname and 
> create a new Classname with just __getattr__.
> 
> Any suggestions ?

Use the new class type, by deriving from some builtin,
and then use the new __getattribute__. For instance

class MyObject(object):
     def __getattribute__(self, name):
         print name
         return getattr(MyObject, name)

But be *very* careful, it is easy to cause endless recursions,
for instance it is not possible to use self.__class__ in
that context.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From danb_83 at yahoo.com  Wed Sep  4 09:53:07 2002
From: danb_83 at yahoo.com (Dan Bishop)
Date: 4 Sep 2002 06:53:07 -0700
Subject: using python to call java code??
References: <3cc90cd2.0209040122.17242a04@posting.google.com>
Message-ID: <ad052e5c.0209040553.5684b133@posting.google.com>

tim.schulz at sdm.de (Tim Schulz) wrote in message news:<3cc90cd2.0209040122.17242a04 at posting.google.com>...
> I am considering to use zope in order to develop an content management
> portal. My problem is, that I have to integrate some ready-built java
> code. Is there any possibility to call java code from python? Is there
> a API for this kind problem?

http://www.jython.org


From cypher_dpg at yahoo.com  Sat Sep  7 00:01:42 2002
From: cypher_dpg at yahoo.com (Jeremy Jones)
Date: Sat, 07 Sep 2002 04:01:42 GMT
Subject: Thinking of joining
References: <3D7970DC.5115BCC8@earthlink.net>
Message-ID: <20020907001139.4daf27a3.cypher_dpg@yahoo.com>

On Sat, 07 Sep 2002 03:18:33 GMT
dark <darksyyyde at earthlink.net> wrote:

> Hello.
> Im currently still learning C, but a friend has been trying to turn me
> on
> to python.

I would consider him/her a very good friend for pointing you to Python.

> From what ive seen it looks pretty good.
> I was just curious if it will do everything PERL will do, as i
> dont care for how PERL reads.

It has done everything for me that Perl can...and then some.  I started learning Perl a few years ago after beating my head against a wall after hitting some limitations with UNIX shell scripts.  Perl was a wonderful breath of fresh air...until I started creating an OO module.  (Actually the syntax started getting to me before I started creating the OO module, but for the sake of semi-brevity, I'll just stick to my gripes with modules.)  The module itself grew to over some 700 lines of code.  Any modification just got more and more painful.  I decided to learn Python.  That was probably the single best career-oriented decision I have ever made.  I re-wrote the module in Python in almost half the lines of code (I think it's right around 400 lines of code right now).  Modifications are simple to make.  Very importantly, I can go months without touching the module and jump right back in with little effort (which I had a difficult time doing the same with the Perl module).  I'll make this statement to keep anyone else from saying it first:  it may well be that my (Perl) code just plain sucks.  It probably does ;-)  But it has become obvious to me that Python is a much more elegant, simple (I don't mean that in a belittling way at all), clear language to work with - unlike Perl.  Anyway, that's my rant.

> ( though a lot of proplr i have talked to try to push PERL on me)
> Im going to try to discipline myself to get C down first, but its
> tempting
> to jump into Python now.

I would recommend jumping into Python.  Depending on what you are doing, you may well never need to even look at C code, let alone write it.  (I know that there are some that would disagree with that.)  And if you don't have to, then don't waste your time.  I started out tinkering with C and I wish I had discovered Python first.  I consider going back and brushing up on my C skills every once and again, but I find myself asking, "Why?  What will getting better at C benefit me?"  And I always wind up sticking pretty much solely to Python because it gets nearly every job done that I need to do.

> I am a monty Python fan of many years, if that helps lol.
> 
> 

Hee hee hee.  That always helps.


Jeremy Jones.


From r.l.richardson at verizon.net  Mon Sep 30 18:50:55 2002
From: r.l.richardson at verizon.net (Rick)
Date: Mon, 30 Sep 2002 15:50:55 -0700
Subject: __call__ question
Message-ID: <3D98D54F.5070200@verizon.net>

If I generate a call method inside a class class with def __call__(self, 
arg1): ...
everything is hunkey dorey.  

However if I generate the class then later try to add the __call__ 
function with setattribute or by create a function that behaves like 
call should and then assigning it to myobj.__class__ I can't call the 
instance of the class like a callable object. It states that that 
instance is not a callable object. I can however:

myobj.__call__(myobj, arg1)
and it works. Anyone know what I have to do to make an object callable 
at runtime?


thanks,

Rick Richardson





From martin at v.loewis.de  Wed Sep 25 02:28:28 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 25 Sep 2002 08:28:28 +0200
Subject: Solaris 64 bit compilation
References: <c2595393.0209201844.2ed62265@posting.google.com> <m3sn03k44m.fsf@mira.informatik.hu-berlin.de> <c2595393.0209210627.44f672a8@posting.google.com> <m3d6r7i9o0.fsf@mira.informatik.hu-berlin.de> <c2595393.0209240941.5f3fc565@posting.google.com> <j4fzvzclfb.fsf@informatik.hu-berlin.de> <c2595393.0209241643.5b276788@posting.google.com>
Message-ID: <m3fzvyoa5v.fsf@mira.informatik.hu-berlin.de>

phoebe_1 at att.net (Holden Caulfield) writes:

> > > Is anyone even interested in the configure.in patches?
> > 
> > I don't really see a need, since it works fine without the patches.
> > 
>   Well, I will have to disagree here. As currently at lease in V2.2.1 vanilla
>   source there is *no* option to built in 64 bit mode, unless you hack the
>   CFLAGS, LDSHARED etc.
>   Unless, you know of way to pass the appropriate flags? 

Setting CC to "cc -xarch=v9", then running configure, has the desired
effect.

>   Because my expectation was that I could build 64bit excutable/object/module
>   by running "configure", currently out of the box I could *not* figure out a
>   way to do this. 

You need to set CC for this.

Regards,
Martin


From shalehperry at attbi.com  Sat Sep  7 20:44:40 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Sat, 7 Sep 2002 17:44:40 -0700
Subject: Larry Wall's comment on python...
In-Reply-To: <oqit1hv1ol.fsf@titan.progiciels-bpi.ca>
References: <mailman.1031377930.4225.python-list@python.org> <200209071429.26401.shalehperry@attbi.com> <oqit1hv1ol.fsf@titan.progiciels-bpi.ca>
Message-ID: <200209071744.40129.shalehperry@attbi.com>

On Saturday 07 September 2002 16:07, Fran?ois Pinard wrote:
> [Sean 'Shaleh' Perry]
>
> > [yka]
> >
> >> [Kow Kuroda]
> >>
> >> > [...] For example, vi and emacs indent by 8 characters by default
> >>
> >> [...] Well at vim :set ts=4 makes more fitting
> >
> > actually emacs defaults to 4 space indenting [...]
>
> The Python mode writers for Emacs managed to automatically recognise the
> indentation in use in a file, and continue editing that file with the same.
> For new files, I get four spaces per level.
>
> However, for Python files indented with TABs only, Emacs cannot
> successfully guess if the author's intent was to expand those TAB with four
> or eight spaces.  By default, I vaguely remember it picks eight.

correct emacs says a tab is a tab and a tab means 8.



From b.maryniuk at forbis.lt  Fri Sep  6 11:08:25 2002
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Fri, 6 Sep 2002 17:08:25 +0200
Subject: ttySx
Message-ID: <200209061708.25677.b.maryniuk@forbis.lt>

Hi, all.
Is anybody know how to check modem status from the Python on Linux?
I need something like send 'ATZ' and get "OK" or something.

I've tried do somethig like:

modem = open('/dev/ttyS0', 'r+')
modem.write('ATZ\r')
modem.read() # Hang here forewer

This is bad. Then I've tried use several modules for serial port, wheere I can 
write, but if I read, always get ''.

Perhaps I should figure out with DTR status and so on.... Any ideas?

-- 
Regards, Bogdan

Anybody who doesn't cut his speed at the sight of a police car is
probably parked.




From aleax at aleax.it  Tue Sep 24 08:10:01 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 24 Sep 2002 12:10:01 GMT
Subject: list problem
References: <mailman.1032818129.8126.python-list@python.org> <mailman.1032828726.20071.python-list@python.org> <amp6i1$7n9bi$1@ID-77524.news.dfncis.de> <Xns929364D72B214duncanrcpcouk@127.0.0.1>
Message-ID: <tCYj9.157570$pX1.5675986@news2.tin.it>

Duncan Booth wrote:

> Thorsten Kampe <thorsten at thorstenkampe.de> wrote in
> news:amp6i1$7n9bi$1 at ID- 77524.news.dfncis.de:
> 
>> * Anthony Tuininga
>>> If you have list comprehensions available, use
>>>
>>> [b for a in x for b in a]
>> 
>> What about:
>> reduce(lambda x, y: x + y, a) ?
>> 
> 
> If you are going to use reduce, then you would be better off writing:
> 
>    import operator
>    reduce(operator.add, x)

Maybe (marginally).  But timing things is often a good idea:

import time, sys, operator

def alaAT(x):
    return [b for a in x for b in a]

def alaTK(a):
    return reduce(lambda x, y: x+y, a)

def alaDB(a):
    return reduce(operator.add, a)

def alaAM(a, add=operator.add):
    return reduce(add, a)


biggie = [ range(120) for i in range(160) ]

def timit(func, N=100):
    repeat = [None] * N
    start = time.clock()
    for x in repeat: func(biggie)
    stend = time.clock()
    return "%2.2f %s" % (stend-start, func.__name__)

print sys.version
for i in range(3):
    for func in alaAT, alaTK, alaDB, alaAM:
        print timit(func)

[alex at lancelot linud]$ python -O timit.py
2.2.1 (#2, Jul 15 2002, 17:32:26)
[GCC 2.96 20000731 (Mandrake Linux 8.1 2.96-0.62mdk)]
1.43 alaAT
2.75 alaTK
2.69 alaDB
2.72 alaAM
1.42 alaAT
2.73 alaTK
2.72 alaDB
2.66 alaAM
[alex at lancelot linud]$


>From this measurement it's unclear whether there's anything to
choose between the versions using reduce, but Thorsten's original
proposal appears to be substantially better than any of them
(for these sizes of lists on this specific machine...) -- almost
a factor of 2, thus almost worth noticing if the performance
of this operation matters for the program being developed.


Alex



From gerhard.haering at opus-gmbh.net  Thu Sep 12 10:23:20 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 12 Sep 2002 14:23:20 GMT
Subject: DB Cursor iterator (was: [ANN] PySQLite 0.3.0 released)
References: <slrnao0e6p.23c.gerhard.haering@haering.opus-gmbh.net> <cL0g9.335846$On.14066007@bin3.nnrp.aus1.giganews.com> <alq6g8$8eb$1@news1.wdf.sap-ag.de>
Message-ID: <slrnao18jr.3as.gerhard.haering@haering.opus-gmbh.net>

Daniel Dittmar wrote:
> Terry Reedy wrote:
>> Does anyone know whether there are any plans to update the Python-DB
>> API to the new iterator protocols, so that the above could be written
>> more simply something like
>>
>> for row in cursor:
>>        print "%14s, %15s, %19s, %8s, %25s" % tuple(row)
> 
> It's already in the 'Optional DB API Extensions' in
> http://www.python.org/peps/pep-0249.html

Yep and these extensions are fully supported by PySQLite and pyPgSQL :-)

-- Gerhard


From david.abrahams at rcn.com  Mon Sep 30 23:02:44 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Mon, 30 Sep 2002 23:02:44 -0400
Subject: Is there a docutils / reStructuredText system for C++ source code available?
References: <3D98D9D9.8000009@attglobal.net> <anarjq$6n3$2@bob.news.rcn.net> <3D9916E2.5030006@attglobal.net>
Message-ID: <anb5iu$b6l$1@bob.news.rcn.net>

"Pierre Rouleau" <pieroul at attglobal.net> wrote in message
news:3D9916E2.5030006 at attglobal.net...
>
> David Abrahams wrote:
> > "Pierre Rouleau" <pieroul at attglobal.net> wrote in message
> > news:3D98D9D9.8000009 at attglobal.net...

> > See http://synopsis.sf.net
> >
> >
> Thanks!
> I also found another tool called Doxigen  at
> http://www.stack.nl/~dimitri/doxygen/
>
> Anyone has experience with it?

Widely used, but it's half-arsed C++ "Parser" is completely inadequate for
the type of code I write. I'm hoping that synopsis will pan out for me since
it is built on a real parser, but so far I haven't had great success. I need
more time to do a good evaluation.

--
-----------------------------------------------------------
           David Abrahams * Boost Consulting
dave at boost-consulting.com * http://www.boost-consulting.com





From pinard at iro.umontreal.ca  Tue Sep  3 07:56:02 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Tue, 03 Sep 2002 07:56:02 -0400
Subject: ANN: Pyrex 0.4.4
In-Reply-To: <mailman.1031050990.31309.clpa-moderators@python.org> (Greg
 Ewing's message of "Mon, 02 Sep 2002 16:38:58 +1200")
References: <mailman.1031050990.31309.clpa-moderators@python.org>
Message-ID: <oqk7m370dp.fsf@titan.progiciels-bpi.ca>

[Greg Ewing]

>   * Tracebacks: Pyrex modules now produce tracebacks that
>     show you where the error occurred in the Pyrex code.

Wow! :-)

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



From xtian at toysinabag.com  Thu Sep 19 20:46:40 2002
From: xtian at toysinabag.com (xtian)
Date: 19 Sep 2002 17:46:40 -0700
Subject: lists in dictonaries ?
References: <mailman.1032426629.1933.python-list@python.org> <slrnaojc79.c05.Andreas.Leitgeb@pc7499.gud.siemens.at>
Message-ID: <d111eafa.0209191646.6429dcb9@posting.google.com>

Andreas.Leitgeb at siemens.at (Andreas Leitgeb) wrote in message news:<slrnaojc79.c05.Andreas.Leitgeb at pc7499.gud.siemens.at>...
> Andrew Thompson <andrew.thompson at ashecastle.com> wrote:
> > You can put any python objects into a python dictionary, even the
> > dictionary itself...
> 
> Oh, can I create a dictionary in Python that contains all those
> dictionaries that don't contain themselves ... ?-)
> 
> *scnr*'ly yours.

Would it contain itself?

xtian


From bokr at oz.net  Thu Sep  5 15:54:02 2002
From: bokr at oz.net (Bengt Richter)
Date: 5 Sep 2002 19:54:02 GMT
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com> <j4it1k4ufy.fsf@informatik.hu-berlin.de> <al886p$lfj$0@216.39.172.122> <m3d6rs46ln.fsf@mira.informatik.hu-berlin.de>
Message-ID: <al8coq$1ql$0@216.39.172.122>

On 05 Sep 2002 20:46:44 +0200, martin at v.loewis.de (Martin v. Loewis) wrote:

>bokr at oz.net (Bengt Richter) writes:
>
>> How about introducing an optional "degree of globalness" for the
>> global declaration? E.g.,
>
>> 
>>     x = 1
>>     def foo():
>>       x = 2
>>       def bar():
>>         x = 3
>>         def baz():
>>           global(1) x
>>           x = 4
>> 
>> would go outwards 1 nested scope and thus target the x in bar,
>> rebinding it from 3 to 4.
>
>That would work - although I'd claim that some users will find that
>notation plain ugly.
>
>It will also have confusing results:
>
>x = 1
>def foo():
>  print x
>  def bar():
>    global(1) x
>    x = 1
>
>At the moment, the "print x" in foo prints the global x, since there
>is no assignment to x, and hence there is no local variable foo. With
>the global statement, you declare that there is a local variable
>foo::x. As a result, since there is no assignment to foo::x, this
>would raise an UnboundLocal error.
>
I Python didn't effectively look ahead of the current line to decide what
scope a symbol belonged to, it wouldn't be a problem. The print x would
find the global x, and if bar were called after print x (and after its
definition, of course), then foo::x would be created, shadowing the global x,
and another print x following that would print the new (indistinguishable above)
value in foo::x.

IMO, that would be more intuitive too. Compare to:

[13:02] C:\pywk\junk>type foo.cpp
#include <cstdio>
int x = 123;
void foo(){
    printf("global x @ %p: %d\n", &x, x);
    int x(x);
    printf(" local x @ %p: %d\n", &x, x);
}
void main(){
    foo();
}

[13:03] C:\pywk\junk>cl foo.cpp
Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 12.00.8168 for 80x86
Copyright (C) Microsoft Corp 1984-1998. All rights reserved.

foo.cpp
Microsoft (R) Incremental Linker Version 6.00.8168
Copyright (C) Microsoft Corp 1992-1998. All rights reserved.

/out:foo.exe
foo.obj

[13:03] C:\pywk\junk>foo
global x @ 00406030: 123
 local x @ 0012FF74: 123

Regards,
Bengt Richter


From gerhard.haering at opus-gmbh.net  Mon Sep 16 07:34:10 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 16 Sep 2002 11:34:10 GMT
Subject: (no subject)
References: <mailman.1032173958.26614.python-list@python.org>
Message-ID: <slrnaobg4m.2b8.gerhard.haering@haering.opus-gmbh.net>

Srijak Rijal wrote:
> Hi all
> 
> I just migrated from perl to python and was wondering if python socket
> can be read line by line too?

You can use makefile() on the socket and then use all the usual file
methods, including readline() afterwards.

-- Gerhard


From spendius at muchomail.com  Mon Sep  2 10:33:09 2002
From: spendius at muchomail.com (Spendius)
Date: 2 Sep 2002 07:33:09 -0700
Subject: Newbie's unsurmountable pb with IO stream trick
Message-ID: <aba30b75.0209020633.26ba4c70@posting.google.com>

I'm about to give up learning Python because I've met
insurmountable problems. I'm a newbie just having begun
teaching myself this language a few days ago, and I 
already stall against unsolvable (at least to me) issues:
so could you please help me and *show me* I'd be wrong
giving up ?

I tried to rewrite something I wrote in Java, trying to
communicate with an OS process (SQL*Plus in this occurence),
but big time bucked against
- the lack of methods of the so-called 'file objects' (if
  there was an equivalent to java.io.InputStream.available() !)
- the fact you don't declare variables: you just assign
  values to them => I immediately hit against the 'global'
  keyword issue (rather a global mess) and gave up...

For those who know SQL*Plus y'all know you get a prompt:
"SQL> " 
(remove the "'s to get the string) and then this frontend
Oracle software waits for your input. This is somewhat tricky 
to deal with, but in Java I managed to find a solution [because
your InputStream blocks, which makes your program hang if
you don't address this 'waiting prompt' problem] thanks to
a sub-thread that tries to read the InputStream every time
I detect that the method available() returns 0 (zero): if
the next character in my stream is easily readable -if it
doesn't block- I return it, if it blocks OK I know I've
reached the final 5 characters 'S', 'Q', 'L', '>', ' ' and
that I'll never get the following one, so I return -1 (EOF
in Java).

So I have a challenge for you guys who are enough acquainted
with Python: how could I perform the same with this language ?

Thanks ever so much to y'all...

          ____________________
         (                    )
         (   operating syst.  )
         (    ___________     )
         (   / SQL*Plus  \    )
         (   \__process__/    )
         (    ?        |      )
         (____|________|______)
              |        |
              |        |
              |        |
              |       \|/
              |      output str. --->--> sent to a tcl window (for ex.)
            input      ('InputStream' in Java -sic !-)
           stream
       (called 'OutputStream'
          in Java !!)
             
Send "select sysdate from dual;" to the input stream & you'll get
the output stream string (remove the "'s):
"\n"
"SYSDATE\n"
"---------\n"
"02-SEP-02\n"
"\n"
"SQL> "


From vetter at lincom-asg.com  Wed Sep 11 09:29:43 2002
From: vetter at lincom-asg.com (Keith)
Date: 11 Sep 2002 06:29:43 -0700
Subject: Regexp Over Multiple Lines
References: <565e316f.0209040740.36573b56@posting.google.com>
Message-ID: <565e316f.0209110529.30e61ea4@posting.google.com>

vetter at lincom-asg.com (Keith) wrote in message news:<565e316f.0209040740.36573b56 at posting.google.com>...
> In Python can I have a regular expression that spans multiple lines...
> e.g. In Perl I may define a regular expression like:
> 
> $depends_def =   qr/
>          [\w_\.]+\s*:                     # simobject:
>          \s*
>          [\w_\.]+                         # job name
>          \(\s*(?:[\d]*)?\s*\)             # arg list for job (optional
> num)
>          \s+
>          depends\s+on                     # keywords "depends on"
>          \s+
>          (?:
>           [\w_\.]+\s*:                    # simobject:
>           \s*
>             (?:
>               [\w_\.]+                    # job name
>               \(\s*(?:[\d]*)?\s*\)        # arg list for job (optional
> num)
>               (?:
>                 (?:
>                  \s+
>                  and\s*[\w_\.]*\s*:       # "and simobject:"
>                  \s*
>                  [\w_]*                   # job name
>                  \(\s*(?:[\d]*)?\s*\)     # arg list for job (optional
> num)
>                 )+
>               )?                          # match optional "and" 1 or
> more times
>             )
>          )
>         /sx;
> 
> then later I can use that regular expression in another regular
> expression.
> 
> It helps to span multiple lines for commenting what the heck the ugly
> regular expression is trying to do...
> 
> BTW: I am investigating Python and have no knowledge of it.  Our Perl
> code has just gotten out-of-hand.
> 
> Thanks,
> Keith

After reading the links you guys suggested... it looks like the python
regular expressions are very powerful.

I am sure that this has been discussed... but in a nutshell can you
tell me where python lacks in power behind perl.  If I go to rewrite
my Perl code, which has gotten really messy, in Python will I be
saying over and over again, "Man, I can't do =that= in Python!"

I am testing the waters before I dive in!

Take care,
Keith


From marklists at mceahern.com  Mon Sep  9 23:36:37 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 9 Sep 2002 22:36:37 -0500
Subject: dynamism
In-Reply-To: <66715c02.0209091923.265fe77e@posting.google.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOEEJADNAA.marklists@mceahern.com>

[Steven Shaw]
> I've been away from Python for awhile and I'm wondering whether Python got
> less "dynamic" while I was away.
>
> I thought you could just extend an object's attributes at
> run-time like this:
>
> >>> p = object()
> >>> p.x = 1
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> AttributeError: 'object' object has no attribute 'x'

object is the base type of new-style classes.  Try something like this
instead:

>>> class foo:pass
...
>>> f = foo()
>>> f.x = 1
>>> f.x
1

Or 'class foo(object):pass' works too.

// m

-




From jkraska at san.rr.com  Thu Sep 26 00:36:41 2002
From: jkraska at san.rr.com (Courageous)
Date: Thu, 26 Sep 2002 04:36:41 GMT
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <3D7958DB.50903@nospam.free.fr> <3D87402E.6020700@thinkware.se> <amsfib$cq1$1@nyheter.chalmers.se> <3D926C1B.7080204@something.invalid>
Message-ID: <8635puklc5oj74sno9emdbl5h0oibcg0g3@4ax.com>

On Thu, 26 Sep 2002 14:08:27 +1200, Greg Ewing <see_reply_address at something.invalid>
wrote:

>Jacob Hallen wrote:
>
>> Please note that I didn't say anything about block
>> delimiters or parentheses in this. I think they were put into programming
>> languages to make it easy to write fast compilers.

Even Python uses block delimeters internally in order to properly
parse the language. They are actually inserted by the _LEXER_, which
performs an elaborate indentation-aware transformation of the code,
inserting INDENT and DEDENT tokens into the stream where they are needed.

Hence, this expression...

	for i in mylist:
		expr

...becomes...

	FOR IDENTIFIER IN IDENTIFIER INDENT expr DEDENT

Well, at least notionally. I'm not highly well-informed on the particulars,
and my knowledge may be a bit out of date.

This is, by the way, about the most elegant design pattern one can use
to solve this particular problem. There are many other approaches which
have heinous hidden pitfalls.

C//



From martin at v.loewis.de  Wed Sep  4 03:01:01 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 04 Sep 2002 09:01:01 +0200
Subject: PyListObject & C Modules
References: <mailman.1031085786.12546.python-list@python.org>
Message-ID: <m3k7m2tf0y.fsf@mira.informatik.hu-berlin.de>

Brian Quinlan <brian at sweetapp.com> writes:

>   if(PyList_Size(list) == 0){
>     PyErr_SetString(PyExc_ValueError, "empty lists not allowed");
>     return NULL;
>   }
>  
>   first = PyList_GetItem(list, 0);
> + if (first == NULL) return NULL;

That can't happen, really - GetItem can fail only if it is not a list,
or if the index is out of range. However, it is a good strategy to
test all results.

> + Py_ADDREF(first)
> - PyList_SetItem(result, 0, first);
> + if (PyList_SetItem(result, 0, first) != 0) {
> +   Py_DECREF(first)
> +   Py_DECREF(result);
> +   result = NULL;
> + }

I see: yes, this is needed. I wasn't really aware that SetItem steals
a reference. This is counter-intuitive - it might be better to use
PySequence_SetItem instead.

Regards,
Martin


From sismex01 at hebmex.com  Mon Sep 30 13:59:03 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Mon, 30 Sep 2002 12:59:03 -0500
Subject: Decimal arithmatic, was Re: Python GUI app to impress the bos
	s?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233026682A0@mail.hebmex.com>

> 
> Steve Holden wrote:
> 
> ><sismex01 at hebmex.com> wrote...
> >  
> >
> >[ ... ]
> >OK, so don't use whole pennies (or smallest currency) as your
> >base, use 10,000ths of a penny, but still use integer arithmetic
> >for your calculations.
> >
> Normal Integers aren't big enough.  2^31 / 10K == 215K.
> 
> --jb
> 
> -- 
> James J. Besemer		503-280-0838 voice
> 2727 NE Skidmore St.		503-280-0375 fax
> Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
> 				http://cascade-sys.com	
> 

"Integer arithmetic" != "32-bit integer arithmetic". The first
may be based on the second but they are definitely not the
same thing. Python longs are memory-limited, yet they
are also based on machine integers.  Does that make them
unsuitable?

-gustavo










Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From dkuhlman at rexx.com  Tue Sep 24 13:12:55 2002
From: dkuhlman at rexx.com (Dave Kuhlman)
Date: Tue, 24 Sep 2002 10:12:55 -0700
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com> <amet8v$e50$1@scotsman.ed.ac.uk> <L6Gi9.119475$ub2.2533804@news1.tin.it> <3D8B3250.BA964BCF@ObjectWay.it>
Message-ID: <amq6e5$85gpo$1@ID-139865.news.dfncis.de>

Nicola Musatti wrote:

> 
> 
> Alex Martelli wrote:
> [...]
>> I think you need substantial capacity and appetite for abstract thought
>> to really _enjoy_ SICP -- it doesn't spoon-feed you, nor does it use
>> baby-talk, lots of somewhat-repetitive examples, and other standard
>> didactical techniques.  But if the book's style does match your own
>> character and inclination, you're likely to fall in love with it.
> 
> Or, to put it in another way, gurus like this book. ;-)

Not meaning to show my ignorance here, but ...

This book is a huge time sink.  One section shows you how to fake object 
oriented programming in a dialect of lisp with no O-O support.  After 
working through that section, I asked myself, Why do I need to know this?  
What good is it going to do me, unless I'm using a lisp dialect that does 
not support object-oriented programming?

And, a large part of the book describes how to program without variables and 
state.  That's weird and interesting, just like the Elephant Man was weird 
and interesting.  But, after that, then what?  Am I supposed to try to 
program that way?

What is the lesson to be learned from "Structure and interpretation of 
Computer Programs"?  Maybe it's main purpose is to provide an enjoyable 
read for those who have still not given up on lisp.

Seriously, what was I supposed to have gained from all that time I spent on 
this book.  As you may suspect, it captivated me, too.

It did not teach me good Python programming style.  I could not hardly read 
my own code with all those many layers of parentheses.

It did not teach me how to use Python data structures.  I found myself doing 
too many screwy things with lists.

I was supposed to learn something more theoretical, right?  What was it?

Or, was this book just a devious way to trick more people into using lisp?

  - Dave

-- 
Dave Kuhlman
dkuhlman at rexx.com
http://www.rexx.com/~dkuhlman



From tismer at tismer.com  Tue Sep 17 16:42:24 2002
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 17 Sep 2002 22:42:24 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com>	<mailman.1032107006.6411.python-list@python.org>	<am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net>	<j4znuio0kr.fsf@informatik.hu-berlin.de>	<mailman.1032204706.14737.python-list@python.org>	<m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de>	<1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <am8259$ce0$1@main.gmane.org>
Message-ID: <3D8793B0.3020906@tismer.com>

Michael Chermside wrote:
...

> Okay, let's see if we can settle this... because I think it has real 
> potential.
> 
> I am going to pose a hypothetical situation... please don't argue about 
> whether these conditions are true now, or whether they will be true, but 
> take these as assumptions:

Thanks a lot for this great summary.
I'd just like to do a small addition:

...

>  (3) Someone (probably NOT Christian) is willing to take on the job of
>      merging it into the core and maintaining it afterward.

I would of course not refuse to keep maintaining it, let it be
in the trunk or not. I was just rying to say that I'm just
not willing to fight for it by discussion any longer. I'm a coder.

Stackless is my heart code and I will do for it whatever it needs.
...
> ... leave Christian to do what he's doing well, and let someone else 
> worry about possible integration.

And that would be the only contribution I'm asking for.

thanks again - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From quiteblack at yahoo.com  Thu Sep 19 06:07:14 2002
From: quiteblack at yahoo.com (black)
Date: Thu, 19 Sep 2002 03:07:14 -0700 (PDT)
Subject: where is the __dict__ ?
Message-ID: <20020919100714.69003.qmail@web21304.mail.yahoo.com>


Howdy~

  my book said all dictionaries have __dict__ method but couldnt find it by using dir(), where is it  ?



---------------------------------
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020919/09d9a469/attachment.html>

From mongo57a at comcast.net  Fri Sep 27 23:54:52 2002
From: mongo57a at comcast.net (mongo57a at comcast.net)
Date: Sat, 28 Sep 2002 03:54:52 GMT
Subject: Help please with menubars
Message-ID: <3d9516b3$1_7@goliath.newsgroups.com>

I have a built a menubar and I can do pulldowns. Lots of examples of that.
But if you want to directly execute a function based on selection of a
menubar item, how do you do that? Specifically: I have an "exit" item on the
menubar and if the user clicks on "exit", the window should be closed. What
I don't want is a pulldown with "exit" as a button option. Anybody know how
to do this? Thanks.


-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From claird at starbase.neosoft.com  Tue Sep 17 10:25:26 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 17 Sep 2002 09:25:26 -0500
Subject: concurrent programming
References: <fe31142f.0209130052.33174fcc@posting.google.com> <B7A63460808730BA.82D6BAEAFA5E4E78.4731C570A5FC8A2B@lp.airnews.net> <fe31142f.0209152321.341fee8d@posting.google.com>
Message-ID: <DA269F68A704CC7A.09043F9C0CB90704.0E096CD4CE5C1493@lp.airnews.net>

In article <fe31142f.0209152321.341fee8d at posting.google.com>,
Chris <cs26 at tecco.at> wrote:
>Thank you for the advice, but I have the instruction to write it in C++, so
>this is a point I cannot change. Only for the scripts, Python should be
>used.
>Well, I know a little bit about how to do concurrent programming in Python,
>and in C++ too, but my problem is that I don't know how to combine these two
>technologies for the PARBEGIN, PAREND block.

Now I'm even more confused.  It sounds as though you're 
constrained to do all the scheduling in C++, so I don't
understand the pertinence of Python's concurrency idioms.
They don't matter.  You have to schedule the parallelism
with C++, so Python is just a passive interpreter embed-
ded into the larger C++ scheduling application, right?
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From yaipa at aol.com  Fri Sep 20 16:17:34 2002
From: yaipa at aol.com (yaipa)
Date: 20 Sep 2002 13:17:34 -0700
Subject: linking extension modules w/ "ld" on Solaris - to be expected?
References: <mailman.1032393445.324.python-list@python.org>        <8d148763.0209190901.6276b37e@posting.google.com> <mailman.1032463585.12006.python-list@python.org>
Message-ID: <8d148763.0209201217.36cf7080@posting.google.com>

Skip Montanaro <skip at pobox.com> wrote in message news:<mailman.1032463585.12006.python-list at python.org>...
> Alan> I've tried compiling both Python 2.1.1 and 2.2.x on Solaris 8 and
>     Alan> 9 with gcc 2.9x and 3.x in my path. I consistently get the same
>     Alan> build error each time, so it is likely that there is some *switch*
>     Alan> I am not setting somewhere.  
> 
>     ...
> 
>     Alan> gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
>     Alan> -DHAVE_CONFIG_H  -DBUILD=`cat buildno` -o Modules/getbuildinfo.o
>     Alan> ./Modules/getbuildinfo.c
>     Alan> rm -f libpython2.2.a
>     Alan> ar cr libpython2.2.a Modules/getbuildinfo.o
>     Alan> gmake: ar: Command not found
> 
> You need /usr/ccs/bin in your PATH.  That's where ar and several other
> compilation- and link-related commands live, not just compilers proper.

Bing! Da-Bang! Da-Boom!  Worked like a champ.  Thanks Skip.


From skip at pobox.com  Mon Sep 16 21:27:03 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 16 Sep 2002 20:27:03 -0500
Subject: Python equivalent to perldoc?
In-Reply-To: <Olvh9.7898$Lg2.1389903@news2.news.adelphia.net>
References: <Olvh9.7898$Lg2.1389903@news2.news.adelphia.net>
Message-ID: <15750.34023.116179.238352@12-248-11-90.client.attbi.com>

    Bob> In Perl you can run "perldoc warnings" and get the definition of
    Bob> what "warnings" does. Is there a equivalent in Python?

Try "pydoc warnings".

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From Inyeol_lee at yahoo.com  Thu Sep 19 12:37:36 2002
From: Inyeol_lee at yahoo.com (Inyeol Lee)
Date: 19 Sep 2002 09:37:36 -0700
Subject: Minimal debug/rep functionality
References: <mailman.1032406044.16132.python-list@python.org> <Bcci9.8985$t%6.5877@nwrddc02.gnilink.net>
Message-ID: <db354fce.0209190837.4f92ebd4@posting.google.com>

I was writing similar tools, and found strange things related to f_lineno.
For multi-line statements, python 2.2.1 sometimes sets f_lineno to the
start line, and sometimes to the current line. This is not related to -O 
optimization.
I reported this to SF bug 607092, but this case is now closed with
"won't fix" status. Anyway, there seems to be an extensive change to
f_lineno in 2.3, so I'm just waiting for it.

--Inyeol


From tismer at tismer.com  Wed Sep 18 13:01:13 2002
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 18 Sep 2002 19:01:13 +0200
Subject: Copy with __slots__
References: <mailman.1032358481.7625.python-list@python.org> <3K1i9.110352$ub2.2335897@news1.tin.it>
Message-ID: <3D88B159.3080202@tismer.com>

Alex Martelli wrote:
> Griebel, Peer wrote:


<snip>

> class withslots(yetanother):
>     __slots__ = ('s2',)
>     def __copy__(self):
>         result = self.__class__.__new__(self.__class__)
>         for base in self.__class__.__mro__:
>             for s in getattr(base, '__slots__', []):
>                 setattr(result, s, getattr(self, s))
>         return result
> 
> the slowdown is even more severe -- to 0.62.  Still nothing
> like the 1.04 you pay for NOT having the slots at all, of
> course, but it's starting to undermine performance improvement,
> the only real rationale for having __slots__ at all.  Still,
> you pays your money, and you makes your choices.

You can get this to the single slot-level speed by doing
the slot grabbing phase of above once, after the class
has been created. I'd add an all_slots class attribute
with all collected slots and then iterate over it.
I think this is quite safe, since when you are using slots,
you are most probably *not* assuming that the classes
will change dynamically. If so, the all_slots need
to be re-built.

ciao - chris

p.s.: I didn't do any testing at all, but I guess some more
speed can be found by not caching the __slots__ but their
get/set wrappers and then just calling these with the new
instance...
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From wlfraed at ix.netcom.com  Thu Sep 19 17:20:28 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Thu, 19 Sep 2002 14:20:28 -0700
Subject: monetary applications [was: Python GUI app to impress the bos s?]
References: <31575A892FF6D1118F5800600846864DCBD817@intrepid> <mailman.1032362843.10509.python-list@python.org> <23891c90.0209190014.26c5bb9@posting.google.com>
Message-ID: <s2fdma.pp3.ln@ix.netcom.com>

Paul Boddie fed this fish to the penguins on Thursday 19 September 2002 
01:14 am:


> It could well be that not many people do financial database
> programming in Python, bizarre and unlikely as that may seem. Taking a
> look at pyPgSQL, as the only database module I have installed that
> actually readily documents such things (and that should tell you
> something), it would seem that monetary amounts are converted to
> floating point numbers - surely the first thing one would want to
> avoid in manipulating financial data. Perhaps people use integers, or
> perhaps other modules provide better decimal support.
>
        If I recall from the c.l.ada group, GNAT's "fixed point" data type is 
implemented as scaled integers, so even that might not pass the spec 
for a monetary type. My VB references are on loan (or currently stuck 
in the Windows environment and unreadable on Linux) so I can't recall 
what VB6 had for fixed/monetary.

        The only language I've ever encountered that had a truely monetary 
safe fixed point data type is COBOL -- which has always used packed BCD 
formats (though where the sign "bit" is stored may vary). COBOL was 
created from the beginning for "business" applications (heck: COmmon 
Business Oriented Language) and was spec'd by a government team. My 
college mainframe required /four/ of the sixteen registers to hold 
/one/ BCD number (32 BCD digits, 2 digits/byte, 16bytes @ 4/register).
 

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From remi at cherrypy.org  Mon Sep  2 04:51:06 2002
From: remi at cherrypy.org (Remi Delon)
Date: Mon, 2 Sep 2002 10:51:06 +0200
Subject: [ANN] CherryPy 0.6 released (SSL support)
Message-ID: <mailman.1031050991.31314.clpa-moderators@python.org>

We're pleased to announce the release of CherryPy-0.6.

This release adds SSL support to CherryPy (based on PyOpenSSL).

Many thanks to Martin Sj?gren for his PyOpenSSL package and to Mark Wormgoor
for his contribution to the CherryPy SSL code.

------------------------------------------
About CherryPy:

CherryPy is a Python-based tool for developing dynamic websites. It
uses many powerful concepts together, which makes it unique in its
approach to website development.
CherryPy sits between an application server and a compiler. You write
source files, compile them with CherryPy and CherryPy generates an
executable containing everything to run the website (including an HTTP
server).

Key properties/features of CherryPy are:
   - Based exclusively on Python (runs everywhere Python runs)
   - Delivers fast, robust, and scalable websites
   - Developers can use OOP as well as AOP (Aspect Oriented
Programming) concepts to develop websites
   - True separation of content and presentation
   - Simple but powerful templating language
   - "HTML editor safe" templating language (templates can go back and
forth between designers and developers)
   - Powerful standard libraries to make your life easy

Other properties/features are:
   - Can be linked to many databases (Oracle, Sybase, MySql,
PostgreSql, ...)
   - Can run behind another webserver (Apache, ...)
   - Easy clustering and load-balancing set up for high-traffic
websites
   - Built-in caching capability
   - SSL support

Remi.

http://www.cherrypy.org






From ugoertz at mi.uni-koeln.de  Sun Sep 29 09:32:34 2002
From: ugoertz at mi.uni-koeln.de (Ulrich Goertz)
Date: 29 Sep 2002 15:32:34 +0200
Subject: Tkinter: bind Alt + mouse click
Message-ID: <i0kznu0j4zx.fsf@Thales.MI.Uni-Koeln.DE>

Hi,

I am puzzled why I cannot bind events of the type "Alt" button press plus
mouse click. In other words, the following script which just opens a window
and should react on Alt-clicks by printing 'click', doesn't work for me:

# ---------------------------------
from Tkinter import *

def p(event): print 'click'

root = Tk()

f = Frame(root, width=100, height=100)
f.pack()
f.bind('<Alt-1>', p)

root.mainloop()
# ---------------------------------

(The window pops up, but nothing happens upon clicks.)

Everything works as expected if I replace <Alt-1> by <1>, <Control-1> or
<Shift-1>. I tried this on a Linux box with Python 2.0, 2.1 and 2.2, and
Tk 8.3. Any hints?

Ulrich



From loewis at informatik.hu-berlin.de  Mon Sep  9 05:20:06 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 09 Sep 2002 11:20:06 +0200
Subject: successor to htmllib
References: <mailman.1031345949.12743.python-list@python.org>
Message-ID: <j4bs77bjuh.fsf@informatik.hu-berlin.de>

Erik Price <erikprice at mac.com> writes:

> I noticed that the htmllib module is really best suited to HTML 2.0
> documents.  I was wondering if there was a newer (4+) HTML or even
> XHTML parsing module in development right now.  (By XHTML parsing lib
> I mean perhaps an XML parser that is specifically written for the
> XHTML 1.0 DTD.  I realize that any XML parser -should- work.)

PyXML offers xml.dom.ext.reader.HtmlLib, and xml.dom.ext.reader. Both
create HTML DOM trees, one from a HTML document, the other from a
XHTML document.

HTH,
Martin



From tjreedy at udel.edu  Thu Sep 19 10:45:53 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 19 Sep 2002 14:45:53 GMT
Subject: strong/weak - dynamic/static [Was: Getting started]
References: <uojinmcfapjf45@corp.supernews.com> <tyf8z1y5cp5.fsf@pcitapi22.cern.ch>
Message-ID: <Bqli9.36669$216.1428116@bin4.nnrp.aus1.giganews.com>

"Jacek Generowicz" <jacek.generowicz at cern.ch> wrote in message
news:tyf8z1y5cp5.fsf at pcitapi22.cern.ch...
> A quick google search suggests that Bruce Eckel really claims that
> python is weakly typed (at least more weakly typed than C++).

To my commonsense idea of strong and weak, this is exactly backwards.

1. Python is not 'typed', at least not in the sense meant here.  Nor
is any other computer language ;-).  Languages (or programmers using
them) type 'things' within the language.

2. Unlike many other languages, Python does *not* type names.  Names
in Python are typeless - unless you want to claim that they take on a
type be virtue of being bound.  Python *does* type objects with
values, so you can also say that it types values.  On the other hand,
values in, for instance, C, can be regarded as typeless bit sequences
(which, at a hardware level, they are).

3. The typing that Python does do is *stronger* than that of C; no
type-changing casting allowed.  In some languages, you can, for
instance, treat a four-byte integer as an array of four characters,
but not in Python.  You can explicitly extract a byte with bit
shifting and making and use chr() to make a new object, but that is
different from treating the byte in place as a char.  [Discussion of
old and new classes and modification of instance.__class__ omitted.]

> Could someone provide a serious reference

I'll leave this for someone else.

>explaining or defining the concepts of ... dynamic and static typing.

Python objects are typed at creation.  Is that dynamic or static?  I'm
not sure the dichotomy applies to Python very well since it partly
depends on a distinction between declaration and definition that does
not exist in Python and a distinction between compile time and run
time that is somewhat different in Python.

Terry J. Reedy




From brueckd at tbye.com  Fri Sep 27 01:02:33 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Thu, 26 Sep 2002 22:02:33 -0700 (PDT)
Subject: How do you pronounce "__" (double underscore)?
In-Reply-To: <amvn2k$gd1$1@news.wrc.xerox.com>
Message-ID: <Pine.LNX.4.44.0209262202150.24123-100000@ns0.tbye.com>

On 26 Sep 2002, Mark Jackson wrote:

> > > Pat Notz wrote:
> > >> @ = "at"
> > >> ! = "bang"
> > >> . = "dot"
> > >> * = "star"
> > >> 
> > >> Is there a short (one or two syllable) pronounciation for "__" (double
> > >> underscore).  "Underscore underscore" and "double underscore" are kind
> > >> of long.  Somthing like "splat"?  Anyone?
> > > 
> > > _ = 'under'
> > > 
> > > __ = 'under-under'
> > 
> > 
> > How about "underder"??
> 
> How about "dunder" (double under)?

How about "score" and "scorescore"?




From mhammond at skippinet.com.au  Thu Sep 12 19:05:03 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 12 Sep 2002 23:05:03 GMT
Subject: [win32all] Out parameter in outgoing interface
References: <u8z27x7s3.fsf@online.no>
Message-ID: <z49g9.26293$_l1.73186@news-server.bigpond.net.au>

Syver Enstad wrote:
> It does not seem possible use out parameters in an outgoing interface.
> 
> Scenario:
> 
> outgoing dispinterface, atl implementation of server, makepy support
> generated, python client which implements the outgoing interface.
> 
> Tried with both win32com.client.getevents and
> win32com.client.DispatchWithEvents
> 
> Debugged through it, the event gets fired from atl code the python
> code gets called, I thought maybe returning a tuple (the return value
> + the outgoing value) would work but the out parameter, a BSTR* (ie a
> OLECHAR**) is untouched back in ATL land.
> 
> I guess this is unsupported, or??
> 

This does work.  Try it first with a simple type.  For a string, you may 
be expecting the string to be modified in place, but in general, a new 
BSTR will be allocated and returned.

Mark.



From theller at python.net  Mon Sep  9 11:18:53 2002
From: theller at python.net (Thomas Heller)
Date: Mon, 9 Sep 2002 17:18:53 +0200
Subject: Windows/DOS: double clicking a .py file
References: <just-A936EC.17062909092002@news1.xs4all.nl>
Message-ID: <alie4t$1q4jol$1@ID-59885.news.dfncis.de>

"Just" <just at xs4all.nl> wrote in message news:just-A936EC.17062909092002 at news1.xs4all.nl...
> Sorry if this is a Windows FAQ, I'm a newbie to Windows, but not to
> Python.
>
> If I double-click a Python script (or use the start command in the dos
> prompt) a new dos prompt window is opened and my program is run. The
> window goes away when the program is done. However, the window _also_
> goes away when the program raises an exception, making it virtually
> impossible to actually _see_ the exception... Is there a way to keep the
> dos prompt window in this case?
One solution is outlined here:
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/65287
This is useful for the development machine. It keeps the DOS window
open, and leaves you in the Python debugger, where you can play around
even more to find out about the state of your program.
>
> A py2exe-fied version of the program shows the same behavior, which
> leads my users to report problems like "the output file is mysteriously
> truncated" instead of "I'm getting a traceback. Here, have a look."...
>
> Any suggestions?

For py2exe-packages apps you should redirect sys.stdout and sys.stderr
into a log-file.

Thomas




From martin at v.loewis.de  Sat Sep 21 12:34:23 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 21 Sep 2002 18:34:23 +0200
Subject: Solaris 64 bit compilation
References: <c2595393.0209201844.2ed62265@posting.google.com> <m3sn03k44m.fsf@mira.informatik.hu-berlin.de> <c2595393.0209210627.44f672a8@posting.google.com>
Message-ID: <m3d6r7i9o0.fsf@mira.informatik.hu-berlin.de>

phoebe_1 at att.net (Holden Caulfield) writes:

> > Please try to report failures carefully. It does not say sizeof(LONG),
> > instead, it says SIZEOF_LONG. Which gives rise to the question: What
> > value does SIZEOF_LONG have, in pyconfig.h?
> > 
> cc -c -DNDEBUG -g -Xc xarch=v9 -xcode=pic32 -I. -I./Include
> -DHAVE_CONFIG_H  -o Modules/p
> ython.o Modules/python.c
> "./Include/pyport.h", line 480: #error: "LONG_BIT definition appears
> wrong for platform (bad gcc/glibc config?)."

So what value has SIZEOF_LONG in pyconfig.h?

How did you invoke configure? How did you manage to get -xarch=v9 into
the compiler command line?

Regards,
Martin


From loewis at informatik.hu-berlin.de  Mon Sep  2 12:56:17 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 02 Sep 2002 18:56:17 +0200
Subject: compiling extensions on linux
References: <3D738D4E.35B6C712@mill.co.uk> <j44rd8jqu9.fsf@informatik.hu-berlin.de> <3D739730.DEB80478@mill.co.uk>
Message-ID: <j4lm6kib4e.fsf@informatik.hu-berlin.de>

Joe Connellan <joec at mill.co.uk> writes:

> it does have a initmyextension declared - Here's a simple example of my
> problem - I've taken all c++ related stuff out so I can use gcc

However, you are still compiling it as a C++ program, since it has the
.cpp extension. You either need to rename the file (to .c), or use the
-x c command line option. If there was C++ stuff to take out, you best
declare the function as extern "C".

Regards,
Martin


From henrik.motakef at web.de  Fri Sep 20 06:07:51 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 20 Sep 2002 12:07:51 +0200
Subject: What the heck? open() requires an integer?
References: <FCBi9.13802$Zk4.708378@news2.telusplanet.net>
Message-ID: <87y99xrn2g.fsf@pokey.henrik-motakef.de>

"Fausto Arinos de A. Barbuto" <fbarbuto at telusplanet.net> writes:

>     The open() function is returning me a strange error.
>     An attempt to create an output file with
> 
>             f = open("output.txt","w")
> 
>     yields the following error:
> 
>     Traceback (most recent call last):
>     File "<interactive input>", line 1, in ?
>     TypeError: an integer is required

Did you by chance import * from os? Looks as if you get os.open, that
indeed wants an integer.

hth
Henrik


From gbreed at cix.compulink.co.uk  Mon Sep 16 09:23:05 2002
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: Mon, 16 Sep 2002 13:23:05 +0000 (UTC)
Subject: "if x == None" raises "'NoneType' object is not callable"
References: <af84f376.0209160501.4d21500@posting.google.com>
Message-ID: <am4lvp$i5j$1@thorium.cix.co.uk>

Luc Saffre wrote:

> Strange! Nobody told you to call 'None'! Who knows an explanation?

You have an error here:

>       except KeyError,e:
>          AttributeError,str(e)

You presumably intend to raise that AttributeError.  Here's a corrected 
version with some error reporting:

>>> class Row:
...     def __init__(self):
...             self.__dict__["_values"] = {}
...     def __getattr__(self, name):
...             try:
...                     return self._values[name]
...             except KeyError, e:
...                     print "tried to get", name
...                     raise AttributeError, str(e)
...     def __setattr__(self, name, value):
...             self.__dict__["_values"][name] = value
... 
>>> row = Row()
>>> row is None
0
>>> row == None
tried to get __eq__
tried to get __coerce__
tried to get __cmp__
0

Without the "raise", __getattr__ returns None for __eq__, and the 
interpreter then tries to call it.  Note that "is None" doesn't have this 
problem.


                      Graham


From martin at v.loewis.de  Thu Sep  5 19:42:00 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 06 Sep 2002 01:42:00 +0200
Subject: German Umlaut Problem with raw_input in IDLE
References: <3D77C66F.4040305@aon.at>
Message-ID: <m3k7m0ngvr.fsf@mira.informatik.hu-berlin.de>

Gregor Lingl <glingl at aon.at> writes:

> H?? Tsch?s
[...]
> TypeError: object.readline() returned non-string
[...]
> I must confess, I do not understand the error message.

Apparently, IDLE returns a Unicode object, not a string object.

Regards,
Martin



From bobx at linuxmail.org  Wed Sep 18 13:56:22 2002
From: bobx at linuxmail.org (Bob)
Date: 18 Sep 2002 10:56:22 -0700
Subject: Python equivalent to perldoc?
References: <mailman.1032302122.25446.python-list@python.org>
Message-ID: <1001ff04.0209180956.359ce174@posting.google.com>

"DG" <david at fielden.com.au> wrote in message news:<mailman.1032302122.25446.python-list at python.org>...
> The contents of my f:\python22\pydoc.bat file:
> 
> @f:\python22\python f:\python22\lib\pydoc.py %1 %2 %3 %4 %5 %6 %7 %8 %9
> 
> Rowdy
Very nice...thanks it was just the bang in the head I needed.

Bob


From peter at engcorp.com  Sat Sep  7 22:51:36 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 07 Sep 2002 22:51:36 -0400
Subject: Larry Wall's comment on python...
References: <mailman.1031349792.2374.python-list@python.org> <3d796c18@news.sentex.net> <3D7A9E1A.1171C857@ihug.co.nz>
Message-ID: <3d7abb20$1@news.sentex.net>

Gib Bogle wrote:
> Peter Hansen wrote:
> 
>>mmaddox at hcsmail.com wrote:
>>>You must agree
>>>that Python can lead you to horizontal scrolling in your editor - the bane
>>>of user interface design.
>>
>>I definitely don't understand how Python, more or less than any other
>>language, leads to more horizontal scrolling.
> 
> Maybe he is referring to an effect of default editor tab size.

Apparently not, according to an email, which just said that
Python leads one to do this more than with other languages.

My point is that indentation (the way I've done it anyway) looks
*exactly* the same in C, Pascal, Python, or any other block-structured
language, so I still have no idea why Python would feel any different
in this respect for someone.

Clearly I'm missing some point, but I still don't know what it is.

C:

if (expression)
     {
     statements;
     }

Python:

if expression:
     statements

Pascal:

if (expression)
     begin
     statements;
     end

(or whatever... haven't written any for three years).

The only thing that would make me exceed 80 columns with a language
like this is if I couldn't define subroutines and call them, to
keep the nesting from growing too large.  Since all these languages
allow that, I have no idea what is special about Python that leads
to excessive indentation.

Oh well, whatever.

-Peter



From paul at boddie.net  Fri Sep 20 05:05:38 2002
From: paul at boddie.net (Paul Boddie)
Date: 20 Sep 2002 02:05:38 -0700
Subject: monetary applications [was: Python GUI app to impress the bos s?]
References: <31575A892FF6D1118F5800600846864DCBD817@intrepid> <mailman.1032362843.10509.python-list@python.org> <23891c90.0209190014.26c5bb9@posting.google.com> <amcnee$c21$1@panix1.panix.com>
Message-ID: <23891c90.0209200105.52bf42fa@posting.google.com>

aahz at pythoncraft.com (Aahz) wrote in message news:<amcnee$c21$1 at panix1.panix.com>...
> In article <23891c90.0209190014.26c5bb9 at posting.google.com>,
> Paul Boddie <paul at boddie.net> wrote:
> >
> >In that case, what do you think of this module? ;-)
> >
> >  http://www.boddie.org.uk/python/downloads/ScaledDecimal.py
> >
> >I'll gladly release this into the public domain, but some commentary
> >would be nice - I fear that I may not have implemented it as well as
> >the "numeric gods" might like.
> 
> Can you clearly explain the differences between your module and
> FixedPoint.py?  If not, the numeric gods probably won't accept it.  ;-)

Well, it doesn't matter so much to me whether the numeric gods accept
it or not, since I wrote it partly for my own amusement and partly to
use it in other things, but the only thing that I can think of is that
you can specify a negative "scale" in the 'get_scaled_decimal'
function and have the resulting object round to the nearest 10**n,
where n is positive. For example:

  get_scaled_decimal("270", -2) -> 300

Apart from that, it's all aesthetics and design decisions. It would be
nice if FixedPoint had a 'round' method which returns another object,
for example, but then it would be nice if I had been consistent with
the parameters to 'get_scaled_decimal' and the 'ScaledDecimal'
constructor: one uses a precision-like thing whereas the other uses an
exponent:

  ScaledDecimal(1002, -1) -> 1002 * (10 ** -1) : real
                          -> 100.2 : real
  get_scaled_decimal("100.2", 1) -> 100.2 : real

I can't remember clearly enough why I didn't use FixedPoint, but it
may have been in an earlier state when I investigated it before
writing ScaledDecimal - this was around the time of the integer
division wars, I think.

Paul


From huaiyu at gauss.almadan.ibm.com  Fri Sep 20 14:34:41 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Fri, 20 Sep 2002 18:34:41 +0000 (UTC)
Subject: [Help]: mailbox classes
References: <jylimd-DAD12F.17305418092002@skates.gsfc.nasa.gov>
Message-ID: <slrnaomqi1.eh6.huaiyu@gauss.almadan.ibm.com>

J. Li <jylimd at yahoo.com> wrote:
>Hi,
>
>I am trying to extract a message from a unix mailbox file. Python does 
>provide few classes for easy and uniform access. So it says! After 
>having read the libiary references, I went nowhere. 
>
>Can anyone give me an example as how to use these mailbox classes and 
>what is it relationship with email.message classes.
>
>thanks

I just figured this out yesterday following examples in spambayes project.
The documentation is very fuzzy, but this works for me:

from mailbox import PortableUnixMailbox as MB
from email import message_from_file as get_msg

def get_mbox(mfile):
	return MB(file(mfile), get_msg)

if __name__ == '__main__':
	mbox = get_mbox('test.mail')
	msgs = list(mbox)
	msg = msgs[2]
	print msg['Subject']
	print msg

Huaiyu


From marklists at mceahern.com  Mon Sep 16 09:14:19 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 16 Sep 2002 08:14:19 -0500
Subject: "if x == None" raises "'NoneType' object is not callable"
In-Reply-To: <af84f376.0209160501.4d21500@posting.google.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOKEPADOAA.marklists@mceahern.com>

Comparison to None should generally be comparison by identity rather than
equality:

  if x is None:

or:

  if x is not None:

not:

  if x == None:

nor:

  if x != None:

What you might consider is adding a __len__ magic method to your Row class
like so:

  def __len__(self):
    return len(self.values)

Then you can do this:

  r = Row()
  if r:
    ...
  else:
    ...

Cheers,

// mark

-




From aleax at aleax.it  Wed Sep 18 09:57:08 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 18 Sep 2002 13:57:08 GMT
Subject: try except ImportError failing to catch error!
References: <mailman.1032341542.17022.python-list@python.org> <3d888514$1@news.sentex.net>
Message-ID: <UC%h9.126664$pX1.4541518@news2.tin.it>

Peter Hansen wrote:
        ...
> catch it.  If you want to catch *any* error in the imported
> module, you'll have to do "except Exception, e" instead.
> Unfortunately, then you'll also have ImportErrors in the mix,
> so you won't know if the module could be found but had its
> own error, or if the call to __import__ failed...

That's trivially easy to remedy, thanks to the fact that you
may have multiple except clauses in a try statement, and they're
conceptually tried in order:

    try:
        themod = __import__(whatever)
    except ImportError, blah:
        print "Can't import (%s): %s" % (whatever, blah)
    except Exception, bluh:
        print "Some other problem: %s" % bluh

The only problem is, the ImportError might not be due to
the __import__ call itself -- it COULD be propagating
from some errant import that the module 'whatever' names
is trying to perform!  Discriminating between the two
cases does need using the traceback module, as you
suggest for other purposes in another part of this post.


Alex



From Padraig at Linux.ie  Wed Sep 25 10:46:55 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Wed, 25 Sep 2002 14:46:55 GMT
Subject: Newbie References Question
References: <3d91c72f$0$8558$79c14f64@nan-newsreader-02.noos.net>
Message-ID: <3D91CC1C.8040405@Linux.ie>

Guy Rabiller wrote:
> Hi,
> 
> let say I have:
> i1 = 1
> i2 = 2
> and
> p = [i1,i2]
> 
> How can I have:
> p=[*i1,*i2]
> rather than
> p=[**i1,**i2] as it is curently ?
> 
> ( Sorry for this nasty hybrid syntax )

I know what you mean, and being a newbie myself
I thought it would do what you want by default?
If you wrap them in lists then it does as you want:

i1=[1]
i2=[2]
p=[i1,i2]
i1[0]="one"
 >>> p
[['one'], [2]]

But I'm confused also as to why this doesn't happen
by default for base types? I suppose when you do:
i1=4 you're creating a new object for the i1 name
to reference, so the p[0] name keeps the reference
to the old object. Hmm I'm confused...

> Textualy, how if I want that p[0] refere to the 'i1' container, and not to
> the refererence it contains ?
> 
> What I want is that if now I set:
> i1 = 4
> that automaticaly:
> p -> [4,2]
> and not keeping [1,2]
> 
> Thanks in advance.



From thomas.heller at ion-tof.com  Thu Sep 26 08:15:50 2002
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: 26 Sep 2002 14:15:50 +0200
Subject: Python threading?
References: <f9Pi9.10275$Lg2.2191294@news2.news.adelphia.net> <amogmi$r1l$1@panix1.panix.com> <T5Rj9.13058$6g7.39977@news-server.bigpond.net.au> <MyUj9.134183$ub2.2937335@news1.tin.it> <AGpk9.11786$Lg2.3232428@news2.news.adelphia.net>
Message-ID: <ptv1kkuh.fsf@ion-tof.com>

"Robert Oschler" <Oschler at earthlink.net> writes:

> Having written _many_ state machines in my life, which async programming
> boils down to in one form or another, I feel it depends on the task.  When
> you have an async event driven model, the code that services a
> "transaction-package" ends up spreading functionality that services the
> "transaction-package" application's code landscape.  This can frequently be
> a headache.  As an example of a "transaction-package", take all the steps
> involved in servicing an incoming TCP connection, with custom authentication
> of a user, and perhaps a header and data object transacttion or two,
> finishing with a proper ending and cleanup of the transaction, and with
> subsequent notification of all necessary modules and data structures of the
> results of the transaction.  For dessert add a centralized clean error
> handler that can log errors propertly and gracefully recover the app
> regardless of where in the distributed event flow the error took place.
> 
> Contrast this with a simple procedure call where a linear series of function
> calls are made, each with appropriate error handling and status adjustment,
> in a pleasant centrally located manner.
> 

I wonder if generators can come to a rescue here?
Still trying to get them into my head...

Thomas


From sross at connectmail.carleton.ca  Thu Sep  5 17:57:22 2002
From: sross at connectmail.carleton.ca (Sean Ross)
Date: 5 Sep 2002 21:57:22 GMT
Subject: Python Source Code Documentation
References: <al69st$n1f$1@driftwood.ccs.carleton.ca> <G5Bd9.427560$m91.16660188@bin5.nnrp.aus1.giganews.com> <m3bs7cubp3.fsf@mira.informatik.hu-berlin.de>
Message-ID: <al8k02$186$1@driftwood.ccs.carleton.ca>

Terry: Yes, I did mean the Python interpreter. And I will give your
suggestion a try.

Martin: I will check out both of those documents.

Thank you both for your suggestions.




From mertz at gnosis.cx  Sun Sep 29 12:47:20 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Sun, 29 Sep 2002 12:47:20 -0400
Subject: Python's import: why doesn't it work?
In-Reply-To: <9ac02e81.0209290839.4f4d6475@posting.google.com>
References: <9ac02e81.0209290839.4f4d6475@posting.google.com>
Message-ID: <Y6yl9kKkXUZL092yn@gnosis.cx>         

|Once again my attempt to profit from Python's vaunted productivity
|increase has come to a screeching halt because some innocent thing
|that one would reasonably expect to work in one line, intuitively, and
|without even having to think about, turns into a nightmare already
|lasting more than an hour, requiring numerous experiments, searching
|through the documentation - both manually and using my editor's
|find-in-files ...

Perhaps you would be happier using a different programming langauge (one
that works without consulting documentation, apparently).  I hear there
are a number of other languages out there.

|>>> import templates.intro

Perhaps you should look into the standard documentation on import paths
(PYTHONPATH).  I'm not really sure why you would imagine for a moment
that any of those things you suggest would have any good effect....
Python is a programming language, not magical incantations that fulfills
vague wishes.

|Abject Defeat, Ask on Internet, Be Ridiculed by Peers for Not Managing
|Basic Operation

I would say that your post deserves ridicule.  Not because you ask about
"basic operation", but for its obnoxious and self-important tone.  You
catch more flies with honey than vinegar, y'know.

|import intro
|... and it would work. For real. The symbols were there. Not a peep
|out of the interpreter.

Yep. Almost surely that's because you tried it at some point with the
CWD in the same place as intro.py.

Yours, Lulu...

--
Keeping medicines from the bloodstreams of the sick; food from the bellies of
the hungry; books from the hands of the uneducated; technology from the
underdeveloped; and putting advocates of freedom in prisons.  Intellectual
property is to the 21st century what the slave trade was to the 16th.



From timr at probo.com  Thu Sep 12 01:04:23 2002
From: timr at probo.com (Tim Roberts)
Date: Wed, 11 Sep 2002 22:04:23 -0700
Subject: Destructor never called ???
References: <76710511.0209111450.40999dd8@posting.google.com>
Message-ID: <tp70ou0r0j11as6enhpsaqul4oqa3mlu6t@4ax.com>

muk_pr at yahoo.com (vb) wrote:

>Hi, 
>  I have the following code my problem is when I call any methods of  
>       TCVector_Object the destructor never executes.
>
>I mean
>
>v = module.CreateTCVector()
>del v;
>//destructor runs. message shown .OK
>
>but 
>in
>
>v = module.CreateTCVector()
>v.Append(123);
>del v;//or closing python interpreter
>
>destructor never gets called. Any ideas?

Yes.  Don't put critical code in a destructor.

When you delete "v", it removes the name "v" from your global namespace,
and decrements the reference count on your TCVector_Objet to 0.  That means
that no one is currently referencing the object.

However, Python is a garbage-collected system.  Unlike in COM, an object is
NOT destroyed as soon as its reference count goes to 0.  Your object will
only be destroyed if the garbage collector happens to run, and decides that
it might be a convenient time to actually free up the space your object
occupies.

If the garbage collector doesn't run before Python terminates, there is no
NEED to clean up memory, so the destructor will NEVER run.

Do a google search for "deterministic destructor python"; you'll see the
previous threads on this.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From chrism3 at cox.net  Thu Sep 12 22:04:34 2002
From: chrism3 at cox.net (Chris McDonough)
Date: Fri, 13 Sep 2002 02:04:34 GMT
Subject: Coding Style: Quotes
References: <bt1g9.17429$Pf7.590136@news1.west.cox.net> <20020912144710.GA78280@hishome.net> <mailman.1031861356.30411.python-list@python.org>
Message-ID: <3D8094C3.9040603@cox.net>

I use single quotes most of the time, but that's just because I'm lazy 
and if I use single quotes I don't need to press the shift key so much. ;-)

Michael Chermside wrote:
> Oren Tirosh wrote:
> 
>> On Thu, Sep 12, 2002 at 02:25:11PM +0000, John Waycott wrote:
>>
>>> I'm curious if others have adopted any standards for choice of single 
>>> vs.
>>> double quotation marks to delimit strings. [...]
>>
>>
>> The default repr of a string uses single quotes unless the string 
>> contains a single quote. Most of the time I use the same convention 
>> but I'm not 100% consistent.
>>
>> For docstrings I always use triple double-quotes even if the docstring 
>> fits in a single line.
>>
>> I use triple single-quotes to comment out blocks of code. This is very 
>> useful since "if 0:" would require re-indentation.
>>
>>     Oren
> 
> 
> Encouragingly, I use EXACTLY the same conventions and practices that 
> Oren does, having developed them independently for the same reasons.
> 
> -- Michael Chermside
> 
> 
> 
> 



From propadovic.nenad at debitel.net  Thu Sep 19 18:13:37 2002
From: propadovic.nenad at debitel.net (Nenad Propadovic)
Date: Fri, 20 Sep 2002 00:13:37 +0200
Subject: Need advice on choosing the technology for intranet based application
References: <3d88fd1e@news.ivm.net> <mailman.1032457645.15159.python-list@python.org>
Message-ID: <3d8a4a75@news.ivm.net>

Dear everybody,
thank you for all the good advice. I think I now have all the hints I need
(or, at least, most of it).
I will definitely start reading Steve Holdens book for some more basics. I
will NOT think about Zope at the moment (seems to be the one opinion in
common), but I will do a short evaluation on the other recomended
techniques & technologies.
This is a fun place to be! :-)
Nenad








From theoryboy at my-deja.com  Wed Sep  4 10:09:37 2002
From: theoryboy at my-deja.com (Peter Saffrey)
Date: 4 Sep 2002 07:09:37 -0700
Subject: efficient list merging
References: <ced73313.0209031012.13624f3@posting.google.com> <al3bfi$2kh$0@216.39.172.122>
Message-ID: <ced73313.0209040609.1c184ef8@posting.google.com>

Sorry, there's a mistake in my previous posting. The merged listing should be:

mergelol = [['a'], ['a', 'b', 'c'], ['d', 'e', 'f'], ['g', 'h', 'i',
'j'], ['k', 'l', 'm'], ['d'], ['f', 'g']]
                               ^^^^^^^^
I'd missed the last non-repeated sub-list from the second list.

Peter


From peter at engcorp.com  Thu Sep 26 08:38:09 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 26 Sep 2002 08:38:09 -0400
Subject: html in xml problem
References: <9a1d8b0f.0209260409.3b759cba@posting.google.com>
Message-ID: <3D92FFB1.5133DEC7@engcorp.com>

walter hoolwerf wrote:
> 
> I need to place some HTML code in an XML file en process it with
> python.

The only relatively safe way to do something like this, and even then you can
get in trouble, is to use the <![CDATA["put all the nasty HTML here"]]>
construct.

The "nasty HTML" cannot contain the sequence ]]> but as long as you can
guarantee that, this should be safe for embedding HTML, or almost anything,
in XML.

-Peter


From aleax at aleax.it  Wed Sep 18 10:00:10 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 18 Sep 2002 14:00:10 GMT
Subject: minimizing Tk programs to the taskbar
References: <3D8872F1.87850856@optushome.com.au> <Vz_h9.126208$pX1.4528221@news2.tin.it> <656913E3D19D40D2.0831BD5165F06A87.FDA870F87AEF2207@lp.airnews.net>
Message-ID: <KF%h9.126691$pX1.4542200@news2.tin.it>

Cameron Laird wrote:

> In article <Vz_h9.126208$pX1.4528221 at news2.tin.it>,
>
> Alex Martelli  <aleax at aleax.it> wrote:
> >Richard Tardif wrote:
> >> Is there any way to minimize programs to the taskbar? I'm talking about
> >> windows here. By taskbar I mean the area down in the bottom right
> >> corner, near the clock, that shows icons for what proggies you've got
        ...
> ?  Alex and Richard, I think you'll want to experiment with
>   root = Tkinter.Tk()
>     ...
>   Tkinter.Wm.iconify(root)

Nope: this (at least on Win/98 -- I have no other Win release at hand to 
try,
but I don't think the difference matters here) minimizes the window in the
normal way, NOT to the system-tray, which Richard described as "the
area down in the bottom right" and (probably mis-)named "taskbar".

> Alex' intuition that a lot of the operations people say they
> want are platform- or window-manager-specific is certainly
> apt.

Sure, but Richard did say he meant Windows (though he probably should
have uppercased that) -- that's pretty much an acknowledgment of
being platform-specific.  On Win in particular, crowding the system tray
with a zillion icons was all the rage once -- so many programs did it,
particularly device drivers and various daemons, as to make things more
or less unusable.  Win/XP now tries to help by hiding most of the system
tray except icons used recently, unless one specifically 'opens' the tray.
It's still somewhat of a mess (and becomes worse, the more programmers
learn to use Shell_NotifyIcon:-), but that's another issue.


Alex



From rajarshi at presidency.com  Fri Sep 20 14:53:07 2002
From: rajarshi at presidency.com (Rajarshi Guha)
Date: Fri, 20 Sep 2002 14:53:07 -0400
Subject: mod_python and authentication
Message-ID: <pan.2002.09.20.14.53.06.183620.1805@presidency.com>

Hi,
  when I want to use the Publisher method from mod_python I set up the
directory entry (in http.conf) like so:

    SetHandler python-program
    PythonHandler mod_python.publisher

However, I would like to add authentication. So I add the lines:

    PythonAuthenHandler entry
    AuthType Basic
    AuthName "DANGER!!!"

And then have a file called entry.py which has the function
authenhandler, the username and passwd dialog popup from the browser, but
after it authenticates it says that 

Directory index forbidden by rule: /var/www/secured/todo/

Ideally I would like the authenhandler to serve up  a page (say by using
abc.xyz.com:/form.py/details ).


I have also tried this: before returning apache.OK I simply
write the HTML using req.write(). However then I get the error:

Directory index forbidden by rule: /var/www/secured/todo/

in the logfile. But I also get the page rendered along with the error
message at the bottom of the page.

Is this the right way to do it? Or is this possible using the Publisher
method - if so, what prevents a user from simply typing the url and
bypassing the authentication function?

Thanks,


From see_reply_address at something.invalid  Wed Sep 18 19:52:53 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Thu, 19 Sep 2002 11:52:53 +1200
Subject: Getting data into an array in NumPy from an other object?
References: <am98u9$1mr$1@news.tudelft.nl>
Message-ID: <3D8911D5.8050000@something.invalid>

Maurice van de Rijzen wrote:

> Dear all,
> I've encountered a problem.
> In one object I've a function:
> 
> class blah:
>     ...
>     ...
>     def GetData(self):
>         return self.data
> 
> 
> Now I want to put his data into an instance of Numeric.array.


You'll have to tell us more about the nature of your
data object before we can give you much help!

If it's a sequence (i.e. a list or tuple) or something
which behaves like a sequence (i.e. len(data) and data[i]
work) you should be able to just pass it directly to the
Numeric.array constructor.

If not, you'll have to tell us more about it.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From kowk at earthlink.net  Tue Sep 17 02:24:29 2002
From: kowk at earthlink.net (Kow K)
Date: Mon, 16 Sep 2002 23:24:29 -0700
Subject: Python on MacOS X 10.2: a warning and instructions
In-Reply-To: <am5log$1ip2$1@nntp6.u.washington.edu>
Message-ID: <1F30F09A-CA06-11D6-8521-00050287C0A6@earthlink.net>

Oh my heck,

I probaby won't migrate to Jaguar for a while.
I can't trust Apple these days. Now, Apple is such a
shameless company that they don't care privileges
of previous users. Why do 10.1.x users have to pay as
much as $129??? No matter how tight their budget is,
that sucks.

Well, I will upgrade when it becomes Tiger.

Sorry about this posting, which is totally unrelated to Python.

Kow

On Monday, September 16, 2002, at 03:25 PM, Russell E. Owen wrote:

> MacOS X 10.2 (Jaguar) breaks the fink installation of Python (due to
> moving the ncurses package).
>
> MacOS X 10.2 does include Python (at least if you install the
> Developer's CD, not sure otherwise) but it has no readlines support
> (i.e. no command recall) and no Tkinter support.
>
> I upgraded before I realized this, and needed python too much to wait
> for the fink binary, so I spent a day and a half figuring out how to do
> the installation (with help from the MacPython list). Instructions are
> at <http://arcserve.astro.washington.edu/InstallingPythonOnJaguar.html>
> if you are unfortunate enough to need them.
>
> These instructions install a version that uses xfree86 (an X server) to
> display windows. However, a release version of the native aqua version
> of tcl/tk now exists (version 8.4.0) but so far the only binary
> installer is for 8.4a4, an older version. The source installer comes
> with no instructions and what I tried failed. If anybody gets Python
> with Tkinter to worth with the aqua version of tcl/tk 8.4.0 under MacOS
> 10.2, I would love instructions.
>
> -- Russell
> --
> http://mail.python.org/mailman/listinfo/python-list
>




From hst at empolis.co.uk  Tue Sep  3 05:05:24 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Tue, 3 Sep 2002 10:05:24 +0100
Subject: string reading bug in ActiveState Python 2.2 ?
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220CFB@hendrix.empolisuk.com>

Axel Kowald wrote:[mailto:kowald at molgen.mpg.de]
> 
> I'm using ActiveStates Python 2.2.1 under win2000 to read in the 
> attached text file and it seems that python is only reading 
> part of the 
> string. After
> 
> fp = open("bla.txt")
> z = fp.read()
> print len(z), z
> 
> I see that only 11 of the 15 characters have been read. I 
> know they are 
> non-printable characters, but that shouldn't matter, should 
> it ?  Using 
> another python on one of our unix machines everything works 
> as expected 
> and all 15 chars are read in.  Is that there something special about 
> windows that I should know, or is this a bug in ActiveStates Python ?
> 
> Many thanks,
> 
>                    Axel Kowald
> 
The 12th of the fifteen characters is \x1a which is EOF on Windows in text mode (the default).
Use fp = open("bla.txt", 'rb') and you will read all 15 characters.

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.



From greek_bill at yahoo.com  Wed Sep 18 14:16:29 2002
From: greek_bill at yahoo.com (Morfeas)
Date: Wed, 18 Sep 2002 19:16:29 +0100
Subject: SWIG and C++
Message-ID: <Xo3i9.3567$8t.155902@newsfep1-win.server.ntli.net>

I don't know if this is the right place to post this, but I couldn't find 
anywhere better.

I'm trying to use SWIG to interface C++ and python, but I'm getting an 
error. Here's what I have :

// Hello.h
class Hello
{
public:
        Hello(double aa, int bb);
        double GetA();
        double GetB();

private:
        double a;
        int b;
};


// Hello.i
%module Hello
%{
        #include "Hello.h"
%}

%include "Hello.h"


Then execute :
swig -proxy -c++ -python Hello.i
 g++ -c Hello_wrap.cxx -I/usr/include/python2.2/ 
 g++ --shared Hello_wrap.o Hello.o -o Hellomodule.so 
-I/usr/include/python2.2/

Python 2.2.1 (#1, Apr  9 2002, 13:10:27)
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import Hello
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ImportError: dynamic module does not define init function (initHello)
>>>

Any clues? SWIG generates a Hello_wrap.cxx, which has a 
#define SWIG_init    init_Hello

is this related? Is there any way to fix this?

Thanks!
Morfeas

-- 
www.freecfm.com/m/morfeas


From blackhole at notformail.invalid.de  Tue Sep 24 11:46:23 2002
From: blackhole at notformail.invalid.de (Volker Apelt)
Date: 24 Sep 2002 17:46:23 +0200
Subject: callback + __call__ + TypeError: unbound method; on non method
References: <lg8z1rtvzi.fsf@mu.biosolveit.local> <JiZj9.157856$pX1.5687433@news2.tin.it>
Message-ID: <lgwupbs84w.fsf@mu.biosolveit.local>

Alex Martelli <aleax at aleax.it> writes:

> Volker Apelt wrote:
> 
> > But I can't remove the .__call__ method from my
> > Callback class. The method is essential for
> 
> The problem is not with method __call__, as it gets called
> with an instance.  Rather, it's right here:
> 
> > class Callback:
> >     """
> >     """
> >     _default_Func = call_me # class wide default callback function
> >     
> >     def __init__(self):
> >         """ """
> >         self._m_func = Callback._default_Func
> 
> Here, self._m_func is NOT a function -- it's an unbound method,
> thus it's not callable unless the first argument is an instance
> of the class (or any subclass thereof).
> 
> When you fetch from a class an attribute that is a Python coded
> function object, what you get is an unbound method object.  This
> behavior is intrinsic to Python's OO mechanisms.
> 
> Remedy: have Callback._default_Func be just about ANY callable
> except a Python-coded function object:-).  In Python 2.2, the
> built-in staticmethod is made just for the purpose:
> 
> class Callback:
>      """
>      """
>      _default_Func = staticmethod(call_me) # class wide default callback 
> 
> and the rest can stay unchanged (haven't examined it in detail, but
> this is the surely-buggy spot in your code and explains the traceback
> that you report).

Let me rephrase what you stated above.  Callback._default_Func is 
*always* a method. Thats the way python understands this code. 
So Python below v2.2 can not store a function reference in a 
class static variable (c++ speak) and use it later as for function
call. 

So, I have to upgrade to v2.2 and write:

class Callback:
    """
    """
    _default_Func = staticmethod(call_me)# class wide default callback function
   
    def __init__(self):
         """ """
         self._m_func = staticmethod(Callback._default_Func)

    def __call__(self):
         self._m_func() # << this will call call_me()  ???


Thank you Alex,


Volker
-- 
Volker Apelt                   


From aglyportREMOvethispart at nospam.yahoo.com  Wed Sep  4 18:21:07 2002
From: aglyportREMOvethispart at nospam.yahoo.com (Anton Graph)
Date: Wed, 04 Sep 2002 15:21:07 -0700
Subject: Memory fault - core dumped   while compiling 2.2.1 on SCO
References: <3D757EFA.9050401@nospam.yahoo.com> <yu99sn0qqw1r.fsf@europa.research.att.com> <m3bs7etem9.fsf@mira.informatik.hu-berlin.de> <3D765A05.6030004@nospam.yahoo.com> <j48z2h61j9.fsf@informatik.hu-berlin.de>
Message-ID: <3D768753.8030302@nospam.yahoo.com>

Martin v. L?wis wrote:
> Anton Graph <aglyportREMOvethispart at nospam.yahoo.com> writes:
> 
> 
>>Did not even take nearly as long for it to coredump. With optionless
>>configure build was going at least for a half an hour and this one
>>bailed out pretty much right away...
> 
> 
> Can you please try to run the resulting Python binary in a debugger
> (say, dbx), and report a backtrace?
> 

python -E setup.py build in gdb:

(gdb) where
#0  0x80016d06 in do_reloc () from /usr/lib/libc.so.1
#1  0x80016bec in _relocate () from /usr/lib/libc.so.1
#2  0x800181fd in _rtld () from /usr/lib/libc.so.1
#3  0x80015569 in dlopen () from /usr/lib/libc.so.1
#4  0x808e315 in _PyImport_GetDynLoadFunc ()
#5  0x8086ec0 in _PyImport_LoadDynamicModule ()
#6  0x8085372 in load_module ()
#7  0x8085ffb in import_submodule ()
#8  0x8085c01 in load_next ()
#9  0x808583e in import_module_ex ()
#10 0x8085988 in PyImport_ImportModuleEx ()
#11 0x80b7a61 in builtin___import__ ()
#12 0x80b629a in PyCFunction_Call ()
#13 0x80736dd in eval_frame ()
#14 0x80748b6 in PyEval_EvalCodeEx ()
#15 0x8075794 in fast_function ()
#16 0x80737a6 in eval_frame ()
#17 0x80748b6 in PyEval_EvalCodeEx ()
#18 0x80ab7cd in function_call ()
#19 0x809d8fc in PyObject_Call ()
#20 0x80a3225 in instancemethod_call ()
#21 0x809d8fc in PyObject_Call ()
#22 0x80759fc in do_call ()
---Type <return> to continue, or q <return> to quit---
#23 0x80737c2 in eval_frame ()
#24 0x80748b6 in PyEval_EvalCodeEx ()
#25 0x8075794 in fast_function ()
#26 0x80737a6 in eval_frame ()
#27 0x80748b6 in PyEval_EvalCodeEx ()
#28 0x8075794 in fast_function ()
#29 0x80737a6 in eval_frame ()
#30 0x80748b6 in PyEval_EvalCodeEx ()
#31 0x8075794 in fast_function ()
#32 0x80737a6 in eval_frame ()
#33 0x80748b6 in PyEval_EvalCodeEx ()
#34 0x8075794 in fast_function ()
#35 0x80737a6 in eval_frame ()
#36 0x80748b6 in PyEval_EvalCodeEx ()
#37 0x8075794 in fast_function ()
#38 0x80737a6 in eval_frame ()
#39 0x80748b6 in PyEval_EvalCodeEx ()
#40 0x8075794 in fast_function ()
#41 0x80737a6 in eval_frame ()
#42 0x80748b6 in PyEval_EvalCodeEx ()
#43 0x8075794 in fast_function ()
#44 0x80737a6 in eval_frame ()
#45 0x80748b6 in PyEval_EvalCodeEx ()
---Type <return> to continue, or q <return> to quit---
#46 0x8075794 in fast_function ()
#47 0x80737a6 in eval_frame ()
#48 0x80748b6 in PyEval_EvalCodeEx ()
#49 0x8075794 in fast_function ()
#50 0x80737a6 in eval_frame ()
#51 0x80748b6 in PyEval_EvalCodeEx ()
#52 0x8071542 in PyEval_EvalCode ()
#53 0x808bb8e in run_node ()
#54 0x808bb46 in run_err_node ()
#55 0x808bb0f in PyRun_FileExFlags ()
#56 0x808aee1 in PyRun_SimpleFileExFlags ()
#57 0x808a9f2 in PyRun_AnyFileExFlags ()
#58 0x8052a64 in Py_Main ()
#59 0x8052300 in main ()
#60 0x805224b in _start ()

Noone around here admits making any changes in hardware or OS to the box 
I'm building on.



From b.maryniuk at forbis.lt  Mon Sep 16 04:10:19 2002
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Mon, 16 Sep 2002 10:10:19 +0200
Subject: CPAN for Python
Message-ID: <200209161010.19478.b.maryniuk@forbis.lt>

Hi.
I "heard" that Python will have something like CPAN for Perl soon. Is there is 
something new and moving forward in this area?

-- 
Regards, Bogdan

"I know you believe you understood what you think I said, but I am not
sure you realize that what you heard is not what I meant!"
                             -- G.Romney




From guy at lightwork.co.uk  Thu Sep 19 11:34:56 2002
From: guy at lightwork.co.uk (Guy)
Date: 19 Sep 2002 08:34:56 -0700
Subject: python on the mac.
Message-ID: <c07d0edc.0209190734.44c7f41f@posting.google.com>

Hi

I just managed to get sys admin to put an uptodate version of python
on our macs. I normally stay well away from the things as I'm not use
to there strange way of doing stuff (or should I stay different way of
doing stuff.)

I have a script which I created on a win32 machine, I now want to port
it to a mac sys9x, python version 2.2.1.

I've got the scripts over on to the mac with there modules and they
run.

Firstly when I was doing this I got an error which allowed me to open
up the debugger (which I got the hang of, and quite liked the look
of).

Question : How do I open the debugger without having to get an error
to start with ? The must be something quite obvious that I'm missing,
perhaps the afternoon is not helping (dragging on and on.)

I'm also having problems with command line arguments. Which I've been
reading a lot of posted messages on here about them.

Say I am running my script like this on win32 :

dist_set.py -V 3 -L c:\error.log           etc etc

can, is it possible to run this under mpw like this ?
One of the messages I read on the newsgroup said hold down the options
button on the mac and double click on the script, to produce a window
which can take unix style command line options. This only works if the
script has been converted to a applet. Is there any way of running the
python text file in python using the same idea ? (As the script I am
wanting to run won't convert to an applet, comes up with an error and
quits. no hint no nothing.)

If I could run my script in the debugger, using correct command line
args, then I might have a chance to see whats going on.

Many Thanks
Guy


From magnus at thinkware.se  Fri Sep 20 11:28:32 2002
From: magnus at thinkware.se (=?ISO-8859-1?Q?Magnus_Lyck=E5?=)
Date: Fri, 20 Sep 2002 17:28:32 +0200
Subject: [newbie] Calling Python from C++
References: <8L5i9.841$Ew4.62552@newsfep2-gui>
Message-ID: <3D8B3EA0.6050500@thinkware.se>

Morfeas wrote:
> What I want to do though, is to execute a python script that operates on an 
> existing object (that may or may not have been instantiated from Python). 
> So my question is : How do I pass an object into the python script?
...
> I'm kind of relying on SWIG to do all the data conversion to and from 
> python, so I don't really know much on this topic!

I'm really as clueless as you are about this, but as far
as I understand, there are better tools than SWIG to
connect Python with C++. SWIG is really a C thing rather
than a C++ thing, right?

Boost and CXX comes to mind. See
http://www.thinkware.se/cgi-bin/thinki.cgi/UsingPythonWithOtherLanguages

-- 
Magnus Lycka, Thinkware AB
Alvans vag 99, SE-907 50 UMEA, SWEDEN
phone: int+46 70 582 80 65, fax: int+46 70 612 80 65
http://www.thinkware.se/  mailto:magnus at thinkware.se



From bloke at ii.net  Sat Sep 14 23:10:24 2002
From: bloke at ii.net (Rob Hall)
Date: Sun, 15 Sep 2002 11:10:24 +0800
Subject: confused with class inheritance
Message-ID: <3d83fa2c$0$24041@echo-01.iinet.net.au>

In a sample app, I have the following:

>>> class a(sgmllib):
 pass

>>> a
<module '?' (built-in)>
>>> dir(a)
[]
>>> dir(sgmllib)
['SGMLParseError', 'SGMLParser', 'TestSGMLParser', '__all__',
'__builtins__', '__doc__', '__file__', '__name__', 'attrfind', 'charref',
'commentclose', 'endbracket', 'entityref', 'incomplete', 'interesting',
'markupbase', 'piclose', 're', 'shorttag', 'shorttagopen', 'starttagopen',
'tagfind', 'test']
>>> b=a
>>> dir(b)
[]
>>>

Why doesn't dir(b) show the same as dir(sgmllib) ?


Rob




From ben at remove.Stanford.EDU  Wed Sep 11 13:00:02 2002
From: ben at remove.Stanford.EDU (Ben Escoto)
Date: Wed, 11 Sep 2002 10:00:02 -0700
Subject: Large file support and stat vs stat64 in extension modules
References: <pan.2002.09.10.16.03.39.756929.4320@remove.Stanford.EDU> <mailman.1031750594.11844.python-list@python.org>
Message-ID: <pan.2002.09.11.10.00.02.210522.16129@remove.Stanford.EDU>

On Wed, 11 Sep 2002 06:22:08 -0700, Skip Montanaro wrote:

> Did you #include <sys/stat.h>?

Yes, but perhaps I'm not including something else.  Here is a short
program (test.c):

#include <sys/stat.h>

main()
{
  struct stat st;
  if (stat("/home/ben/work/big/bigfile", &st) != 0)
      perror("Error");
  else printf("OK\n");
}

when I "gcc test.c; ./a.out" it says: "Error: Value too large for defined
data type".  I assume in my extension module the same thing is happening. 
What needs to be defined so that this works?  Looking through
posixmodule.c it doesn't seem to define __USE_FILE_OFFSET64 or
__USE_LARGEFILE64, yet it gets things right.

> Note that in posixmodule.c references are made to a
> HAVE_LARGEFILE_SUPPORT macro.  When #defined, it causes the output
> struct to be interpreted differently (the sizes of some fields change).

> You might scan through your configure.in file.  If "long long" support
> is present and sizeof(off_t) is greater than sizeof(long) and
> sizeof(long long) is greater than or equal to sizeof(off_t), large file
> support is assumed.

I am using distutils so don't have a Configure.in.  But I think there are
two separate problems:  firstly getting the stat() call to work without
raising an error, and secondly interpreting the filled in stat structure.
Perhaps I will have problems with #2 later, but I'm still on #1 for now.
As far as I can tell HAVE_LARGEFILE_SUPPORT, off_t, and long long don't
affect this first step.

    Thanks a lot for your help!


-- 
Ben Escoto


From sadams123 at optushome.com.au  Mon Sep 30 19:23:07 2002
From: sadams123 at optushome.com.au (Steven)
Date: Tue, 1 Oct 2002 09:23:07 +1000
Subject: Id, or stinky primo?
References: <7A4m9.25416$V64.5149538@newsfeed.slurp.net>
Message-ID: <3d98dcdc$0$23174$afc38c87@news.optusnet.com.au>

"Bruce Edge" <edgebruce at yahoo.com> wrote in message
news:7A4m9.25416$V64.5149538 at newsfeed.slurp.net...
> Is the Id it really worth the $$$$?
> I've been looking at a kona stinky primo, and it seems like a comparable
> bike for less dough.
> What's the general opinion out there?
> Why is the Id better?

I think you posted to the wrong group

who makes the Id? what are you intending on doing with the bike? the the
Kona doesn't look like something you'd want to take on a downhill run.

Steven




From jb at cascade-sys.com  Tue Sep 10 17:12:25 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 10 Sep 2002 14:12:25 -0700
Subject: "str"(expr)"str" interpolation syntax
References: <Pine.GSO.4.33.0209102221070.10002-100000@techst02.technion.ac.il>
Message-ID: <3D7E6039.9060809@cascade-sys.com>

I don't see that your proposals materially improve upon

    "abc" + `x*3` + "def"

which already works.

Having previously used a number of languages where string concatination 
is implied by adjacent expressions and where numeric expressions in this 
context are automatically coerced to strings, I agree that having to 
explicitly spell out "+" and `` or str() seems a BIT tedious, at least 
at first.  
However, this is a case where I have to agree that it's no big deal and 
after using it for a while you get used to it.

Regards

--jb


Beni Cherniavksy wrote:

>Hi all.
>
>I was searching for a new interpolation syntax that would use quotes
>for starting/ending the interpolated part (I don't feel comfortable
>putting code inside unclosed quotes...)
>
>After contemplating some horrors like $"""str"expr"str""" or
>$("str" expr "str"), I noticed that currently both  "(  and  )"  are
>illegal (first because strings are not callable, second is a syntax
>error).  So I propose that a parenthesized expresion near a string would
>imply concatenation.  In other words,
>	"foo"(x * 3)"bar"
>would be equivallent to
>	"foo" + str(x * 3) + "bar".
>This works of course with only one side, e.g. "foo"(x) or (y)"bar" but in
>longer cases it would be good style to add empty strings on either end,
>e.g. ""(y)"bar".  And of course it all works with single quotes too.  One
>of the benefits over current interpolation PEPs is that you can change the
>quote type in the middle...
>
>It is slightly more verbose (especially when interpolating into tripple
>quote strings: """foo"""(x)"""bar""")  but it is more readable and
>understandable (just extends the current concatenation of constant
>strings), allows clear nesting of strings into the expr:
>	"foo"('%3d' % x)"bar"
>and most importantly, you never need to remember arbitrary rules where the
>interpolation ends...
>
>What do you think?
>





From llothar at web.de  Sun Sep  1 10:32:29 2002
From: llothar at web.de (Lothar Scholz)
Date: Sun, 01 Sep 2002 16:32:29 +0200
Subject: Python GUI
References: <3d8d3c0.0208262339.e7bc02c@posting.google.com> <uniomu8pdi7tm169b4g6225d5oqdrkj2go@4ax.com> <akobqe$4en$1@usenet.Stanford.EDU> <3D6FBB3B.5090700@river-bank.demon.co.uk>
Message-ID: <r794nugam0su9metpaas0nobn4jsm84j3j@4ax.com>

On Fri, 30 Aug 2002 19:36:43 +0100, Phil Thompson
<phil at river-bank.demon.co.uk> wrote:

>Or take the Qt (and PyQt) approach and implement all styles on all 
>platforms and let the user decide on a global or per application basis 
>with a simple command line argument.

Thats completely wrong it the user selects a style from the OS style
manager. At least for MacOS Users this is very popular and then
your QT application looks again like an alien. Same is true for WinXP


From sfeil at io.com  Fri Sep 20 11:51:06 2002
From: sfeil at io.com (Steven Feil)
Date: Fri, 20 Sep 2002 15:51:06 GMT
Subject: [Jython] serialization transient
Message-ID: <KtHi9.161792$z91.6588460@bin3.nnrp.aus1.giganews.com>

I would like to prevent Jython from attempting to write out
unserializable variable within a class.  Is there a Jython equivalent
to the java keyword transient?

========================================================================
 Steven Feil               | Gram-pa, back at the turn of the      .~.  
 Programmer/Developer      | century, why did people use an        /V\  
 sfeil at io.com              | operating system, when they were not // \\ 
                           | allowed to see the source code?      (X_X) 
========================================================================


From jb at cascade-sys.com  Wed Sep 25 12:37:33 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Wed, 25 Sep 2002 09:37:33 -0700
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <mailman.1031377930.4225.python-list@python.org> <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com> <3D8E6353.7030303@something.invalid> <dvbtouciidaufjneja3s4t89ds0r7jd76k@4ax.com> <3D911A1B.1000109@something.invalid> <a6f9355c.0209250720.4512b3c3@posting.google.com>
Message-ID: <3D91E64D.8030802@cascade-sys.com>

Alan Daniels wrote:

>Sometimes I wonder if Guido would have saved a lot of debate by declaring,
>long ago, that Python code would never allow tabs in it, period. Which may
>be a bit extreme, I'll admit. :-)
>
Ironically, Guido originally ruled (in the style guide, IIRC) that thou 
shalt use tabs and tabs are 8 spaces, "as god intentended them to be."

Unfortunately, he later changed position to the present, unsatisfying, 
compromise position.

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From phr-n2002b at NOSPAMnightsong.com  Fri Sep 20 08:44:36 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 20 Sep 2002 05:44:36 -0700
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amabpb$5rm$1@news.nems.noaa.gov> <3D8B1537.4010803@thinkware.se>
Message-ID: <7xn0qcvnij.fsf@ruckus.brouhaha.com>

Magnus Lyck? <magnus at thinkware.se> writes:
> Ok, you might say that it's not a big thing to write
> 
> x = 5 * Decimal(2.43)
> 
> instead, but it's not quite the same thing.

It would have to be x = 5 * Decimal('2.43')


From aldo at nullcube.com  Sun Sep 15 00:40:24 2002
From: aldo at nullcube.com (Aldo Cortesi)
Date: Sun, 15 Sep 2002 14:40:24 +1000
Subject: two or more whitespace regular expression?
In-Reply-To: <am12a3$qo0$1@newsreader.mailgate.org>
References: <am12a3$qo0$1@newsreader.mailgate.org>
Message-ID: <20020915044024.GA21615@nullcube.com>

Thus spake eugene kim (eugene1977 at hotmail.com):

> i used this..
> 
> words = string.split(line, '['+string.whitespace 
> +']'+'['string.whitespace+']+')


The string "split" method does not take regular expressions
- you need to use the re module for that. Try the following:

import re
line = "a b  c   d    e"
print re.split(r"\s+", line)


Read the re module documentation for more information on
regular expression syntax. 




Cheers,


Aldo






-- 
Aldo Cortesi
aldo at nullcube.com
http://www.nullcube.com



From aleax at aleax.it  Mon Sep 23 03:54:30 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 23 Sep 2002 07:54:30 GMT
Subject: Decimal arithmetic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amgpm4$5si22$1@ID-125932.news.dfncis.de> <7xwupf52jx.fsf@ruckus.brouhaha.com> <TXOVyjACLDj9EwHM@jessikat.fsnet.co.uk>
Message-ID: <WMzj9.150692$pX1.5406331@news2.tin.it>

Robin Becker wrote:
        ...
> rules evolved before modern computers are mostly computable using
> fingers and counting. In particular almost all the calculations have to
> be with pre-specified precision and rounding rules.

Right.

> EG a bond pays fixed amounts of money at pre-specified intervals, no
> calculations are involved at all. Of course when I want to compute the
> yield or npv I need to do some arithmetic and FP is usually easiest and
> faster as well.

Wrong!  At least, if you have to follow legally mandated procedures
to post the yield, etc -- and at least in the EU there are quite a
few cases where you do.  For example, whenever you offer something
for sale to the public on any kind of instalment-payment plan, you
have to publish TWO different values for the interest rate that the
buyer ends up paying, exactly because there are subtly different
ways to compute such things and the legislators wanted to make VERY
certain that no seller tries to skew the computation and thereby
misinform buyers.  The two ways of computing interest use the
acronyms TAEG and TAEV in Italy -- I'm not sure what they refer to,
exactly: in practice, everybody looks at the higher one of the two
before entering into any financially substantial transaction (and/or
for a LARGE discrepancy between them, to take as an indication that
there is something fishy about the whole deal -- but fractions of a
percent of difference are quite typical, and the numbers are posted 
with specified accuracy of two decimal digits, e.g. "TAEG 10.23%").

I don't think there's any degree of freedom about how you can perform
these calculations -- the whole purpose of the legislation seems to
be exactly to ensure that sellers do NOT get any freedom to choose
what numbers to display in order to mislead buyers in any way.  In
particular, I think rounding procedures are strictly specified (but
I don't know the exact details).


> Anyhow it seems to me that BCD is pretty good only for additive type
> calculations. I can't see that BCD is any better than FP at representing
> 1/3 unless we go fully rational, but that's another story.

Yes, but I don't see legal specifications to use such fractions; rather,
I see specifications such as "18.75% of the sum" -- always a finite
number of decimal digits.  Your quoted paragraph at the start may
explain why -- things must in principle be computable with "fingers
and counting"... and, of course, "pre-specified precision and
rounding rules", which are never expressed in terms compatible with
binary floating-point hardware... always in terms of somebody
conducting a decimal computation with finite numbers of digits.


Alex



From jh at web.de  Fri Sep 27 03:46:35 2002
From: jh at web.de (Juergen Hermann)
Date: Fri, 27 Sep 2002 09:46:35 +0200
Subject: [XML-SIG] well-formed xml
In-Reply-To: <JHEOKEOOLIGLDHCMAHMOAEOPEBAA.mark@mceahern.com>
Message-ID: <E17uppg-0007LF-00@smtp.web.de>

On Thu, 26 Sep 2002 19:08:48 -0500, Mark McEahern wrote:

>I'm obviously missing something because this seemingly innocent chunk of
>xhtml:
>
>  from xml.dom import minidom
>
>  s = "<a href='http://google.com/search?hl=en&q=foobar'>search</a>"

This is not well-formed and thus not XHTML.

... en&amp;q= ...

is the correct form.


Ciao, J?rgen

--
J?rgen Hermann, Developer
WEB.DE AG, http://webde-ag.de/





From e_fax_t at hotmail._ZAPME_.com  Wed Sep 25 12:06:10 2002
From: e_fax_t at hotmail._ZAPME_.com (Justin)
Date: Wed, 25 Sep 2002 18:06:10 +0200
Subject: readlines() and "binary" files
References: <3d90cacf$0$30457$afc38c87@sisyphus.news.be.easynet.net> <9r8k9.46422$V7.11170409@twister.socal.rr.com>
Message-ID: <3d91de2d$0$30457$afc38c87@sisyphus.news.be.easynet.net>

> I think readlines() is just a shortcut for a very common task. Since your
> task isn't quite as common, I think it would be a better idea to use
> read() to read the whole thing, splitting the lines up by the 0x0d 0x0a
> pair (CR NL).

Good suggestion, Jeff, thanks. I did not not realise that even data
read from a 'binary' stream could be split()'ed.
Maybe someone else will have a use for this, and I wanted to correct
what I had said: it's extra newlines and not carriage returns that
the data contains.

        f = open('testdata.csv','rb')
        # swallow everything in memory
        block = f.read()
        lines = block.split( "\r\n" )
        f.close()
        # free our in-core file copy asap
        del block

        for line in lines:
            # get rid of the odd extra NL
            line = line.replace( "\n", "_" )
            fields = line.split(';')
            print len(fields)    # now constant

My biggest data file is about 8 megs, so reading it in one swoop
is doable. Still I only actually need to look at about 120 bytes
at a time, so that's rather overkill. If anyone can think of a more
economical way of doing it (short of defining my own iterator), I'd
be interested.

Thanks.




From magnus at thinkware.se  Fri Sep 20 08:31:51 2002
From: magnus at thinkware.se (=?ISO-8859-1?Q?Magnus_Lyck=E5?=)
Date: Fri, 20 Sep 2002 14:31:51 +0200
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amabpb$5rm$1@news.nems.noaa.gov>
Message-ID: <3D8B1537.4010803@thinkware.se>

ChrisBarker wrote:
> The only reason I can see
> to not writing your class in Python is performance, and unless you are 
> doing a lot of calculations, that is probably a non-issue.

Performance issues might well matter in large financial
systems, but that's not the only reason. If we want a
statement like

x = 5 * 2.43

to yield an exact decimal number, and not a float, we can't
get away with a Python class.

Ok, you might say that it's not a big thing to write

x = 5 * Decimal(2.43)

instead, but it's not quite the same thing. And if someone
would write

x = Decimal(9.999999999999999) - Decimal(9.9999)

they would not get what they expected (unless they
expected 0.000099999999998). Ok, they could write

x = Decimal('9.999999999999999') - Decimal('9.9999')

but now it's starting to look pretty ugly compared to

x = 9.999999999999999 - 9.9999

Python is claimed to use the principle of least surprise:
Thus 1 / 2 is being changed from meaning 0 to meaning 0.5.
Quite a few people have expressed their surprise over things
like

 >>> 0.9
0.90000000000000002

and reported it as a bug. Floating point arithmetic is
surprising to a lot of people. It's annoying when people
report this behaviour as a bug. It's worse if it leads
them to make buggy programs. Particularly if they are
used at your bank!

I've seen ambitious and appreciated programmers with a
non-technical background doing things like
"if (a == (b / 5))" in tools that create C++ using
doubles for a and b.

Python is in many ways an excellent language for people
who are beginners at programming, or not full time
programmers. It would be great if we could remove some
more of the remaining stumbling blocks.

What you see as a natural restriction, completely logic,
is an evil, subtle and completely absurd bug to some people.

I don't want to break a lot of existing code though.

It's boring if we have to change a lot of 3.14 to 3.14e0.
That's hardly acceptable before Python 3.0.

But perhaps a command line argument or something like
"import exact_arithmetic" could change the behaviour to
using exact instead of floating point arithmetic. I guess
we'd need a module anyway, to specify things. Like:

 >>> import exact_arithmetics as exact
 >>> exact.default_rounding(exact.UP)
 >>> exact.default_fractions(2)
 >>> exact.max_precision = 20
 >>> 7.5/20
0.38
 >>>





From ianb at colorstudy.com  Tue Sep 17 01:53:33 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 17 Sep 2002 00:53:33 -0500
Subject: Stackless Python, eventual merge?
In-Reply-To: <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de>
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com>
	<mailman.1032107006.6411.python-list@python.org>
	<am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net>
	<j4znuio0kr.fsf@informatik.hu-berlin.de>
	<mailman.1032204706.14737.python-list@python.org> 
	<m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de>
Message-ID: <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com>

On Mon, 2002-09-16 at 16:50, Martin v. Loewis wrote:
> I looked at the old Stackless patches once, and spotted significant
> amounts of unexplicable code - code that did not serve any obvious
> purpose, was not documented, or clearly should have been removed.  So
> the old Stackless patch was incomplete for integration into Python -
> it would have taken efforts to complete it, and to work with Python
> maintainers to integrate it. Nobody went through this effort (and
> again, I don't blame anybody for not doing that).
> 
> >From a shallow inspection, the code merely looked complex - but it
> really was complicated, and complex is better than complicated.

But it doesn't seem like complexity was ever the fundamental problem --
the complexity could have been dealt with, but I don't believe there was
ever any indication that if it was simplified or documented that it
would get into standard Python.  

It has been simplified, but I don't think that changes anything with
regard to getting it into standard Python -- the same barriers still
exist.  It's simplification just means that it might be easier (and more
viable) to continue to maintain a forked version.

  Ian





From jkraska at san.rr.com  Sat Sep 28 21:45:57 2002
From: jkraska at san.rr.com (Courageous)
Date: Sun, 29 Sep 2002 01:45:57 GMT
Subject: What does Python fix?
References: <e15b6fa3.0209271418.14cd5642@posting.google.com> <flaapucbduj6g4ifrdf3oq99hdrepi1o3v@4ax.com> <3d95fb09@news.vip.fi>
Message-ID: <1hmcpusq28e8fnbpkashc5coh5n2iot98d@4ax.com>

>After studying Python for 2 years I myself have
>recently switched to DrScheme because I have no interest
>relearning C to write extensions to Python.
>
>What an eye opener it has been to study scheme...

Scheme continuations are pretty cool.

A fair criticism of Python. "Python is not for that," applies to the
degree that you can't get Python to work without extensions for certain
classes of compute-bound problems. OTOH, I've prototyped whole systems
under Python in order to get complex design right. One such system, later
ported fully to C++, actually really wasn't all that much faster. It was
fast enough to matter (2 hours of compute time is indeed better than
5), but for small tasks, the limit formula pushes the practical difference
to zero. What's the difference between 20 and 4 milliseconds? In *user
time*, that's no difference.

C//



From andrewa at calibredigital.com  Mon Sep 23 14:23:53 2002
From: andrewa at calibredigital.com (Andrew Alzner)
Date: Mon, 23 Sep 2002 14:23:53 -0400
Subject: parsing tabs in a string
References: <JHEOKEOOLIGLDHCMAHMOOELOEAAA.marklists@mceahern.com>
Message-ID: <008701c2632e$5f54e5c0$025e10ac@calibredd.com>

Thanks everyone for the quick response! This helps out a lot. I will be
using the string.split to make it more manageable.

Andrew

----- Original Message -----
From: "Mark McEahern" <marklists at mceahern.com>
To: "Andrew Alzner" <andrewa at calibredigital.com>; <python-list at python.org>
Sent: Monday, September 23, 2002 1:18 PM
Subject: RE: parsing tabs in a string


> [Andrew Alzner]
> > I'm trying to parse information extracted from a spreadsheet
> > which is tab separated. Here is a simplified version of what I have:
> >
> > list=['"103-01a-17"\t2\t33\t256\t227\n']
> > y=list[0]
> > print y.find("\t")
> > print y.find("\t",1)
>
> Don't use list as a variable name, you'll stomp all over the builtin list.
> Rather than finding each tab, why not split on tabs:
>
> a = ['"103-01a-17"\t2\t33\t256\t227\n']
> y = a[0]
> for x in y.split("\t"):
>   print x
>
> // m
>




From ods at strana.ru  Tue Sep 24 11:43:58 2002
From: ods at strana.ru (Denis S. Otkidach)
Date: Tue, 24 Sep 2002 19:43:58 +0400 (MSD)
Subject: hex numbers??
In-Reply-To: <574747556.1032881448716.JavaMail.nobody@webmail2.brturbo.com>
Message-ID: <Pine.LNX.4.44L.0209241942540.31141-100000@ods.pp.ru>

On Tue, 24 Sep 2002 jubafre at brturbo.com wrote:

> >>> y=hex(x)
> >>> y
> '0xa'
> >>> type(y)
> <type 'str'>
>
> the hex() function put the hexadecimal number in a string type
> like this '0xa', but the real hexadecimal number of 10 is 'a',
> why python puts a '0x',  and how i can get just the real
> hexadedcimal number, for example i want to write in a file 10
> in hexadecimal, but if i use hex(10) the result is '0xa' as a
> string, is not the same as 'a'.
>
> '0xa' != 'a'
>
> x = [12,13,14]
> y = ['0xc', '0xd', '0xe']
>
> hex() create a list of string,  how i can write in a file just
> the real value of hexadecimal number?

>>> hex(10)[2:]
'a'
>>> '%x' % 10
'a'
>>> '%X' % 10
'A'

-- 
Denis S. Otkidach
http://www.python.ru/      [ru]
http://diveinto.python.ru/ [ru]




From dan65536 at hotmail.com  Wed Sep 18 22:20:03 2002
From: dan65536 at hotmail.com (daniel w. moore)
Date: 18 Sep 2002 19:20:03 -0700
Subject: conditional expressions
Message-ID: <33b491f.0209181820.17b6ea70@posting.google.com>

It's occurred to me after some brainstorming on the subject that the
C/perl construct

  a ? b : c

can be arranged like this in python:

  (a and [b] or [c])[0]

or, putting into lambda form, and using tuples for a slight efficiency
gain:

  cond_switch = lambda a,b,c : (a and (b,) or (c,))[0]

Has anyone put this to good use, i wonder?
Or have I just made a daring and heroic discovery in my own particular
idiom ;)
(http://www.stone-dead.asn.au/movies/holy-grail/scene-15.html)


From kp at kyborg.dk  Fri Sep 20 04:31:34 2002
From: kp at kyborg.dk (Kim Petersen)
Date: Fri, 20 Sep 2002 10:31:34 +0200
Subject: Python embedded - linking problem on Linux - RESOLVED
References: <3D57BD22.8090305@kyborg.dk>
Message-ID: <3D8ADCE6.5020404@kyborg.dk>

Thank you to everyone who helped me on this one....

The trouble *was* the Cobol call to dlopen() ... it did _not_ have the 
RTLD_GLOBAL flag set on their load of my extention (their own way of 
extending). (After some quaralling over dlopen() i finally got them to 
send me a patched version of their runtime with RTLD_GLOBAL to test).

Conclusion: in the *first* call to a chain of dynamic objects, it is 
neccesary to do dlopen() with RTLD_GLOBAL - otherwise python will not be 
able to export its symbols to its own symbols.

-- 
Med Venlig Hilsen / Regards

Kim Petersen - Kyborg A/S
IT - Innovationshuset
Havnepladsen 2
7100 Vejle
Tlf. +4576408183 || Fax. +4576408188



From davids at simutility.com  Thu Sep 19 15:40:28 2002
From: davids at simutility.com (Dave Swindell)
Date: 19 Sep 2002 12:40:28 -0700
Subject: Controlling IE 5.x from Python: HTTP Status Codes
Message-ID: <db2ceea9.0209191140.3e5b126e@posting.google.com>

I'm trying to control IE 5.5 from Python and ham having some problems.
Using PythonWin with Mark Hammond's COM interface extensions, I've
been able to launch IE and navigate to a specified website using the
following code:

from win32com.client import constants, Dispatch, DispatchWithEvents

ie=Dispatch("InternetExplorer.Application)
ie.Visible=1
ie.Navigate("http://www.python.org",2)

This all works fine, except that I'd like to know the HTTP Status Code
that resulted from the Navigate automation interface in IE. Using the
DispatchWithEvents call, I've been able to trap OnDocumentComplete and
OnNavigate2Complete events, but these don't appear to provide you with
access to the HTTP status codes resulting from the Navigate call.

IE 6.0 has a NavigateError event which, according to the
documentation, will allow you to access the HTTP status code, but the
project I'm working on needs to use IE 5.0/5.5.

Anyone else run into this issue or even better, found a way around it?

Thanks,

    -- Dave Swindell
       davids at simutility.com


From mstenner at phy.duke.edu  Thu Sep 12 09:32:21 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Thu, 12 Sep 2002 09:32:21 -0400
Subject: how can i make a executable program with a python2's script
In-Reply-To: <dd083388.0209111829.671b2a1b@posting.google.com>; from jingguoming@yahoo.com.cn on Wed, Sep 11, 2002 at 07:29:00PM -0700
References: <dd083388.0209060003.1188000c@posting.google.com> <3D7924BF.805@motorola.com> <mailman.1031351050.14744.python-list@python.org> <dd083388.0209111829.671b2a1b@posting.google.com>
Message-ID: <20020912093221.A6377@phy.duke.edu>

On Wed, Sep 11, 2002 at 07:29:00PM -0700, jgm2163 wrote:
> Michael Stenner <mstenner at phy.duke.edu> wrote in message news:<mailman.1031351050.14744.python-list at python.org>...
> > On Fri, Sep 06, 2002 at 04:57:19PM -0500, Stephen Boulet wrote:
> > > You can put
> > > 
> > >     #/usr/bin/env python
> > > 
> > > at the top of the file and do a chmod +x file. Is this the answer to the 
> > > right question?
> > 
> > Although since you're trying to use python 2.2.1 on a RHL 7.2 box, you
> > may need to make that
> > 
> >  #!/usr/bin/env python2
> > 
> > Otherwise, you'll get python 1.5.2, which can lead to confusing errors.
> > 
> > 					-Michael
> > > jgm2163 wrote:
> > > > os     :RedHat 7.2
> > > > Python :Python2.2.1
> > > > 
> > > > now,i had writen a python's script (extend with .py)
> > > > how can i make a executable program on RedHat 7.2?
>
> sorry,i want to make a release program on the computer that donot
> install python2.1 (because that python1.5 was installed on the redhat
> 7.2 defaultly).

Well, the most practical way is to just go ahead and install the
python2 package under RHL 7.2 (have your program "Require" it in its
rpm).  

If that's not feasible for you, then probably the easiest thing is to
backport your program to 1.5.2.

After that, the next easiest (and basically only other) thing you can
do is to include python 2.1 (or later) in your program.  I don't know
of any tricky ways to do that.  You would just have to include python
and all of it's files (or all the ones YOU use anyway) in your
program's rpm in such a way that they don't conflict with any other
programs/rpms.  Realistically, this is just silly.  Your admin should
probably be willing to install python2.

					-Michael
-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From phr-n2002b at NOSPAMnightsong.com  Mon Sep 23 23:53:23 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 23 Sep 2002 20:53:23 -0700
Subject: Python threading?
References: <f9Pi9.10275$Lg2.2191294@news2.news.adelphia.net> <amogmi$r1l$1@panix1.panix.com> <T5Rj9.13058$6g7.39977@news-server.bigpond.net.au>
Message-ID: <7xfzw0avrg.fsf@ruckus.brouhaha.com>

Mark Hammond <mhammond at skippinet.com.au> writes:
> A "one thread per connection" model is doomed from the start -
> therefore, in most applications, I would consider 5 threads a large
> number - particularly when all are actively doing something.  OTOH, if
> you are willing to accept that your first version will accept a max of
> ~20 connections, the brain-power saved by using threads beats the
> additional OS overhead by a mile ;)

The new Linux native thread implementation was recently benchmarked
running 100,000 simultaneous threads on a 1 GB machine.  Starting all
the threads and closing them took around 2 seconds.  


From fperez528 at yahoo.com  Sat Sep  7 17:50:20 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Sat, 07 Sep 2002 15:50:20 -0600
Subject: Smoothing a discrete set of data
References: <3csmdso5.fsf@morpheus.demon.co.uk> <B99F9C2F.1238E%pereira@cis.upenn.edu> <mailman.1031432711.19023.python-list@python.org>
Message-ID: <aldsas$b8$1@peabody.colorado.edu>

Tim Churches wrote:

> AFAIK, there is no native Python package which
> implements a
> comprehensive range of regression and smoothing techniques, but I can
> highly
> recommend the R package for statistics and statistical graphics - this
> mature, free, open
> source package has more statistical facilities than you are ever likely
> to need.

While not as comprehensive as R in its statistical coverage, SciPy probably 
has enough in it for what the OP wants (which sounded fairly basic). Quick 
linear (in the parameters) fits are easily done with scipy, plotting, simple 
smoothing (if he wants to), etc. I know R is industrial-strength statistics, 
but it sounds like more than what the OP needed.

cheers,

f.


From sperry at newcreations.info  Wed Sep 25 13:05:41 2002
From: sperry at newcreations.info (Sperry Russ)
Date: Wed, 25 Sep 2002 13:05:41 -0400
Subject: Python and Scite
Message-ID: <200209251305410290.00B482C8@newcreations.info>

Hello,
 
I am new to programming.  I have downloaded Python 2.2.1 and Scite 1.48 to learn how to program.  I am trying to configure my Scite so that it will run the python files by using go.  When I try to do this i get a message in the second window at the bottom -- unable to find file.  I have tried configuring the global options and have failed can anyone help me? I need a step by step example if you can because I cannot understand it according to the documentation that came with Scite
 
I installed scite like this: D:\Scintilla Text Editor\SciTE because I keep programs in a seperate partition.
 
Thanks,
Sperry Russ




From polux2001 at wanadoo.fr  Mon Sep  2 09:52:21 2002
From: polux2001 at wanadoo.fr (polux)
Date: Mon, 02 Sep 2002 15:52:21 +0200
Subject: Does IDLE handle unicode?
References: <QdIc9.2412$Ak.1132837@kent.svc.tds.net>
Message-ID: <3D736D15.5040800@wanadoo.fr>

Edward K. Ream wrote:
> IDLE took an exception writing a .py file containing the copyright
> character.  This happened a while ago; IIRC, I lost the file.
> 
> Is this a known problem?  I see nothing about this in the FAQ.
> 
> Edward
> --------------------------------------------------------------------
> Edward K. Ream   email:  edream at tds.net
> Leo: Literate Editor with Outlines
> Leo: http://personalpages.tds.net/~edream/front.html
> --------------------------------------------------------------------
> 
> 
> 

Read in the FAQ :
----------------------------
4.102. UnicodeError: ASCII [decoding,encoding] error: ordinal not in 
range(128)
This error indicates that your Python installation can handle only 7-bit 
ASCII strings. There are a couple ways to fix or workaround the problem.

If your programs must handle data in arbitary character set encodings, 
the environment the application runs in will generally identify the 
encoding of the data it is handing you. You need to convert the input to 
Unicode data using that encoding. For instance, a program that handles 
email or web input will typically find character set encoding 
information in Content-Type headers. This can then be used to properly 
convert input data to Unicode. Assuming the string referred to by 
"value" is encoded as UTF-8:

     value = unicode(value, "utf-8")

will return a Unicode object. If the data is not correctly encoded as 
UTF-8, the above call will raise a UnicodeError.

If you only want strings coverted to Unicode which have non-ASCII data, 
you can try converting them first assuming an ASCII encoding, and then 
generate Unicode objects if that fails:

     try:
         x = unicode(value, "ascii")
     except UnicodeError:
         value = unicode(value, "utf-8")
     else:
         # value was valid ASCII data
         pass

If you normally use a character set encoding other than US-ASCII and 
only need to handle data in that encoding, the simplest way to fix the 
problem may be simply to set the encoding in sitecustomize.py. The 
following code is just a modified version of the encoding setup code 
from site.py with the relevant lines uncommented.

     # Set the string encoding used by the Unicode implementation.
     # The default is 'ascii'
     encoding = "ascii" # <= CHANGE THIS if you wish

     # Enable to support locale aware default string encodings.
     import locale
     loc = locale.getdefaultlocale()
     if loc[1]:
         encoding = loc[1]
     if encoding != "ascii":
         import sys
         sys.setdefaultencoding(encoding)

Also note that on Windows, there is an encoding known as "mbcs", which 
uses an encoding specific to your current locale. In many cases, and 
particularly when working with COM, this may be an appropriate default 
encoding to use.

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


try to use mbcs on windows
i did it and it works



From gherron at islandtraining.com  Tue Sep 10 21:28:02 2002
From: gherron at islandtraining.com (Gary Herron)
Date: Tue, 10 Sep 2002 18:28:02 -0700
Subject: Maybe a dumb question
In-Reply-To: <3D7E6B0E.4060709@caramail.com>
References: <3D7E6B0E.4060709@caramail.com>
Message-ID: <200209101828.02617.gherron@islandtraining.com>

On Tuesday 10 September 2002 02:58 pm, Fabien HENON wrote:
> I have written an editor for Povray using Python, Tkinter and Pmw
>
> It is available at :
>
> http://pyvon.sourceforge.net
>
> The software includes one main file and 6 others which are for the
> language customization. People can choose between French, English,
> German,...
> All the translations for these languages are stored in files called
> menu_francais.py, menu_english.py, menu_deutsch.py ....when the editor
> is started.
>
> I would like to clean the directory a bit and put all these 'languages
> files' into a sub-directory called Lang.
> How do I tell python to look into a particular sub-directory when
> importing a file.
>
> I tried the os.path.join syntax, but to no avail.

Two possibilities:

1.  You can manipulate the search path for imports thus:

    import sys
    langDir = '...whatever...'
    sys.path.append(langDir)
    import Whatever

  where the calculation of the path to the language directory is
  something like:

    thisPyFile = os.path.join(os.getcwd(), sys.argv[0])
    thisDir = os.path.normpath(os.path.dirname(thisPyFile))
    langDir = os.path.join(thisDir, 'Lang')


2.  Use the package interface for the directory Lang:

  Create an empty file Lang/__init__.py.  Then you can import
  Lang/menu_francais.py like this:

    import Lang.menu_francais.py

  Look through the standard library for many examples of (non empty)
  __init__.py files.

Gary Herron




From mak at trisoft.com.pl  Mon Sep 30 17:46:42 2002
From: mak at trisoft.com.pl (Grzegorz Makarewicz)
Date: Mon, 30 Sep 2002 23:46:42 +0200
Subject: Printing with python
References: <Mahogany-0.64.2-1396-20020930-214624.00@>
Message-ID: <3D98C642.3090003@trisoft.com.pl>

Florian Schulze wrote:
> Hi!
> 
> I would like to know what solutions exist for printing (on paper) with
> python. Preferable would be a crossplatform approach, but only for Windows
> would be good enough. I also thought about using the reportlab pdf lib and
> using the free Acrobat Reader for preview and printing. It has a nice API
> and I don't think it will be too hard to send it to Acrobat Reader.
> 
> Any suggestions or thoughts about this?
> 
> Regards,
> Florian Schulze
> 
> 
> 

try DDE connection to AR:

import os
import win32ui
import dde

fname=os.path.join(os.getcwd(),'demo.pdf')
server = dde.CreateServer()
server.Create("PyAcroview")

conversation = dde.CreateConversation(server)

conversation.ConnectTo("acroview", "control")
conversation.Exec("[DocOpen(%s)]" %fname)

--
You can print, select pages and documents, ...

mak




From Dick.Zantow at lexisnexis.com  Tue Sep 17 11:01:54 2002
From: Dick.Zantow at lexisnexis.com (rzed)
Date: Tue, 17 Sep 2002 11:01:54 -0400
Subject: Good book choice?
References: <mailman.1032271021.9221.python-list@python.org> <3D873E8F.7060501@diespammerdieuselesspython.com>
Message-ID: <am7g7a$eg4$1@mailgate2.lexis-nexis.com>

"Rob Andrews" <rob at diespammerdieuselesspython.com> wrote in message
news:3D873E8F.7060501 at diespammerdieuselesspython.com...
> > <vent>
> > This isn't directed at the book mentioned (which is why I didn't
quote
> > its name in this rant), rather at all books which come with CDs.
I hate
> > it when books come with CDs.  I'm perfectly capable of downloading
the
> > source code used in the book, and in fact I prefer to hand-type it
while
> > reading from the book anyway, to make sure I see every part of the
> > code.  When a book comes with a CD, it is cumbersome to read
portably,
> > which is the whole advantage of books, unless you take the CD out.
And
> > if I take the CD out, then there is no doubt that it will get
lost, or I
> > will be unable to return the book if dissatisfied, etc.  Plus, you
can't
> > tell me that the inclusion of a CD and the envelope/publishing
used to
> > include the CD doesn't add a few units of currency to the book's
cost.
> >
> > Poke this opinion full of holes, but when I buy a book, I want a
book,
> > not a software package.  The book should be good enough to stand
on its
> > own without the "bonus" material on CD.  I think twice about
buying a
> > book that comes with a CD, to tell you the truth.
> > </vent>
>
> All perfectly valid sentiments, of course. But for the village in
Nepal
> lacking bandwidth because they are still six months away from having
a
> phone line (not as far-fetched an example as it may seem), an
entirely
> different set of considerations may be found.
>
> I find myself on the middle path on this issue, since I rarely have
> personal use for the CDs, but find that they do make it easier to
> distribute Python in my fits of zealous advocacy. I have bandwidth,
but
> no burner, and some are more interested in hearing me out if they
don't
> have to face a long (in some cases, expensive) download before they
even
> get to see if they would be interested.
>

The further aspect to it is that code written in a book may be
dependent on the environment the author is working with. Some or all
of it may still be valid a couple of releases later, but if things are
not working, it is useful to be able to set up the *same* environment
the author had to verify that the code is not the problem. Otherwise
you're left translating to the current version of things, and then
debugging code that doesn't correspond to what the author wrote to
begin with.

--
rzed




From sismex01 at hebmex.com  Thu Sep 26 17:32:13 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Thu, 26 Sep 2002 16:32:13 -0500
Subject: Who knows somefunction?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F0623302604FAC@mail.hebmex.com>

> 
> Hi Pythonistas!
> 
> Is there a function <somefunction> in Python
> for which
> 
> something is <somefunction>(id(something))
> 
> returns invariably true?
> 
> Gregor
> 

Woo, this is dangerous, kinda like turning an arbitrary
(*void) into whatever you like (well, you can).

So, this function would be the equivalent of a
type cast, only in Python. Nope, I don't like it.

Why'dyawanna something like this, anyhow?

:-)

-gus







Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From fperez528 at yahoo.com  Fri Sep 27 21:55:07 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Fri, 27 Sep 2002 19:55:07 -0600
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <amvcev$cf6$1@s1.read.news.oleane.net> <7nHk9.172208$pX1.6117207@news2.tin.it> <ofLk9.304267$z91.13145351@bin3.nnrp.aus1.giganews.com> <dITk9.175519$pX1.6209886@news2.tin.it> <t_3l9.334924$AR1.14880141@bin2.nnrp.aus1.giganews.com> <mailman.1033165663.5015.python-list@python.org>
Message-ID: <an325r$hir$1@peabody.colorado.edu>

Mike Rovner wrote:

> 
> "Terry Reedy" <tjreedy at udel.edu> wrote in message
> news:t_3l9.334924$AR1.14880141 at bin2.nnrp.aus1.giganews.com...
>> na=1; np=3; nl=1; ng=0
>> print "Found: %d apple%s, %d plum%s, %d lemon%s, and %d grape%s." %\
>> (na, na!=1 and 's' or '', np, np!=1 and 's' or '', nl, nl!=1 and 's'
>> or '', ng, ng!=1 and 's' or '')
>>
>> Found: 1 apple, 3 plums, 1 lemon, and 0 grapes.
>>
>> Alternatives to this 'risky' construct:
>> 1. forgo the nicety of correct English output
>> 2. nest if/else four levels deep to choose correct variant of 16 print
>> statements (ugh)
>> 3. precalculate 4 suffixes and store in 4 temp vars with 4 if/else's
>> (8 or 16 lines);
>>    though clearly superiour to 2), this is still tedious with some
>> risk of typo error.
>>
> 4. Try:
> use_s=('','s')
> print "Found: %d apple%s, %d plum%s, %d lemon%s, and %d grape%s." %\
> (na, use_s[na!=1], np, use_s[np!=1], nl, use_s[nl!=1], ng, use_s[ng!=1])

Why not just a simple function?

In [23]: plur=lambda s,n:'%s %s' % (n,s+('','s')[n!=1])

In [24]: plur('cat',0)
Out[24]: '0 cats'

In [25]: plur('cat',1)
Out[25]: '1 cat'

In [26]: plur('cat',9)
Out[26]: '9 cats'

In [27]: na=1; np=3; nl=1; ng=0

In [28]: print "Found %s, %s, %s, %s" % (plur('apple',na),plur('plum',np),
   ....: plur('lemon',nl),plur('grape',ng))
Found 1 apple, 3 plums, 1 lemon, 0 grapes

If the list is big the above can easily be automated with a map or a list 
comprehension.

Cheers,

f.


From cbbrowne at acm.org  Thu Sep  5 16:24:18 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: 5 Sep 2002 20:24:18 GMT
Subject: Graham's spam filter
References: <mailman.1029992314.27908.python-list@python.org> <3D647B4B.28D9BF04@alcyone.com> <mailman.1030050253.11330.python-list@python.org> <3D655AE6.A097AA25@alcyone.com> <ak44j1$1f2ft4$2@ID-125932.news.dfncis.de> <c2ef5940.0209041912.16e102fb@posting.google.com> <3D770156.2E55864C@alcyone.com> <al7tg4$1o5a9k$5@ID-125932.news.dfncis.de> <3D77AC75.B9B7F27@alcyone.com>
Message-ID: <al8ehh$1nt6ua$1@ID-125932.news.dfncis.de>

Quoth Erik Max Francis <max at alcyone.com>:
> Christopher Browne wrote:
>> Have you considered simply replacing strings that appear to be
>> base64-encoded with a token like "base64-text"?
>> 
>> That allows the database to at least be aware that the spam
>> commonly contains base64 data.
>
> Well, that really depends on what your goal is.  Again, if you're
> one of those people that has a very tight circle of email buddies
> and so essentially any unsolicited email is by definition spam, then
> you can tighten down your spam filter in all kinds of very powerful
> ways.
>
> I, as I've mentioned before, receive unsolicited email from my Web
> sites and various projects, and so unfortunately don't have the
> luxury of doing this.  So I need to support receiving email from
> faraway lands and unknown email addresses as well as trying to
> vigorously filter spam.
>
> Fact is, unfortunately, lots of people send legitimate email that is
> MIME encoded.

No, I am certainly _not_ defining "all unsolicited email" as being
spam.  Quite to the contrary, I receive quite a lot of interesting
email from unexpected sources.  Very little of it, statistically
speaking, is heavily MIME encoded, mind you...

The MIME encoded stuff does _not_ solely consist of "base64" text; it
also has header information that at least _suggests_ file type info.

A recent virus email contained:

Content-Type: application/octet-stream;
        name=snoopy.exe
Content-Transfer-Encoding: base64
Content-ID: <UCuk0QbULj2h8t9F>

another had:

Content-Type: audio/x-wav;
        name=bgcolor.exe
Content-Transfer-Encoding: base64
Content-ID: <W0H1pml7>

I get legitimate mail that contains base64 material; it _never_, in my
experience, consists solely of base64 material.

It always contains _some_ sort of commentary, and whether that
commentary came as text or as HTML, it's quite nicely sufficient to
distinguish the "unexpected resumes from Russia" from the email
viruses.

>> -> Supposing there is interesting text encoded (such as source code
>>    for a virus) inside the base64 stuff, it _would_ be useful to
>>    decode it;
>> 
>> -> Supposing the base64 stuff is basically just a GIF/JPEG/PNG, or
>>    something else that doesn't contain "interesting text," you'll
>>    have not much of value from the decoding process.
>> 
>> Making the "tokenizing" step a tad smarter (e.g. - recognizing "this
>> is likely base 64" and collecting stats on numbers of lines of base64
>> material) requires minimal added effort, and I expect it would buy you
>> _most_ of the benefits of decoding.
>
> Spammers are hitting upon the strategy, though, of sending emails in
> which the body consists of nothing but a completely encoded base64
> MIME part.  So in that case, the entire body of your message would
> consist solely of your "base64encoded" token.  So in the general
> case of any kind of spam filter (not just limited to a Graham
> filter), it's questionable how useful this will be, unless you plant
> to always filter against that token, presuming it to always indicate
> spam.

I've been using naive Bayesian filtering for years; I don't assume
that _any_ particular token indicates _any_ particular result.

I'm not interested in the "rule-based" stuff, only in the schemes
based on statistical analysis.

And the body of the message would most certainly NOT consist solely of
a "base64encoded" token.

The body portion would consist of:

- Various "Content-foo" tokens
- The header information that these documents _do_ contain; they
  normally contain an HTML header.
- Not "solely a base64encoded token," but rather some sort of count
  involving  _many_ base64encoded tokens.

The notion that it's "solely one token" is in your imagination, not in
reality.  There are _no_ "presumptions" being made here.

What I'm saying, that apparently isn't being read, is that I expect
that collecting stats on the numbers of "base64 lines" is likely to be
_nearly_ as useful as decoding the contents, and that it's _certainly_
simpler and faster.

If it _proves_ insufficient as a discriminator (please feel free to
direct any nonsense about 'presuming anything to always indicate
spam' to /dev/null), then it might prove necessary to _try_ to analyze
the contents.

_Trying_ to decode and analyze the contents may still prove a futile
exercise.  You won't get much useful material out of such common MIME
contents as graphics, PDFs, ZIP files, and audio files, without going
to even _more_ gratuitous lengths to analyze them that might very well
make you vulnerable to DOS attacks directed against the mail filter
itself.
-- 
(reverse (concatenate 'string "gro.mca@" "enworbbc"))
http://www.ntlug.org/~cbbrowne/nonrdbms.html
"I will not send lard through the mail" ^ 100 -- Bart Simpson


From mark.charsley at REMOVE_THIS.radioscape.com  Wed Sep 11 07:04:00 2002
From: mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley)
Date: Wed, 11 Sep 2002 12:04 +0100 (BST)
Subject: Embedded Python implementation
References: <MAfe9.266658$v53.14092531@news3.calgary.shaw.ca>
Message-ID: <memo.20020911120426.1292A@a.a>

In article <MAfe9.266658$v53.14092531 at news3.calgary.shaw.ca>, 
tagith1394 at hotmail.com (Evan) wrote:

> Hello,
> 
> I am currently trying to embed Python in an application I am writing, 
> and all is going well, except for one inexplicable bug..
> 
> When I pass a variable (pointer to a string) to PyRun_String, it will 
> always error on line 2, character 5.. regardless of what is actually on 
> the line in that spot (for that matter, even if there is no line 2)..

When I had a problem like that it was because an error had occurred in a 
previous call into the python DLL, and hadn't been cleared. As a result my 
call to python also failed, but the error message I retrieved was to do 
with the prvious call.

-- 

Mark - personal opinion only, could well be wrong, not representing 
company, don't sue us etc. etc.


From tismer at tismer.com  Sat Sep 21 00:22:08 2002
From: tismer at tismer.com (Christian Tismer)
Date: Sat, 21 Sep 2002 06:22:08 +0200
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amabpb$5rm$1@news.nems.noaa.gov> <3D8B1537.4010803@thinkware.se> <ad052e5c.0209201128.1a53108@posting.google.com>
Message-ID: <3D8BF3F0.4000204@tismer.com>

Dan Bishop wrote:
> Magnus Lyck? <magnus at thinkware.se> wrote in message news:<3D8B1537.4010803 at thinkware.se>...
> 
>>ChrisBarker wrote:
>>
>>>The only reason I can see
>>>to not writing your class in Python is performance, and unless you are 
>>>doing a lot of calculations, that is probably a non-issue.
>>
>>Performance issues might well matter in large financial
>>systems, but that's not the only reason.
> 
> 
> Please name one other real-world situation in which decimal arithmetic
> is actually *needed*.

There is no reason to argue about that.
Decimal arithmetic is exact for most of the operations
bankers require. They have been demanding for it since
decades, and they will not drop it. Having "no rounding
issue at all" is better than any consideration about
neglectible errors, it cannot be beaten, since *they*
understand it.
Cannot be discussed. They pay the bill.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From fperez528 at yahoo.com  Mon Sep 23 21:43:22 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 23 Sep 2002 19:43:22 -0600
Subject: e vs exp()? / financial applications
References: <mailman.1032801271.6597.python-list@python.org> <QgPj9.253053$AR1.10476629@bin2.nnrp.aus1.giganews.com>
Message-ID: <amofvr$mei$1@peabody.colorado.edu>

Terry Reedy wrote:

> Although w appears to be exactly represented, it must not be.  Cute
> example.
> 

In this case it's not a problem of the exactness of w's representation, but 
the exactness of the intermediate results. It turns out that basic properties 
of common arithmetic operations, such as the associativity of addition, are 
NOT valid in floating point:

(a+b)+c != a+(b+c)

in a computer. It may appear to be true in many cases, but it is not in 
general, as the above example showed. A classic textbook example:

In [29]: eps = 1e-16

In [30]: (1+eps)+eps
Out[30]: 1.0

In [31]: 1+(eps+eps)
Out[31]: 1.0000000000000002

Any decent text on numerical analysis discusses this at length. While for 
'common' day to day things one can gloss over this, there are many instances 
where it will bite you hard.

Cheers,

f.


From tdelaney at avaya.com  Thu Sep 19 03:02:25 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 19 Sep 2002 17:02:25 +1000
Subject: Stackless Python, eventual merge?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5B5@natasha.auslabs.avaya.com>

> From: martin at v.loewis.de [mailto:martin at v.loewis.de]
> 
> It is better than that: it does not need to be "switched off". Even if
> switched on (at build time), it won't start copying slices of the
> stack. An application actively has to import one of the modules, and
> to call a function in those modules, before you need to get nervous.

The problem of course is if you import a module which itself uses stackless.

Currently, there are limited ways in which simply importing a module can
kill Python. This would add another way. Obviously all such modules should
be documented, but ...

I'm not saying I'm against Stackless - quite on the contrary. However, it is
something which needs to be considered *very* carefully.

At the very least, PyChecker should warn about Stackless (even if it never
gets into the core ;)

Tim Delaney



From grabiller at 3dvf.net  Sun Sep  8 06:45:32 2002
From: grabiller at 3dvf.net (Guy Rabiller)
Date: Sun, 8 Sep 2002 12:45:32 +0200
Subject: untrusted dynamic module: win32api
References: <unk63h7q26lefe@news.supernews.com> <OAwe9.5047$_l1.12416@news-server.bigpond.net.au>
Message-ID: <unmai51quucsde@news.supernews.com>

[Mark]
> It appears that SoftImage is configuring ActiveScripting for "untrusted"
> scripts - ie, SoftImage is requesting the scripting engine switch into
> untrusted mode.  Thus, you really need to ask SoftImage is the
> application can be configured to not request restricted execution mode
> of the language

After a closer look, it seems that this has nothing to do with Softimage,
but with MSIE.

Even if I include ( for exemple ):

<SCRIPT Language="Python">
import win32com.client
w = win32com.client.Dispatch("Photoshop.Application")
</SCRIPT>

in a simple html file, then launching it in MSIE, totaly outside from
Softimage, it doesn't work, I still have the same error message.
( it works in VBScript and JScript )

I think this has to do with restricted execution mode, but I have no clue
how to customize it for allowing the loading of 'untrusted dynamic modules'.

Any help on this would be greatly apreciated.


Guy.


3d animator / td
grabiller at 3dvf.net
http://grabiller.3dvf.net


"Mark Hammond" <mhammond at skippinet.com.au> wrote in message
news:OAwe9.5047$_l1.12416 at news-server.bigpond.net.au...
> > Hi,
> >
> > I'm a Softimage|XSI user, a 3D graphic software that support ActiveX
> > Scripting on win32 platforms.
> >
> > This software integrate a 'NetView', an integrated internet explorer.
> >
> > To be able to communicate between the NetView and XSI I have to create
the
> > XSI.Application object.
> >
> > Works fine with VBScript and JScript, however I'm running into a problem
> > trying to make it work with Python.
> >
> > Basicaly I have to include in my html page the following code:
> >
> > <SCRIPT Language="Python">
> > import win32com.client
> > w = win32com.client.Dispatch("XSI.Application")
> > </SCRIPT>
> >
> > But when I load this page into the NetView ( or even in an external
> > brother ) I have this error message:
> >
> > 'ImportError: untrusted dynamic module: win32api'
> >
> > Is there any way to 'force' the import ?
> >
> > Sorry if this has been discussed earlier, I'm relatively new to Python.
>
> It appears that SoftImage is configuring ActiveScripting for "untrusted"
> scripts - ie, SoftImage is requesting the scripting engine switch into
> untrusted mode.  Thus, you really need to ask SoftImage is the
> application can be configured to not request restricted execution mode
> of the language.
>
> Mark.
>




From lozinski at openstepnews.com  Tue Sep 24 11:49:24 2002
From: lozinski at openstepnews.com (lozinski at openstepnews.com)
Date: Tue, 24 Sep 2002 08:49:24 -0700 (PDT)
Subject: Lots of European Zope Contractors Available
Message-ID: <1020924084924.8727AAC/i.lozinski@maya>

At:

http://zope.jobmart.com/Resumes/ShowEuropeans

Including 11 British Citizens who are python and Zope Developers.

Please let me know which candidates look interesting, and I will go ahead and
post their resumes on the public web site.

The great thing about having a large number of candidates immediately available,
is that you can select just the right person for your requirements.

If you send me your phone number, I can call you to discuss the general Zope
marketplace.

And don't forget to check out my resume at:
http://zope.jobmart.com/Resumes/1154/Resume

Regards
Chris

1-510-740-6486
lozinski at openstepnews.com
lozinski at jobmart.com



From spoermeg at voldelig.com  Tue Sep 17 19:36:08 2002
From: spoermeg at voldelig.com (Terje Johan Abrahamsen)
Date: Tue, 17 Sep 2002 18:36:08 -0500
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <mailman.1032152058.11043.python-list@python.org> <m3admizdzq.fsf@mira.informatik.hu-berlin.de> <37fh9.112911$pX1.4044383@news2.tin.it> <7xk7lml7px.fsf@ruckus.brouhaha.com>
Message-ID: <am8eq1$3ml4d$4@ID-53055.news.dfncis.de>

"Paul Rubin" <phr-n2002b at NOSPAMnightsong.com> wrote in message
news:7xk7lml7px.fsf at ruckus.brouhaha.com...
> Alex Martelli <aleax at aleax.it> writes:
> > > You'll have to give your customers a sealed computer and no permission
> > > to install additional software, or communicate with the internet, if
> > > you want to reliably prevent them from analysing your program.
> >
> > Incidentally, the latter option can be feasible in certain situations:
some
> > key parts of your program's functionality might reside on a computer you
> > control, accessible to the rest of the application (installed more
> > traditionally on users-controlled computers) only as a black box via the
> > net.  XML-RPC, SOAP, and other distributed-computing approaches
> > such as Corba, make implementing this particularly easy these days.
>
> It doesn't have to be via the net: you can literally deliver a sealed
> computer to the customer, with the sealing enforced by contracts and
> inspections (this is quite normal, e.g. some financial applications
> are delivered as a rack-mount box that you install in your machine
> room) or by tamper resistant hardware (for higher security
> applications).  For VERY high security, of course you can't let
> hostile parties touch the hardware.  Nonetheless, the pay TV industry
> deploys millions of sealed computers (smart cards) every year to
> customers, every single one of which is considered an attacker.  While
> that industry had spectacular early security failures, these days it's
> a fairly solved problem and they limit losses to acceptable levels.
>
> It doesn't sound like the OP was asking for this type of solution though.

Nope.... Then I assume I had been able to pay for a security consultant to
do that part of the work. Not plundering to do it myself with variable
success...




From mgerrans at mindspring.com  Thu Sep 26 01:31:01 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Wed, 25 Sep 2002 22:31:01 -0700
Subject: a python puzzle
References: <mailman.1033002153.8360.python-list@python.org>
Message-ID: <amu63d$89m$1@slb6.atl.mindspring.net>

> Keeping medicines from the bloodstreams of the sick; food from the bellies
of
> the hungry; books from the hands of the uneducated; technology from the
> underdeveloped; and putting advocates of freedom in prisons.  Intellectual
> property is to the 21st century what the slave trade was to the 16th.

Good quote!   (Although, unfortunately -- and perhaps more so, if it's a
predictor for "intellectual property" -- slave trade persisted long past the
16th century).




From bkc at Murkworks.com  Tue Sep 24 13:21:14 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Tue, 24 Sep 2002 13:21:14 -0400
Subject: ANNOUNCE: wxPython 2.3.3.1
References: <mailman.1032486266.20015.python-list@python.org>
Message-ID: <3d909c63_6@goliath.newsgroups.com>

Sorry this is OT.

Robin, any news on wxEmbedded for Windows CE?  The embedded version for
CE hasn't made any (visible) progress on the wxWindows website, and I can't
find an email address for the person/org claiming to do the CE port.







-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From ods at strana.ru  Thu Sep 26 12:37:01 2002
From: ods at strana.ru (Denis S. Otkidach)
Date: Thu, 26 Sep 2002 20:37:01 +0400 (MSD)
Subject: design question
In-Reply-To: <1ha6pukqdqell8f31q1k7rtor2rv5ogjs0@4ax.com>
Message-ID: <Pine.LNX.4.44L.0209262034570.31859-100000@ods.pp.ru>

On Thu, 26 Sep 2002, Gon?alo Rodrigues wrote:

GR> class Wrapper1(object)
GR>     def __init__(self, <more args>):
GR>         self.__wrap = Class1()
GR>         <more initialization code>
GR>
GR> But then if I did the same with Wrapper2, it would be at
GR> least quirky
GR> since it inherits from Wrapper1 but it wraps Class1
GR> <scratches head in
GR> puzzled bewilderment>

class Wrapper1(object):
    _wrapped_class = Class1
    def __init__(self, ...):
        self._wrap = self._wrapped_class()
        ...

class Wrapper2(Wrapper1):
    _wrapped_class = Class2
    ...

-- 
Denis S. Otkidach
http://www.python.ru/      [ru]
http://diveinto.python.ru/ [ru]




From storedmails at gmx.net  Mon Sep 23 09:11:00 2002
From: storedmails at gmx.net (beachboy)
Date: 23 Sep 2002 06:11:00 -0700
Subject: Session handling
Message-ID: <c5b3549a.0209230511.5bbe6849@posting.google.com>

Hi!
Please can anyone help me.
My problem:
I'm trying to program a web application with python. This tool has to
provide diffrent languages. The user has to choose the language at the
beginning. So i have to save the information which language the user
chooses. I think this is a perfect example for using sessions.
The only problem is I've got no idea how to realise this in python.
Please help me as soon as you can because I have to solve this problem
very quickly.

hoping to hear from you soon
beachboy


From mhammond at skippinet.com.au  Tue Sep 17 06:59:22 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 17 Sep 2002 10:59:22 GMT
Subject: Calling Python from C# - please help
References: <6de16ac9.0209130508.5d1b56c6@posting.google.com> <j4elbyt3yw.fsf@informatik.hu-berlin.de> <ur8fyxau0.fsf@online.no> <0iCg9.2350$Ee4.6516@news-server.bigpond.net.au> <3D82FA5B.3050003@mxm.dk>
Message-ID: <eWDh9.15201$Ee4.45064@news-server.bigpond.net.au>

Max M wrote:
> Mark Hammond wrote:
> 
>> I have no examples of this, but I should.  I'm co-authoring a .NET 
>> book which is nearly out :) - 
> 
> 
> 
> That sounds great. I enjoyed the win32 book, and still use it as a 
> reference.

Thanks.  Me too - it is amazing how much I have forgotten since writing 
it ;)

> Are you by any chance still developing Pyhton .NET, or has it completely 
> stopped?

Pretty much stopped.  No one is funding it, and no one has stepped up to 
the plate and offered to help.  It is too much for me to do alone, 
especially on my own time.

I'd love to pick it back up, but unfortunately, I really can't see who's 
interest it is in to fund this thing - Microsoft really have no further 
interest which IMO is quite reasonable - they have generously funded 
enough to prove that Python could work, but there is no reason they 
should further fund Python over any (and therefore every) other open 
source language, especially when they have their own commercial offerings.

Maybe Python should place an ad in the .singles newsgroups looking for a 
sugar-daddy <0.1 wink>

Mark.



From mihan at foss.kharkov.ua  Wed Sep  4 04:36:59 2002
From: mihan at foss.kharkov.ua (Mykhaylo Sorochan)
Date: Wed, 4 Sep 2002 11:36:59 +0300
Subject: embedding python (Makefile creation) problem
Message-ID: <al4gm7$3u7$1@hyppo.gu.net>

Hi!

I have tryed to implement some examples from Very High Level Embedding from
Python documentation but hadn't found examples or rules of Makefile
creation.
Can anybody point me to the answer or describe such Makefiles creation.
I need Makefile for Unix(FreeBSD).

Thanks.

--
Regards, Mykhaylo Sorochan





From mike at bindkey.com  Tue Sep  3 18:44:13 2002
From: mike at bindkey.com (Mike Rovner)
Date: Tue, 3 Sep 2002 15:44:13 -0700
Subject: Drawing graphs/newtorks with Python
References: <Pine.OSF.4.10.10209031748180.23796-100000@wilde.oit.umass.edu>
Message-ID: <al3dsn$opv$1@main.gmane.org>

"Alexander Stepanov" <stepanov at oitunix.oit.umass.edu> wrote in message
news:Pine.OSF.4.10.10209031748180.23796-100000 at wilde.oit.umass.edu...
>
> Hi Pythonians,
>
> I work on program which performs some calculations over graphs.
> (shortest paths). I would like to draw an image/ps_document with
> Python. Does anybody use some kind of modules to draw graphs?
> I will appreciate if you can point me out to the source.

Take a look on GraphViz http://www.research.att.com/sw/tools/graphviz/ or
http://www.graphviz.org/








From gumuz at looze.net  Wed Sep 11 04:58:54 2002
From: gumuz at looze.net (Gumuz)
Date: Wed, 11 Sep 2002 10:58:54 +0200
Subject: blocking socket dilemma
Message-ID: <3d7f046f$0$230$4d4ebb8e@news.nl.uu.net>

Hi all,

I am a little bit stuck with my design (again!).

I have this server class, which creates a session object for every accepted
incoming connection. The session object has a RequestHandler function which
is then run in a new thread.

all fine, but...

This RequestHandler function actually does a infinite loop. The loop wil
receive data and take some action and it will send data which is piled up in
this session's queue. The problem that i'm having is that the loop will hang
on socket.recv if the client is not sending anything. Ofcourse this is very
obvious and logical, but i can't think of another way of constructing my
design, so that this problem doesn't occur.

I hope i've made my problem clear to you guys.

thanx,

gumuz




From joec at mill.co.uk  Mon Sep  2 13:05:53 2002
From: joec at mill.co.uk (Joe Connellan)
Date: Mon, 02 Sep 2002 18:05:53 +0100
Subject: compiling extensions on linux
References: <3D738D4E.35B6C712@mill.co.uk> <j44rd8jqu9.fsf@informatik.hu-berlin.de> <3D739730.DEB80478@mill.co.uk> <j4lm6kib4e.fsf@informatik.hu-berlin.de>
Message-ID: <3D739A71.F8650187@mill.co.uk>

of course ;) that fixed it - I've now put the c++ stuff back in and
declaring it as an extern "C" function

Cheers

Joe





"Martin v. L?wis" wrote:

> Joe Connellan <joec at mill.co.uk> writes:
>
> > it does have a initmyextension declared - Here's a simple example of my
> > problem - I've taken all c++ related stuff out so I can use gcc
>
> However, you are still compiling it as a C++ program, since it has the
> .cpp extension. You either need to rename the file (to .c), or use the
> -x c command line option. If there was C++ stuff to take out, you best
> declare the function as extern "C".
>
> Regards,
> Martin



From erick_papadakis at REMOVEFORSPAM.yahoo.com  Sun Sep 15 21:42:09 2002
From: erick_papadakis at REMOVEFORSPAM.yahoo.com (Erick)
Date: Mon, 16 Sep 2002 10:42:09 +0900
Subject: can wxpython do this?
References: <737aefee.0209151702.3f735252@posting.google.com> <spcaou804m5oo6gah31oqaeuqsuj60h0ip@4ax.com>
Message-ID: <am3d0b$jj$1@news01bc.so-net.ne.jp>

Thank you very much Lothar.


> There is no problem to setup an installer easily that takes
> Python,WXPysthon and whatever you need. It simply blowes up the
> installer package size.


Isn't this the way to distribute wxpython apps? Because most users will not
have python installed on their machines, much less the wxpython libraries?
Or is it that after compilation users do not need Python installed at all?


  > >for instance, can i minimize my application to a user's desktop's
  > >taskbar and have a right-click context menu on that taskbar entry?

> I think its not comfortable but possible. Look at the API to find out.


I searched the html help for "taskbar" and there is only one function for
Win95/NT (wxTaskBarIcon). Looks like this feature is not yet supported?

Actually I am interested -- is there any popular utility written in
wxpython? One of those things that are downloadable from Download.com or
something?



> There is a really simple WxHTML Widget and at least a C++ version of a
> wrapped IExplorer control.


The wxHTML widget supports only some very basic HTML tags, but I guess
that'll do for now. What I really wanted to know was if there is a way for
me to connect to the Internet, make a call to a webserver as a browser would
do, retrieve data and display it inside my application (with or without
HTML...does not have to be HTML at all, actually).

Perhaps wxSocket is what I am talking about. But I did not see anything in
the "demo" that has much to do with sockets. Would anyone know an example of
using sockets in wxpython?

Thanks a lot for the reply, I am already drawn!

.ep





From gua81 at XXXyahoo.com  Thu Sep 26 06:13:45 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Thu, 26 Sep 2002 20:13:45 +1000
Subject: Getting last char in string.
References: <mailman.1033027776.19152.python-list@python.org>
Message-ID: <amumk5$p52$1@lust.ihug.co.nz>

then what does
s[-1:] do?

"Quinet, Joel" <joel.quinet at be.unisys.com> wrote in message
news:mailman.1033027776.19152.python-list at python.org...
> Here under an example:
>
> s = 'abcd'
> print s[-1]
> d
> print s[-2]
> c
>
> Joel
> -----Original Message-----
> From: CheapSkate [mailto:gua81 at XXXyahoo.com]
> Sent: jeudi 26 septembre 2002 9:49
> To: python-list at python.org
> Subject: Getting last char in string.
>
>
> Hi, im writing a simple method that takes in a string input and check if
the
> last charcter is an '/'
> then it will do A else do B
>
> my Q now is how to get the last char???
>
> Thanks
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>




From michaels at one.net  Sun Sep 15 23:09:49 2002
From: michaels at one.net (Michael Schneider)
Date: Sun, 15 Sep 2002 23:09:49 -0400
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net>
Message-ID: <3D854B7D.1060209@one.net>



David Abrahams wrote:

>"Christian Tismer" <tismer at tismer.com> wrote in message
>news:mailman.1032107006.6411.python-list at python.org...
>
>>The changes to Python are also so small now, that
>>it doesn't matter if it will be merged or not.
>>Porting Stackless to a new Python is a matter of a few
>>hours. You may rely on getting a Stackless version
>>of any new Python version in almost no time.
>>
>
>I think it still matters. You don't want people who need stackless'
>capabilities to have to use a non-standard Python forever, do you?
>

I am a little confused by this.   I use threads, but threads must be 
compiled in.
This has been that the same since I downloaded my first python in '93 (Very
good language Guido :-)


Could someone please explain to a "confused" python user. Why the 
current version of
stackless python would not be distributed like threads???



Thanks to all for a great language!!!!!!!!!!!!!

Mike


>
>--
>-----------------------------------------------------------
>           David Abrahams * Boost Consulting
>dave at boost-consulting.com * http://www.boost-consulting.com
>
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020915/7c6ef49c/attachment.html>

From pyth at devel.trillke.net  Mon Sep  2 06:45:05 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 2 Sep 2002 12:45:05 +0200
Subject: Why it is so dramatical?
In-Reply-To: <200209021228.58453.b.maryniuk@forbis.lt>; from b.maryniuk@forbis.lt on Mon, Sep 02, 2002 at 12:28:58PM +0200
References: <mailman.1030957743.4021.python-list@python.org> <Xns927D6DBDE418Cduncanrcpcouk@127.0.0.1> <200209021228.58453.b.maryniuk@forbis.lt>
Message-ID: <20020902124505.J3893@prim.han.de>

Bo M. Maryniuck wrote:
> On Monday 02 September 2002 11:57, Duncan Booth wrote:
> > None of this is specific to Python, the same issues arise in almost any
> > programming language (although not always the same solutions).
> Almost... Keep in mind, that here is the range 0-0xffffff, what is quite
> bigger that 0-0xfff:
> ----------------------8<---------------------------
> #!/usr/bin/perl
> 
> $b = '';
> $data = 'AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA';
> for $a (0..0xffffff) {
>     $b += $data;
> }

That should result in a 1.3 GIGABYTE string.  
Watching the perl-process i see that perl doesn't use more than 1MB.
So probably perl does some optimization for this in-fact meaningless code.

It's much more interesting to look at real performance limitations
instead of some theoretical pusing-the-limits loops.  

    holger



From max at alcyone.com  Mon Sep 23 02:41:45 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 22 Sep 2002 23:41:45 -0700
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <%8ci9.138477$5r1.5541622@bin5.nnrp.aus1.giganews.com> <amltd3$3k$1@glue.ucr.edu> <3D8EB212.1000104@cygnus-software.com>
Message-ID: <3D8EB7A9.F7CDE16@alcyone.com>

Bruce Dawson wrote:

> I find the (a and [b] or [c])[0] construct to obscure to be useful -
> the
> space it saves is wasted by the comment I have to put in to explain
> it.

Precisely.  The purpose of a conditional expression is to compress a
trivial if/else statement into an expression in a readable manner. 
While you can use short-circuiting operators like and, or and lambda
expressions to get this effect, it defeats the readability purpose and
so is essentially useless.  The shorter, more readable forms, tend to be
subject to corner cases having different behavior than a standard p ? a
: b conditional expression.

> It occurs to me that the bool() built in function allows for what is -
> in my opinion - a cleaner solution. Easier for mortals to comprehend.

Well, you could already use operator.truth for that.

> Basically, construct a two entry list of the two items and then index
> that with your expression, converted to bool. The bool function is
> essential since there was previously no generic way of getting the
> true/false value of an expression.

I use this technique frequently enough, but this isn't really what a
conditional expression does.  A conditional expression only _evaluates_
one of its consequents, depending on the value of the test expression. 
The (a, b)[operator.truth(p)] evaluates both.

This is fine in many cases, of course -- such as when you're selecting
between two strings -- but it isn't the full power of a true conditional
expression.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Many situations can be clarified by the passing of time.
\__/ Theodore Isaac Rubin
    CSBuddy / http://www.alcyone.com/pyos/csbuddy/
 A Counter-Strike server log file monitor in Python.


From see_reply_address at something.invalid  Sun Sep 29 21:01:28 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Mon, 30 Sep 2002 13:01:28 +1200
Subject: Larry Wall's comment on python...
References: <mailman.1032972272.26262.python-list@python.org>
Message-ID: <3D97A268.6040603@something.invalid>

sismex01 at hebmex.com wrote:

>>Alan Daniels wrote:
>>
>>>Sometimes I wonder if Guido would have saved a lot of debate 
>>>by declaring, long ago, that Python code would never allow
>>>tabs in it, period. Which may be a bit extreme, I'll admit. :-)
> 
> Actually, it's the only universally-workable solution
> to this "problem".


This is not a workable solution, because there are
some environments in which tabs-only is the only
sane thing to use, e.g. BBEdit and many other code
editors on Mac.

Mandating either tabs-only or spaces-only would be
fine, though.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From loewis at informatik.hu-berlin.de  Tue Sep 24 09:53:14 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 24 Sep 2002 15:53:14 +0200
Subject: Python.Net ?
References: <ampnua$nti$00$1@news.t-online.com>
Message-ID: <j4bs6nebp1.fsf@informatik.hu-berlin.de>

"Achim Domma" <achim.domma at syynx.de> writes:

> a german IT magazin (iX) mentioned a .Net version of python, developed by
> ActiveState. My last information was, that there will be no .Net version
> with produces IL code, but only an extension for the new Visual Studio.
> Am I missing something or is the article simply wrong?

There is a product called VisualPython, which integrates with
Microsoft Visual Studio .NET. Depending on what .NET is for you, this
could be called a .NET version of Python (once Microsoft .NET Server
is out, any software that works on it could be called "works with
.NET"). But yes, this version does not produce MSIL, and does not
interoperate with the .NET runtime.

Regards,
Martin



From paul at boddie.net  Mon Sep 23 04:14:23 2002
From: paul at boddie.net (Paul Boddie)
Date: 23 Sep 2002 01:14:23 -0700
Subject: PyQt setup on Linux Mandrake 8.2
References: <3D8B4898.20808@nospam.attbi.com> <3d8e7141$1@news.nz.asiaonline.net> <3D8E74A4.7000404@nospam.attbi.com>
Message-ID: <23891c90.0209230014.56a45933@posting.google.com>

djw <dwelch91 at nospam.attbi.com> wrote in message news:<3D8E74A4.7000404 at nospam.attbi.com>...
> Thanks to you and the others, I finally got this to work. My remaining 
> question/problem/puzzlement is what to do about the fact I now have two 
> Python 2.2 installs on my system.

Don't worry about this: I run Red Hat 7.3 and have at least two Python
installations present. It doesn't make any difference in practice,
provided you know which one you're using.

>                                   Mandrake installed everything during 
> its setup into /usr and everything I built/installed went into 
> /usr/local. I was going to rip out the version that Mandrake layed down 
> in /usr and just use the /usr/local copy, but when I ran the package 
> manager and told it to uninstall all the various py* pieces, the list of 
> dependencies it wanted to also delete was scary (stuff like filesystem 
> modules).

I'm sure that RPM and other package management systems look great on
paper (although I'm not entirely convinced by RPM), but the more
irritating problems with them seem to arise when the people who put
the packages together decide to split them up into "runtime" and
"devel", regardless of whether this makes any real sense, and then
leave out "Python.h" or whatever other essential file might be
required by any software that needs to make use of (for example)
Python. I suppose you're really supposed to install an RPM of that
software, prepared by the same person who made the original Python
packages, if only because that person has the luxury of having built
from source.

>           Unfortunately, when I try to install new py modules, they all 
> want to install into /usr instead of /usr/local, since /usr/bin/python 
> is ahead of /usr/local/bin/python on $PATH, causing sys.prefix to point 
> to /usr, I suppose.

Solution: set your PATH appropriately for all users who want the new
installation of Python. There shouldn't be any major danger of
inadvertently using the wrong versions of other software, at least on
Linux systems, because /usr/local is probably pretty empty after
installing a mainstream Linux distribution - it's not like mixing up
/usr/ccs and /usr/ucb on Solaris and then finding out that you're
running some "other" version of your favourite tools.

> So, this leaves me wondering... its not possible to install Mandrake 
> without Python (as evidenced by the dependencies), but the location it 
> is installed at is "wrong" by default - so, what is the "proper" way to 
> handle this situation? (It would seem nice to install Linux without 
> Python and compile and install it afterwards).

Ignore the distribution's installations and install your own in
/usr/local, which helpfully happens to be the default "prefix", and
then complain to your distribution vendor. What then seems to happen,
based on experiences made public with Red Hat, is that they then
ignore those suggestions whilst issuing vague hints that their chosen
Python strategy makes complete sense, and that the entire Python
community is "missing something".

>                                                I can't tell from your 
> list of commands where python is located, but I assume its in /usr?

It usually is.

> Also, I noticed that the copy of python that was installed by the system 
> lacked Python.h, whereas the copy I compiled and installed had this 
> file. Are you supposed to "overlay" the system copy in /usr with the 
> extra files that things like sip/pyqt need (i.e., Python.h)?

You're not supposed to do things like that, especially since there
could always be versioning issues in doing so, but Python.h does at
least seem to have the same contents regardless of how your own Python
installation was configured.

Paul


From nospam at horvath.com  Thu Sep 26 08:26:47 2002
From: nospam at horvath.com (Bob Horvath)
Date: Thu, 26 Sep 2002 12:26:47 GMT
Subject: combining mailbox and email modules
References: <NPck9.581867$UU1.104088@sccrnsc03> <slrnap41re.1mh.matt@happy-hour.t-island.com> <LAxk9.79207$gA4.27843@sccrnsc02> <u4euma.fc4.ln@ix.netcom.com>
Message-ID: <b2Dk9.7969$wH.339@sccrnsc01>

Dennis Lee Bieber wrote:
> Bob Horvath fed this fish to the penguins on Wednesday 25 September 
> 2002 11:14 pm:
> 
> 
> 
>>... and here it the message...
>>
>> From mickey at verizon.net  Tue Aug 20 10:03:52 2002
>>Return-Path: <mickey at verizon.net>
>>Delivered-To: bob at horvath.com
> 
> 
>         Is that indent on the "From ..." line really in the file, or just some 
> artifact of pasting it into your post? I just checked my Eudora .mbx 
> files and they don't show an indent on the "From ..." line
> 

Doh!  Sorry, cut and paste error. No, the space before the From is 
not there.



From pknotz at sandia.gov  Fri Sep 27 08:18:19 2002
From: pknotz at sandia.gov (Pat Notz)
Date: 27 Sep 2002 05:18:19 -0700
Subject: well-formed xml
References: <mailman.1033085436.9260.python-list@python.org>
Message-ID: <525eaba.0209270418.911d329@posting.google.com>

You have to be careful of non-alphanumeric characters.  Try replacing
'&' with &amp;  I beleive there is an escape() function in one of the
xml modules that will do that kind of escaping on strings so thay may
safely be placed into XML.


From bobx at linuxmail.org  Thu Sep 12 17:07:44 2002
From: bobx at linuxmail.org (Bob X)
Date: Thu, 12 Sep 2002 21:07:44 GMT
Subject: [FEEDBACK] Is this script efficient...is there a better way?
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <L9Rf9.168314$l_4.127784@atlpnn01.usenetserver.com> <4ZRf9.1782$Lg2.248733@news2.news.adelphia.net> <3hSf9.168947$l_4.98525@atlpnn01.usenetserver.com> <8PSf9.1796$Lg2.257378@news2.news.adelphia.net> <_oUf9.170870$l_4.126599@atlpnn01.usenetserver.com>
Message-ID: <Am7g9.2344$Lg2.402676@news2.news.adelphia.net>

Steve Holden wrote:
> "Bob X" <bobx at linuxmail.org> wrote ...
> [ ... ]
> 
>>Thank you very much...your shorter script did work. The keywords are all
>>"normal" english words that would not have any special characters. I
>>only added a re.IGNORECASE (since case doesn't matter to me) after a
>>quick search of the python docs.
>>
> 
> Was there any significant difference in the timings for the two different
> methods, and how does the number of different keywords affect the timings?
> 
I haven't tested...is there an easy way to time it...besides my going 
"that one ran faster"?

However, I had a chance to run your RE version today and it did very 
well. BTW..the line that looked like a debug statement was just a visual 
for me to see the script was actually doing something. I put in a 
counter that prints "lines written:", a where "a" is the counter. I do 
have one question on that. It writes each line sequentially like:

lines written: 1
lines written: 2
lines written: 3

Can I just get it to do it in place? Where the numbers are just replaced 
instead of scrolling? Just a thought...

Bob


Bob



From neilhart at attbi.nospam.com  Mon Sep  2 14:52:21 2002
From: neilhart at attbi.nospam.com (Neil Everhart)
Date: Mon, 02 Sep 2002 18:52:21 GMT
Subject: Dead Man Timer?
References: <fE6c9.145308$aA.31724@sccrnsc02> <pan.2002.08.26.17.48.00.759341@effectif.com> <cb035744.0209011247.81b42dc@posting.google.com> <3D728103.5060308@itamarst.org_NOSPAM>
Message-ID: <FrOc9.110025$_91.144030@rwcrnsc51.ops.asp.att.net>

Itamar,

Thank you for the suggestion.  I will test the timeoutsocket.py code
tomorrow
on my target setup.  I have incorporated it into a version of my script but
can not really evaluated it until I am running in the terminal server
environment.

Neil.

--


"Itamar Shtull-Trauring" <itamar at itamarst.org_NOSPAM> wrote in message
news:3D728103.5060308 at itamarst.org_NOSPAM...
> 1. Use timeoutsocket.py - this lets you set timeouts on socket operations
> (find it on http://py.vaults.ca).
>
> 2. Use an event driven networking architecture - don't know of any that
have
> telnet client support though.
>
> Option 1 is your easiest solution.
>
> --
> Itamar Shtull-Trauring    http://itamarst.org/
> Available for Python, Twisted, Zope and Java consulting
>




From loewis at informatik.hu-berlin.de  Mon Sep 23 05:47:56 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 23 Sep 2002 11:47:56 +0200
Subject: UpToDate SOAP-module ??
References: <ATAj9.21315$0p1.376966@news2.ulv.nextra.no>
Message-ID: <j4fzw1ghpv.fsf@informatik.hu-berlin.de>

"Thomas Weholt" <2002 at weholt.org> writes:

> I've looked around for a SOAP-module to use in a WebService-based
> project, but the SOAP-implementations in python seem to be a bit ...
> unstable ?? Some of them haven't been updated for a while etc.

I'd say that the qualifications contradict: If the module is not
changing, it seems it is "stable" - it may be buggy or incomplete, but
I associate stability with lack of change, or constance of behaviour
in the presence of change.

> Any opinions, hints or ideas are welcome. Thoughts on ease of use,
> list of features, comparisons etc. are also interesting.

My personal recommendation is to ignore SOAP.

Regards,
Martin


From aleax at aleax.it  Wed Sep 25 10:54:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 14:54:24 GMT
Subject: Newbie References Question
References: <3d91c72f$0$8558$79c14f64@nan-newsreader-02.noos.net>
Message-ID: <A6kk9.142235$ub2.3104522@news1.tin.it>

Guy Rabiller wrote:

> Hi,
> 
> let say I have:
> i1 = 1
        ...
> Textualy, how if I want that p[0] refere to the 'i1' container, and not to

There is no 'i1' CONTAINER.  'i1' is just a NAME.  That's what variables
are in Python -- they don't contain anything: they're like post-it labels,
attached to an object or to another.

> the refererence it contains ?
> 
> What I want is that if now I set:
> i1 = 4
> that automaticaly:
> p -> [4,2]
> and not keeping [1,2]

There is no way to do exactly what you require.  There may be ways to
do roughly-equivalent things, if you help us help you by explaining
your application context a bit more widely.


Alex



From peter at engcorp.com  Mon Sep  2 14:43:59 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 02 Sep 2002 14:43:59 -0400
Subject: RedHat 7.3 and Python2.2.1
References: <88bc63c6.0209020932.54723791@posting.google.com>
Message-ID: <3D73B16F.76B355DB@engcorp.com>

Doug Farrell wrote:
> 
> I'm trying to install the latest version of Python
> (python2-2.2.1-1.i386.rpm) from the www.python.org download page, and
> I'm having problems. When I run
> 
> rpm -i python2-2.2.1-1.i386.rpm
> 
> I get all kinds of conflict messages like this:
> 
> file /usr/bin/pydoc2 from install of python2-2.2.1-1 conflicts with
> file from package python2-2.2-16
> 
> probably hundreds of them in fact. Can anyone tell me what is going on
> here, as the www.python.org site says the RPM I'm using is for RH7.3,
> which is what I'm using. If you can tell me what I'm doing wrong and
> how to address this, that would be great too! My ultimate goal is to
> also install wxPython for the python2.2.1 installation, so if anyone
> has some 'heads up' on that, I'd appreciate it.

Sounds like you already have a previous version of Python2.2 installed.
You should probably be using the "-U" option to update instead of
the "-i" option to install... or uninstall the previous one first.

-Peter


From from_deja at alandaniels.com  Mon Sep 23 09:34:57 2002
From: from_deja at alandaniels.com (Alan Daniels)
Date: 23 Sep 2002 06:34:57 -0700
Subject: Universally unpacking tuples
References: <158d3913.0209222233.123a9675@posting.google.com>
Message-ID: <a6f9355c.0209230534.5e6dd620@posting.google.com>

peter.bittner at gmx.net (Peter Bittner) wrote in message news:<158d3913.0209222233.123a9675 at posting.google.com>...
> Hi there!
> 
> I have problems universally unpacking tuples. I would like to take the
> first element of a tuple and remove it (thus creating a new tuple with
> one element less)...

Just use the slice notation. A slice of a list is a new list, and a slice
of a tuple is a new tuple. In the example below, where y is a tuple, the
"y[1:]" means "a slice out of y, from element 1 all the way to the end".

[Example...]
y = (100, 200, 300, 400, 500)
while len(y) > 0:
    (head, y) = (y[0], y[1:])
    print head, y

[This will print...]
100 (200, 300, 400, 500)
200 (300, 400, 500)
300 (400, 500)
400 (500,)
500 ()


From michael.weigend at fernuni-hagen.de  Fri Sep  6 12:56:15 2002
From: michael.weigend at fernuni-hagen.de (Michael Weigend)
Date: 6 Sep 2002 09:56:15 -0700
Subject: question: iso 14977 ebnf grammar for Python?
Message-ID: <77977a28.0209060856.6974e6eb@posting.google.com>

Hello!

Is there a ebnf grammar for Python written according to the ISO 14977
standard somewhere in the www?

Michael asks


From tjreedy at udel.edu  Wed Sep 18 01:18:20 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 18 Sep 2002 05:18:20 GMT
Subject: Splitting a list of strings
References: <mailman.1032309381.3626.python-list@python.org> <am8pvo$dcn$1@driftwood.ccs.carleton.ca>
Message-ID: <w0Uh9.116763$5r1.4544705@bin5.nnrp.aus1.giganews.com>

"Sean Ross" <sross at connectmail.carleton.ca> wrote in message
news:am8pvo$dcn$1 at driftwood.ccs.carleton.ca...
> Okay. Here is some representative sample data:
> [snip]
> @attribute 'cap-shape' { 'b', 'c', 'f', 'k', 's', 'x'}
> [snip]
>
'x','s','n','t','p','f','c','n','k','e','e','s','s','w','w','p','w','o
','p',
> 'k','s','u','p'

Unless you plan to exec parts of the lines as if they were Python
code, I believe you would be better off without all the quotes.  You
could then split on ',' and have the 1-char strings you want..

Terry J. Reedy




From bdesth at nospam.free.fr  Sun Sep  8 23:21:23 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Sun, 08 Sep 2002 23:21:23 -0400
Subject: Python plugins
References: <alf7rc$1p06l1$1@ID-41653.news.dfncis.de>
Message-ID: <3D7C13B3.2070606@nospam.free.fr>

Tobias Klausmann wrote:
> Hi there,
> 
> For an open source project that needs to be extensible by its
> users, I would like to create a plugin architecture. Mind you,
> I'm not trying to use Python as a plugin language only like XChat
> or Gnumeric do. Let me illustrate:

[snip]

> But now I have two
> problems: first, I have no idea how to call, say
> pi.someplugin.dostuff(args) except constructing each and every
> call as a string and exec() it - which is both error-prone and
> awkward IMHO.


Just a snippet without any other testing code or whatsoever...

module[index] = __import__(module_name, globals(), locals(), [])
(...)
module[index].dostuff(args)

HTH
laotseu



From fromclay at maxitec.co.za  Wed Sep  4 03:42:49 2002
From: fromclay at maxitec.co.za (Morne)
Date: Wed, 4 Sep 2002 09:42:49 +0200
Subject: help on log
Message-ID: <009201c253e6$b4f47ba0$2500a8c0@maxitec.co.za>

Could someone please help me with this? Im really new to any kind of programming .If posible would you be
 able to answer in detail.

Thanx
 
 HERE IS WHAT I HAVE:

 import re
 import string

 secondsMatcher=re.compile("Connect time: (\d+) secs")

 def getStuff(file):
     closeline = 0
     errorcount = 0
     secondscount = 0
     connectcount = 0
     data_outcount = 0
     data_incount = 0
     ppplist = file.readlines()

     for line in ppplist:

       if string.find(line,"LayerUp") >= 0:
           print "found connect"
           connectline = string.split(line)
           connectcount = connectcount + 1

       elif string.find(line,"ERROR") >= 0:
           print "found error"
           errorcount = errorcount + 1

        elif string.find(line,"Connect time") >= 0:
           print "found close"
connectSecs=secondsMatcher.search(line).group(1)
           print "Connected for",connectSecs
          closeline = string.split(line)
           secondscount = secondscount + string.atoi(connectSecs)

 def main():
    f=open("C:/my documents/ppp.log","r")
     getStuff(f)
     return None

 if __name__=="__main__":
     main()

 THIS IS WHAT I NEED THE PROGRAM TO DO:

 I want the program to read the attached ppp.log file and give me the
 following type of report:


 PPP Log file summary for 5 Jul 2002
 -----------------------------------
 Jul  5 18:15:01 Connected to internet (IP 155.239.150.146)
 Jul  5 18:18:08 Closed connection. ( Connect time: 197 secs: 5091 octets
in,
 1864 octets out )
 -----------------------------------
 Successful connections: 5
 Failed connections: 2
 Total time online: 456 seconds Total data in: 66576 octets
 Total Data out: 66349 octets

HERE IS part of the olg file seems noone wants to touch the Attachment(lol)

> Jul  5 18:00:00 maxigate newsyslog[87300]: logfile turned over
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: Connected to local client.
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Command: /var/run/internet: quit
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: /var/run/internet: Client
connection dropped.
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: Connected to local client.
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Command: /var/run/internet: set
timeout 180
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Command: /var/run/internet: dial
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: bundle: Establish
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: deflink: closed -> opening
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: deflink: Connected!
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: deflink: opening -> dial
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Chat: Phone: 0860007249
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Chat: deflink: Dial attempt 1 of
1
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: /var/run/internet: Client
connection closed.
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Chat: Send: AT^M
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Chat: Expect(5): OK
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Chat: Received: AT^M^M
Jul  5 18:15:01 maxigate ppp[34607]: tun0: Chat: Received: OK^M
Jul  5 18:15:01 maxigate ppp[34607]: tun0: Chat: Send: ATE1Q0^M
Jul  5 18:15:01 maxigate ppp[34607]: tun0: Chat: Expect(5): OK
Jul  5 18:15:01 maxigate ppp[34607]: tun0: Chat: Received: ATE1Q0^M^M
Jul  5 18:15:01 maxigate ppp[34607]: tun0: Chat: Received: OK^M
Jul  5 18:15:01 maxigate ppp[34607]: tun0: Chat: Send: ATx1DT0860007249^M
Jul  5 18:15:03 maxigate ppp[34607]: tun0: Chat: Expect(40): CONNECT
Jul  5 18:15:22 maxigate ppp[34607]: tun0: Chat: Received:
ATx1DT0860007249^M^M
Jul  5 18:15:22 maxigate ppp[34607]: tun0: Chat: Received: CONNECT
33600/ARQ/V34/LAPM/V42BIS^M
Jul  5 18:15:22 maxigate ppp[34607]: tun0: Phase: deflink: dial -> carrier
Jul  5 18:15:23 maxigate ppp[34607]: tun0: Phase: deflink: /dev/cuaa0: CD
detected
Jul  5 18:15:23 maxigate ppp[34607]: tun0: Phase: deflink: carrier -> login
Jul  5 18:15:23 maxigate ppp[34607]: tun0: Phase: deflink: login -> lcp
Jul  5 18:15:23 maxigate ppp[34607]: tun0: LCP: FSM: Using "deflink" as a
transport
Jul  5 18:15:23 maxigate ppp[34607]: tun0: LCP: deflink: State change
Initial --> Closed
Jul  5 18:15:23 maxigate ppp[34607]: tun0: LCP: deflink: State change
Closed --> Stopped
Jul  5 18:15:24 maxigate ppp[34607]: tun0: LCP: deflink: LayerStart
Jul  5 18:15:24 maxigate ppp[34607]: tun0: LCP: deflink: SendConfigReq(179)
state = Stopped
Jul  5 18:15:24 maxigate ppp[34607]: tun0: LCP:  ACFCOMP[2]
Jul  5 18:15:24 maxigate ppp[34607]: tun0: LCP:  PROTOCOMP[2]
Jul  5 18:15:24 maxigate ppp[34607]: tun0: LCP:  ACCMAP[6] 0x00000000
Jul  5 18:15:24 maxigate ppp[34607]: tun0: LCP:  MRU[4] 1500
Jul  5 18:15:2



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020904/be17b564/attachment.html>

From DGustafson at attbi.com  Thu Sep 19 14:15:58 2002
From: DGustafson at attbi.com (Dave Gustafson)
Date: Thu, 19 Sep 2002 18:15:58 GMT
Subject: Creating a typelibrary for calling Python COM server from c#
Message-ID: <yvoi9.516983$UU1.86221@sccrnsc03>

There was a thread recently that suggested creating a typelibrary in order
to access a Python based COM server from C# (or .NET).  Can anyone please
suggest an easy way to create such a library???




From root at [127.0.0.1]  Mon Sep 16 18:37:39 2002
From: root at [127.0.0.1] (nobody)
Date: Tue, 17 Sep 2002 08:37:39 +1000
Subject: stdin on Win32
Message-ID: <3d865d34$0$22177$afc38c87@news.optusnet.com.au>

This works on Linux, but not Win32.  Win32 just gives me EOF..

# echo foo|test.py

where test.py is:

import sys

print sys.stdin.readline()


WTF?





From gmuller at worldonline.nl  Sun Sep 22 14:29:27 2002
From: gmuller at worldonline.nl (GerritM)
Date: Sun, 22 Sep 2002 20:29:27 +0200
Subject: Smalltalk vs. Python and COM was Re: Byte Magazine and Smalltalk
References: <MXFh9.23885$jF4.2037258@twister.southeast.rr.com> <3d87a686@tobjects.> <rd8fougqbgdivimn7vuest3ugoj57lvrr6@4ax.com> <Xns928CA588EF17Dtrippyparcplacecom@64.164.98.7> <3d87dd7c$1@tobjects.> <g9Vh9.313655$kp.1041470@rwcrnsc52.ops.asp.att.net> <Xns928D58DF09B6trippyparcplacecom@64.164.98.6> <20020918.115128.1605894428.1527@cybermesa.com> <amkpk0$or2$1@fe2.cs.interbusiness.it> <amkqfs$ldg$1@reader1.tiscali.nl> <amkup0$qtb$1@fe2.cs.interbusiness.it>
Message-ID: <aml2a5$q19$1@reader1.tiscali.nl>

"Pino Gargiulo" <jk3380 at naida.org> schreef in bericht
news:amkup0$qtb$1 at fe2.cs.interbusiness.it...
<...snip...>
>
> Thanks for your advice and prompt reply!
>
> If I understand you correctly you are suggesting that Python is the way to
> go
> for COM m$ API rather than Smalltalk .. again Id like to hear suggestions
> on which product. My dabbling experience is just with VWNC
>
> TNX again
>
> Pino
>
What is VWNC? What end level application do you want to support and/or
build?

The ActiveState distribution I suggested is a Python distribution
("batteries included") which allow you to do nearly everything. Additional
packages can be found at many places a.o. in the vaults of Parnassus.

regards Gerrit
--
www.extra.research.philips.com/natlab/sysarch/





From duncan at NOSPAMrcp.co.uk  Tue Sep 24 05:07:31 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 24 Sep 2002 09:07:31 +0000 (UTC)
Subject: list problem
References: <mailman.1032818129.8126.python-list@python.org> <3D8F9F5E.DC48B13F@pop.ntlworld.com> <mailman.1032828726.20071.python-list@python.org> <amp6i1$7n9bi$1@ID-77524.news.dfncis.de>
Message-ID: <Xns929364D72B214duncanrcpcouk@127.0.0.1>

Thorsten Kampe <thorsten at thorstenkampe.de> wrote in news:amp6i1$7n9bi$1 at ID-
77524.news.dfncis.de:

> * Anthony Tuininga
>> If you have list comprehensions available, use
>>
>> [b for a in x for b in a]
> 
> What about:
> reduce(lambda x, y: x + y, a) ?
> 

If you are going to use reduce, then you would be better off writing:

   import operator
   reduce(operator.add, x)

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From bruce-news at hartweg.us  Mon Sep 30 00:58:14 2002
From: bruce-news at hartweg.us (Bruce Hartweg)
Date: Mon, 30 Sep 2002 04:58:14 GMT
Subject: idea for a much needed application
References: <576c1752.0209291326.56dcdf31@posting.google.com> <pasepuc0r73mh5vlcsu1l7erl4e2b17qvi@4ax.com> <an831g$bfrqn$1@ID-157863.news.dfncis.de>
Message-ID: <3D97D9E8.5080008@hartweg.us>


mike henley wrote:
> "TGOS" <tgos at invalid.invalid> wrote in message
> news:pasepuc0r73mh5vlcsu1l7erl4e2b17qvi at 4ax.com...

> 
> "other than the ones already present on a web-server", exactly! i want to be
> ably to use whatever language, modules or applications i choose, and over
> the year use others, freely. if i choose perl, php, or even python, ruby or
> tcl, then i want to be able to do that. if i want to use mysql or something
> else that i prefer, then i want to have the freedom to do that. i do not
> wish to be restricted by the "ones already present on a web-server". i do
> not want to end up in a catch 22 situation, should i choose a good deal on a
> host, and then choose applications within the limits of what it allows me,
> or choose applications, and try to find a host that'll provide a good deal.
> 

The "freedom" to do all that is there, but freedom ain't neccessarily
free.

Besides, I think you are thinking too hard on this. If you want to run
your own site off your own PC & aren't on-line all the time, then so be it.

There are dynamic DNS services (even free ones!) and people can go to 
your page - if your not on-line they won't get there, if you are, they 
will. The auto-notification is mostly overkill, If they aren't around
to see it what goo dis it, if they are around they can alsways just try.

Also you could do a 2 tiered approach where you have much of your
content on a real hosted site (free, or paid) and then have a link from
there to you home machine for the sections you want to have all the 
goodies on - you could even auto-update that page when you connect
or disconnect to only have the link when you are on-line.

Bruce



From jb at cascade-sys.com  Tue Sep 17 01:11:14 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 16 Sep 2002 22:11:14 -0700
Subject: doh
References: <3d83fa2c$0$24041@echo-01.iinet.net.au> <mailman.1032061218.4602.python-list@python.org> <3d84a3ed$0$24051@echo-01.iinet.net.au> <3D8659FE.D406C4F8@pop.ntlworld.com>
Message-ID: <3D86B972.9060209@cascade-sys.com>

a.clarke11 wrote:

>>A deer...
>>    
>>
A Female deer...

---jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From claird at starbase.neosoft.com  Thu Sep 19 08:44:50 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 19 Sep 2002 07:44:50 -0500
Subject: linux python ideas
References: <3D88E578.6010001@diespammerdieuselesspython.com> <amapge$22r$1@peabody.colorado.edu> <23891c90.0209182329.52b55550@posting.google.com>
Message-ID: <8D080F827517373B.5119372FF30E25F4.FCB4F73B5EFD9166@lp.airnews.net>

In article <23891c90.0209182329.52b55550 at posting.google.com>,
Paul Boddie <paul at boddie.net> wrote:
			.
		[several apt and
		accurate points]
			.
			.
>Of course, stuff like "network programming" is fairly easy with other
>languages, but unless Perl has spawned an interactive mode, for
>example, you'll be showing people things they haven't seen before if
>you show some experimentation at the command prompt. You might also
			.
			.
			.
Has Perl spawned an interactive mode?  Read
<URL: http://phaseit.net/claird/comp.lang.perl.misc/perl_interactive.html >
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From sholden at holdenweb.com  Sat Sep  7 15:00:39 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 7 Sep 2002 15:00:39 -0400
Subject: Bind threads to addresses -- Windows & urllib?
References: <mailman.1031182087.6726.python-list@python.org>
Message-ID: <_2se9.16902$l_4.8964@atlpnn01.usenetserver.com>

"Nick Arnett" <narnett at mccmedia.com> wrote ...
>
>
> --
> Nick Arnett
> Phone/fax: (408) 904-7198
> narnett at mccmedia.com
>
> > That's not gonna help you at all, because all threads will
> > be capped by the interface's top throughput, or by your
> > machine's processing power.
>
> You're making an incorrect assumption about the purpose of using multiple
> addresses.  It has nothing to do with my end of the connection; it is to
> cope with servers that regard a reasonably well-behaved spider (in my
> opinion, at least) as a denial-of-service attack.  If the server operators
> would reveal what they regard as well-behaved, I wouldn't have to resort
to
> this, but as one might expect, nobody wants to disclose the parameters of
> their DOS defenses.  I can't get the relevant sites to even respond to
> inquries... and they don't even have a robots.txt file.
>
> On the other hand, the more I think about this, the less interested I am
in
> bothering, since it would surely be easy for a server to block a range of
> addresses.
>
> I'm also working the other obvious solution, heuristics for the spider to
> set its own speed so that it won't trigger  defenses -- but it appears
> there's something more than simple rules at the other end.  The robot wars
I
> expected years ago have arrived...
>
> No lectures on what consists good robot behavior, please -- I've been
> operating *the* list on that subject for years
> (http://www.mccmedia.com/mailman/listinfo/robots)
>

Nick:

Have you considered that it may simply be the lack of the usual headers
produced by browser that causes the remote servers to reject your traffic?

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                        pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From buzzard at urubu.freeserve.co.uk  Tue Sep 10 14:47:42 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Tue, 10 Sep 2002 19:47:42 +0100
Subject: Graph Layout Algorithms
References: <alksio$d2l$1@newsg3.svr.pol.co.uk> <pSmf9.3468$cP3.6944@news.iol.ie>
Message-ID: <alleev$vah$1@newsg2.svr.pol.co.uk>

"Padraig Brady" <Padraig at Linux.ie> wrote in message
news:pSmf9.3468$cP3.6944 at news.iol.ie...
> Duncan Smith wrote:
> > Hello,
> >          Just wondering if anyone has implemented anything in Python.  I
> > need to lay out my graphs (directed acyclic / undirected / trees) in a
> > reasonably clear way.  So if anyone has already implemented anything, or
if
> > anyone has advice on appropriate algorithms, I'd be glad to hear about
it.
> > Cheers.
> >
> > Duncan
>
> Do you mean graphically represent or internally in data structures?

Represent; the data structures are done (although might be rethought at some
stage).  What I'm after is an algorithm that will calculate suitable
coordinates for the nodes based on indegree / outdegree etc.  Currently I
just arrange them in a circle, but that doesn't look too good.  I suppose
there might be cases  where I'd like to show structures within the graphs as
clearly as possible too (ancestral graphs / cliques / separators / maximal
prime subgraphs).  But some basic algorithms to get me started would be
useful.  I'll check out the links.  Cheers.

Duncan





From bokr at oz.net  Tue Sep  3 18:01:22 2002
From: bokr at oz.net (Bengt Richter)
Date: 3 Sep 2002 22:01:22 GMT
Subject: efficient list merging
References: <ced73313.0209031012.13624f3@posting.google.com>
Message-ID: <al3bfi$2kh$0@216.39.172.122>

On 3 Sep 2002 11:12:38 -0700, theoryboy at my-deja.com (Peter Saffrey) wrote:

>I have two lists of lists of strings which I would like to merge
>efficiently without repeats. I assume that each list currently
>contains no repeats. At the moment, I have two possible
>implementations:
>
They aren't implementations. Come back when you've at least tried
to compile the code you're posting ;-/

Post a short example with data. That is the clearest way to show
what you are trying to do. Also say something about how you got
your lists of lists of strings.

Regards,
Bengt Richter


From yka at mbnet.fi.invalid  Sat Sep  7 16:48:55 2002
From: yka at mbnet.fi.invalid (yka)
Date: Sat, 07 Sep 2002 20:48:55 GMT
Subject: Larry Wall's comment on python...
References: <mailman.1031353212.13037.python-list@python.org>
Message-ID: <1112946.EjPAhmYJ08@cartman>

Mark McEahern wrote:

>> Of course, most of my current code has one big block, main(), and I just
>> cut-and-paste the rest.
> 
> Your message brought me smiles, tears, and laughter--I haven't had that
> much fun since watching Heehaw.
> 

yes I'm astonished by this revelation too
-- 
Haloo?


From aleax at aleax.it  Mon Sep 30 10:03:01 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 30 Sep 2002 14:03:01 GMT
Subject: help with writing extension
References: <inHl9.24658$Ii4.986202@bin2.nnrp.aus1.giganews.com>
Message-ID: <pQYl9.190868$pX1.6828893@news2.tin.it>

<posted & mailed>

Eric Hagemann wrote:

> I have been handcrafting an extension module creating a new type
> 
> I am making use of the numeric methods (+,-,* etc)
> 
> I solved the problem of the object interating with other types by working
> through the __coerce__() function and have no issue with operations like
> <obj>+2 but I am getting an error on 2-<obj>.

As I understand, __coerce__ is not recommended for new code.  You should,
I believe, add Py_TPFLAGS_CHECKTYPES to your tp_flags slot, and be ready
in your numeric functions to deal with types as they come.  Sometimes
coercion can be handier, but in general avoiding it is better.


> In Beazley's book there are descriptions of functions like __radd__() when
> defining a new type in python, but I cannot find reference to these
> functions in C ?

If you grep Python's C sources (2.2 or better), you'll find some such
references.  Slots nb_add of the PyNumberMethods structure to which
slot tp_as_number of your PyTypeObject structure point supplies both
__add__ and __radd__ (when you call PyType_Ready on your type, suitable
wrappers on that slot are exposed with those two names).


> I have checked the doc's (not to say that I did not miss something ;) )

I think you still need to complement existing docs with some little
study of the Python sources when you want to design rich and complete
extension types -- and I think that dealing with __radd__ etc in the
right way is part of a "rich and complete" (rather than just a
really minimal) extension type.  I think this will still be true
when my "Python in a Nutshell" is finally out, because the chapter
on extending and embedding with C is limited to 40-50 pages (can't
be more precise right now because the book's being edited and I'm
not sure how many of my examples and details the editor will want
to keep -- or if even more will be added), and doing complete justice
to the subject would require about 10 times as much space.

But I also think that asking here about such issues is just right!
Best place to get help -- and this way the help gets archived (thanks
to google groups and other archivers) and can be pointed to in the
future (which alas doesn't work for other mailing lists such as
python-help, nor, of course, for private mail:-).


An example can help -- sorry if it's not very concise, but the
PyTypeObject structure is so large that any examples of its use
are also large, even though most fields are 0.  Say I have a
directory ~/extype with a setup.py file such as:

from distutils.core import setup, Extension

setup(name = "samp1",
    version = "1.0",
    description = "Sample extension type",
    maintainer = "Alex Martelli",
    maintainer_email = "aleaxit at yahoo.com",
       
    ext_modules = [ Extension('samp1', sources=['samp1.c']) ]
)

and a samp1.c file such as:

#include "Python.h"

typedef struct {
    PyObject_HEAD
} samp1;

static void
samp1_dealloc(PyObject *op)
{
    op->ob_type->tp_free(op);
}

static PyObject *
samp1_amethod(PyObject *self)
{
    Py_INCREF(Py_None);
    return Py_None;
}

static PyObject *
samp1_add(PyObject *self, PyObject *other)
{
    PyObject* result = PyTuple_New(2);
    if(!result)
        return NULL;
    Py_INCREF(self);
    PyTuple_SET_ITEM(result, 0, self);
    Py_INCREF(other);
    PyTuple_SET_ITEM(result, 1, other);
    return result;
}

static PyMethodDef samp1_methods[] = {
    {"amethod", (PyCFunction)samp1_amethod, METH_NOARGS},
    {NULL,  NULL}           /* sentinel */
};

static PyNumberMethods samp1_as_number = {
    samp1_add,
};

static PyTypeObject samp1_t = {
    PyObject_HEAD_INIT(0)
    0,
    "samp1",
    sizeof(samp1),
    0,
    samp1_dealloc,
    0,
    0,
    0,
    0,                                  /* tp_compare */
    0,                                  /* tp_repr */
    &samp1_as_number,                   /* tp_as_number */
    0,                                  /* tp_as_sequence */
    0,                                  /* tp_as_mapping */
    0,                                  /* tp_hash */
    0,                                  /* tp_call */
    0,                                  /* tp_str */
    0,                                  /* tp_getattro */
    0,                                  /* tp_setattro */
    0,                                  /* tp_as_buffer */
    Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE|Py_TPFLAGS_CHECKTYPES,
    "sample extension type",
    0,                                  /* tp_traverse */
    0,                                  /* tp_clear */
    0,                                  /* tp_richcompare */
    0,                                  /* tp_weaklistoffset */
    0,                                  /* tp_iter */
    0,                                  /* tp_iternext */
    samp1_methods,                      /* tp_methods */
    0,                                  /* tp_members */
    0,                                  /* tp_getset */
    0,                                  /* tp_base */
    0,                                  /* tp_dict */
    0,                                  /* tp_descr_get */
    0,                                  /* tp_descr_set */
    0,                                  /* tp_dictoffset */
    0,                                  /* tp_init */
    PyType_GenericAlloc,                /* tp_alloc */
    PyType_GenericNew,                  /* tp_new */
    _PyObject_Del,                      /* tp_free */
};

static PyMethodDef no_methods[] = { {0} };

void
initsamp1(void)
{
    PyObject* self;
    PyType_Ready(&samp1_t);
    self = Py_InitModule("samp1", no_methods);
    PyObject_SetAttrString(self, "samp1", (PyObject*)&samp1_t);
}


Now of course running python setup.py in this dictionary builds
and installs extension module samp1.  Let's look at it in action:

[alex at lancelot extype]$ python
Python 2.2.1 (#2, Jul 15 2002, 17:32:26)
[GCC 2.96 20000731 (Mandrake Linux 8.1 2.96-0.62mdk)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import samp1
>>> examp = samp1.samp1()
>>> examp + 'ciao!'
(<samp1 object at 0x8160f28>, 'ciao!')
>>> 'salute...' + examp
('salute...', <samp1 object at 0x8160f28>)
>>>

See what's going on?  Despite the (traditional) argument names,
samp1_add is being called with the LEFT operand of + as the first
argument ('self'), the RIGHT operand as the second argument
('other') -- your C code can test the types directly to do the
right thing in both the __add__ and __radd__ cases...!

Sure, you CAN make do with coercion (so have all Python extension
types until about last year), but bypassing it makes things a bit
clearer, I believe.  If you do want coercion in some but nor all
of your numeric-methods you can deal with it yourself -- call your
own "coercion-like" function from your methods' code as and where
appropriate.  I do think it's simpler than having Python do it for
you and sometimes having to "bypass" it, so to speak... this way
you can still supply a function to be called when the user does an
explicit coerce call, and yet have whatever precise behavior you
want for all operators that look like numeric operators ("look like",
because, for example, sequence concatenation isn't really numeric --
it just LOOKS LIKE it is, and has to live in PyNumberMethods, but
that's not quite where one might like to place it...).


Feel free to keep posting such questions to this list if there is
anything unclear in the above, and/or if you need any more help!


Alex



From mike at bindkey.com  Wed Sep 25 13:48:05 2002
From: mike at bindkey.com (Mike Rovner)
Date: Wed, 25 Sep 2002 10:48:05 -0700
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <%8ci9.138477$5r1.5541622@bin5.nnrp.aus1.giganews.com> <amltd3$3k$1@glue.ucr.edu> <3D911FE1.1050801@something.invalid>
Message-ID: <amssru$pt$1@main.gmane.org>

"Greg Ewing" <see_reply_address at something.invalid> wrote in message
news:3D911FE1.1050801 at something.invalid...
> thp at cs.ucr.edu wrote:
>
> Personally, I rather like
>
>     a if b else c
>
> but Guido doesn't like that either, unfortunately.
>
> Perhaps I'll slip it into Pyrex when no-one's looking...:-)

It's not as readable as one think ;)

I, personally, read this as 'take a if b is true else take c' which
corresponds to 'b? a: c' in C.
So, please, don't create ambiguity from nowhere.

gratefully-plex-er-ly yours,
Mike







From timr at probo.com  Fri Sep 27 00:53:57 2002
From: timr at probo.com (Tim Roberts)
Date: Thu, 26 Sep 2002 21:53:57 -0700
Subject: problem with dictionaries within lists
References: <pan.2002.09.25.05.21.07.477231.167@ihatespam.shaw.ca>
Message-ID: <pso7pu0ob92v95oqbf6nccvf39960vd0n5@4ax.com>

Grant Hallam <gupsmail at ihatespam.shaw.ca> wrote:
>
>#Compiled regular expressions
>#Used in web parser
>rcN = re.compile('Name')
>rcL = re.compile('Location')
>rcT = re.compile('Transit')
>rcTN = re.compile('Technical_Knowledge')
>rcLS = re.compile('Like_Site')
>rcRT = re.compile('Resolve_Tech_Issues')
>...many more deleted...

On an unrelated note, I hope you won't mind if I comment on this part of
your code.

You would get better performance and less resource use out of this if you
replaced these regular expression matches with simple
list[0].find('XxxXxx') calls.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From marklists at mceahern.com  Wed Sep  4 17:38:56 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 4 Sep 2002 16:38:56 -0500
Subject: str() of a tuple
In-Reply-To: <H1xnHJ.1zE1@beaver.cs.washington.edu>
Message-ID: <JHEOKEOOLIGLDHCMAHMOEEGMDMAA.marklists@mceahern.com>

> I'm sure this is a FAQ[...]

It is:

http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&selm=slrnacjga3.2uv.huaiy
u%40gauss.almadan.ibm.com

> Is there a convenient way to print out a tuple of floats so that the
> output is actually readable?

Subclass tuple and do your own __str__:

$ python
>>> class mytuple(tuple):
...     def __repr__(self):
...             strings = [str(x) for x in self]
...             s = ','.join(strings)
...             return "(%s)" % s
...
>>> t = mytuple((2, 0.4))
>>> t
(2,0.4)
>>> print (2, 0.4)
(2, 0.40000000000000002)

// m

-




From news at datatailors.com  Thu Sep 19 10:55:01 2002
From: news at datatailors.com (Peter van Kampen)
Date: Thu, 19 Sep 2002 16:55:01 +0200
Subject: Need advice on choosing the technology for intranet based application
References: <3d88fd1e@news.ivm.net> <Gn7i9.1769$ik2.430@fe02>
	<108666F172046BF0.AFF78CED7027291C.080DCC403BC7489C@lp.airnews.net>
Message-ID: <slrnaojpa4.foj.news@datatailors.com>

In article <108666F172046BF0.AFF78CED7027291C.080DCC403BC7489C at lp.airnews.net>, 
Cameron Laird wrote:
> In article <Gn7i9.1769$ik2.430 at fe02>,
> Steve Holden <sholden at holdenweb.com> wrote:
>>"Nenad Propadovic" <propadovic.nenad at debitel.net> wrote ...
> 			.
> 		[much thoroughly
> 		sensible advice]
> 			.
> 			.
>>> The GUI has a part I consider complex: I have to present some statistics
>>in
>>> form of charts or graphs.
>>
>>The Python Image Library (PIL) is very good for creating graphics. You can
>>create them "on the fly" in your web server and then serve back HTML
>>documents with references to the created graphics.
> 			.
> It strikes me that Mr. Propadovic is tackling
> several new techniques and technologies at once.
> Given that, I have an alternative to suggest for
> "some statistics in form of charts or graphs".
> 
> There's a trick for making making simple bar
> charts which requires no extra modules or pro-
> gramming.  It's so simple, in fact, that it makes
> barcharting easy in shell or assembler or C.
> 
> Apparently I've only documented it in Tcl <URL:
> http://wiki.tcl.tk/barchart > (see references
> there, though, for Python hints).  If you need
> a sample dilation in Python, just say so here.
> 
> Also, Mr. Propadovic:  take Steve's advice to 
> heart.  Even the CGI about which he writes need
> not be as challenging as your comments suggest
> you fear it to be.

There's even a simpler method (for a browser-based app): CSS. The
example below works with most (all?) modern (graphical) browsers (I
checked with IE5.5, Opera 6.01 on Win2K and Mozilla 1.0.0 on linux). The
vertical bars are a little harder but not much.

I hope the HTML doesn't mess everyones clients up.

<div style="background-color:red  ;height:16px;width:100px;"></div>
<div style="background-color:blue ;height:16px;width: 50px;"></div>
<div style="background-color:green;height:16px;width: 70px;"></div>
<br />
<div style="position:relative;background-color:red;
	width:16px;height:100px;"></div>
<div style="position:relative;background-color:blue;
	width:16px;height: 50px;top: -50px;left: 20px;"></div>
<div style="position:relative;background-color:green;
	width:16px;height: 70px;top:-120px;left: 40px;"></div>

PterK


From mhammond at skippinet.com.au  Mon Sep 23 23:37:23 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 24 Sep 2002 03:37:23 GMT
Subject: Python threading?
References: <f9Pi9.10275$Lg2.2191294@news2.news.adelphia.net> <amogmi$r1l$1@panix1.panix.com>
Message-ID: <T5Rj9.13058$6g7.39977@news-server.bigpond.net.au>

Aahz wrote:
> In article <f9Pi9.10275$Lg2.2191294 at news2.news.adelphia.net>,
> Robert Oschler <Oschler at earthlink.net> wrote:
> 
>>Also, what is a reasonable number of threads to expect to be able to run
>>before context switching overhead becomes a problem (I'm using a PIII 500
>>Mhz with 512MB ram if that helps).
> 
> 
> It depends.  For pure Python-based computation, more than one thread
> will cost you more than you gain.  If it's almost pure I/O (file or
> socket stuff), you might be able to use as many as a couple of hundred
> threads.  Mostly, you're probably best off somewhere between five and
> thirty threads, depending on what you're doing.

I would have thought that a brave call.  If you are performing lots of 
IO and need many many connections, you are almost certainly better off 
using an "asynchronous" approach, much like asyncore.  With a large 
number of threads (where "large" can be as low as single figures) the OS 
overhead of switching will be greater than your internal overhead of a 
more complicated design.

A "one thread per connection" model is doomed from the start - 
therefore, in most applications, I would consider 5 threads a large 
number - particularly when all are actively doing something.  OTOH, if 
you are willing to accept that your first version will accept a max of 
~20 connections, the brain-power saved by using threads beats the 
additional OS overhead by a mile ;)

Mark.



From sadams123 at optushome.com.au  Wed Sep  4 03:33:49 2002
From: sadams123 at optushome.com.au (Steven)
Date: Wed, 4 Sep 2002 17:33:49 +1000
Subject: maximum number of threads?
Message-ID: <3d75b761$0$26216$afc38c87@news.optusnet.com.au>

Is there a maximum number of threads that can be running 'simultaneously' in
Python 2.2 on Win2k/XP?

I'm choosing between a select or threads based design, threads would be
easier, but wasn't sure whether I'll exceed some maximum number of threads.

Basically its a voice server, or rather a voice to text server, so
information produced might be sporadic, or fairly continuous. An arbitrary
number of subscribers subscribe to the system and get voice information sent
out as it comes in (and if it matches the language the subscriber wants),
threads make it simple since each subscription could be separate.

Is there a maximum number of threads, and any opinions on whether the GIL
and context switching in the interpreter is liable to make it unworkable (or
a case of 'build and see')?

TIA

Steven




From mstenner at phy.duke.edu  Fri Sep 27 15:26:50 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Fri, 27 Sep 2002 15:26:50 -0400
Subject: watching mutables?
In-Reply-To: <Xns9296D3AB53777cliechtigmxnet@62.2.16.82>; from cliechti@gmx.net on Fri, Sep 27, 2002 at 08:48:10PM +0200
References: <mailman.1033137396.14482.python-list@python.org> <an211j$oi$1@news.hccnet.nl> <Xns9296C58C9ADFCcliechtigmxnet@62.2.16.82> <k7l7qpvf.fsf@ion-tof.com> <Xns9296D3AB53777cliechtigmxnet@62.2.16.82>
Message-ID: <20020927152650.A17611@phy.duke.edu>

On Fri, Sep 27, 2002 at 08:48:10PM +0200, Chris Liechti wrote:
> Thomas Heller <thomas.heller at ion-tof.com> wrote in
> news:k7l7qpvf.fsf at ion-tof.com: 
> > Chris Liechti <cliechti at gmx.net> writes:
> [triggered by *any* change of object m]
> >> see EventObj:
> >> http://oomadness.tuxfamily.org/p-pyobjtools.php
> >> 
> > This is what I get when clicking on the download link:
> > <title>404 Not Found</title>
> 
> works for me.

I find that rather surprising since that file isn't there.  Are you
sure you clicked on the download link?  For your convenience, I
provide that link here:

http://oomadness.tuxfamily.org/downloads/EventObj-0.2.tar.gz

I am also interested in this module (curiosity, mostly) and would
appreciate a working link if you have one.

					-Michael

-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From james.kew at btinternet.com  Thu Sep 26 04:00:21 2002
From: james.kew at btinternet.com (James Kew)
Date: Thu, 26 Sep 2002 09:00:21 +0100
Subject: Getting last char in string.
References: <amue3q$h59$1@lust.ihug.co.nz>
Message-ID: <amueqq$94qgr$1@ID-71831.news.dfncis.de>

"CheapSkate" <gua81 at XXXyahoo.com> wrote in message
news:amue3q$h59$1 at lust.ihug.co.nz...
> Hi, im writing a simple method that takes in a string input and check if
the
> last charcter is an '/'
> then it will do A else do B
>
> my Q now is how to get the last char???

if stuff[-1:] == '/':

or

if stuff.endswith('/'):

--
James Kew
james.kew at btinternet.com





From ChandrS at telkom.co.za  Thu Sep  5 10:24:08 2002
From: ChandrS at telkom.co.za (Sathish Chandrasekaran (S))
Date: Thu, 5 Sep 2002 16:24:08 +0200
Subject: What database should I use
Message-ID: <CC2C254F4F467345A0A5754CA6E25F0B2B9426@PRTRRA06-XS03.telkom.co.za>

In that case MS-access will be better on the size issue. But on the
other hand if u want ur application to be flexible and ease of use I
would rather suggest FoxPro (though it's it considered obsolete
now-a-days).


Here Is A Test To Know Whether The Mission Of Your Life Is Over .
If You Are Alive It Isn't .

Sathish C
Infra-Amdocs 


-----Original Message-----
From: A [mailto:printers at sendme.cz]
Sent: 05 September 2002 15:21
To: python-list at python.org; tutor at python.org;
activepython at listserv.ActiveState.com; python-help at python.org
Cc: db-sig at python.org; wx-users at lists.wxwindows.org
Subject: What database should I use


Hello,
I am going to program a small application (in Python and wxPython)
which consists about 
2000 - 3000 records . Can you please recommend which database is the
best for that. I 
use Windows.
Maybe MS Access  or FoxPro or MySQL or better?
I want to keep the program as small as possible.
Thanks
Ladislav

_______________________________________________
ActivePython mailing list
ActivePython at listserv.ActiveState.com
To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
Other options:
http://listserv.ActiveState.com/mailman/listinfo/ActivePython




From max at alcyone.com  Thu Sep 19 17:35:39 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 19 Sep 2002 14:35:39 -0700
Subject: lists in dictonaries ?
References: <mailman.1032425484.16513.python-list@python.org>
Message-ID: <3D8A432B.DBCB3EF5@alcyone.com>

Manuel Hendel wrote:

> Is it possible to put lists in dictionaries?

I presume you mean as dictionary keys, since any Python object can be a
dictionary value.  Dictionary keys must be immutable objects (they must
define a hash function); lists are not immutable, but tuples are.

> And is it also possible
> to put multiple lists in a dictionary for one key?

I'm not sure what this means.  Tuples can contain tuples as elements, so
you could always make a higher-level tuple out of the tuples you want to
has as "one key," if that's what you mean.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ People are taught to be racists.
\__/ Jose Abad
    Fauxident / http://www.alcyone.com/pyos/fauxident/
 A "faux" ident daemon in Python


From gerhard.haering at gmx.de  Sun Sep  1 22:37:58 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Mon, 2 Sep 2002 04:37:58 +0200
Subject: curses: AttributeError: initscr
In-Reply-To: <un5iq3ht257s22@corp.supernews.com>
References: <un5dhs4f7qlte8@corp.supernews.com> <un5iq3ht257s22@corp.supernews.com>
Message-ID: <20020902023758.GA5333@lilith.ghaering.test>

* Cesar <g0w1s0 at yahoo.ca> [2002-09-01 22:22 -0400]:
> in the same directory where my script live, there was an old file named 
> curses.py. Then the instruction : "import curses" imported the bad one...

You're not the first one to have this kind of problem.

A way to debug such problems is to start Python with the option "-v" or
even "-vv", which helps you track down where modules are really loaded
from.

-- Gerhard



From alet at unice.fr  Wed Sep 18 08:19:09 2002
From: alet at unice.fr (Jerome Alet)
Date: Wed, 18 Sep 2002 14:19:09 +0200
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com>
Message-ID: <3D886F3D.9AE251F9@unice.fr>

Lee Gray wrote:
> 
> I'm a fairly new programmer with experience in ASP/VBScript/DHTML.
> Other than small utilities, so far I've mostly written ASP apps to
> talk to an Oracle database on HP/UX, with Windows NT/2000 clients, and
> I've done maintenance work on VB and Oracle Forms apps.

Perhaps this is really silly, but why don't you show Zope to your boss ?

http://www.zope.org/

not really a GUI, though

hth

Jerome Alet


From yuba at cyberback.com  Sat Sep 21 14:33:33 2002
From: yuba at cyberback.com (Greg & Janet LINDSTROM)
Date: Sat, 21 Sep 2002 13:33:33 -0500
Subject: mySQL and Python 2.2
Message-ID: <000801c2619d$64c3aba0$f8ffa8c0@lindstrom>

Hello-
I am trying to convert from gadfly to mySQL for my database needs on 
my Windows 98 box. I have installed mySQL and it works (I have 
created my tables and can manipulate them manually), and have 
downloaded mySQL-python-0.9.1, but when I python setup.py build, I 
get an error about cl.exe not existing (and, sure enough, it 
doesn't :-).

I'm running Windows 98 with Activestate Python2.2. Any help you 
could supply (including code snippets of how to get connected to 
MySQL once I get it running) would be appreciated.

--greg


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020921/4b08b417/attachment.html>

From dag at animagicnet.no  Fri Sep 20 04:37:43 2002
From: dag at animagicnet.no (Dag)
Date: 20 Sep 2002 08:37:43 GMT
Subject: finding last match in a file
References: <slrnaojlin.8re.dag@omally.animagicnet.no> <87r8fplvq4.fsf@smtp.blueyonder.co.uk>
Message-ID: <slrnaolc45.8re.dag@omally.animagicnet.no>

In article <87r8fplvq4.fsf at smtp.blueyonder.co.uk>, Keith O'Connell wrote:
> Dag <dag at animagicnet.no> writes:
> 
>> I have a number of very large files where I have to find the last line
>> which matches a certain experssion.  Currently I'm simply opening the
>> file and looping through it noteing each time a line matches and at the
>> end of the loop I see which was the last match.  However this is very
>> slow and inefficient.  Is there some way to read through the file
>> backwards and simply find first match starting from the end of the file.
>> The files are really too big to read the entire thing into memory.
> 
> 
> Would you not be able to use grep with the "-n" option. It will list
> all your matches with the lines numbered so you can see the last match
> having the highest line number. If you then use wc on the file you will
> get the total number of lines. Take one from the other and you can see
> how far from the end your last match is.
> 
> That seems to easy - Have I misunderstood you question?

The problem is that greping through large files with a fairly complex
regexp is taking far too long, esp. when I know that the last occurance is
almost always within the last 100 lines.

Dag


From delaray at hotmail.com  Mon Sep 16 19:00:02 2002
From: delaray at hotmail.com (Raymond de Lacaze)
Date: Mon, 16 Sep 2002 23:00:02 GMT
Subject: International Lisp Conference 2002
Message-ID: <am5m8l$ai8$1@news.franz.com>

==========================================================

               ASSOCIATION OF LISP USERS (ALU)

    INTERNATIONAL LISP CONFERENCE 2002 (ILC 2002)


             HOLIDAY INN - FINANCIAL DISTRICT

                 SAN FRANCISCO, CALIFORNIA

     SUNDAY, OCT. 27TH - THURSDAY, OCT. 31ST


===========================================================

            CALL FOR PARTICIPATION

===========================================================

 Registration deadline:  September 31st, 2002

NEW: Competitions

     1. Dynamic Languages RoboCup Simulation League Competition

     2. Chatterbot Competion


Website: http://www.international-lisp-conference.org/

===========================================================


The Conference
-------------------

The Asociation of Lisp Users (ALU) is proud to announce the
International Lisp Conference 2002. The conference will take
place at the Holiday Inn in San Francisco's Financial District
from October 27th through October 31st.

At the start of the 21st century we see the challenges ahead:
Applications will be more dynamic than ever and will be part of the
fibers of the Web, software and hardware robots will be our
assistants, and biology will be source of both computational problems
and programming metaphors. Applications will be customized to
individuals who happen upon them, our needs and desires will be
served by artificial beings, and our software may more accurately
resemble a termite nest than Rube Goldberg inventions.

Will our mainstream concepts, approaches, and languages hold up?
Where might we turn for progress, inspiration, and technology? Was it
right in front of us all along? The International Lisp Conference
addresses these questions and more.

This four-day conference will bring together a mesh of brilliant minds
from  a diverse international community that involves or borders the
Lisp family of languages. The conference will be uniquely multi-themed in
nature, and each day will paint a different theme. The themes  are  Lisp
and Functional Languages, Web Applications, Robotics and Bioinformatics.
We have  gathered an impressive list of distinguished speakers from each
of these communities that promise to deliver a  captivating a stimulating
conference experience.


==============================================================

REGISTRATION

Registration information is available at:

http://www.international-lisp-conference.org/Overview/LC02-Registration.html


==============================================================

MORE INFORMATION

For more information, again, please visit the website:

http://www.international-lisp-conference.org/

or contact Raymond de Lacaze at delaray at hotmail.com


================================================================








From emile at fenx.com  Sun Sep 22 13:34:49 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 22 Sep 2002 10:34:49 -0700
Subject: Smalltalk vs. Python and COM was Re: Byte Magazine and Smalltalk
References: <MXFh9.23885$jF4.2037258@twister.southeast.rr.com> <3d87a686@tobjects.> <rd8fougqbgdivimn7vuest3ugoj57lvrr6@4ax.com> <Xns928CA588EF17Dtrippyparcplacecom@64.164.98.7> <3d87dd7c$1@tobjects.> <g9Vh9.313655$kp.1041470@rwcrnsc52.ops.asp.att.net> <Xns928D58DF09B6trippyparcplacecom@64.164.98.6> <20020918.115128.1605894428.1527@cybermesa.com> <amkpk0$or2$1@fe2.cs.interbusiness.it> <amkqfs$ldg$1@reader1.tiscali.nl>
Message-ID: <amkut6$c5f$1@main.gmane.org>

GerritM:
> but also described extensively in "Python programming on Win32" by
Mark
> Hammond and Andy Robinson, ISBN 1-56592-621-8. I am using the
ActiveState
> distribution, which is a complete distribution including this win32
stuff.
>
> In short highly recommended!
>

Seconded.

If you've got the ActiveState distribution installed, take a look at
PYTHONHOME\Lib\site-packages\win32com\HTML\QuickStartServerCom.html


--

Emile van Sebille
emile at fenx.com

---------






From erict at millfilm.co.uk  Tue Sep 17 04:32:08 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Tue, 17 Sep 2002 09:32:08 +0100
Subject: how to cast an instance id
Message-ID: <3D86E888.90AF569@millfilm.co.uk>

It there a way to create a new reference of  an object
from the string representation of its id?
Thanks for any help.

class foo:
    def info(self):
        print self

a = foo()
a.info()
ccc = "newa = 0x%x" % id(a)
exec(ccc)





From h.g.normann at gmx.de  Sat Sep 28 13:50:38 2002
From: h.g.normann at gmx.de (Hans - G. Normann)
Date: Sat, 28 Sep 2002 19:50:38 +0200
Subject: how to create QLineEdit.focusOutEvents?
References: <an2eos$9v7m1$1@ID-143433.news.dfncis.de> <mailman.1033203218.22387.python-list@python.org>
Message-ID: <an4nna$avk1k$1@ID-143433.news.dfncis.de>

Phil Thompson wrote:
> 
> Sub-class QLineEdit and re-implement focusOutEvent() in your sub-class.
> 

OK, nice to hear, but it is my first project. Normally im coding in Delphi. 
Who can give an very small example?

from qt import *

class Form1(QDialog):
    def __init__(self,parent = None,name = None,modal = 0,fl = 0):
        QDialog.__init__(self,parent,name,modal,fl)

        if name == None:
            self.setName('Form1')

        self.resize(216,77)
        self.setCaption(self.tr('Form1'))

        self.LineEdit1 = QLineEdit(self,'LineEdit1')
        self.LineEdit1.setGeometry(QRect(60,25,116,22))
        #this shoul hav an focusOutEvent



Hans



From jepler at unpythonic.net  Wed Sep 11 18:46:01 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Wed, 11 Sep 2002 17:46:01 -0500
Subject: [FEEDBACK] Is this script efficient...is there a better way?
In-Reply-To: <200209111453.23751.shalehperry@attbi.com>
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <200209111453.23751.shalehperry@attbi.com>
Message-ID: <20020911174559.A1555@unpythonic.net>

On Wed, Sep 11, 2002 at 02:53:23PM -0700, Sean 'Shaleh' Perry wrote:
> As a final comment you need to be aware that the more keywords you look for 
> the slower this will be.  However there is not a way to get around that, it 
> is just something to keep in mind.

With a good, simple Regular Expression like
    literal1|literal2|...
the time spent searching one string for the compiled RE should be only
proportional to the number of characters in the scanned string (It's just a
big table of state transitions).  This is more cycles per character than
a simple comparison, but for some number of comparisons N you start to win.

I hope that Python's implementation of REs has this property, but I'm not
certain.  Needing to support things like backrefs and lookahead-assertions
blows this nice simple model of REs out of the water.

Jeff



From nmichael at yahoo.com  Fri Sep  6 14:35:14 2002
From: nmichael at yahoo.com (Neophytos Michael)
Date: 6 Sep 2002 11:35:14 -0700
Subject: Computing directory sizes problem
Message-ID: <d32d294c.0209061035.3e6fa400@posting.google.com>

Dear Python experts,
I wrote the following function to compute the size of a directory
(being new to python there maybe better ways to do this - if so please
advice).  It works quite well except when the directory in question
contains files whose names that are not made up of ascii (english)
characters (this typically happens on my machine for my non-english
mp3 files).  For those files os.path.isfile returns false and so their
sizes are not added in the total sum.  All this is on win 2k.

Any help would be greatly appreciated.

Thank you,
Neophytos

--------------------------------------------------------------------
import os;
import os.path;

def dirSize(p):
   def visit(arg, dirname, names):
      paths = map(lambda x: os.path.join(dirname, x), names);
      files = filter(os.path.isfile, paths);
      arg[0] += reduce(lambda x,f: x + os.path.getsize(f), files, 0);

   siz = [0];
   os.path.walk(p, visit, siz);
   return siz[0];
--------------------------------------------------------------------


From loewis at informatik.hu-berlin.de  Mon Sep 16 03:28:20 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 16 Sep 2002 09:28:20 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net>
Message-ID: <j4znuio0kr.fsf@informatik.hu-berlin.de>

Michael Schneider <michaels at one.net> writes:

> Could someone please explain to a "confused" python user. Why the
> current version of stackless python would not be distributed like
> threads???

Very easy: because nobody makes it so. That, in turn, is because
nobody works on making it so. 

Christian (rightfully) points out that this is not his obligation:
this is free software, he is not obliged to do anything - unless you
pay him to do something. Nobody else is obliged to incorporate
Stackless into Standard Python either. If you want this to happen,
contribute.

Regards,
Martin


From kkto at csis.hku.hk  Sun Sep  8 09:39:30 2002
From: kkto at csis.hku.hk (Isaac To)
Date: 08 Sep 2002 21:39:30 +0800
Subject: python program to watch a file - example.
References: <mailman.1031464092.30848.python-list@python.org>
Message-ID: <7ielc48ust.fsf@enark.csis.hku.hk>

>>>>> "Wayne" == Wayne R <wayne at tbnets.com> writes:

    Wayne> I am trying to write a python program that will run as a deamon
    Wayne> and watch a file for data to arrive and then wake up and perform
    Wayne> duties on that data.  I found ways to make a python program run
    Wayne> as a deamon but I am at a loss to find anything related to
    Wayne> watching a file for data so that it can then process the data.  I
    Wayne> didn't really want to do a sleep loop or something like, I am
    Wayne> sure there is a better way.
 
    Wayne> Any ideas or links to more info?
 
See this:

                      http://oss.sgi.com/projects/fam/

Basically, the established interface to access such functionalities in OS,
if implemented, is the C++ library called FAM (File Alteration Monitor).
Perhaps you have to write a module to make it accessible from within Python.
Internally, IRIX use the /dev/imon device to provide such functionality.

             http://www.mcsr.olemiss.edu/cgi-bin/man-cgi?imon+7

Linux has the fcntl called "F_NOTIFY" (see Documentation/dnotify.txt in your
kernel source).  Other OS probably has no interface to get it done, so you
are back to a timed loop.

Regards,
Isaac.


From jkraska at san.rr.com  Sun Sep 29 12:34:35 2002
From: jkraska at san.rr.com (Courageous)
Date: Sun, 29 Sep 2002 16:34:35 GMT
Subject: What does Python fix?
References: <e15b6fa3.0209271418.14cd5642@posting.google.com> <vcmcpusnjqvpj9lavtqlqa7cs01h4i1iij@4ax.com> <87znu1rugt.fsf@pokey.henrik-motakef.de> <mailman.1033302218.21848.python-list@python.org>
Message-ID: <qmaepu07vf83sjv1vq1hpue26gv4ekoq6e@4ax.com>

>> No, I'm talking about how newbies tend to react when you tell them
>> that block structure in Python is determined by indentation.
>
>What do newbies care? I'm sure if they have a problem with indentation for 
>blocks, they're going to have more trouble with remembering to put additional 
>punctuation either side of the blocks, in the right places...
>
>Unless, of course, your "newbie" is actually someone who has had the C-style 
>brace crap engraved on their mind before encountering Python...

When he's referring to "newbies" he means people who already know how to
program. Either that, or he's misinformed. Real newbies aren't baffled at
all by using indentation for scope. 

C//



From bokr at oz.net  Fri Sep  6 16:30:38 2002
From: bokr at oz.net (Bengt Richter)
Date: 6 Sep 2002 20:30:38 GMT
Subject: Interrupting Python
References: <yz5e9.4237$rU2.145619@news4.srv.hcvlny.cv.net>
Message-ID: <alb39e$ers$0@216.39.172.122>

On Fri, 6 Sep 2002 13:26:27 -0400, "Bob Easton" <bob at eleaston.com> wrote:

>I have a script that can run, accessing network resources, for several days.
>Since the script does not normally need keyed input, exception processing
started by crond or such?
>does not raise the keyboard exception until after the program ends normally.
>I would like to be able to interrupt if from the keyboard, but have not
>learned the trick.  How can I do this?
>
Doesn't this work on your system? Or is all the time spent outside the python
interpreter where the SIGINT signal might go elsewhere? Will Ctrl-Break do anything?
Look into the signal module for ways to write your own handlers etc.
man kill if you're on unix.

[13:25] C:\pywk\pi>type catchctlc.py
i=x=0
def foo(n):
   global i,x
   for i in xrange(n):
       x = i**100

if __name__ == '__main__':
    import sys
    try:
        foo(int(sys.argv[1]))  # call your thing here instead
    except KeyboardInterrupt:
        print 'Interrupted by Ctrl-C'
    else:
        print 'Exited w/o interrupt'
    print 'FHOI, the last i**100 x value: %s**100 = %s' % (i, x)

[13:32] C:\pywk\pi>catchctlc.py 1000
Exited w/o interrupt
FHOI, the last i**100 x value: 999**100 = 904792147113709042032214606239950347800488416333
469929276204638572786486592967687651442293753075422163470827543775910358772483632664400945
560381166977421367930719070025493287934646681492648403959754575431541487824089366478208362
425806884425205853497846463239410463810703487931177116401063304949900001

--- here I hit Ctrl-C pretty quickly after starting the program:

[13:32] C:\pywk\pi>catchctlc.py 1000
Interrupted by Ctrl-C
FHOI, the last i**100 x value: 214**100 = 688589417268581026354513628323939985152292357234
549348854713726783654348445214825147860269910421564660573125261812017623244964892234617683
915075029738282499638533043036366826058088415904236592773165115757274885063575866471121351
34001


Regards,
Bengt Richter


From dsavitsk at e-coli.net  Tue Sep  3 00:13:42 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Tue, 03 Sep 2002 04:13:42 GMT
Subject: opening 'file open' dialog from console
Message-ID: <WFWc9.6430$yt3.2976494@newssrv26.news.prodigy.com>

I would like to open a file open dialog from a console on windows (2000). I
do this now using the line

>>> from tkFileDialog import *
>>> fileName = askopenfilename(filetypes=[(".txt", ".txt")], title="Select
the file")

but this opens a second window and leaves it there.  what is the best way
(tkinter or MFC) to do this w/o the second window?

thanks,

-doug




From shalehperry at attbi.com  Wed Sep 11 10:56:21 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 11 Sep 2002 07:56:21 -0700
Subject: deleting the first and the last character of a string
In-Reply-To: <20020911105818.GC2054@partagas.as.de.cw.net>
References: <20020911105818.GC2054@partagas.as.de.cw.net>
Message-ID: <200209110756.21495.shalehperry@attbi.com>

On Wednesday 11 September 2002 03:58, Manuel Hendel wrote:
> I got a file with fields seperated by a "|". The problem is that there
> is also a "|" at the beginning and at the end of each line although
> there no fields in front or behind these "|". Now I want to delete
> these two "|". Is there anything which says delete the first or last
> character of a string? I checked the string module on the web, but I
> couldn't find anything.
>

>>> s = '|f|o|o|'
>>> s.strip('|')
'f|o|o'

and of course the equivalent strip call from the string module.



From pipa at rush.avalon.hr  Mon Sep 23 03:31:01 2002
From: pipa at rush.avalon.hr (rush)
Date: Mon, 23 Sep 2002 09:31:01 +0200
Subject: Smalltalk vs. Python and COM was Re: Byte Magazine and Smalltalk
References: <MXFh9.23885$jF4.2037258@twister.southeast.rr.com> <3d87a686@tobjects.> <rd8fougqbgdivimn7vuest3ugoj57lvrr6@4ax.com> <Xns928CA588EF17Dtrippyparcplacecom@64.164.98.7> <3d87dd7c$1@tobjects.> <g9Vh9.313655$kp.1041470@rwcrnsc52.ops.asp.att.net> <Xns928D58DF09B6trippyparcplacecom@64.164.98.6> <20020918.115128.1605894428.1527@cybermesa.com> <amkpk0$or2$1@fe2.cs.interbusiness.it>
Message-ID: <ammfvg$30i7$1@as201.hinet.hr>

"Pino Gargiulo" <jk3380 at naida.org> wrote in message
news:amkpk0$or2$1 at fe2.cs.interbusiness.it...

> Anybody care to give some advice? I played around with DST but never
looked
> much at the existance of a COM wrapper/interface.

Dolphin Smalltalk has nice interface to windows (and COM), and it is very
comfortable to work with.

rush
www.rush.avalon.hr





From max at alcyone.com  Thu Sep  5 18:10:31 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 05 Sep 2002 15:10:31 -0700
Subject: Graham's spam filter
References: <mailman.1029992314.27908.python-list@python.org> <3D647B4B.28D9BF04@alcyone.com> <mailman.1030050253.11330.python-list@python.org> <3D655AE6.A097AA25@alcyone.com> <ak44j1$1f2ft4$2@ID-125932.news.dfncis.de> <c2ef5940.0209041912.16e102fb@posting.google.com> <3D770156.2E55864C@alcyone.com> <al7tg4$1o5a9k$5@ID-125932.news.dfncis.de> <3D77AC75.B9B7F27@alcyone.com> <al8ehh$1nt6ua$1@ID-125932.news.dfncis.de>
Message-ID: <3D77D657.DE280373@alcyone.com>

Christopher Browne wrote:

> Quoth Erik Max Francis <max at alcyone.com>:
>
> > Spammers are hitting upon the strategy, though, of sending emails in
> > which the body consists of nothing but a completely encoded base64
> > MIME part.  So in that case, the entire body of your message would
> > consist solely of your "base64encoded" token.  So in the general
> > case of any kind of spam filter (not just limited to a Graham
                                     ^
> > filter), it's questionable how useful this will be, unless you plant
> > to always filter against that token, presuming it to always indicate
> > spam.
> 
> I've been using naive Bayesian filtering for years; I don't assume
> that _any_ particular token indicates _any_ particular result.

I thought I made it clear that I was discussing spam filters in general,
not just Graham/Bayesian filters.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ There is nothing so subject to the inconstancy of fortune as war.
\__/ Miguel de Cervantes
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From gmcm at hypernet.com  Sat Sep 21 12:09:42 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 21 Sep 2002 16:09:42 GMT
Subject: Creating a standalone embedded Python app.
References: <f0eefec9.0209210210.56067a8f@posting.google.com>
Message-ID: <Xns92907BB72242Bgmcmhypernetcom@199.171.54.214>

Allen Unueco wrote:

> I'm looking for a way to somehow bundle the Python modules that my
> embedded application uses.  I would like to be able to distribute an
> executable file that would work without having Python installed on the
> system.
> 
> I would rather have everything in one file but having the modules in a
> separate bundle (pickle?) would be workable.

People have used Installer for this.

  http://www.mcmillan-inc.com/installer_dnld.html

You'll need to dive under the covers. You'll need to
figure out how to create an archive that has the
modules you need, and you'll need to look at 
source/common/launch.c & fit that in with however
you're starting Python.

-- Gordon
http://www.mcmillan-inc.com/
 


From Janssen at rz.uni-frankfurt.de  Tue Sep 17 04:55:25 2002
From: Janssen at rz.uni-frankfurt.de (Michael Janssen)
Date: Tue, 17 Sep 2002 10:55:25 +0200
Subject: problems getting a single item out of a list
References: <mailman.1032247223.29574.python-list@python.org>
Message-ID: <3d86ee83@nntp.server.uni-frankfurt.de>

fields[2] for item with *index*-nr 2

Manuel Hendel wrote:

> I try to get a single item out of a list, but after this, the single
> item is stil a list item.
> 
> domainfield = fields[3:4]
> 
> I just want to get a new variable with a the value of the 3rd item of
> the list fields.
> 
> Can anyone help?
> 
> Thanks,
>     Manuel
> 



From chris.gonnerman at newcenturycomputers.net  Thu Sep 19 08:13:05 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Thu, 19 Sep 2002 07:13:05 -0500
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <bd751532.0209082142.54236307@posting.google.com> <allrrs$6hm$2@solaris.cc.vt.edu> <ambpgq$dre$1@nntp9.atl.mindspring.net>
Message-ID: <006501c25fd5$ecc80740$0101010a@local>

----- Original Message ----- 
From: "Matt Gerrans" <mgerrans at mindspring.com>


> One of the niftiest things about Python is you never have to 
> argue about whether you should put your curlies in the wrong 
> place, like this:
> 
> void foo() {
>    //...
> }
> 
> ...or whether you should place them correctly, like this:
> 
> void foo()
> {
>     //...
> }

Them's fightin' words, son...   :-)
 
> It saves the world from a lot of unnecessary creation and 
> release of hot air (so Python has probably done more to help 
> reduce the greenhouse effect than any other software 
> technology!).

I just like that I don't have to type the #!$&@!! things 
anymore.

No smiley this time... I really hate typing curly after curly.
It makes me wonder that anyone *likes* the freakin' things so
much that they named a LANGUAGE after them.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From jarober at gosmalltalk.com  Sun Sep 22 14:33:19 2002
From: jarober at gosmalltalk.com (James A. Robertson)
Date: Sun, 22 Sep 2002 18:33:19 GMT
Subject: Smalltalk vs. Python and COM was Re: Byte Magazine and Smalltalk
References: <MXFh9.23885$jF4.2037258@twister.southeast.rr.com> <3d87a686@tobjects.> <rd8fougqbgdivimn7vuest3ugoj57lvrr6@4ax.com> <Xns928CA588EF17Dtrippyparcplacecom@64.164.98.7> <3d87dd7c$1@tobjects.> <g9Vh9.313655$kp.1041470@rwcrnsc52.ops.asp.att.net> <Xns928D58DF09B6trippyparcplacecom@64.164.98.6> <20020918.115128.1605894428.1527@cybermesa.com> <amkpk0$or2$1@fe2.cs.interbusiness.it> <amkqfs$ldg$1@reader1.tiscali.nl> <amkup0$qtb$1@fe2.cs.interbusiness.it>
Message-ID: <s53soucsjj28a0um62dgmvregr2lccqpqu@4ax.com>

VW does support being a COM server - we have customers using it that
way.  Load COMAll and look at the examples

On Sun, 22 Sep 2002 19:30:57 +0200, "Pino Gargiulo" <jk3380 at naida.org>
wrote:

>
>"GerritM" <gmuller at worldonline.nl> wrote in message
>news:amkqfs$ldg$1 at reader1.tiscali.nl...
>> Long long ago I worked in a large scale Objective-C project. Objective-C
>> inherited many of its good characteristics from Smalltalk. Now I am using
>> Python quite a lot. It has again a lot of the good (dynamic)
>> characteristics, with on top an excellent set of predefined modules. One
>of
>> them are the win32 modules by Mark Hammond, documented on line of course,
>> but also described extensively in "Python programming on Win32" by Mark
>> Hammond and Andy Robinson, ISBN 1-56592-621-8. I am using the ActiveState
>> distribution, which is a complete distribution including this win32 stuff.
>>
>> In short highly recommended!
>
>Gerrit
>
>Thanks for your advice and prompt reply!
>
>If I understand you correctly you are suggesting that Python is the way to
>go
>for COM m$ API rather than Smalltalk .. again Id like to hear suggestions
>on which product. My dabbling experience is just with VWNC
>
>TNX again
>
>Pino
>
>
>

<Talk Small and Carry a Big Class Library>
James Robertson
Product Manager, Cincom Smalltalk


From rm at no-mail.com  Thu Sep 12 13:07:40 2002
From: rm at no-mail.com (Roy Marteen)
Date: Fri, 13 Sep 2002 02:07:40 +0900
Subject: Larry Wall's comment on python...
References: <mailman.1031339711.632.python-list@python.org>
	<pc5bla.off.ln@boundary.tundraware.com>
	<roy-780F0D.21015006092002@reader1.panix.com>
	<vb2cla.6di.ln@boundary.tundraware.com>
Message-ID: <20020913020740.2dae5bb6.rm@no-mail.com>

On 07 Sep 2002 05:25:07 GMT
Tim Daneliuk <tundra at tundraware.com> wrote:

> Roy Smith wrote:
> <SNIP>
> 
>   Writing for a machine which only has 256 bytes of RAM gives you a
> > somewhat different outlook on programming :-)
> 
> The Microchip PIC 16F84A chip I have been using as *68* bytes of RAM -
> well RISC register file, really, but its the same thing -  32 bytes of
> EEPROM, and a whopping 1K of flash memory for storing the program
> itself. And yes, it does give you a whole different perpective on
> programming...

Well... it reminds me when I was young, and we had to program using
ABACUS. It was really somewhat different outlook on programming.


From aleax at aleax.it  Fri Sep 20 05:20:57 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 09:20:57 GMT
Subject: What the heck? open() requires an integer?
References: <FCBi9.13802$Zk4.708378@news2.telusplanet.net>
Message-ID: <ZLBi9.136908$pX1.4869246@news2.tin.it>

Fausto Arinos de A. Barbuto wrote:

>     The open() function is returning me a strange error.
>     An attempt to create an output file with
> 
>             f = open("output.txt","w")
> 
>     yields the following error:
> 
>     Traceback (most recent call last):
>     File "<interactive input>", line 1, in ?
>     TypeError: an integer is required
> 
>     That happens when I am using either Pythonwin or
>     IDLE 0.8, but not in interactive mode. In this latter
>     case I can open, write to and close a file OK.
> 
>     What's the matter? Thanks in advance!

I'd be willing to bet that you did what you SHOULDN'T do,
specifically a statement
        from somewhere import *
before this point in your interactive session.  I'd lay
good odds that the 'somewhere' is os.

Don't do it.  "from ... import *" tramples all over your
namespace: after you use it, how can you know WHAT
function object is referred to by any given name?

Well, if you "print open" you could be able to tell, at
least in Python 2.2...:

>>> print open
<type 'file'>
>>> from os import *
>>> print open
<built-in function open>
>>>

So, when you're calling open after "from os import *",
you're really calling os.open -- and THAT one does indeed
require an integer (as its second argument).


A suggestion: forget all about the existence of statement
"from" until you feel you have totally mastered Python.  I
think you'll have an easier time learning, using, and
mastering Python, if you ALWAYS use the import statement
when you need to do imports, rather than using from.  (The
"from ... import *" is the main issue, but there are others:
e.g., built-in reload won't do what you expect if you've
used statement from -- only if you've used statement import).


Alex



From erict at millfilm.co.uk  Tue Sep 17 11:15:22 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Tue, 17 Sep 2002 16:15:22 +0100
Subject: how to cast an instance id
References: <3D86E888.90AF569@millfilm.co.uk> <hl4eou0q4e8ovjs9942m271r000jrvd1pm@4ax.com>
Message-ID: <3D87470A.CBD72239@millfilm.co.uk>

"Gon?alo Rodrigues" wrote:

> AFAIK there isn't - You probably have to go down the bowls of the C
> implementation to do it. But why would you need such a contortionist's
> act?

Yes I know it sound a little bit twisted.


> If you explain your problem maybe the people here in the list can
> figure out a Pythonic way to do it.

Here is why (simplification)
I designed a generic class which is customizable through a python like
scripting.
Script let you instantiate, add, set/get attribute and much more...

script example:
taskName = "ExecBatch" # default name
ARGUMENT : taskName = '%(taskName)s'
taskRef = Fac.createInstance( "progCommand" ,taskName)
print taskRef,"created",taskRef.getName()
taskRef.Set(command = "someExe -batch")
top.addChild(taskRef)


Now, I would like to give the user some way to save preference, and for
consistency I tought it will be nice to be able to have a script of the
same
kind:

ataskid = 0
ARGUMENT : 'ataskid = 0x%(taskName)x'
if id != 0:
   ataskRef = Fac.getInstance(id)
   if ataskRef != None:
         ataskRef.Set('whoToTell',value = ['someOne','someoneElse'])
         ataskRef.Set('sendEmail',value = 1)




>
> Best regards,
> Gon?alo Rodrigues
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020917/009d993f/attachment.html>

From martin at v.loewis.de  Tue Sep 10 03:00:16 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 10 Sep 2002 09:00:16 +0200
Subject: Python 2.3
References: <3d7cfc19_5@news.newsgroups.com> <3D7D7410.4070000@earthlink.net>
Message-ID: <m3y9aal473.fsf@mira.informatik.hu-berlin.de>

Hans Nowak <wurmy at earthlink.net> writes:

> 
> There's a PEP, but it doesn't look like this will be in 2.3:
> 
>    http://www.python.org/peps/pep-0239.html
> 
> I don't know what the status of this PEP is, BTW... it's not deferred
> or rejected, which suggests this is still under consideration.

Yes, it is. However, it is no target for Python 2.3, either - just
look at the list of open issues. Contributions are welcome, but some
of these issues are really tricky.

Regards,
Martin


From pinard at iro.umontreal.ca  Tue Sep  3 15:19:44 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Tue, 03 Sep 2002 15:19:44 -0400
Subject: efficient list merging
In-Reply-To: <ced73313.0209031012.13624f3@posting.google.com> (theoryboy@my-deja.com's
 message of "3 Sep 2002 11:12:38 -0700")
References: <ced73313.0209031012.13624f3@posting.google.com>
Message-ID: <oqu1l66fu7.fsf@titan.progiciels-bpi.ca>

[Peter Saffrey]

> I have two lists of lists of strings which I would like to merge
> efficiently without repeats.  [...] Or is there a better way?

I am not sure about the relative speed of everything, but I would be
tempted to try:

   dict(zip(list1 + list2, [None] * (len(list1) + len(list2)))).keys()

It seems to work:

>>> list1 = list('abcdebxyz')
>>> list2 = list('pqqr')
>>> list1
['a', 'b', 'c', 'd', 'e', 'b', 'x', 'y', 'z']
>>> list2
['p', 'q', 'q', 'r']
>>> dict(zip(list1 + list2, [None] * (len(list1) + len(list2)))).keys()
['a', 'c', 'b', 'e', 'd', 'q', 'p', 'r', 'y', 'x', 'z']

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



From tjabo.kloppenburg at unix-ag.org  Wed Sep 11 05:34:03 2002
From: tjabo.kloppenburg at unix-ag.org (Tjabo Kloppenburg)
Date: Wed, 11 Sep 2002 11:34:03 +0200
Subject: ANN: lfm 0.9
In-Reply-To: <1031161776.1238.0.camel@lethe.infernus.org>
References: <1031161776.1238.0.camel@lethe.infernus.org>
Message-ID: <20020911113403.0d36d9c5.tjabo.kloppenburg@unix-ag.org>

I?igo Serna <inigoserna at terra.es> wrote:

IS> After 6 months since last release, I bring you a new version
IS> (0.9) of lfm. Read more and download it in next link:...

I tried it, but had some problems because lfm tried to use "python2" as interpreter in some files (the #!/usr/bin/env line).

Normally I'ld expect a program to use the interpreter I use to call setup.py:
  "python2.2 setup.py build"

I missed an uninstall feature aswell, because the files in /usr/bin were not easy to find (files copied with old file dates?).

tk.
--
Tjabo Kloppenburg
  --=- In 103.54 Tagen ist Weihnachten! -=--

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020911/014bb6e5/attachment.sig>

From ny_r_marquez at yahoo.com  Tue Sep 10 10:18:00 2002
From: ny_r_marquez at yahoo.com (R.Marquez)
Date: 10 Sep 2002 07:18:00 -0700
Subject: Windows/DOS: double clicking a .py file
References: <mailman.1031613313.24086.python-list@python.org> <3d7d6f26@news.sentex.net>
Message-ID: <8a27e309.0209100618.46640eaa@posting.google.com>

What I do is the following:

try:
    something()
except:
    print "Program encountered an error:"
    print sys.exc_type
    print sys.exc_value
    os.system("pause")

-Ruben

Peter Hansen <peter at engcorp.com> wrote in message news:<3d7d6f26 at news.sentex.net>...
> David LeBlanc wrote:
> > cmd.exe will accept a flag telling it to stay open after something is run.
> > You can create a shortcut along the lines of:
> > 
> > cmd /K python myscript.py <args>
> [snip rest]
> 
> Although, unfortunately for the OP who has users on Win98, cmd.exe does
> not exist except in NT/2K/XP.
> 
> -Peter


From aahz at pythoncraft.com  Thu Sep 12 12:20:11 2002
From: aahz at pythoncraft.com (Aahz)
Date: 12 Sep 2002 12:20:11 -0400
Subject: class variable won't icrement!
References: <3d8054ef$0$234$4d4ebb8e@news.nl.uu.net> <alpner$bm$1@wanadoo.fr> <alq9hf$bn$1@panix1.panix.com> <PQ2g9.191834$l_4.58187@atlpnn01.usenetserver.com>
Message-ID: <alqerr$sgs$1@panix1.panix.com>

In article <PQ2g9.191834$l_4.58187 at atlpnn01.usenetserver.com>,
Steve Holden <sholden at holdenweb.com> wrote:
>"Aahz" <aahz at pythoncraft.com> wrote in message
>news:alq9hf$bn$1 at panix1.panix.com...
>> In article <alpner$bm$1 at wanadoo.fr>,
>> Eric Brunel  <eric.brunel at pragmadev.com> wrote:
>>>
>>>I always found that inheritance of class attributes to instance attributes
>>>was really misleading, and I even think it would be better if it were
>>>simply removed from Python. IMHO, it would far more conform to the
>>>"explicit is better than implicit" rule that is one of the basis of Python.
>>
>> If I think of it as "scope resolution" rather than "inheritance", I find
>> that it makes more sense.  It works exactly the same way as hiding a
>> global with a local.
>
>Well, not quite exactly the same way. If a local hides a global
>then there's no way to access the global without explicit use of a
>namespace.

Exactly!  The only difference is that it's an error in a function to
attempt to read a global and then later modify it (in the absence of the
global statement).

>A class variable is available by instance-qualified name up until the time
>the instance variable with the same name is created, which is why statements
>like
>
>    self.var += 1
>
>can be so confusing. Especially when you compare operations that give
>different results for mutable and immutable objects.

It might make sense to tighten the rules for modifying class variables,
just the way this gives an error:

x = 1
def foo():
    print x
    x = x + 1

The mutable/immutable thing is exactly the same for functions and
instances.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From kowald at molgen.mpg.de  Tue Sep  3 04:40:29 2002
From: kowald at molgen.mpg.de (Axel Kowald)
Date: Tue, 03 Sep 2002 10:40:29 +0200
Subject: string reading bug in ActiveState Python 2.2 ?
Message-ID: <3D74757D.1010405@molgen.mpg.de>

Hello,

I'm using ActiveStates Python 2.2.1 under win2000 to read in the 
attached text file and it seems that python is only reading part of the 
string. After

fp = open("bla.txt")
z = fp.read()
print len(z), z

I see that only 11 of the 15 characters have been read. I know they are 
non-printable characters, but that shouldn't matter, should it ?  Using 
another python on one of our unix machines everything works as expected 
and all 15 chars are read in.  Is that there something special about 
windows that I should know, or is this a bug in ActiveStates Python ?

Many thanks,

                   Axel Kowald

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: bla.txt
URL: <http://mail.python.org/pipermail/python-list/attachments/20020903/89d88d07/attachment.txt>

From viking_kiwi at yahoo.com  Fri Sep  6 14:25:55 2002
From: viking_kiwi at yahoo.com (Rod Stephenson)
Date: 06 Sep 2002 20:25:55 +0200
Subject: Larry Wall's comment on python...
Message-ID: <wkr8g7rn4c.fsf@yahoo.com>

Slashdot has a list of questions posed to Larry Wall (perl). When
asked his thoughts on other scripting languages, he makes the
following observation about python

"Python is cool to look at small bits of, but I think the "outline"
syntax breaks down with larger chunks of code. I'm with Aristotle on
the structure of discourse--a story should have a beginning, and
middle, and an end. So should blocks"

I'm not quite sure what he's trying to get at here - I guess that for
a long heavily indented chunk of code, you could lose track of the
overall structure, but I don't write code this way.

Any comments?


From dsiroky at email.cz  Sun Sep 29 09:30:10 2002
From: dsiroky at email.cz (David Siroky)
Date: Sun, 29 Sep 2002 15:30:10 +0200
Subject: Graphics library
Message-ID: <pan.2002.09.29.13.30.10.255573@email.cz>

Hi!

I'd like to know, which library for creating images (like libgd) is "the
best". I want to use it with Python.

Thanx

David


From chironsw at swbell.net  Wed Sep 18 16:58:16 2002
From: chironsw at swbell.net (Greg Goodman)
Date: Wed, 18 Sep 2002 20:58:16 GMT
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <mailman.1032152058.11043.python-list@python.org> <am8ept$3ml4d$1@ID-53055.news.dfncis.de>
Message-ID: <3D88E8FC.C8D51FC@swbell.net>

> If I am going to sell my programs, I do not want somebody else to take my
> sourcecode, change a few things, and sell it

If you copyright your work and release it under a restrictive license,
then someone might pirate your work, but it would be illegal and you
would have recourse through the courts.

If you obfuscate and compile your work, then a determined pirate can
*still* steal your work.  Unless you copyright it and release it under a
restrictive license, that "piracy" may not be illegal and you might have
no effective recourse.

In my estimation, the risk to your potential income will be reduced more
by legitimate legal protections than by a technological barrier,
particularly a weak one.  (And, as has been noted before, virtually all
the technological protections for software are weak.)

Regards,
Greg Goodman


From lordmacro at hotmail.com  Thu Sep 19 21:45:05 2002
From: lordmacro at hotmail.com (Tom Widrick)
Date: Thu, 19 Sep 2002 20:45:05 -0500
Subject: implementing reduce protocol in C
Message-ID: <3D8A7DA1.7090506@hotmail.com>

I have an extension type I'm trying to implement the reduce protocol 
with. I have successfully gotten type instances to pickle but cannot 
unpickle them. I get the following error.

newRoot = cPickle.loads(d)
TypeError: ('argument list must be a tuple', <type 'myobj'>, None)

By following the instructions at 
http://www.python.org/doc/current/lib/node65.html I have the following C 
code.

static PyObject *
myobj_reduce(self)
	MyObject *self;
{
	PyObject *args;
	PyObject *result;

	args = PyTuple_New(5);
	if(args == NULL)
		return NULL;

	PyTuple_SET_ITEM(args, 0, self->po_name);
	PyTuple_SET_ITEM(args, 1, self->po_parents);
	PyTuple_SET_ITEM(args, 2, self->po_children);
	PyTuple_SET_ITEM(args, 3, self->po_doc);
	PyTuple_SET_ITEM(args, 4, self->po_dict);

	result = Py_BuildValue("(OOO)", self->ob_type, Py_None, args);
	return result;
};

The type constructor wouldn't have worked as a call to unpickle the 
object so I've used basicnew instead.

static PyObject *
myobj_basicnew()
{
	MyObject *self;

	self = PyObject_New(MyObject, &MyObject_Type);
	return (PyObject *)self;
}

and a setstate method which doesn't matter because the previous error 
occurs before accessing it.

What is it that I am missing?

Tom




From mhammond at skippinet.com.au  Wed Sep 18 22:08:30 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 19 Sep 2002 02:08:30 GMT
Subject: COM object registration
References: <afph9.9468$yt3.4669302@newssrv26.news.prodigy.com> <T_Rh9.17227$Ee4.51953@news-server.bigpond.net.au> <uCTh9.10343$yt3.4866032@newssrv26.news.prodigy.com> <4J_h9.19439$Ee4.59690@news-server.bigpond.net.au> <JY2i9.10441$yt3.4917953@newssrv26.news.prodigy.com>
Message-ID: <ykai9.21321$Ee4.64449@news-server.bigpond.net.au>

dsavitsk wrote:
> "Mark Hammond" <mhammond at skippinet.com.au> wrote in message
> news:4J_h9.19439$Ee4.59690 at news-server.bigpond.net.au...
> 
>>dsavitsk wrote:
>>
>>>>>>a = win32com.client.Dispatch('ecp.MODULE_REGISTER2')
>>>>>
>>>Traceback (most recent call last):
>>>  File "<interactive input>", line 1, in ?
>>>  File "C:\Python\win32com\client\__init__.py", line 92, in Dispatch
>>>    dispatch, userName =
>>>dynamic._GetGoodDispatchAndUserName(dispatch,userName,clsctx)
>>>  File "C:\Python\win32com\client\dynamic.py", line 81, in
>>>_GetGoodDispatchAndUserName
>>>    return (_GetGoodDispatch(IDispatch, clsctx), userName)
>>>  File "C:\Python\win32com\client\dynamic.py", line 72, in
>>
> _GetGoodDispatch
> 
>>>    IDispatch = pythoncom.CoCreateInstance(IDispatch, None, clsctx,
>>>pythoncom.IID_IDispatch)
>>>com_error: (-2147467259, 'Unspecified error', None, None)
>>
>>It is likely that an exception is being encountered that we can get at
>>pretty easy.
>>
>>Add a named param "debug=1" to all UseCommandLine calls.  Then, on both
>>machines, start Pythonwin and select the debug trace collector.  Rerun
>>your tests and see if any additional interesting exceptions are reported.
>>
>>Mark.
>>
> 
> 
> This is the traceback, but I'm afraid it doesn't mean much to me.  It is the
> error that I got when using a VB client and where I get the 'localserver'
> notion.
> 
> in _Invoke_ with 1000 0 2 ()
> in _Invoke_ with 1000 0 1 (u'MODULE_REGISTER2',)
> Registered: ecp.MODULE_REGISTER2 (for debugging)
> Object with win32trace dispatcher created (object=None)
> Traceback (most recent call last):
>   File "C:\Python\win32com\server\dispatcher.py", line 27, in
> _CreateInstance_
>     self.policy._CreateInstance_(clsid, reqIID)
>   File "C:\Python\win32com\server\policy.py", line 199, in _CreateInstance_
>     myob = call_func(classSpec)
>   File "C:\Python\win32com\server\policy.py", line 634, in call_func
>     return apply(resolve_func(spec), args)
>   File "C:\Python\win32com\server\policy.py", line 624, in resolve_func
>     return getattr(module, fname)
> AttributeError: 'localserver' module has no attribute 'regmodule2'

The problem is that for some reason, the registration expects 
localserver.py to have a class named 'regmodule2'.  I'm afraid I am not 
too sure exactly what is going on here and dont have time to setup a 
similar test environment - but after registration, you should be able to 
look in the registry for the objects you registered.  These will 
nominate the module and class names used by this class.  See what you 
can track down there.

Mark.



From kokohh at hotmail.com  Thu Sep  5 19:15:53 2002
From: kokohh at hotmail.com (koko)
Date: Thu, 05 Sep 2002 23:15:53 GMT
Subject: web crawler help?
Message-ID: <JARd9.6927$yt3.3340577@newssrv26.news.prodigy.com>

is there any sample for basic web crawler, that ask for a starting url and
log the url and extract the hyperlinks?
thx




From emile at fenx.com  Thu Sep 26 19:19:13 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 26 Sep 2002 16:19:13 -0700
Subject: Who knows somefunction?
References: <664e469667.69667664e4@icomcast.net> <3D9382AF.5070800@aon.at> <20020926221714.GA11579@sferacarta.com>
Message-ID: <an04jc$7ei$1@main.gmane.org>

Marco Mariani
> Now you can harm yourself. :-)
> 
> Tested with python 2.2.1
> 

;-)

FWIW, against current cvs it crashes when you do:

from pointer import pointer
x = range(3)
y = id(x)
del x
x1 = pointer(y)
print x1

Where's Houdini when you need him...

 
--

Emile van Sebille
emile at fenx.com

---------






From andrew.thompson at ashecastle.com  Tue Sep 24 05:12:52 2002
From: andrew.thompson at ashecastle.com (Andrew Thompson)
Date: Tue, 24 Sep 2002 10:12:52 +0100
Subject: instance creation
In-Reply-To: <265e380f.0209240052.4c2988f3@posting.google.com>
Message-ID: <014901c263aa$904f6e30$4802a8c0@blair>

Apologies for

'

type(arg1)==type(string)...

should read

type(arg1)==type('abc')

'


-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]
On Behalf Of Paul MG
Sent: 24 September 2002 09:52
To: python-list at python.org
Subject: instance creation

Hi

I am dabbling with Python and have run into a bit of an oddness. I
have a class Appointment, with fields Title and Details. I want to
allow two means of construction:

1) Providing Title and Details strings explicitly, to create a new
instance.
2) Providing an ID, which causes that Appointment to be loaded from a
backing store.

Now in C++ or smalltalk, I would do one of

- Provide two constructors:

    def __init__(self, name, details):
      self.name = name
      self.details = details

    def __init__(self, id):
      load(id)

    def load(self, id):
      i = open("appointment"+str(id))
      self = pickle.load(i)

- Provide a class method (creation method) to instantiate:

    def create(id):
      i = open("appointment"+str(id))
      return pickle.load(i)

However, much of this seems not to work.

- Having two constructors seems to be illegal - or at least the second
overwrites the first.
- Assigning to 'self' in the load method seems to really not work!
- Putting the create() method outside of the class works fine - but I
want it to be a class method on Appointment, goddammit!:) Is there no
way to do this?

Any help would be appreciated, basically i am looking for the python
idiom to do this, as my C++ and Smalltalk idiom's clearly aren't in
the grain of the language!

Thanks very much
Paul MG
-- 
http://mail.python.org/mailman/listinfo/python-list




From loewis at informatik.hu-berlin.de  Thu Sep 12 06:12:45 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 12 Sep 2002 12:12:45 +0200
Subject: Pickling slotted instances
References: <mailman.1031805375.9030.python-list@python.org>
Message-ID: <j4fzwf349u.fsf@informatik.hu-berlin.de>

Andrew McNamara <andrewm at object-craft.com.au> writes:

> Instances using __slots__ can't be pickled - is there a fundemental reason
> why this is so, or is it simply yet to be implemented?

See python.org/sf/520644 for a longer discussion.

> At first glance, a __getstate__ for a slotted class could be as simple as:
> 
>     def __getstate__(self):
>         attrs = [getattr(self, attr, None) for attr in self.__slots__]
>         return self.__slots__, attrs
> 
> What am I missing?

That won't work: it won't access the slots of the base type. Also, you
are missing a __setstate__ implementation.

Regards,
Martin


From sismex01 at hebmex.com  Mon Sep  9 10:54:11 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Mon, 9 Sep 2002 09:54:11 -0500 
Subject: Larry Wall's comment on python...
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062330237C5B2@mail.hebmex.com>

Hmmm...

I've found that block-delimiting-chars become like a
drug to the brain; you don't really think about them
until they're not available anymore, then the pain
(well, kinda) starts, and there's whines ("whyyyy
can't I use such-and-such delimiters for my python
programs pleeze???"), and all that.

Once you wean your brain off block-delimiter-chars,
you begin to code more freely, much happier. Really!

Another drug which is taken in great amounts by
C++ and Java programmers are "private:" and
"protected:" and "friends:" stuff, really hideous
and much more sneaky, because they first convince
you that you *need* them in order to make good
programs, and it takes a bit more effort to wean
young minds off them.

;-)

-gus
--

Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From user at domain.invalid  Sun Sep 29 17:30:17 2002
From: user at domain.invalid (user at domain.invalid)
Date: Sun, 29 Sep 2002 21:30:17 GMT
Subject: tkinter & serial port
Message-ID: <IhKl9.45766$bX.8904@sccrnsc02>

Please excuse this if it is a FAQ.

I'm new with Python/Tkinter, but have written LOTS of X windows code.
This may be the root of my problem.

My program needs to talk over a serial port to a device AND update
the display. I get about 10 lines of text every second or so.
Under X Windows, I would simply open the serial device
and add it to the select list which would then cause Xt to give
me a callback whenever there is input available. I can set the I/O
port to understand lines which makes the whole thing neat and clean.

How do I do this or its moral equivalent under Python so that it
runs without modification under both BSD and Windows? The closest
that I've been able to figure out is to poll the serial device
a few times per second. This is ugly, involves overhead and generally
not up to the standard of Python -- clearly I don't yet 'see the light.'

Suggestions? Pointers? Ideas?



From just at xs4all.nl  Mon Sep  9 11:06:29 2002
From: just at xs4all.nl (Just)
Date: Mon, 09 Sep 2002 17:06:29 +0200
Subject: Windows/DOS: double clicking a .py file
Message-ID: <just-A936EC.17062909092002@news1.xs4all.nl>

Sorry if this is a Windows FAQ, I'm a newbie to Windows, but not to 
Python.

If I double-click a Python script (or use the start command in the dos 
prompt) a new dos prompt window is opened and my program is run. The 
window goes away when the program is done. However, the window _also_ 
goes away when the program raises an exception, making it virtually 
impossible to actually _see_ the exception... Is there a way to keep the 
dos prompt window in this case?

A py2exe-fied version of the program shows the same behavior, which 
leads my users to report problems like "the output file is mysteriously 
truncated" instead of "I'm getting a traceback. Here, have a look."...

Any suggestions?

This is with Python 2.2.1 on Windows 98, but a report like above came 
also froma WinXP user.

Just


From jmdeschamps at cvm.qc.ca  Thu Sep 19 16:55:06 2002
From: jmdeschamps at cvm.qc.ca (jmdeschamps)
Date: 19 Sep 2002 13:55:06 -0700
Subject: Tkinter: scrolling the canvas items AND widgets
Message-ID: <3d06fae9.0209191255.63def2eb@posting.google.com>

I'm trying to use canvas as a scrollable region to make dynamic entry
forms (variable number of Entry fields).
I would like to be able to scroll down the canvas if the fields go
outside the visible region using scrollbars :)
This code lets me scroll the graphical items but the widgets stay put
using PythonWin 2.2.1 on Windows 2000
*********
import Tkinter

root = Tkinter.Tk()

# Create widgets
canvas = Tkinter.Canvas(root)
canvas.pack(side=LEFT)
# Vertical scroll bar
verticalScrollbar = Tkinter.Scrollbar(root)                       
verticalScrollbar.pack(side=LEFT,expand=1,fill=Y)          
                                                           
# Configure
root.rowconfigure(0, weight=1)
root.columnconfigure(0, weight=1)
canvas.config(   height = 300,                    
    scrollregion = (0, 0, 100, 1000),
    yscrollcommand=verticalScrollbar.set )
verticalScrollbar.config(
    orient=Tkinter.VERTICAL,
    command = canvas.yview)
                                
# Draw on canvas
canvas.create_polygon(10, 10, 300, 50, 10, 100)
#Add widgets
for i in range(1):
    f=Entry(canvas)
    f.place(x=10,y=10)

root.mainloop()
**** end code *******

With this, I can scroll the polygon but not the Entry field ???

Thanks in advance,

Jean-Marc


From Michael.Sparks at rd.bbc.co.uk  Thu Sep 12 12:18:01 2002
From: Michael.Sparks at rd.bbc.co.uk (Michael Sparks)
Date: Thu, 12 Sep 2002 17:18:01 +0100
Subject: wrapping yield ?
Message-ID: <alqehl$9qv$1@nntp0.reith.bbc.co.uk>

Hi,

I'd like to be able to wrap "yield"  in a function - is that
possible, and if not (my eading of stuff  I've seen so far
says not) are there any plans to make it possible?

Whilst it may seem odd, I'd like to be able to do this:

def my_generator(arg):
    dostuff....
    while(1):
       waitForSomethingCheckingPeriodically()
       processit

where

def waitForSomethingCheckingPeriodically():
   found = 0
   while(not found):
       x = checkTheThing()
       if foundRightThing(x):
           found = 1
      else:
         yield "foo"
   return x

The reason for wanting to wrap this up is due to the
waitForSomethingCheckingPeriodically is a sequence of
code that I'd like to have in several different generators.
The problem of course here is that "my_generator"
doesn't create a generator due to not containing a "yield"
keyword, whereas waitForSomethingCheckingPeriodically
generates a generator, even though I don't want it to be
one...

The context here is of a piece of network server code,
and wanting to be able to enforce relinquishing of control
in the pieces of code that perform reads/writes, as well as
connection acceptance etc. (basically I want to wrap up
the yield call into the piece of code that should yield control,
rather than rely on the user of the code being relied upon
to yield in the right places)

Having read: http://www.python.org/peps/pep-0255.html
I suspect that what I'd like to do here wouldn't work, and
what I really want is something more like:

def_generator my_generator(arg):
...
def waitForSomethingCheckingPeriodically():
...

And have the keyword (or something other than "yield")
indicate that the definition is a generator, since that would
solve this completely - which I not has been ruled out in that
document, and I'm hoping things may've changed since
then :-)

If this is the wrong place to ask this q - apologies, I'd
appreciate being pointed in the right direction :-)

Regards & thanks in advance,


Michael.





From james.kew at btinternet.com  Thu Sep 26 18:00:46 2002
From: james.kew at btinternet.com (James Kew)
Date: Thu, 26 Sep 2002 23:00:46 +0100
Subject: Searching for and sorting files by date and compiling stats in Windows
References: <MOKk9.66578$U7.23791943@twister.socal.rr.com>
Message-ID: <an002g$9us6o$1@ID-71831.news.dfncis.de>

"A. Dimsdale and A. Moser" <fetus at AUTO> wrote in message
news:MOKk9.66578$U7.23791943 at twister.socal.rr.com...
> How can I do the following in Python:
> * Search through networked Windows drives for the string "~w*.tmp" (*
meaning
> wildcard of course) in the filename

os.path.walk and match a regular expression against each file in names on
each callback.

> * Calculate how much disk space is taken up by Word temp files by date
modified
> (i.e., files with "~w*.tmp" as their filenames with date modified in 2001
on
> drive G: on SDspam01 use 75 MB total, files with "~w*.tmp" as their
filenames with date modified in August 2002 on drive H: on DCspam04 use 60
MB total, etc.)

os.stat on each matching file.

> * Optimally, automagickally put this data into an MS Excel spreadsheet
file

win32com.client, but that's the easy part; the difficult part is working out
how to use Excel's automation interfaces to do what you want to do. Dumping
the data into a CSV file might be a half-reasonable substitute.

James




From skoranda at uwm.edu  Thu Sep 26 13:32:39 2002
From: skoranda at uwm.edu (Scott Koranda)
Date: 26 Sep 2002 10:32:39 -0700
Subject: fcntl module on Linux and nonblocking I/O
Message-ID: <7e9d8e38.0209260932.ed2c78@posting.google.com>

Hello,

I just installed python 2.2.1 on my RH 7.3 Linux box and noticed
that the fcntl module does not appear to have O_NDELAY, FNDELAY,
or anything that appears similar.

Python 2.2.1 (#1, Sep 26 2002, 11:55:24) 
[GCC 2.96 20000731 (Red Hat Linux 7.3 2.96-110)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import fcntl
>>> dir(fcntl)
['DN_ACCESS', 'DN_ATTRIB', 'DN_CREATE', 'DN_DELETE', 'DN_MODIFY',
'DN_MULTISHOT', 'DN_RENAME', 'FD_CLOEXEC', 'F_DUPFD', 'F_EXLCK',
'F_GETFD', 'F_GETFL', 'F_GETLEASE', 'F_GETLK', 'F_GETLK64',
'F_GETOWN', 'F_GETSIG', 'F_NOTIFY', 'F_RDLCK', 'F_SETFD', 'F_SETFL',
'F_SETLEASE', 'F_SETLK', 'F_SETLK64', 'F_SETLKW', 'F_SETLKW64',
'F_SETOWN', 'F_SETSIG', 'F_SHLCK', 'F_UNLCK', 'F_WRLCK', 'LOCK_EX',
'LOCK_MAND', 'LOCK_NB', 'LOCK_READ', 'LOCK_RW', 'LOCK_SH', 'LOCK_UN',
'LOCK_WRITE', '__doc__', '__file__', '__name__', 'fcntl', 'flock',
'ioctl', 'lockf']

Without such a constant defined in the fcntl module how can I make a
pipe
non-blocking?

Sincerely,

Scott


From bokr at oz.net  Mon Sep 16 15:50:53 2002
From: bokr at oz.net (Bengt Richter)
Date: 16 Sep 2002 19:50:53 GMT
Subject: read() file??
References: <mailman.1031684894.10361.python-list@python.org> <kuvg5c1r7v.fsf@lasipalatsi.fi>
Message-ID: <am5cmt$m68$0@216.39.172.122>

On 11 Sep 2002 18:27:48 +0300, Erno Kuusela <erno-news at erno.iki.fi> wrote:

>In article <mailman.1031684894.10361.python-list at python.org>,
>jubafre at brturbo.com writes:
>
>| I have a file with assembler instruction and i want read the lines
>| of the file and put in a list of string, this is easy to do, but i
>| don?t want to consider the caracteres "/" and the othes after him.
>
>| file.asm:
>| ADD D1  /coment1
>| LDA D2  /coment2
>
>try something like...
>
>def read_asm(filename):
>    out = []
>    for line in open(filename).readlines():
>        out.append(line.split('/')[0])
         out.append(line.split('/')[0].rstrip())  # so as to get ['ADD D1', ...] not ['ADD D1  ', ...]
                                      
>    return out
>
A nit ;-)

Alternatively, if you have 2.2,

 >>> [x.split('/',1)[0].rstrip() for x in file('file.asm')]
 ['ADD D1', 'LDA D2']


where the file is
 >>> file('file.asm').readlines()
 ['ADD D1  /coment1\n', 'LDA D2  /coment2']

Oops. Well, the missing final \n didn't hurt.

Regards,
Bengt Richter


From gmurali at cs.uchicago.edu  Tue Sep 17 11:52:15 2002
From: gmurali at cs.uchicago.edu (Murali Krishnan Ganapathy)
Date: 17 Sep 2002 08:52:15 -0700
Subject: Windows Registry Manipulation
Message-ID: <83f35bd6.0209170752.64e8119a@posting.google.com>

I have seen people in this news group asking about a higher level
access to the windows registry. I have written one. If people are
interested, I can post the code here. This uses _winreg, so does not
provide some things like Checking Access Permissions on Registry keys.
This is not object oriented. But it is easy to build an OO interface
on top of this.

Some sample code:

import winreg

# Add a new acronym to the database
winreg.SetAcronym('WIN',r'MICR\Windows\CurrentVersion')
# MICR automatically expands to HKLM\SOFTWARE\Microsoft
# and HKLM expands to HKEY_LOCAL_MACHINE

# change a value
winreg.SetValue(r'WIN\Explorer\Advanced\DontPrettyPath',1)

# This will prevent explorer from showing files like "ALLCAPS" as
"Allcaps"

- Murali


From jepler at unpythonic.net  Sat Sep 21 18:05:08 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Sat, 21 Sep 2002 17:05:08 -0500
Subject: Math optimization
In-Reply-To: <6T5j9.145247$pX1.5195840@news2.tin.it>
References: <6T5j9.145247$pX1.5195840@news2.tin.it>
Message-ID: <20020921170507.A1641@unpythonic.net>

Did you bother benchmarking?  I'd be surprised to see a huge difference,
since the overhead of the bytecode loop should be much much greater than
the cost of a single division operation.

Except for 'if 0' and a special variable (__debug__?) which can be 0, there
is no constant folding in the compilation step.  You could fake this by
using
    def f(a, factor=3*5/7/2):
	return a/factor
instead of 
    def f(factor):
	return factor/(3*5/7/2)
(the value of default arguments is computed once, when the function is
created)

Jeff



From aleax at aleax.it  Mon Sep 16 03:41:04 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 16 Sep 2002 07:41:04 GMT
Subject: possibly OT - delete char?
References: <3pch9.10253$Ee4.26056@news-server.bigpond.net.au>
Message-ID: <kWfh9.113071$pX1.4053630@news2.tin.it>

:B nerdy wrote:

> im making a console app. i wish to clear the line. ive got 2 ideas
> 
> - does python have a clear line method to clear the current text of line?
> - is there a character which deletes the previous printed character when
> printed?

There is no cross-platform way to control text-mode consoles, as
far as I know.  For Unix-like platforms, your best best is most
likely to use curses.  For Windows, the effbot's Console package,
http://effbot.org/efflib/console/ .


Alex



From rob at diespammerdieuselesspython.com  Tue Sep 17 10:34:18 2002
From: rob at diespammerdieuselesspython.com (Rob Andrews)
Date: Tue, 17 Sep 2002 14:34:18 GMT
Subject: Good book choice?
References: <mailman.1032271021.9221.python-list@python.org>
Message-ID: <3D873E8F.7060501@diespammerdieuselesspython.com>

> <vent>
> This isn't directed at the book mentioned (which is why I didn't quote 
> its name in this rant), rather at all books which come with CDs.  I hate 
> it when books come with CDs.  I'm perfectly capable of downloading the 
> source code used in the book, and in fact I prefer to hand-type it while 
> reading from the book anyway, to make sure I see every part of the 
> code.  When a book comes with a CD, it is cumbersome to read portably, 
> which is the whole advantage of books, unless you take the CD out.  And 
> if I take the CD out, then there is no doubt that it will get lost, or I 
> will be unable to return the book if dissatisfied, etc.  Plus, you can't 
> tell me that the inclusion of a CD and the envelope/publishing used to 
> include the CD doesn't add a few units of currency to the book's cost.
> 
> Poke this opinion full of holes, but when I buy a book, I want a book, 
> not a software package.  The book should be good enough to stand on its 
> own without the "bonus" material on CD.  I think twice about buying a 
> book that comes with a CD, to tell you the truth.
> </vent>

All perfectly valid sentiments, of course. But for the village in Nepal 
lacking bandwidth because they are still six months away from having a 
phone line (not as far-fetched an example as it may seem), an entirely 
different set of considerations may be found.

I find myself on the middle path on this issue, since I rarely have 
personal use for the CDs, but find that they do make it easier to 
distribute Python in my fits of zealous advocacy. I have bandwidth, but 
no burner, and some are more interested in hearing me out if they don't 
have to face a long (in some cases, expensive) download before they even 
get to see if they would be interested.

Rob Andrews



From markus.vonehr at ipm.fhg.de  Mon Sep  9 11:35:46 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Mon, 09 Sep 2002 17:35:46 +0200
Subject: compile error on SWIG file
References: <3D7C55EE.D9278B1C@ipm.fhg.de> <3D7CB26B.2040706@users.sourceforge.net>
Message-ID: <3D7CBFD2.4E7DFD1F@ipm.fhg.de>

Hi,

I only extended the std_vector example from the distribution with
one function, see example.i included (there are templates inside).
When I run SWIG with -c++ option, the warnings disappear in fact.
But the error from the compiler is the same...

Thanks,

Markus


/* File : example.i */
%module example

%{
#include "example.h"
%}

%include stl.i
/* instantiate the required template specializations */
namespace std {
    %template(IntVector)    vector<int>;
    %template(DoubleVector) vector<double>;
}

/* Let's just grab the original header file here */
%include "example.h"


Lyle Johnson schrieb:
> 
> Markus von Ehr wrote:
> 
> > compiling my c-file generated from SWIG
> > (swig -c -python -o mvelib.c mvelib.i)
> > with a setup-file, I receive the error:
> >
> > c:\programme\vc98\include\eh.h(32) : fatal error C1189: #Error: "eh.h
> > is only for C++!"
> >
> > Anybody knows what I have to change?
> 
> Is it a C++ library that you're wrapping (sounds that way)? If so, you'd
> want to run SWIG with the -c++ option to tell it that.


From rpm1deleteme at frontiernet.net  Fri Sep 27 19:58:09 2002
From: rpm1deleteme at frontiernet.net (RPM1)
Date: Fri, 27 Sep 2002 19:58:09 -0400
Subject: How do you pronounce "__" (double underscore)?
References: <525eaba.0209260956.72bf6786@posting.google.com> <20020926164641.1753ebbb.ckasso@sprynet.com> <an1n0c$3sda$1@news3.infoave.net>
Message-ID: <up9s1033mua980@corp.supernews.com>

> Chris Kassopulo <ckasso at sprynet.com> writes:
> 
> 
>>On 26 Sep 2002 10:56:45 -0700
>>pknotz at sandia.gov (Pat Notz) wrote:
>>
>>
>>>@ = "at"
>>>! = "bang"
>>>. = "dot"
>>>* = "star"


Well, since ! is not pronounced "exclamation", why should __ be 
pronounced "underscore"?  How about just a "sh" sound, since it kind of 
means hidden?

"sh" init "sh" == __init__

or even

"sh" init == __init__

I need another beer.



From max at alcyone.com  Thu Sep  5 15:01:18 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 05 Sep 2002 12:01:18 -0700
Subject: Privacy and Inheritance
References: <mailman.1031180226.13060.python-list@python.org> <al7a71$hnh$1@wanadoo.fr>
Message-ID: <3D77A9FE.68176A7F@alcyone.com>

Eric Brunel wrote:

> I just meant that Python has no way of enforcing "protectedness", as
> leading double underscores (kind of) enforce privateness. As you say,
> using
> a single leading underscore in protected attribute names is merely a
> convention, even if it's a strong one: Python won't prevent any use of
> the
> attribute, neither from sub-classes, nor from the "outside world".
> Anybody
> not knowing the convention can use these attributes in any way without
> any
> problem.

That's right.  Python takes the approach that "We're all adults here,"
and presumes that people will behave.  In many languages the concept of
restricted access can be circumvented in varying ways with increasing
levels of ugliness, so you can think of restricted access as really a
hint (albeit a very strong one in those languages where it is very
awkward).

Python just leaves these things at the hint level (with the exception of
double underscore name mangling, which is a fairly recent addition).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ There is nothing so subject to the inconstancy of fortune as war.
\__/ Miguel de Cervantes
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From merman at snafu.de  Tue Sep 17 10:20:50 2002
From: merman at snafu.de (T. Kaufmann)
Date: Tue, 17 Sep 2002 16:20:50 +0200
Subject: newbie-question: private attributes
Message-ID: <3D873A42.1070400@snafu.de>

Hello everybody,

I know that Python has no modifiers like Java/ C++: keywords like private, public and so on.

What is the difference between the two attributes in this class?

class Test:

     def __init__(self):

         self.__var1 = 1 # I think this is a private attribute
         self._var2  = 2 # ... and this? Maybe protected?

Thanks for help.

o-o

Thomas



From grabiller at 3dvf.net  Wed Sep 25 10:24:56 2002
From: grabiller at 3dvf.net (Guy Rabiller)
Date: Wed, 25 Sep 2002 16:24:56 +0200
Subject: Newbie References Question
Message-ID: <3d91c72f$0$8558$79c14f64@nan-newsreader-02.noos.net>

Hi,

let say I have:
i1 = 1
i2 = 2
and
p = [i1,i2]

How can I have:
p=[*i1,*i2]
rather than
p=[**i1,**i2] as it is curently ?

( Sorry for this nasty hybrid syntax )

Textualy, how if I want that p[0] refere to the 'i1' container, and not to
the refererence it contains ?

What I want is that if now I set:
i1 = 4
that automaticaly:
p -> [4,2]
and not keeping [1,2]

Thanks in advance.

--
guy rabiller
3d animator / td
grabiller at 3dvf.net
http://grabiller.3dvf.net




From sholden at holdenweb.com  Sun Sep 29 06:34:26 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 29 Sep 2002 06:34:26 -0400
Subject: Pythonwin (win32all build 148) on Python 2.2.1 on Win2K, bogus syntax e
References: <u47k9.46401$V7.11091779@twister.socal.rr.com> <CJ8k9.16479$6g7.52569@news-server.bigpond.net.au> <sHJk9.66556$U7.23747920@twister.socal.rr.com> <a9Kk9.53648$V7.13380285@twister.socal.rr.com>
Message-ID: <8IAl9.5016$pD4.2297@fe05>

"A. Dimsdale and A. Moser" <fetus at AUTO> wrote ...
> A. Dimsdale Fetus A. Moser <fetus at AUTO> writes:
>
> >"Neil Hodgson" <nhodgson at bigpond.net.au> writes:
>
> >>Aaron Dimsdale and Alex Moser:
>
> >>> Pythonwin barfed on the quote marks marked by carets, returning:
"Failed
> >>to run script -
> >>> syntax error - invalid syntax".
>
> >>   Try turning on View | End-of-line markers to see if you have
non-standard
> >>line endings such as pure carriage returns, which cause just this sort
of
> >>error. On Windows, you want line endings to be carriage return + line
feed.
>
> >>   Neil
>
> [snip]
>
> Well, I went in and manually edited the EOLs on every code line in the
file.
> (Fun fun fun.) As it turns out, the cause of the problem is likely the
> difference between DOS-style and Unix-style EOLs. Alex and I jointly own
an
> account on my Linux box at home, where we keep the latest version of the
script.
> Alex does most of the coding on our script, and I just do other stuff like
> looking over the code when it slaps him, adding fixes and/or minor
features,
> etc. Well, since he does most of the coding, he had the script on his work
comp
> when we were at work on Tuesday, and I had him upload it to my Linux box
so I
> could grab it from my work box by FTP, and through the Win-to-Linux-to-Win
> transfer the EOLs apparently got funkified.
> Thanks again for the EOL tip.
>
For future reference, make sure you use ASCII mode for FTP transfers (of
text files, nach). This gives the receiving system the chance to modify the
line endings on the received file.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From bosahvremove at netscape.net  Sat Sep  7 01:15:49 2002
From: bosahvremove at netscape.net (Bo)
Date: Sat, 07 Sep 2002 05:15:49 GMT
Subject: Zope Status?
References: <Xns9281221A07167bosahvnetscapenet@66.185.95.104> <3d78a657$1@news.sentex.net>
Message-ID: <Xns9282D14CD9ECbosahvnetscapenet@66.185.95.104>

Peter Hansen <peter at engcorp.com> wrote in
news:3d78a657$1 at news.sentex.net: 

> Very likely, going to http://dev.zope.org would answer these
> questions. 
> 
> 

I looked there but it seems somewhat vague and out of date.

ie.

Project Plan

The Zope 2.6 release is tentatively planned for Q2 2002. Once the 
ProposedFeatures have been solidified into a final feature set, we'll put a 
table here with status and links to project areas for the major components 
of the project.




The site says they are gathering features for zope 2.6 and working on a 
vision statement for 2.7 which will use python 2.2.




I'm just curious if some people involved in the code would share their 
progress\hopes.


From whisper at oz.net  Thu Sep  5 22:15:56 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 5 Sep 2002 19:15:56 -0700
Subject: newbie problem: use socke lib to retrieve one web page:
In-Reply-To: <D613C81A8A3A93B8.5464F563C49215DF.E1058876104EFE5E@lp.airnews.net>
Message-ID: <GCEDKONBLEFPPADDJCOECEDKENAA.whisper@oz.net>

And for real convenience, there's websucker.py in the tools/webchecker
directory of the standard distro.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Cameron Laird
> Sent: Thursday, September 05, 2002 17:24
> To: python-list at python.org
> Subject: Re: newbie problem: use socke lib to retrieve one web page:
>
>
> In article <mailman.1031240588.2234.python-list at python.org>,
> Erik Price  <erikprice at mac.com> wrote:
> >
> >On Wednesday, September 4, 2002, at 11:53  PM, koko wrote:
> >
> >> I write this to retrieve one web page using socket lib.
> 			.
> 			.
> 			.
> >Works for me:
> >
> > >>> import socket
> > >>> host = 'www.uic.edu'
> > >>> port = 80
> > >>> s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
> > >>> s.connect((host, port))
> > >>> header = """HEAD /home/events.shtml HTTP/1.0
> >... From: hh at uic.edu
> >... User-Agent: test/1.0
> >...
> >... """
> > >>> s.send(header)
> >72
> > >>> data = s.recv(4096)
> > >>> print data
> >HTTP/1.1 200 OK
> >Date: Thu, 05 Sep 2002 15:40:00 GMT
> >Server: Apache/1.3.26 (Unix) PHP/4.1.2 mod_perl/1.27 mod_ssl/2.8.10
> >OpenSSL/0.9.6
> >Connection: close
> >Content-Type: text/html
> >
> >
> > >>> s.close()
> >
> >
> >I used the HEAD method instead of GET for brevity.  But GET works too.
> 			.
> 			.
> 			.
> To criticize code that's already giving satisfaction
> is nearly beyond me.  I'll point out, though, that
> the original questioner might consider this alterna-
> tive which has, I believe, evident advantages for
> long-term maintenance:
>   from urllib import urlopen
>
>   URL = "http://www.uic.edu"
>   page = urlopen(URL).read()
>   print page
> Note that urllib is one of the batteries the standard
> Python distribution includes.
>
> My summary:  use higher-order facilities when applicable.
> --
>
> Cameron Laird <Cameron at Lairds.com>
> Business:  http://www.Phaseit.net
> Personal:  http://starbase.neosoft.com/~claird/home.html
> --
> http://mail.python.org/mailman/listinfo/python-list




From kkuroda at crl.ucsd.edu  Mon Sep  2 20:47:16 2002
From: kkuroda at crl.ucsd.edu (Kow Kuroda)
Date: Mon, 2 Sep 2002 17:47:16 -0700
Subject: RedHat 7.3 and Python2.2.1
In-Reply-To: <3D73B16F.76B355DB@engcorp.com>
Message-ID: <B1D4556F-BED6-11D6-B4E4-00050287C0A6@crl.ucsd.edu>

On Monday, September 2, 2002, at 11:43 , Peter Hansen wrote:

> Doug Farrell wrote:
>>
>> I'm trying to install the latest version of Python
>> (python2-2.2.1-1.i386.rpm) from the www.python.org download page, and
>> I'm having problems. When I run
>>
>> rpm -i python2-2.2.1-1.i386.rpm
>>
>> I get all kinds of conflict messages like this:
>>
>> file /usr/bin/pydoc2 from install of python2-2.2.1-1 conflicts with
>> file from package python2-2.2-16
>>
>> probably hundreds of them in fact. Can anyone tell me what is going on
>> here, as the www.python.org site says the RPM I'm using is for RH7.3,
>> which is what I'm using. If you can tell me what I'm doing wrong and
>> how to address this, that would be great too! My ultimate goal is to
>> also install wxPython for the python2.2.1 installation, so if anyone
>> has some 'heads up' on that, I'd appreciate it.
>
> Sounds like you already have a previous version of Python2.2 installed.
> You should probably be using the "-U" option to update instead of
> the "-i" option to install... or uninstall the previous one first.

I just wanted to make a comment on Peter's post.

I think -U or --upgrade option works, but it's dangerious to uninstall 
Python 2.1, which is now part of Red Hat Linux 7.3, unlike in prevous 
releases where Python 1.5.2 or something is used for installation scripts.
  If you try to uninstall 2.1, you'll get a bunch of dependency errors and 
won't succeed. If you happen to be successful, it's pretty likely that you 
get messed up with the stability issues.

If .i386.rpm doesn't install even with -U, the best way to upgrade Python 
2 should be using a .src.rpm instead of an i386.rpm, as usuaul. It is 
explained in the rpm installation page of www.python.org.

Cheers,
Kow




From mennosimons at gmx.net  Tue Sep 24 15:01:44 2002
From: mennosimons at gmx.net (Ulli Stein)
Date: Tue, 24 Sep 2002 21:01:44 +0200
Subject: hex numbers??
References: <mailman.1032881550.14068.python-list@python.org>
Message-ID: <amqcqp$ffa$00$1@news.t-online.com>

jubafre at brturbo.com wrote:

>>>> y=hex(x)
>>>> y
> '0xa'
>>>> type(y)
> <type 'str'>
> 
> the hex() function put the hexadecimal number in a string type like this
> '0xa', but the real hexadecimal number of 10 is 'a', why python puts a
> '0x',  and how i can get just the real hexadedcimal number, for example i
> want to write in a file 10 in hexadecimal, but if i use hex(10) the result
> is '0xa' as a string, is not the same as 'a'.
> 
> '0xa' != 'a'
> 
> x = [12,13,14]
> y = ['0xc', '0xd', '0xe']
> 
> hex() create a list of string,  how i can write in a file just the real
> value of hexadecimal number?
> 
> 
> Juliano Freitas
> www.gebrasil.hpg.com.br

With your nice Python strings:

>>> "%x, %x, %x, %x"%(1,10,14,24234234)
'1, a, e, 171c8fa'

Have fun.


From jimb417 at NO*  Sat Sep 21 22:29:12 2002
From: jimb417 at NO* (Jim)
Date: Sat, 21 Sep 2002 20:29:12 -0600
Subject: PyQt setup on Linux Mandrake 8.2
References: <3D8B4898.20808@nospam.attbi.com> <ddc19db7.0209210259.7beb52d4@posting.google.com> <3D8BC2F4.8080107@nospam.attbi.com>
Message-ID: <amj9ud01bhp@enews3.newsguy.com>

dwelch wrote:


>>>( snipped )
> 
> OK, I found and installed tmake. Now, SIP can't find some of the Qt
> pieces it needs. So, I still don't understand the situation with Qt. Is
> it really possible to install and run my applications that I write with
> say Qt 3.0.5, when KDE on my system is using Qt 2.3.1? How does/would
> KDE know which version to use? How will my apps know what version to
> use? I am assuming there can only be one QTDIR active on a system at a
> time...
> 
> Thanks for the help,
> 
> Don

Hello Don,

You can write and compile apps using Qt 3.0.5 or whatever and still use KDE 
and Qt 2.3.1.

To compile apps with Qt 3.0.5, put this in your /etc/profile file: 

QTDIR=/usr/local/qt305 ( or wherever you put Qt )
PATH=$QTDIR/bin:$PATH
MANPATH=$QTDIR/doc/man:$MANPATH
LD_LIBRARY_PATH=$QTDIR/lib:$LD_LIBRARY_PATH

export QTDIR PATH MANPATH LD_LIBRARY_PATH

I've compiled several apps with 3.0.5 and they have all run on my
KDE desktop ( I'm still using 2.2.2! )

Hope this helps, Jim
-- 
Registered Linux User #269187
http://counter.li.org



From lee100 at sport.rr.com  Tue Sep 17 20:42:42 2002
From: lee100 at sport.rr.com (Lee Gray)
Date: Wed, 18 Sep 2002 00:42:42 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <Xns928CED692541cliechtigmxnet@62.2.16.82>
Message-ID: <6_Ph9.479711$q53.16315681@twister.austin.rr.com>

Chris Liechti wrote:

> lpgray at uop.com (Lee Gray) wrote in
> news:96904b50.0209171305.6d65f574 at posting.google.com:
>> My boss is pretty open-minded, but also needs to see something working
>> to be convinced Python is even a viable platform (he had never heard
>> of it).  Otherwise, .Net is a given, whether it's any good or not,
>> since corporate is MS all the way.
> 
> with python you have to choice of diffrent GUI libs. if you consider
> wxWindows you can play around with the wxPython demo application that
> comes with the download. on the other hand when you support win32 only the
> win32all package has support for all the windows APIs and it has some
> samples that get installed.
> 
> chris

Well, our shop is win32 only, but on a personal level I'd like to be more 
flexible (I mostly use Linux at home).  Isn't Tk still considered the 
"official" GUI lib?

For a newbie, what's easier, Tk or wxWindows?

Thanks,
Lee


From Oschler at earthlink.net  Sun Sep 29 12:57:32 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sun, 29 Sep 2002 16:57:32 GMT
Subject: Linker can't find library (PyProlog setup)
References: <mdGl9.14197$Lg2.4064162@news2.news.adelphia.net>
Message-ID: <0iGl9.14200$Lg2.4065036@news2.news.adelphia.net>

"Robert Oschler" <Oschler at earthlink.net> wrote in message
news:mdGl9.14197$Lg2.4064162 at news2.news.adelphia.net...
<snip>

Ugh, forgot to point out that I am running those commands from within a
script file.

thx





From edwardam at interlix.com  Sun Sep 22 14:18:42 2002
From: edwardam at interlix.com (Edward Muller)
Date: 22 Sep 2002 11:18:42 -0700
Subject: Python HTTPSConnection/HTTPResponse Problem
References: <9abedeb.0209212200.5c669a50@posting.google.com> <mailman.1032691649.13632.python-list@python.org>
Message-ID: <9abedeb.0209221018.75721059@posting.google.com>

It has ssl and I have to use version 2.1.3 because this is, in the end
being done using Zope, which doesn't work with anything newer AFAIK.

Gerhard H?ring <gerhard.haering at gmx.de> wrote in message news:<mailman.1032691649.13632.python-list at python.org>...
> * Edward Muller <edwardam at interlix.com> [2002-09-21 23:00 -0700]:
> > I am trying to use HTTPSConnection and HTTPResponse objects [...]
> 
> Is your Python SSL-enabled?  Check with:
> 
>     import socket
>     hasattr(socket, "ssl")
> 
> Which version of Python are you using? Be sure to use at least version
> 2.2, as it contains many SSL-related bugfixes.
> 
> -- Gerhard


From evan at 4-am.com  Sat Sep 28 21:02:17 2002
From: evan at 4-am.com (Evan)
Date: Sun, 29 Sep 2002 01:02:17 GMT
Subject: Best way to create a class dynamically on the fly?
References: <tFml9.13043$Lg2.3865279@news2.news.adelphia.net>
Message-ID: <3D96511D.2040407@4-am.com>

Robert Oschler wrote:
> What is the best way to create a Python class dynamically at runtime?

ALL Python classes are created dynamically at runtime.  The following is 
not a declaration in the sense that you're accustomed to from C/C++, it 
is normal code executed when the containing file is run (or imported):

class Foo:
   def meth1(self):
     return "I am a method"

Your class definition can include whatever Python code you need, and it 
doesn't have to be at global module level:

def makeFoo(x):
   class Foo:
     sharedValue = x * 2
     if x > 2:
       def manymethod(self, i):
         return i + 3
     else:
       def fewmethod(self, i):
         return i + 1
   return Foo

Foo = makeFoo(4)
f1 = Foo()
print f1.sharedValue, f1.manymethod(2)

Cheers,

Evan @ 4-am



From gerhard.haering at opus-gmbh.net  Thu Sep 12 02:52:37 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 12 Sep 2002 06:52:37 GMT
Subject: [ANN] PySQLite 0.3.0 released
Message-ID: <slrnao0e6p.23c.gerhard.haering@haering.opus-gmbh.net>

NAME:
    pysqlite - Extension module for SQLite databases.

DESCRIPTION:
    A Python-DB API 2.0 compliant extension module that connects to
    SQLite databases.

    SQLite is a powerful, embedded relational database in a compact C library.
    It supports a large subset of SQL92, multiple tables and indices,
    transactions, and triggers. It has a simple C/C++ interface requiring only
    three functions to perform queries. It has TCL bindings and an ODBC
    driver. Sources are uncopyrighted and can be used for any purpose. More
    information can be found at <http://www.hwaci.com/sw/sqlite/index.html>.

USAGE:
    import sqlite

    conn = sqlite.connect("db")
    cursor = conn.cursor()
    SQL = """ select category, family, genus, species
              from calflora order by genus, species limit 10"""

    cursor.execute(SQL)

    for col in cursor.description:
            print "\t %12s - %3s bytes" % (col[0], repr(col[3]))

    row = cursor.fetchone()

    while row != None:
            print "%14s, %15s, %19s, %8s, %25s" % tuple(row)
            row = cursor.fetchone()

    SQL = "insert into calflora (family,genus,species) values(%s,%s,%s)"
    cursor.execute(SQL, ('greenus', 'weedus', 'maximus')) 

    conn.close()

HOMEPAGE:
    <http://pysqlite.sourceforge.net>

DOWNLOAD:
    Source and Windows binary downloads are available at 
    <http://www.sourceforge.net/projects/pysqlite>

AUTHORS:
    Gerhard H?ring <gerhard.haering at gmx.de>
    Michael Owens <mike at mikesclutter.com>
    William Trenker <wtrenker at shaw.ca>

LICENSE:
    Python

NOTES:
PySQLite has been tested on FreeBSD, Linux, and Windows with both Python 2.1 
and 2.2. The latest versions of SQLite (v. 2.6 and greater) are recommended. 

Version 0.3.0 fixed a lot of bugs, while also adding new features. Users of
PySQLite are recommended to upgrade.

The following are some of the changes and enhancements since the last 
release:

- Squashed a few memory leaks.

- Exposed the sqlite_exec, sqlite_last_insert_rowid and
  sqlite_changes functions of SQLite to Python as methods of the
  connection object.

- Add support for Date types, if mxDateTime is
  available.

- Support for optional DB-API extensions from PEP 0249

- Added files for creating a PySQLite Debian package.

- setup.py: - Added Cygwin as platform that's supported by default

- Added third example: program to dump a table in XML format.

- Use bool type and custom bool converter for boolean fields.

- Set cursor.rowcount appropriately after DML (insert/update/delete) 
  statements.

- Fixed a bug with the SQL-quoting of longs.

- Inline documentation improvement.

- Change invocation of ReferenceError to work with Python 2.1

- (really) implemented weak references from cursors to connections

- Added new test suite.

- Adapted test for threadsafety attribute.

- Added checks with user-defined float and string functions.

- Removed Connection.begin(). Transactions are now started
  implicitely.  - Use DB-API quoting instead of manual Python
  quoting.  - Use string methods instead of the string module.

- Added checks for functions/aggreagates that raise exceptions or
  return None (NULL).

- Added tests for proper commit/rollback behaviour. Added tests
  for autocommit feature.

- Implemented autocommit feature. Renamed the parameter of
  Connection.__init__ from "filename" to "db". Commit and rollback
  now check if the connection is open at all. Started adapting the
  weak reference code from pyPgSQL to loosely couple cursors and
  connections. This is needed when you want to check if there are
  any open cursors for a given connection, for example.

- Simplified version numbers, removed the check for equal version
  numbers in _sqlite and sqlite. This looked like overhead to
  me. The version number scheme is now only: major.minor.micro Set
  threadsafety to level 1. I believe we can guarantee than sharing
  the module among threads will do no harm. Sharing the connection
  would.


From darin at naboo.to.org.no.spam.for.me  Mon Sep 30 10:49:54 2002
From: darin at naboo.to.org.no.spam.for.me (Darin McBride)
Date: Mon, 30 Sep 2002 14:49:54 GMT
Subject: idea for a much needed application
References: <576c1752.0209291326.56dcdf31@posting.google.com>
Message-ID: <mwZl9.443653$Ag2.18610813@news2.calgary.shaw.ca>

mike henley wrote:

> I wish i can just use my own computer, with whatever technologies I
> choose, keeping my content that i build up over time safe here on my
> own machine, but unless you have a broadband connection this is not
> possible.

You provide an interesting idea.  The major downfall to the request
comes right here.  This type of technology already has no future. 
Maybe when the web was initially born, this would have had some sort of
future where whoever invented a solution to your problem would be able
to succeed (money if commercial, fame if open source).  Now, however,
with broadband connections becoming almost ubiquitous in North America,
and gaining popularity in other parts of the world, there is little to
be gained for the programmer who comes up with a solution to this -
it's only a matter of time before always-on internet becomes a reality
for anyone who wants it, and there will be no need for any other
solution.

-- 
To reply, please remove the obvious spam filter


From jimb417 at NO*  Sat Sep  7 22:06:30 2002
From: jimb417 at NO* (Jim)
Date: Sat, 07 Sep 2002 20:06:30 -0600
Subject: PyQT picture display
References: <albp27030dq@enews4.newsguy.com> <3D799F42.2020903@SPAMnwinternet.com>
Message-ID: <alebe2021ft@enews4.newsguy.com>

Jim wrote:


> As far as I can tell, nothing's wrong. It works fine
> for me. Are you sure the jpg files actually exist and
> the path to them is correct? The only thing I changed
> in your program was the filenames for the image files.
> 
> I used PyQt-3.3.2 with Qt 2.3.0 on Linux.
> 
> (another) Jim

The jpg files are in the same directory as the program. I'm using Qt 3.0.5 
with PyQt 3.3.2 on Linux. I'm sure that JPEG support is compiled in the Qt 
library as I got this to work using C++. Does JPEG support need to be built 
in when compiling PyQt?  if so, what must I do to accomplish this? Thanks,

Jim 
-- 
Registered Linux User #269187
http://counter.li.org



From oren-py-l at hishome.net  Tue Sep 10 05:23:32 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Tue, 10 Sep 2002 05:23:32 -0400
Subject: dynamism
In-Reply-To: <aljpvd$5qk$1@bob.news.rcn.net>
References: <66715c02.0209091923.265fe77e@posting.google.com> <aljpvd$5qk$1@bob.news.rcn.net>
Message-ID: <20020910092332.GA19346@hishome.net>

On Mon, Sep 09, 2002 at 11:45:35PM -0400, Raymond Hettinger wrote:
> >>> class BetterAttrDict(dict):
>  def __getattr__(self, key):
>   return dict.__getitem__(self, key)
> 
> 
> >>> bad = BetterAttrDict({'a': 1, 'b':2})
> >>> bad.a
> 1
> >>> bad.b
> 2

class EvenBetterAttrDict(dict):
    def __init__(self, d={}, **kw):
        """ Initialize from a mapping or using keyword=value pairs """
        dict.update(self, d)
        dict.update(self, kw)

    def __getattr__(self, key):
        """ Return dictionary items as members """
        return dict.__getitem__(self, key)

    def __setattr__(self, key, value):
        """ Set dictionary items as members """
        try:
            dict.__getattribute__(self, key)
        except AttributeError:
            dict.__setitem__(self, key, value)
        else:
            dict.__setattr__(self, key, value)

    def __delattr__(self, key):
        """ Delete dictionary items as members """
        try:
            dict.__getattribute__(self, key)
        except AttributeError:
            dict.__delitem__(self, key)
        else:
            dict.__delattr__(self, key)

    def __repr__(self):
        """ Display as a constructor call with keyword arguments """
        items = dict.items(self)
        items.sort()
        return "%s(%s)" % (
            self.__class__.__name__,
            ", ".join(
                ["%s=%s" % (k,`v`) for k,v in items]
            )
        )

    def __members__(self):
        """ Support dir(), rlcompleter attribute completion """
        return dict.keys(self)

    __members__ = property(__members__)



>>> class person(EvenBetterAttrDict): pass
...
>>> p = person(name='Oren', born=1971)
>>> p
person(born=1971, name='Oren')
>>> p.likes ='Chocolate!'
>>> p
person(born=1971, likes='Chocolate!', name='Oren')
>>>

Overkill? perhaps.

	Oren




From jhorneman at pobox.com  Wed Sep  4 12:32:58 2002
From: jhorneman at pobox.com (Jurie Horneman)
Date: 4 Sep 2002 09:32:58 -0700
Subject: Which XML parser under Python can handle Unicode?
References: <9ac02e81.0209030214.515d6c0@posting.google.com> <j4vg5ngx42.fsf@informatik.hu-berlin.de>
Message-ID: <9ac02e81.0209040832.32e13bd6@posting.google.com>

Many thanks for your advice.

Jurie.


From spoermeg at voldelig.com  Fri Sep 20 15:59:06 2002
From: spoermeg at voldelig.com (Terje Johan Abrahamsen)
Date: Fri, 20 Sep 2002 14:59:06 -0500
Subject: newbie seeks advise.
References: <77c84db2.0209192137.51167ebb@posting.google.com> <mailman.1032501210.16162.python-list@python.org> <77c84db2.0209200425.4c723344@posting.google.com>
Message-ID: <amfulr$55okv$1@ID-53055.news.dfncis.de>

"akirasugiura" <kodokuchef at yahoo.com> wrote in message
news:77c84db2.0209200425.4c723344 at posting.google.com...
> Thank you very much. I didn't know there is such a "max" function.
> ("max"
>  is a function isn't it). I also tried "min" to find a smallest
> number(lol) and it worked. "list(data)" also worked. Thanks again.

As a fellow beginner, I will also advise you to check out the tutor list on
the python.org site. It is great when learning the language....




From bumm3rm4n at yahoo.com  Mon Sep 30 21:02:01 2002
From: bumm3rm4n at yahoo.com (bumm3rm4n at yahoo.com)
Date: 30 Sep 2002 18:02:01 -0700
Subject: Can python be used to debug other languages?
Message-ID: <4d1e3be8.0209301702.109d2b04@posting.google.com>

I am doing research in school and I was just wondering if python can
be used to debug other languages such as C++ and java. IF it does, can
anybody give me an info how?


From marklists at mceahern.com  Thu Sep 26 14:07:46 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 26 Sep 2002 13:07:46 -0500
Subject: How do you pronounce "__" (double underscore)?
In-Reply-To: <525eaba.0209260956.72bf6786@posting.google.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOMENFEBAA.marklists@mceahern.com>

http://groups.google.com/groups?selm=mailman.1024136779.31270.python-list%40
python.org





From loewis at informatik.hu-berlin.de  Tue Sep 17 16:42:50 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 17 Sep 2002 22:42:50 +0200
Subject: Instantiating and initialising C++ with embedded Python
References: <K0Mh9.2591$pr5.174881@newsfep1-win.server.ntli.net>
Message-ID: <j43cs8e4ad.fsf@informatik.hu-berlin.de>

Morfeas <greek_bill at yahoo.com> writes:

> 1. My main program is in C++, so I need to embed the python interpreter.

Correct.

> 2. I need to allow python access to my C++ classes, therefore I need to 
> extend the embedded interpreter.

Correct.

> a) a brief description of the whole process

You need to provide two things:

- modules that expose "core" functions of your API,
- Python "types" that encapsulate your C++ objects.

The Python interpreter has a global table of all available modules,
Modules/config.c:_PyImport_Inittab. This table lists entry points for
all "builtin" modules. You need to compose a table that has all the
"standard" builtin modules you want to use in your scripts, plus any
extension modules that you implement yourself.

In addition, to wrap C++ objects, you need to define "type objects",
which are structures that essentially contain function pointers to the
"virtual method tables" of a Python type; associated with such a type
is a C structure that defines the layout of actual instances of the
type.

> b) some code examples

See Modules/xxmodule.c for an example of an extension module; I
recommend to also read the "Embedding and Extending" tutorial.

> c) suggestions about useful tools

I'd recommend to write all the necessary code by hand. People will
suggest various tools, specifically:

- SWIG, which can generate wrappers around existing APIs,
- CXX, which exposes the Python API in a C++-ish way,
- Boost Python, to wrap C++ libraries.

HTH,
Martin


From andy at reportlab.com  Mon Sep  9 16:27:08 2002
From: andy at reportlab.com (Andy Robinson)
Date: Mon, 09 Sep 2002 20:27:08 GMT
Subject: jython memory usage
References: <hVm5ydAY2hd9Eww$@jessikat.demon.co.uk><3d76366a$1_5@news.bluewin.ch> <rnLbuKAv5jd9Ewzz@jessikat.demon.co.uk><3d76494d$1_5@news.bluewin.ch> <YR13DVAUNld9EwXH@jessikat.demon.co.uk><3d765ce6_4@news.bluewin.ch> <zzKd7KAqhnd9EwxT@jessikat.fsnet.co.uk><3d767c4a$1_1@news.bluewin.ch> <YWyHyIAJKyd9EwHA@jessikat.demon.co.uk><3d774f40$1_1@news.bluewin.ch> <RTiVzIAJD3d9Ew2t@jessikat.demon.co.uk> <3d7779b0$1_3@news.bluewin.ch>
Message-ID: <wu7f9.6927$DL2.50598960@news-text.cableinet.net>

> > [Loaded java.io.RandomAccessFile from C:\Program
> Files\Java\j2re1.4.0_01\lib\rt.jar]
> > [Loaded sun.reflect.GeneratedMethodAccessor20]
> >
FYI, ReportLab sat down to try our test suite under Jython
this week, and got out-of-memory errors when running
some fairly innocuous code with 1.4.0_01.
We then tried with some earlier JVMs and it was fine.
We sus[pect a bug in that JVM version.

Andy Robinson
ReportLab







From maxm at mxm.dk  Mon Sep  9 12:15:25 2002
From: maxm at mxm.dk (Max M)
Date: Mon, 09 Sep 2002 18:15:25 +0200
Subject: PythonWin bug [very weird]
References: <vU2f9.7489$yt3.3820112@newssrv26.news.prodigy.com>
Message-ID: <3D7CC91D.5000607@mxm.dk>

dsavitsk wrote:
> This is PythonWin 145 (patched for asp), Python 2.1.3, Win2k.  I have no
> means for testing other versions.
> 
> The bug is that pressing the wrong sequence of buttons can render a file
> unsaveable. To reproduce it:
> 1. open a .py file
> 2. make some changes
> 3. press the X to close the file (not to close PythonWin)
> 4. when asked to save (Yes, No, Cancel) press Cancel
> 5. Save the file (File | Save)
> 6. Any new edits to the file are unsaveable (the buton is grayed out). To do
> so, you must copy the entire contents, close the file, re-open it, and paste
> the clipboard.


That is almost as good as when you write a top latin-1 character like 
'?' and you try to delete it via backspace and it only seems to delete 
one of two bytes making it up, and you end up with '?'.

And sometimes if you try and save a file with one of those 'half' 
characters in it it simply deletes your source file :-/ So the only 
version left is the one in the open document, which you cannot save ...

regards Max M



From robin at execulink.com  Mon Sep 16 08:30:48 2002
From: robin at execulink.com (robin at execulink.com)
Date: Mon, 16 Sep 2002 08:30:48 -0400
Subject: threads or queue for this task
References: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com> <GmCg9.2354$Ee4.6690@news-server.bigpond.net.au> <cqp7ou8d0jjcdtrqetu9v9ls9gst2cp4qi@4ax.com> <mailman.1032076038.5654.python-list@python.org>
Message-ID: <h2jbouo4j9f8ee646bhfejt3erqku76d61@4ax.com>

"James J. Besemer" <jb at cascade-sys.com> wrote:

>On the other hand, you probably DO want to eliminate explicit sleeps 
>like this if you can.  E.g., the sleeps force a delay if data arrives to 
>be processed during the sleep.

That's what I was thinking.

>If you are clever, proper design of LookForWork() can automatically 
>suspend the threads when they have nothing to do.

I may sometimes be clever, and can no doubt cobble together code from
reading the docs (which I have been doing so far), but what I don't
know are some of the trade-offs of using certain constructs, or the
optimal way to do things. That is much more difficult to glean without
actual code examples.

For example, I really don't know what "proper design" I should use in
this case. Any sample code?

> E.g., that's one of 
>the great features of using a Queue -- threads are put to sleep if they 
>read an empty queue and they are woken up automatically when something 
>it put into the queue.  Similarly, if your thread is reading a Socket() 
>-- it'll sleep until data is available.  Then you don't need the 
>explicit sleeps.

The interface to Queue is deceptively simple. Are there constructs to
avoid or performance trade-offs to different methods? Again, any
explicit code (or at least explicit help as to which methods are
preferable) would be appreciated.

I am also put off by the description of the Queue Object, particularly
methods like empty() which says: "Return 1 if the queue is empty, 0
otherwise. Because of multithreading semantics, this is not reliable."

If it's not reliable, then what's the point of it existing? Why is it
not reliable? (I imagine simply because the thread doing the checking
could be superceded by the thread that implements the Queue in the
time it takes for the method to return and be acted upon.) Is it just
in certain cases that it's not reliable? How can one code if you don't
reliably know when a Queue is empty or full?

What this boils down to is: I'm sure I can do the job with the great
help I've already received, but do not have the confidence to know if
I'm doing the job the *right* way.

-- robin


From itamar at itamarst.org_NOSPAM  Sun Sep  1 17:05:07 2002
From: itamar at itamarst.org_NOSPAM (Itamar Shtull-Trauring)
Date: Sun, 01 Sep 2002 17:05:07 -0400
Subject: Dead Man Timer?
References: <fE6c9.145308$aA.31724@sccrnsc02> <pan.2002.08.26.17.48.00.759341@effectif.com> <cb035744.0209011247.81b42dc@posting.google.com>
Message-ID: <3D728103.5060308@itamarst.org_NOSPAM>

1. Use timeoutsocket.py - this lets you set timeouts on socket operations 
(find it on http://py.vaults.ca).

2. Use an event driven networking architecture - don't know of any that have 
telnet client support though.

Option 1 is your easiest solution.

-- 
Itamar Shtull-Trauring    http://itamarst.org/
Available for Python, Twisted, Zope and Java consulting



From gerhard.haering at opus-gmbh.net  Tue Sep  3 04:45:52 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 3 Sep 2002 08:45:52 GMT
Subject: Newbie's unsurmountable pb with IO stream trick
References: <aba30b75.0209020633.26ba4c70@posting.google.com> <3D738EE2.3A8756D@engcorp.com> <aba30b75.0209030009.3ccb0207@posting.google.com>
Message-ID: <slrnan8tk8.29s.gerhard.haering@haering.opus-gmbh.net>

"Spendius" wrote:
> Peter Hansen <peter at engcorp.com> wrote in message...
>> And what does not declaring variables have to do with the "global
>> keyword issue" (whatever that is)?
> I kept getting the error message # UnboundLocalError: local variable
> 'cpt' referenced before assignment where I really didn't expect to get
> such an error, and noticed in this forum a lot of guys who know Python
> complain and talk themselves about a "mess" (my 'cpt' var. is a counter
> incremented every time a method gets called and I must sometimes
> re-initialize it to 0 inside an 'if' statement: merely IMPOSSIBLE). 

def setToZero():
    global cpt
    cpt = 0

Merely impossible.

>> If you have an SQL database to access, why not use one of the
>> standard DBAPI modules to get to it?  You might be doing a lot
>> of extra work when you don't have to.
> And why not ? Anyway SQL*Plus contains plenty of built-in commands
> that allow you to format your output how you like:

Probably still easier to do with Python code.

> I'd like to keep that functionality. And let's say that I want to do that
> for the sole purpose of learning the handling of IO streams in Python!

The popen functions should be useful to interface SQL Plus. I just tried to
do so, but didn't succeed.

>> If you're just learning Python, you've taken on a fairly
>> complicated and large task right at the beginning.  Why not
> I admit it. But I did *exactly* the same with Java (writing the app I'm
> trying to port) but never got stuck into problems that blocking...

Having to use multithreading in Java to get the desired behaviour doesn't
sound too sexy for me.

-- Gerhard


From loewis at informatik.hu-berlin.de  Wed Sep  4 14:40:58 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 04 Sep 2002 20:40:58 +0200
Subject: Memory fault - core dumped   while compiling 2.2.1 on SCO
References: <3D757EFA.9050401@nospam.yahoo.com> <yu99sn0qqw1r.fsf@europa.research.att.com> <m3bs7etem9.fsf@mira.informatik.hu-berlin.de> <3D765A05.6030004@nospam.yahoo.com>
Message-ID: <j48z2h61j9.fsf@informatik.hu-berlin.de>

Anton Graph <aglyportREMOvethispart at nospam.yahoo.com> writes:

> Did not even take nearly as long for it to coredump. With optionless
> configure build was going at least for a half an hour and this one
> bailed out pretty much right away...

Can you please try to run the resulting Python binary in a debugger
(say, dbx), and report a backtrace?

Thanks,
Martin


From gerhard.haering at opus-gmbh.net  Mon Sep 16 09:27:46 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 16 Sep 2002 13:27:46 GMT
Subject: "if x == None" raises "'NoneType' object is not callable"
References: <af84f376.0209160501.4d21500@posting.google.com>
Message-ID: <slrnaobmpm.27k.gerhard.haering@haering.opus-gmbh.net>

Luc Saffre wrote:
> row = Row()
> if row == None: # here it happens.
>    print "row instance is None!"

To avoid this and similar problems in the future, test for None using the is
operator, like:

if row is None:
    print "row instance is None!"

This way, row doesn't need to support comparison to None. The same holds true
for all other singletons you have in your app. As an additional plus,
'is'-comparison is faster.

-- Gerhard


From wlfraed at ix.netcom.com  Wed Sep 25 20:16:49 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Wed, 25 Sep 2002 17:16:49 -0700
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com> <3D8E6353.7030303@something.invalid> <amsfnk$cq1$2@nyheter.chalmers.se>
Message-ID: <hljtma.4f3.ln@ix.netcom.com>

Jacob Hallen fed this fish to the penguins on Wednesday 25 September 
2002 07:03 am:


> 
> Exactly the point! My code is my work of art. Why should anybody be
> allowed to change it? We don't allow people to go painting moustaches
> on the Mona-Lisa, do we?
>
        Ah, but if you can persuade the Louvre to /sell/ you the Mona Lisa, 
you are then perfectly free to do what you want with it.

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From Padraig at Linux.ie  Mon Sep 16 05:54:48 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Mon, 16 Sep 2002 09:54:48 GMT
Subject: python performance
References: <3D84E66F.6020701@linux.ie> <zm7h9.51182$AR1.2015178@bin2.nnrp.aus1.giganews.com>
Message-ID: <IThh9.4238$cP3.10401@news.iol.ie>

Terry Reedy wrote:
> "Padraig Brady" <padraig at linux.ie> wrote in message
> news:3D84E66F.6020701 at linux.ie...
> 
>>I was wondering about the performance characteristics
>>of python and ran a simple test. The 2 programs
>>below are functionally equivalent and just read
>>in fields into a list. file.fields contains
>>720 fields in each of 405 lines of the form of
>>repeating: oneoneone twotwo "thre e three"
>>
>>The time to run is shown above each program
>>from which I've inferred the following:
>>
>>1. The function call version is (6.3%) faster because
>>    the cumulative cost of parsing the simpler expressions
>>    and function call overhead is smaller than parsing the
>>    1 single complex expression?
> 
> 
> Calling functions adds overhead.  Running code within functions
> reduces overhead for object access.

Ah OK, this because it has to search a smaller namespace?
Wouldn't compiling overcome this? I suppose python is too
dynamic.

> See end for third version to
> test.

results below.
thanks!
P?draig.

>>Or the function is parsed
>>    only once and doesn't have to be reparsed. This would
>>    suggest that top level code is parsed for each iteration?
> 
> 
> No.  All code is parsed only once.  See above and below.
> 
> 
>>2. Anyway I thought that parsing affects would be removed by doing
> 
> the
> 
>>    parsing only once, i.e. compiling the code to .pyc (I used
>>    py_compile.compile()). However this makes no difference at all?
>>    Surely compiling is not just for code obfuscation.
> 
> 
> No.  The one-time parsing of 10 lines of code is extremely fast.  If
> it were parsed over again for each line, then you would notice the
> parsing time.
> 
> 
> 
>>Note I did do the test several times and averaged the results.
>>----------------------------
>>2.514s
>>----------------------------
>>#!/usr/bin/env python2.2
>>import re
>>
>>reFieldFinder = re.compile('[^ "]+|"[^"]+"') #unquoted|quoted
>>def getFields(line):
>>     fields = reFieldFinder.findall(line)
>>     return [field.replace('"', '') for field in fields]
>>
>>for line in open("file.fields").readlines():
>>     listLine = getFields(line[:-1])
>>
>>----------------------------
>>2.672s
>>----------------------------
>>#!/usr/bin/env python2.2
>>import re
>>
>>reFieldFinder = re.compile('[^ "]+|"[^"]+"') #unquoted|quoted
>>for line in open("file.fields").readlines():
>>     listLine = [field.replace('"', '') for field in
>>reFieldFinder.findall(line[:-1])]
> 
> 
> Try one more test to run at function speed without repeated
> getsfields():
> 

----------------------------
2.498s
----------------------------
> def mytest:
>     reFieldFinder = re.compile('[^ "]+|"[^"]+"') #unquoted|quoted
>     for line in open("file.fields").readlines():
>         listLine = [field.replace('"', '') for field in
>     reFieldFinder.findall(line[:-1])]
> 
> mytest()



From jhalttun at pp.htv.fi  Thu Sep  5 08:43:19 2002
From: jhalttun at pp.htv.fi (janne halttunen)
Date: Thu, 05 Sep 2002 15:43:19 +0300
Subject: How to get os.popen subprocess PID?
Message-ID: <3D775167.6060307@pp.htv.fi>

Hi all,

I have a subprocess opened with os.popen, producing undefined length of 
data.  I want to read maybe some of this data and then get rid of the 
subprocess, not waiting until the end of data.  I have thought of 
os.kill-ing the process.  So far I haven't been able to determine how to 
get the pid of the subprocess, required by os.kill .. Any pointers?


janne



From see_reply_address at something.invalid  Tue Sep  3 21:59:11 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Wed, 04 Sep 2002 13:59:11 +1200
Subject: ANN: Pyrex 0.4.4.1
Message-ID: <mailman.1031253706.7158.clpa-moderators@python.org>

It has been brought to my attention (several times!) that
Plex was missing from the 0.4.4 release. This one fixes
that.

What is Pyrex?
--------------

Pyrex is a new language for writing Python extension modules.
It lets you freely mix operations on Python and C data, with
all Python reference counting and error checking handled
automatically.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg




From dougz at cs.washington.edu  Tue Sep 17 16:05:02 2002
From: dougz at cs.washington.edu (Douglas Zongker)
Date: Tue, 17 Sep 2002 20:05:02 GMT
Subject: subclassing tuple
Message-ID: <H2LMGE.1JBF@beaver.cs.washington.edu>

Is there any way to create a subclass of 'tuple' that has a customized
constructor?  I tried doing the obvious thing, but it doesn't work --
it seems the tuple initializer gets called *before* the one for my
subclass.  This:

    class Foo(tuple):
        def __init__( self, a ):
            print 'self is already:', self
            print 'argument is', repr(a)

    x = Foo('hello')

produces:

    value is already set: ('h', 'e', 'l', 'l', 'o')
    argument is 'hello'

I could do this with lists instead, but that means changing a lot of
existing code that expects tuples (and I don't want these objects to
be mutable!)

Suggestions welcome.

thanks,
dz


From tjlavoie at mb.sympatico.ca  Tue Sep 24 00:26:03 2002
From: tjlavoie at mb.sympatico.ca (Tim Lavoie)
Date: 23 Sep 2002 23:26:03 -0500
Subject: Outlook 2002 and Python
References: <mailman.1032839910.12131.python-list@python.org>
Message-ID: <87ptv4f1yc.fsf@theasylum.dyndns.org>

Well, two things. 

First, please do NOT post to newsgroups or mailing lists in HTML,
particularly that emitted by Microsoft tools. Even if readable in
Outlook, they're horrid to read otherwise, and will often be ignored
by those who might help. I'd repeat it to show you what it looks like
to everyone else, but others have suffered enough.

Second, what do you want your plugin to do? Something useful on the
client side I would guess, but you never said. Regardless of the
language used, deciding what you would like to accomplish is always an
excellent first step.

-- 
I finally went to the eye doctor.  I got contacts.  I only need them to
read, so I got flip-ups.
                -- Steven Wright


From DGustafson at attbi.com  Fri Sep 20 14:53:28 2002
From: DGustafson at attbi.com (Dave Gustafson)
Date: Fri, 20 Sep 2002 18:53:28 GMT
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com> <t2Di9.336220$kp.1062154@rwcrnsc52.ops.asp.att.net> <7xznuc2a4r.fsf@ruckus.brouhaha.com>
Message-ID: <I8Ki9.157253$Jo.48767@rwcrnsc53>

> That sounds like either a "port" or a re-interpretation of the classic
> "The Little Lisper".

Yes, "The Little Schemer" is indeed a re-casting of "The Little Lisper"
which I believe was first published in 1974...  This is the 4th edition, 6th
printing.  Fun stuff, and certainly a decent way to get comfortable with
some of the cool things that can be done in Python.

> You should also of course have read "Structure And Interpretation of
> Computer Programs" by Hal Abelson and Gerry Sussman:
>
>   http://www-mitpress.mit.edu/sicp/
>
> The complete text is on that website now!

WOW and again backwards WOW
the "purple book" is on-line!
hadn't noticed that before...






From robin at jessikat.fsnet.co.uk  Sun Sep  1 05:23:23 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 1 Sep 2002 10:23:23 +0100
Subject: [Q] win32com/Excel
References: <lbzsn0uw5cr.fsf@aqualene.uio.no>
Message-ID: <pZj7hsALydc9Ew59@jessikat.fsnet.co.uk>

In article <lbzsn0uw5cr.fsf at aqualene.uio.no>, Rolf Marvin B?e Lindgren
<roffe at aqualene.uio.no> writes
>I would like to emulate this Visual Basic statement from Python, using
>Com:
>
>        Range(Range("A1"), ActiveCell.SpecialCells(xlLastCell)).Select
>
>this statement selects the active worksheet (which is, incidentally,
>not similar to UsedRange).  
>
>any clues?  
>
>thank you,
from win32com.client import Dispatch
from win32com.client import constants
xl = Dispatch("Excel.Application")
AB = xl.ActiveWorkbook
AS = AB.ActiveSheet
AW = xl.ActiveWindow
R = AS.Range('A1',AW.ActiveCell.SpecialCells(constants.xlLastCell))
print AB.Name, AS.Name, R
-- 
Robin Becker


From gupsmail at ihatespam.shaw.ca  Wed Sep 25 10:51:48 2002
From: gupsmail at ihatespam.shaw.ca (Grant Hallam)
Date: Wed, 25 Sep 2002 14:51:48 GMT
Subject: problem with dictionaries within lists
References: <pan.2002.09.25.05.21.07.477231.167@ihatespam.shaw.ca> <S4fk9.163219$pX1.5840444@news2.tin.it>
Message-ID: <pan.2002.09.25.14.49.37.25889.181@ihatespam.shaw.ca>

I reset the dict keys to blank strings and it works now.  
added the following after webdatabaselist(dict.append)
dict = {'Name': '','Location':'','Transit':'','Position':'','Technical_Knowledge':'','Like_Site':'','Resolve_Tech_Issues':'','Easy_Navigation':'','Publish_Comments':'','Answered_No':'','Find_Support':'','Find_Support_Other':'','Find_Information':'','Which_LEOnet_Site':'','Find_Information_Other':'','Visit_Again':'','Visited_Old_Site':'','Time_Spent':'','Comments':'','B1':''}
and it works ok now.  Seems like a bit of a hack, have to think of a
better way later.

On Wed, 25 Sep 2002 02:11:14 -0700, Alex Martelli wrote:

> Grant Hallam wrote:
>         ...
>>  However, the list seems to get populated with the same dictionary
>         ...
>> webdatabaselist.append(dict)
> 
> Of course -- you ARE appending the same dict over ands over (AFAICT --
> you're probably using tabs for indentation, making your code hard to
> comprehend for those who read news with KDE's KNode [or MS's OE] as tabs
> don't show there... use spaces, NOT tabs...).
> 
> Anyway, you probably need something like:
> 
> webdatabaselist.append(dict.copy())
> dict = {}
> 
> either might suffice, but generally you'll need to copy the current dict
> AND make dict empty for the next leg of the loop, if I get the general
> drift.  If you do rebind name dict to {} for the next leg, then you
> don't need the call to the copy method.
> 
> 
> Alex


From propadovic.nenad at debitel.net  Wed Sep 18 20:00:56 2002
From: propadovic.nenad at debitel.net (Nenad Propadovic)
Date: Thu, 19 Sep 2002 02:00:56 +0200
Subject: Need advice on choosing the technology for intranet based application 
References: <3d88fd1e@news.ivm.net> <Gn7i9.1769$ik2.430@fe02>
Message-ID: <3d89121f@news.ivm.net>

Another answer. WOW!
> ----- Original Message -----
> From: "Steve Holden" <sholden at holdenweb.com>
> Well, assume in that case that you should write for a CGI environment:
> it's
> maybe not the absolute fastest way to go, but it's pretty portable, and so
> you'll be able to make plans ahead of that particular decision being made
> (or do you get to make it?), and you can accommodate almost any server in
> this way.
Whenever I read "CGI" I also read "a pain to maintain" some words later... ?
It's a project with a (modest) possibility to turn into a product, so both
portability and maintanability are issues.
>
> The Python Image Library (PIL) is very good for creating graphics. You can
> create them "on the fly" in your web server and then serve back HTML
> documents with references to the created graphics.
That sounds good, too. I seem to have two solutions for that part allready
:-)
>
> Seems to me that CGI under Apache, while probably not the most exciting
> cxhoice, should handle your requirements quite well.
On the other hand, Mark also meant CGI/Apache, writting Python/Apache,
right?
>
> The Python DB API means you can choose from a wide range of relational
> database back-ends.

Fine.

> Otherwise you get involved with writing Java applets you can send
> the data to, and things can get ugly fast ;-)
No thanks :)
>
> A pleasure. There's a book called "Python Web Programming" by some chap
> called Holden that might give you some useful background ...
>
Oh, if I choose CGI you can bet I'll byu this guy Holden's book :)
>
> > P.S. I'm just introducing Python to my group at BMW (a completely
different
> > project, though), and I'll let you know if there are some bigger success
stories.
> > >
> Let us know what we can do to help!
Not much, really. I was not suposed to write Python, I just started doing
it. I guess at some point they'll just notice half of the SW is in Python,
anyway, so it will become a de facto standard. It's a creeping invasion :) .
Oh by the way, I'm just about to go offline, it's 1:59 in the morning on my
computer watch.
Thanks again, folks.




From tismer at tismer.com  Sun Sep 15 12:22:23 2002
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 15 Sep 2002 18:22:23 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com>
Message-ID: <3D84B3BF.1010309@tismer.com>

Mark McEahern wrote:
> [Robert Oschler]
> 
>>[...] Do you think that it [stackless] eventually
>>will be merged with the main Python code base?
> 
> 
> Note which category PEP 219 is in:
> 
>   http://www.python.org/peps/

This is true (deferred/abandoned), but the PEP was
written for the old, continuation based Stackless
implementation.
The new one is completely different and very easy
to merge with Python. There exists no current PEP
that deals with that.

The changes to Python are also so small now, that
it doesn't matter if it will be merged or not.
Porting Stackless to a new Python is a matter of a few
hours. You may rely on getting a Stackless version
of any new Python version in almost no time.

ciao - chris
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From Oschler at earthlink.net  Wed Sep 25 17:16:10 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Wed, 25 Sep 2002 21:16:10 GMT
Subject: Can't suspend a thread?
References: <QZ4j9.10630$Lg2.2393215@news2.news.adelphia.net> <amognf$r6g$1@panix1.panix.com> <juj2puko63pmq5k4j4tm2eonoadvpup50d@4ax.com>
Message-ID: <uIpk9.11787$Lg2.3232725@news2.news.adelphia.net>

"Tim Roberts" <timr at probo.com> wrote in message
news:juj2puko63pmq5k4j4tm2eonoadvpup50d at 4ax.com...
> aahz at pythoncraft.com (Aahz) wrote:
> >
>
> It is worth pointing out that this "restriction" is not a Python thing; it
> is an operating system thing.
> --
> - Tim Roberts, timr at probo.com
>   Providenza & Boekelheide, Inc.

You mean a Linux/Python restriction?  Windows can suspend threads so I
assume that's what you mean.  Note I'm not saying that you can't suspend
threads on Linux (POSIX), I don't know if that's possible or not, I just
know you can on Windows.

thx





From magnus at thinkware.se  Tue Sep 17 14:56:38 2002
From: magnus at thinkware.se (=?ISO-8859-1?Q?Magnus_Lyck=E5?=)
Date: Tue, 17 Sep 2002 20:56:38 +0200
Subject: Beginners Question - use of variables
References: <1031345108.547.165.camel@martin.kleinerdrache.org> <mailman.1031346013.14271.python-list@python.org>
Message-ID: <3D877AE6.40701@thinkware.se>

> 
> my_list = [
>            ['sean', 'perry', 25, '20020906'],
>            ['Martin', 'Klaffenboeck', 20, '20020826'],
>           ]
> 
> for entry in my_list:
>     print entry[0], entry[2] # prints first string and the int

A more Pythonish solution is to use a list of tuples,
rather than a list of lists. Tuples are a better choice
for data structures that look like records or structs.
A tuple is not mutable, and it is more lightweight. I.e:

my_list = [
            ('sean', 'perry', 25, '20020906'),
            ('Martin', 'Klaffenboeck', 20, '20020826'),
           ]

for entry in my_list:
     print entry[0], entry[2] # prints first string and the int

The most memory compact storage type (in case you have very big lists)
is probably achieved by using the struct module. See the library
reference, section 4.3.



From Justin9688 at yahoo.com  Mon Sep  9 14:30:36 2002
From: Justin9688 at yahoo.com (Justin9688)
Date: Mon,  9 Sep 2002 13:30:36 -0500 (PET)
Subject: Ciencia y Tecnolog
Message-ID: <20020909183036.0BEC99ACD1@smtp2.terra.com.pe>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020909/d032feb0/attachment.html>
-------------- next part --------------

-- Antivirus de Terra ha detectado un grave virus en "este mensaje". Con la finalidad de evitar su contagio, hemos procedido a borrar el virus --

ROWS.pif is removed from here because it contains a virus.

Si tiene alguna consulta o comentario escr?banos a antivirus at terra.com.pe - Terra Networks Peru S.A. -
-------------- next part --------------
A non-text attachment was scrubbed...
Name: socinfo[1].jpg
Type: application/octet-stream
Size: 3884 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020909/d032feb0/attachment.obj>

From yka at mbnet.fi.invalid  Sat Sep  7 16:46:22 2002
From: yka at mbnet.fi.invalid (yka)
Date: Sat, 07 Sep 2002 20:46:22 GMT
Subject: Larry Wall's comment on python...
References: <mailman.1031377930.4225.python-list@python.org>
Message-ID: <1549197.0PZNOebDTH@cartman>

Kow Kuroda wrote:


> 
> Well, Python code may look awkward --- if not awk-like ---
> if you don't know how to customize the tab width in your
> editor of choice. For example, vi and emacs indent by 8
> characters by default, which I suspect is too wide for Python.
> Some code with deep nesting wouldn't look very nice with
> this default.
> 
> Kow

Well at vim :set ts=4 makes more fitting
-- 
Haloo?


From from_deja at alandaniels.com  Fri Sep 20 13:18:50 2002
From: from_deja at alandaniels.com (Alan Daniels)
Date: 20 Sep 2002 10:18:50 -0700
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <bd751532.0209082142.54236307@posting.google.com> <allrrs$6hm$2@solaris.cc.vt.edu> <ambpgq$dre$1@nntp9.atl.mindspring.net>
Message-ID: <a6f9355c.0209200918.5f252c64@posting.google.com>

"Matt Gerrans" <mgerrans at mindspring.com> wrote in message news:<ambpgq$dre$1 at nntp9.atl.mindspring.net>...

> I think "the pound-end operator" is the technical term (I'll have to defer
> to Guido on that).

"Pound-end operator"? Now *there's* an image...


From see_reply_address at something.invalid  Sun Sep 22 20:49:19 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Mon, 23 Sep 2002 12:49:19 +1200
Subject: Destructor never called ???
References: <mailman.1032506132.22396.python-list@python.org> <amfqoq$5iect$1@ID-59885.news.dfncis.de>
Message-ID: <3D8E650F.3030400@something.invalid>

Thomas Heller wrote:

> So what are the recommended practices if my program creates datastructures
> containing cyclic references?
> 
> - avoid using __del__ in these objects by  cleaning them 'manually'?
> - only using weak references to break these cycles?
> 
> - and finally using the debugflags of the gc module to ensure everything
> is cleaned correctly?


If you really must have __del__ methods in cyclic
data structures, you could examine the uncollectable-
garbage list periodically and do something to break
up the cycles. Once you've manually broken the
cycle, normal refcounting will take care of the
rest and the __del__s will be called.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From wlfraed at ix.netcom.com  Sun Sep 29 19:08:32 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Sun, 29 Sep 2002 16:08:32 -0700
Subject: What does Python fix?
References: <e15b6fa3.0209271418.14cd5642@posting.google.com> <87ptuxub5x.fsf@pokey.henrik-motakef.de> <vcmcpusnjqvpj9lavtqlqa7cs01h4i1iij@4ax.com> <87znu1rugt.fsf@pokey.henrik-motakef.de>
Message-ID: <g518na.knj.ln@ix.netcom.com>

Henrik Motakef fed this fish to the penguins on Sunday 29 September 
2002 02:55 am:

> Courageous <jkraska at san.rr.com> writes:
> 
>> >Let's not forget that Python's syntax (most of all the significance
>> >of whitespace) is considered strange at least by about all novices.
>> >In both cases - significant whitespace and lots of irritating
>> >superflous parentheses - people don't just get used to it, but tend
>> >to like it after a while.
>> 
>> I'm fairly sure that you meant to say "Lisp" where you said Python,
>> right?
> 
> No, I'm talking about how newbies tend to react when you tell them
> that block structure in Python is determined by indentation.

        I suspect the confusion was the phrase "significant whitespace and 
lots of irritating superflous parentheses" -- keyword being AND

        Change that to OR: ... significant whitespace (python) OR lots of 
irritating superfluous parentheses (lisp)


 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From bokr at oz.net  Fri Sep 27 12:22:22 2002
From: bokr at oz.net (Bengt Richter)
Date: 27 Sep 2002 16:22:22 GMT
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <amvcev$cf6$1@s1.read.news.oleane.net> <7nHk9.172208$pX1.6117207@news2.tin.it> <ofLk9.304267$z91.13145351@bin3.nnrp.aus1.giganews.com> <dITk9.175519$pX1.6209886@news2.tin.it>
Message-ID: <an20ju$bdk$0@216.39.172.122>

On Fri, 27 Sep 2002 07:23:53 GMT, Alex Martelli <aleax at aleax.it> wrote:
[...]
>Terry Reedy wrote:
>
>> 
>> "Alex Martelli" <aleax at aleax.it> wrote in message
>> news:7nHk9.172208$pX1.6117207 at news2.tin.it...
>>> The reason
>>>         zz = a and b or c
>>> is not the same thing as
>>>         zz = a ? b : c
>>> is that, when a is true and b is false, the former returns c,
>>> while the latter returns b.
>> 
>> In almost all the practical cases where I have thought about using
>> this,  b is *not* false or cannot become false, in which important and
>> real case the two expressions *are* equivalent.
>
>A typical, frequent, practical case is classification, e.g.:
>
>"Append bleep to list truebleeps when bleep is true, append it
>to list falsebleeps instead when false".
>
>If we had a ternary operator, a natural way of expression would be:
>
>    (bleep?truebleeps:falsebleeps).append(bleep)
>
>Of course this would be a disaster with the and/or idiom instead:
>
>    (bleep and truebleeps or falsebleeps).append(bleep)
>
>As list truebleeps starts out empty, thus false, this buggy code
>would NEVER append anything to it.
>
>In this case, like in most others (but far from all!) we don't
>need short-circuiting, so we can perfectly well code:
>
>    (truebleeps, falsebleeps)[not bleep].append(bleep)
>
>This is now correct -- but still nowhere as clear as good old:
It's correct given that truebleeps and falsebleeps are simple references
with no possible side effects from use as above, but I think it's
important to note that the side effect restriction makes it
not semantically identical to
>
>    if bleep: truebleeps.append(bleep)
>    else: falsebleeps.append(bleep)
>
when there is more to it than simple references.
>which to me remains the canonical way to code this in Python.
>Yes, it IS fractionally more verbose, but, so what?  It's
>immediately obvious, simple, and not clever.
>
E.g., if truebleeps and falsebleeps were properties,

    if bleep: objectHavingProperty.truebleeps.append(bleep)
    else: objectHavingProperty.falsebleeps.append(bleep)

would not trigger potential side effects of respective property getter functions, whereas

    (objectHavingProperties.truebleeps,
     objectHavingProperties.falsebleeps)[not bleep].append(bleep)

would. I think it worth repeating that short-circuiting was (ISTM) invented to
allow expressions where the terms preceding others in left-right order
**prevent evaluation** of succeeding terms as soon as they cannot affect the total
boolean value, so we can e.g. avoid exceptions in writing things like

    if i>0 and i<len(aList) and aList[i] is not None: foo(aList[i])
>
The ternary expression also has the feature of preventing evaluation of the
non-selected term. Someone else brought this aspect up before, but I think
it's important enough to repeat.

>> I think this is an elegant hack, and so I shared it in response to the
>> OP.  Others think it ugly.  So be it.
>
>It's _dangerous_ -- to use it, one always has to be fully aware
>of the risk that subexpression b might be false.  I think this
>is a pitfall just waiting for the unwary to stumble.  Not being
>an aesthete, I consider that more important than issues of beauty.
>
I'd like pitfall-free beauty ;-)

Regards,
Bengt Richter


From cliechti at gmx.net  Fri Sep 27 14:48:10 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 27 Sep 2002 20:48:10 +0200
Subject: watching mutables?
References: <mailman.1033137396.14482.python-list@python.org> <an211j$oi$1@news.hccnet.nl> <Xns9296C58C9ADFCcliechtigmxnet@62.2.16.82> <k7l7qpvf.fsf@ion-tof.com>
Message-ID: <Xns9296D3AB53777cliechtigmxnet@62.2.16.82>

Thomas Heller <thomas.heller at ion-tof.com> wrote in
news:k7l7qpvf.fsf at ion-tof.com: 
> Chris Liechti <cliechti at gmx.net> writes:
[triggered by *any* change of object m]
>> see EventObj:
>> http://oomadness.tuxfamily.org/p-pyobjtools.php
>> 
> This is what I get when clicking on the download link:
> <title>404 Not Found</title>

works for me.

chris

-- 
Chris <cliechti at gmx.net>



From ralf_ahlbrink at web.de  Sun Sep 22 14:55:58 2002
From: ralf_ahlbrink at web.de (RA)
Date: Sun, 22 Sep 2002 20:55:58 +0200
Subject: PyQt setup on Linux Mandrake 8.2
References: <3D8B4898.20808@nospam.attbi.com> <ddc19db7.0209210259.7beb52d4@posting.google.com> <3d8c9b92$0$182$9b622d9e@news.freenet.de> <3D8CA781.3010000@nospam.attbi.com>
Message-ID: <3d8e123e$0$178$9b622d9e@news.freenet.de>

On Sat, 21 Sep 2002 17:08:17 +0000, djw wrote:

> <snip>
>> Unfortunately, Mandrake is somewhat more perl than python biased.
>> 
> 
> Do you mean there is more support/newer versions of Perl or is it more 
> of a philosophical thing: There are many ways vs. there is one way?

I don't know if it is a philosophic issue. But nearly all configuration 
software is perl. And they use gtk as gui for those apps. So, don't expect
a rapidly updated PyQt.

> 
> <snip>
>> If you know a bit about RPM, perhaps you, dwelch, can manage
>> to update these packages (rpm -Uhv xx.src.rpm, a few changes ..., 
>> rpm -ba SPECS/sip.spec).
>> Otherwise you could contact me. In about a week I've got time
>> to do this.
> 
> Uhhh... considering my extreme novice status on Linux and especially 
> RPMs (as witnessed by my problems), I don't think I am the right person 
> for this task.
>

Just an excerpt from sip.spec:

%build
mkdir -p $RPM_BUILD_ROOT%{_libdir}/python2.2/site-packages
%{__python} build.py -q %{_libdir}/qt3 -b $RPM_BUILD_ROOT%{_bindir} \
    -d $RPM_BUILD_ROOT%{_libdir}/python2.2/site-packages
%make
cd siplib
%make staticlib


If you don't want to build a RPM, you can nevertheless use the info
from the .spec for your question about the qt issue. So try something like
"python build.py -q /usr/lib/qt3"

> <snip>
>> The RPM  stuff (or DEB) has got real advantages compared to 
>> "setup.exe". 
>> And if the offered versions are outdated, fetch the
>> new source, learn from the (e.g.) sip.spec file somewhat about
>> the compilation, and make a new package!
>> 
> 
> Well, here I have to disagree (at least from a user's perspective). 
> Having come from over 10yrs of Windows development experience (including 
> writing Windows installers from scratch), I find RPM package management 
> (can't say about DEB format, never used it) to be very complex, 
> difficult, and it rarely works for me.

I haven't got any experience with building Windows installers. So, I can't
compare both procedures. And I don't know, how you felt at the beginning
of your Windows installer exp.
I meant the _end_ users comfort. Perhaps you know, what files get
installed and which registry entries were set. But I believe (not sure)
that you have to use some third party "auditing" software. With rpm or deb
you've got a database for installed files. With linux you often use
more than one version of a library. The packages should "know" which
version they need. If you want to use another version (library or main 
package) you or someone else has to rebuild the package.

Now I will stop with this (somewhat) offtopic stuff.

> 
> Case in point is two systems I tried to install at work... one Mandrake 
> 8.x and one Windows 2000. The packages I was installing were Python 
> 2.2.1 with a bunch of extensions like MySQLdb and mxDateTime, MySQL 4.0, 
> Jabber x.xx (can't remember version) and an FTP daemon. The Windows box 
> was up and running in less than an hour after running a handful of 
> setup.exe-type programs, and the Linux box still wasn't fully installed 
> after two days of trying. I completely gave up trying to upgrade a copy 
> of MySQL 3.23 to 4.0 after every way I tried failed. In the end, the 
> system became corrupted that I had to re-install Linux (probably a real 
> Linux person would laugh at this, but after a while /usr/bin got so 
> filled with junk that I didn't know what I had anymore and wanted to 
> start over "clean").

It would be certainly better if you install self compiled stuff in 
/usr/local/ (or control the installation with rpm).

> 
> Not to say that this sort of thing doesn't happen in Windows-land, but 
> setup.exe works on 99% of the packages I have tried 99% of the time. 
> Don't get me wrong, I really like Linux, but this is one area where I 
> believe that Linux still has a ways to go to "catch up" in the 
> ease-of-use department.
> 
> Don

Ralf.



From jb at cascade-sys.com  Mon Sep 16 14:50:16 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 16 Sep 2002 11:50:16 -0700
Subject: threads or queue for this task
References: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com> <GmCg9.2354$Ee4.6690@news-server.bigpond.net.au> <cqp7ou8d0jjcdtrqetu9v9ls9gst2cp4qi@4ax.com> <mailman.1032076038.5654.python-list@python.org> <h2jbouo4j9f8ee646bhfejt3erqku76d61@4ax.com>
Message-ID: <3D8627E8.1010903@cascade-sys.com>

robin at execulink.com wrote:

>I may sometimes be clever, and can no doubt cobble together code from
>reading the docs (which I have been doing so far), but what I don't
>know are some of the trade-offs of using certain constructs, or the
>optimal way to do things. That is much more difficult to glean without
>actual code examples.
>
>For example, I really don't know what "proper design" I should use in
>this case. Any sample code?
>
The following illustrates my understanding of what you're trying to do.

    import thread
    import Queue

    # three queues connect four threads

    CmdQ1 = Queue.Queue()
    CmdQ2 = Queue.Queue()
    CmdQ3 = Queue.Queue()

    # Reader thread reads commands from stdin and
    # dispatches to appropriate thread via queue

    def Reader():
        while 1:
            line = stdin.readline()
            if not line:
                return

            if line[0] == '1':
                CmdQ1.put( line[1:])
            elif line[0] == '2':
                CmdQ2.put( line[1:])
            elif line[0] == '3':
                CmdQ3.put( line[1:])
            else:
                continue    # or signal error

    # Command threads process commands as they arrive via queue

    def Cmd1():
        while 1:
            cmd = CmdQ1.get()
            process1( cmd )

    def Cmd2():
        while 1:
            cmd = CmdQ2.get()
            process2( cmd )

    def Cmd3():
        while 1:
            cmd = CmdQ3.get()
            process3( cmd )

    def main():
       
        thread.start_new_thread( Cmd1, ())
        thread.start_new_thread( Cmd2, ())
        thread.start_new_thread( Cmd3, ())

        # reader thread runs in mainline
        Reader()


Frankly, it doesn't get any simpler than this.

With threads as nodes and Queues as pipes, you can create practically 
any data flow topology you wish.

>The interface to Queue is deceptively simple. Are there constructs to
>avoid or performance trade-offs to different methods? Again, any
>explicit code (or at least explicit help as to which methods are
>preferable) would be appreciated.
>

It's not deceptively simple.  It's merely simple.  Sometimes a Queue is 
simply a Queue.

>I am also put off by the description of the Queue Object, particularly
>methods like empty() which says: "Return 1 if the queue is empty, 0
>otherwise. Because of multithreading semantics, this is not reliable."
>

The answer is accurate when it is returned to you but due to the nature 
of multithreading semantics, the value also is subject to change without 
notice.  So if you read it and see a 0 or a 1, this is a true historical 
fact but you can't rely on the value being true at arbitrary points in 
the future so you can't really base any decisions on it.  

This is a function that was trivial and natural for the class designer 
to implement.  It probably IS used internal to the class in order to 
decide whether or not to suspend or resume a thread.  But is of little 
practical value to the end user.  Fageddabouit.

>How can one code if you don't
>reliably know when a Queue is empty or full?
>

Generally, Queues never get "full" and you generally don't care if 
they're empty.  

The import thing about Queues is that the reader thread will 
automatically be suspended when reading from an empty queue and 
furthermore it will automatically wake up when the queue becomes non empty.

All you need from Queues is to create them and to get and put data.

>What this boils down to is: I'm sure I can do the job with the great
>help I've already received, but do not have the confidence to know if
>I'm doing the job the *right* way.
>

If your code does what you want and doesn't take an unreasonable amount 
of time to run on your computer then that's 80% of the "*right*" way. 
 I'll defer to others to argue the remaining 20%.

There really aren't as many alternatives here as you seem to suspect.

You probably should look for a book on concurrent sequential processes. 
 Most of the paradigms and pitfalls are pretty standard and would be 
spelled out in the book.  Unfortunately, I am at a loss to make a 
specific recommendation.  Perhaps somebody else has an idea.

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From SBrunning at trisystems.co.uk  Tue Sep  3 09:19:29 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 3 Sep 2002 14:19:29 +0100 
Subject: How to avoid freeing memory on exit?
Message-ID: <31575A892FF6D1118F5800600846864DCBD710@intrepid>

	> os._exit

> From:	Gumuz [SMTP:gumuz at looze.net]
> so what does 'one' underscore prefixing a methodname mean?
 
Proceed with caution.

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




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



From wlfraed at ix.netcom.com  Sat Sep  7 13:45:41 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Sat, 07 Sep 2002 10:45:41 -0700
Subject: Thinking of joining
References: <3D7970DC.5115BCC8@earthlink.net>
Message-ID: <50edla.pu2.ln@ix.netcom.com>

dark fed this fish to the penguins on Friday 06 September 2002 08:18 pm:


> I was just curious if it will do everything PERL will do, as i
> dont care for how PERL reads.

        FORTRAN will /do/ everything perl does -- you just wouldn't want to 
have to write it...

        Assuming some linkage to OS system calls, an ability to do comparison, 
branching, and simple arithmetic, ALL languages are equivalent in 
capability -- they only differ expression (usually targeted to an 
application domain -- I'd hate to consider writing a satellite tracking 
program in perl... well, to be honest, I'd hate to write /anything/ in 
perl).

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From tdelaney at avaya.com  Wed Sep 11 19:43:10 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 12 Sep 2002 09:43:10 +1000
Subject: I must be an idiot, please pity me!
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A591@natasha.auslabs.avaya.com>

> From: newbie [mailto:sam at localhost.localdomain]
> 
> Perhaps I wasn't clear, but I don't believe I mentioned anything not
> functioning properly. I did say "All works fine". I don't have failing
> code, I simply suspect (hope) the current implentation isn't 
> very good.
> The pseudo code listed was what I was epecting python to 
> allow me to do,

Pseudocode is useless for optimisation purposes. Python code is very similar
to pseudocode - post your exact code if you want help.

Tim Delaney



From erikprice at mac.com  Tue Sep 10 10:10:02 2002
From: erikprice at mac.com (Erik Price)
Date: Tue, 10 Sep 2002 10:10:02 -0400
Subject: web crawler help?
In-Reply-To: <JARd9.6927$yt3.3340577@newssrv26.news.prodigy.com>
Message-ID: <FFD5FC58-C4C6-11D6-A54B-00039351FE6A@mac.com>

On Thursday, September 5, 2002, at 07:15  PM, koko wrote:

> is there any sample for basic web crawler, that ask for a starting url 
> and
> log the url and extract the hyperlinks?
> thx

If you just want to learn the basics of making a web crawler, there is 
a great tutorial on page 187 of "Python Web Programming" by Steve 
Holden.  There are more fully-featured spiders available in Python, but 
this one presents the core principles nicely.



Erik



--
Erik Price                                                  (zombies 
roam)

email: erikprice at mac.com
jabber: erikprice at jabber.org




From erno-news at erno.iki.fi  Wed Sep 11 11:27:48 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 11 Sep 2002 18:27:48 +0300
Subject: read() file??
References: <mailman.1031684894.10361.python-list@python.org>
Message-ID: <kuvg5c1r7v.fsf@lasipalatsi.fi>

In article <mailman.1031684894.10361.python-list at python.org>,
jubafre at brturbo.com writes:

| I have a file with assembler instruction and i want read the lines
| of the file and put in a list of string, this is easy to do, but i
| don?t want to consider the caracteres "/" and the othes after him.

| file.asm:
| ADD D1  /coment1
| LDA D2  /coment2

try something like...

def read_asm(filename):
    out = []
    for line in open(filename).readlines():
        out.append(line.split('/')[0])
    return out

  -- erno


From quiteblack at yahoo.com  Fri Sep  6 06:33:33 2002
From: quiteblack at yahoo.com (black)
Date: 6 Sep 2002 03:33:33 -0700
Subject: how could we know which module is imported already ?
Message-ID: <7153651c.0209060233.2a72caaa@posting.google.com>

want to know how many and what the module is imported~


From bokr at oz.net  Mon Sep  2 15:30:46 2002
From: bokr at oz.net (Bengt Richter)
Date: 2 Sep 2002 19:30:46 GMT
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <un50ccotaak86b@corp.supernews.com>
Message-ID: <al0e96$oms$0@216.39.172.122>

On Sun, 01 Sep 2002 21:07:24 -0000, "A.M. Kuchling" <amk at amk.ca> wrote:

>In article <3d725881.345921 at news.t-online.de>, Gerson Kurz wrote:
>> in the match or find function. So basically, couldn't one come up with
>> a *human readable* syntax for re, and compile that instead? Python
>
>Maybe Ka-Ping Yee's rxb?  http://web.lfw.org/python/rxb15.py
>
>The problem with a new syntax is that no one else would be using it, so
>you'd still need to learn the existing syntax for use with grep, vi, Perl,
>&c.  (It wouldn't surprise me if Perl 6's revised regexes run into this very
>difficulty and don't gain much adoption.)
>
I agree about new syntax, but I wouldn't mind having a re.help(regexp) function
for interactive use that would just explain in 'English' what the regexp expression
stands for. It would be a nice easy double check on whether I wrote what I meant,
and helpful for understanding someone else's magic. It shouldn't be that hard to do.

Regards,
Bengt Richter


From peter at engcorp.com  Wed Sep 18 19:00:09 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 18 Sep 2002 19:00:09 -0400
Subject: Python-style license?
References: <jikm5-p5l.ln1@junker.stroeder.com> <mailman.1032367164.17304.python-list@python.org>
Message-ID: <3d89056c$1@news.sentex.net>

Michael Chermside wrote:
> Michael Str?der wrote:
> 
>> HI!
>>
>> In some Python modules there's only a note about a "Python-style 
>> license". Now I wonder what that means. Any clues?
>>
>> Ciao, Michael.
>>
> 
> http://www.python.org/2.2/license.html

And more generally, they are generally talking about licenses with
the characteristic of allowing one to use the software even for
commercial purposes, provided appropriate attribution is made etc.

Another way of looking at it is that they're gnot GNU licenses.

-Peter



From aleax at aleax.it  Sat Sep 14 15:45:27 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 14 Sep 2002 19:45:27 GMT
Subject: Converting a string to time - simple task but don't know how to do it!
References: <7A6D2E53461F2BF2.AC83A3D5E1A01468.48CC780BD6530537@lp.airnews.net>
Message-ID: <rlMg9.108608$pX1.3818369@news2.tin.it>

Richard Kessler wrote:

> I have a string returned from a database expressing the time i.e.
> 2002-09-14
> 10:30:34.  I need to compare this value to the current time to determine
> the amount of time elapsed between the two. I get the current time using
> time.time() but I cannot find a way to convert the string from the
> database into a time value (seconds from epoch) to do the simple
> subraction to find seconds elapsed.

What you want is function strptime of module time.  Unfortunately, up
to Python 2.2 it's platform-dependent whether module time does supply
function strptime or not (in Python 2.3, you'll be able to use that
function on any platform).  There are pure-Python implementations
of strptime out on the net, e.g. 
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/56036


Alex



From shalehperry at attbi.com  Thu Sep 12 11:30:14 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 12 Sep 2002 08:30:14 -0700
Subject: "str.contains(part)" or alternatives?
In-Reply-To: <6g2g9.295995$aA.51455@sccrnsc02>
References: <mailman.1031787694.27175.python-list@python.org> <6g2g9.295995$aA.51455@sccrnsc02>
Message-ID: <200209120830.14264.shalehperry@attbi.com>

On Thursday 12 September 2002 08:19, Pete Shinners wrote:
> Delaney, Timothy wrote:
> > In fact, this was the deciding factor in Guido changing his position on
> > 'if s1 in s2'. A string has sequence-like behaviour (e.g. it is iterable,
> > individual characters can be indexed), but also has its own behaviour
> > (e.g. substring containment, all elements are single-character strings).
>
> does this mean we can also now do this? if so, nifty
>
>  >>> 1,2,3 in range(10)
>
> 1
>
>  >>> 3,2,1 in range(10)
>
> 0

I hope not.  1,2,3 should evaluate to a tuple and the tuple (1,2,3) is not in 
[1,2,...,10].



From robin at jessikat.fsnet.co.uk  Fri Sep  6 05:38:40 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 6 Sep 2002 10:38:40 +0100
Subject: jython memory usage
References: <hVm5ydAY2hd9Eww$@jessikat.demon.co.uk>
 <3d76366a$1_5@news.bluewin.ch> <rnLbuKAv5jd9Ewzz@jessikat.demon.co.uk>
 <3d76494d$1_5@news.bluewin.ch> <YR13DVAUNld9EwXH@jessikat.demon.co.uk>
 <3d765ce6_4@news.bluewin.ch> <zzKd7KAqhnd9EwxT@jessikat.fsnet.co.uk>
 <3d767c4a$1_1@news.bluewin.ch> <YWyHyIAJKyd9EwHA@jessikat.demon.co.uk>
 <3d774f40$1_1@news.bluewin.ch> <RTiVzIAJD3d9Ew2t@jessikat.demon.co.uk>
 <3d7779b0$1_3@news.bluewin.ch>
Message-ID: <Iz7TraAgeHe9EwPO@jessikat.demon.co.uk>

In article <3d7779b0$1_3 at news.bluewin.ch>, Samuele Pedroni
<pedronis at bluewin.ch> writes
.....
>
>Have you tried with an older JVM? anyway is the code and minimal
>instructions to reproduce the "disaster"
>available ?
>
>regards.
OK my colleague ran with sun jdk 1.3.1 and the test worked! So I guess
the problem lies with Marius Gedminas' TTFont class vs 1.4.0_01
grrhhhhh!
-- 
Robin Becker


From peter at engcorp.com  Wed Sep 18 00:57:15 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 18 Sep 2002 00:57:15 -0400
Subject: Python comment stripper
References: <16Nh9.248$Wc6.17087796@newssvr13.news.prodigy.com>
Message-ID: <3d8807b1$1@news.sentex.net>

Dennis Reinhardt wrote:
> Does anyone have a Python source comment stripper or can anyone recommend a
> parser that could be adapted for same?
> 
> Why?  ... well there is a better than 2:1 byte size reduction in both raw
> and compressed .py files if comments are stripped.  

This doesn't seem like a good reason to me.  Why not just use .pyc files
all the time and be done with it?

The only reason to use .py instead of .pyc if size is such an issue
(and with 60GB drives cheap I can't imagine why it would be), is that
you want to maintain access to the source to make changes.  In other
words, to maintain the code.  And maintaining code that has the comments
stripped seems pretty counterproductive, doesn't it?

Or am I missing something?

-Peter



From eric.brunel at pragmadev.com  Wed Sep  4 11:56:38 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Wed, 4 Sep 2002 17:56:38 +0200
Subject: efficient list merging
References: <mailman.1031150946.17547.python-list@python.org>
Message-ID: <al5a1g$hat$1@wanadoo.fr>

Mark McEahern wrote:
[snip]
> Anyway, here's yet another approach:
> 
> #! /usr/bin/env python
> 
> lol1 = [['a'], ['a', 'b', 'c'], ['d', 'e', 'f'], ['g', 'h', 'i', 'j']]
> lol2 = [['d', 'e', 'f'], ['k', 'l', 'm'], ['d'], ['f', 'g'], ['a']]
> 
> def mergelists(*lists):
>     d = {}
>     for each_list in lists:
>         for item in each_list:
>             key = str(item)
>             if key not in d:
>                 d[key] = item
>     ret = d.values()
>     ret.sort()
>     return ret
> 
> mergelol = mergelists(lol1, lol2)
> 
> expected = [['a'], ['a', 'b', 'c'], ['d'], ['d', 'e', 'f'], ['f', 'g'],
>             ['g', 'h', 'i', 'j'], ['k', 'l', 'm']]
> 
> assert mergelol == expected
> 
> // m

To avoid explicit for loops, you may write:

l = lol1 + lol2
mergelol = dict(zip(map(str, l), l)).values()
mergelol.sort()         # if needed...


Maybe doing:

mergelol = dict(zip(map(tuple, l), l)).values()

can be a little faster: converting a list to a tuple *seems* easier than 
converting it to a string. But I may be wrong...

Anyway, any method involving a dictionary will require to convert your 
lists to something else: list are mutable, and cannot be used as dictionary 
keys...
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From mgarcia at cole-switches.com  Fri Sep 20 20:06:10 2002
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: 20 Sep 2002 17:06:10 -0700
Subject: What the heck? open() requires an integer?
References: <FCBi9.13802$Zk4.708378@news2.telusplanet.net> <ZLBi9.136908$pX1.4869246@news2.tin.it> <MTIi9.14708$Zk4.751630@news2.telusplanet.net>
Message-ID: <fec163ad.0209201606.6fff11c0@posting.google.com>

"Fausto Arinos de A. Barbuto" <fbarbuto at telusplanet.net> wrote in message news:<MTIi9.14708$Zk4.751630 at news2.telusplanet.net>...
>     Also, it is a good idea to learn more Python before
>     venturing to develop code. Or to adapt someone
>     else's source code, which was exactly the case.

Please don't be too intimidated by Python.  It is fairly typical to
write a non-trivial Python program while learning the language.  For
example, it is surprising how straightforward the trickiest of OO
programming is in Python.

O'Reilly publishes two excellent books for people who wish to learn
Python and write applications quickly: Programming Python, 2nd
Edition, and Python Cookbook.

Also the Python newsgroup is very friendly to people new to the
language


From sholden at holdenweb.com  Wed Sep 18 18:49:03 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 18 Sep 2002 18:49:03 -0400
Subject: Need advice on choosing the technology for intranet based application 
References: <3d88fd1e@news.ivm.net>
Message-ID: <Gn7i9.1769$ik2.430@fe02>

"Nenad Propadovic" <propadovic.nenad at debitel.net> wrote ...
> Hello pythonistas,
> I'm about to write an application related to injury prevention in big
> industrial companies. It's supposed to be intranet-based (because the
admins
> are reluctant to install it on x (where x>>1) clients). So: the client is
a
> web-browser. I still don't know whether I will get a server on my own, or
> I'll have to fit into an unknown web server environment, both options
being
> realistic.

Well, assume in that case that you should write for a CGI environment: it's
maybe not the absolute fastest way to go, but it's pretty portable, and so
you'll be able to make plans ahead of that particular decision being made
(or do you get to make it?), and you can accommodate almost any server in
this way.

> By now it seems the logic of the application will be simple. It's going to
> be a relational database with several tables, a modest number of users
with
> write priviledges, some more with read priviledges only. Performance is
not
> a big issue.

Great. Almost any language could eat this part up for breakfast :-)

> The GUI has a part I consider complex: I have to present some statistics
in
> form of charts or graphs.

The Python Image Library (PIL) is very good for creating graphics. You can
create them "on the fly" in your web server and then serve back HTML
documents with references to the created graphics.

If you need help, you can buy the professional PIL with a support contract,
I believe. The guy who wrote that software shows every appearance of knowing
what he's doing.

> Now I've never done any web based application in my life, have a very
> rudimentary knowledge of HTML, have been doing som database hacking (O.K.,
> engineering :-) ) with MS Access and Ingress/ Open Road, have once written
a
> socket-based combined telnet/ftp client.

You probably have enough programming experience for the task, although you
might need some help with the basics of the technologies.

> So I kindly ask for knowledgeable opinions on the following question:
> -which web/application server? I'm considering Zope (seams rather BIG when
> you download & unpack it, and it seams all possible web applications have
> been written for it) and twisted (smaller, seams lower lewel and somewhat
> easier to learn), and I'm wondering if there is something else out there,
> which is both powerfull and flexible and easy to learn and complete. Oh
yes,
> I have both a time constraint (mid november, having only my spare time for
> the project), and don't want end up rewritting the stuff under another
> system, because the chosen one isn't flexible enough. Database included
> would be nice.

Seems to me that CGI under Apache, while probably not the most exciting
cxhoice, should handle your requirements quite well. You're corect about
Zope: the learning curve is pretty high. Twisted is still a bit
experimental, and because of its somewhat unconventional architecture and
limited documentation that too might take a bit of learning. Both are
excellent products, but your constraints seem to rule them out (unless you
learn much faster than me).

The Python DB API means you can choose from a wide range of relational
database back-ends.

> -I assume the simplest way to present a graph/chart kind of graphic is to
> create a jpg/giff/tiff on the server side, and just show it on the client?

Correct. Otherwise you get involved with writing Java applets you can send
the data to, and things can get ugly fast ;-)

> Thanks to everybody taking the time to read the message, and even more to
> those who answer it :-)

A pleasure. There's a book called "Python Web Programming" by some chap
called Holden that might give you some useful background ...

> P.S. I'm just introducing Python to my group at BMW (a completely
different
> project, though), and I'll let you know if there are some bigger success
> stories.
>
Let us know what we can do to help!

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From mcfletch at rogers.com  Wed Sep 25 00:31:01 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Wed, 25 Sep 2002 00:31:01 -0400
Subject: Removing all occurences of a character from a string.
References: <amr8nh$rd8$1@driftwood.ccs.carleton.ca> <DL9k9.201596$216.7001230@bin4.nnrp.aus1.giganews.com> <amrav1$4rk$1@driftwood.ccs.carleton.ca>
Message-ID: <3D913C05.3080706@rogers.com>

Under later Python versions (such as 2.2.1), you can do the following:

attr = [[item.replace("'","") for item in element] for element in attr]

Which is rather elegant IMO.  It also has the side-effect of _not_ 
rewriting the data in-place in your original list, instead just 
replacing that list when we are finished creating the new list, which 
may or may not be desirable for you.

HTH,
Mike

PS: Always nice to hear from another Canadian Pythonista :)

Sean Ross wrote:
> Thank you Justin.
> 
> How about this, then:
> 
> I have a list of lists of these strings and I want to apply replace to each
> string in the list of lists.
> 
> So, say:
> 
> attr = [ ["'gill-attachment'", "'a'", "'d'", "'f'",
> "'n'"],["'gill-spacing'", "'c'", "'d'", "'w'"], ["'gill-size'", "'b'",
> "'n'"]]
> 
> I can do the following:
> for i in xrange(0, len(attr)):
>     for j in xrange(0,len(attr[i])):
>          attr[i][j] = attr[i][j].replace("'","")
> 
> to get what I want; which is:
> [ ['gill-attachment', 'a', 'd', 'f', 'n'],['gill-spacing', 'c', 'd', 'w'],
> ['gill-size', 'b', 'n']]
> 
> Is there a nicer way to do this?
> Sean
...




From Oschler at earthlink.net  Sat Sep 14 20:48:12 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sun, 15 Sep 2002 00:48:12 GMT
Subject: Stackless Python, eventual merge?
Message-ID: <gNQg9.6056$Lg2.882744@news2.news.adelphia.net>

I've been reading a lot of good things about Stackless Python, especially
when it comes to speed and multi-threading.  Do you think that it eventually
will be merged with the main Python code base?  I'm thinking of trying it
out, but I'm worried about ending up with a heavy code fork that I'll end up
regretting.

thx





From rsrchstr at hotmail.com  Fri Sep 27 18:11:36 2002
From: rsrchstr at hotmail.com (mike henley)
Date: 27 Sep 2002 15:11:36 -0700
Subject: windows xp command prompt
Message-ID: <576c1752.0209271411.6d7bb4c4@posting.google.com>

i have installed the activestate perl, python, tcl, and pragmatic
programmer's ruby. when i go to the command prompt on windows xp and
type perl for example... (or python, ruby... etc) i get the following
message.

'perl' is not recognized as an internal or external command,
operable program or batch file.

in the control panel| system | advanced | environment variables | Path
i have

C:\ruby\bin;C:\Python22;C:\Perl\bin;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem

does anyone know what i should do to fix this problem? (apart from the
obvious of restarting the machine)


From syver-en+usenet at online.no  Fri Sep 20 19:56:56 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Fri, 20 Sep 2002 23:56:56 GMT
Subject: Python Compiling
References: <amfveg$5g64k$1@ID-53055.news.dfncis.de>
Message-ID: <uelbo5i6k.fsf@online.no>

"Terje Johan Abrahamsen" <spoermeg at voldelig.com> writes:

> First of all, I am going to warn you. I will ask some stupid
> questions. I
> 
> haven't programmed too much, and don't really know how it all links
> together.
> 
> But, I guess my question boils down to, why is not Python a compiled
> language?
> 
> To my best knowledge, the computer understand 0s and 1s. So, any
> programming language must be brought down to that.

Maybe it helps to think of it this way. The interpreter is a compiled
program. Python programs then, are a kind of meta program, a program
that runs in another program.

-- 

Vennlig hilsen 

Syver Enstad


From loewis at informatik.hu-berlin.de  Sun Sep 15 14:04:04 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 15 Sep 2002 20:04:04 +0200
Subject: Python11
References: <xwMg9.1433$Dy4.464@nwrddc03.gnilink.net> <m3admk47pf.fsf@mira.informatik.hu-berlin.de> <Xr0h9.2585$Dy4.200@nwrddc03.gnilink.net>
Message-ID: <j47khnruy3.fsf@informatik.hu-berlin.de>

"Jeff Sasmor" <jeff at sasmor.com> writes:

> One hopes that Python11 won't be lost in the much larger OSCON..

There are also plans to have YAS (Yet Another Society) organize a
Python-dedicated conference, at a much lower price than OSCON, at GWU
(George Washington University). Plans for that are not final, either.

Regards,
Martin



From loewis at informatik.hu-berlin.de  Thu Sep 19 12:35:32 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 19 Sep 2002 18:35:32 +0200
Subject: Remarks to Python 2.3
References: <20020918222421.328088eb.use-net@schabi.de> <amannn$p4b$1@panix1.panix.com> <m3u1knj8j9.fsf@mira.informatik.hu-berlin.de> <20020919152834.592eaa39.leafnode-usenet@schabi.de>
Message-ID: <j4r8fqq6nf.fsf@informatik.hu-berlin.de>

Markus Schaber <leafnode-usenet at schabi.de> writes:

> I'm not really shure whether my remarks are about "bugs" - I just wanted
> to collect other opportunities here to help me decide whether one can
> regard this as bugs or not, and eventually submit bug reports.

There is always the potential danger that a newsgroup message gets
ignored, and the poster gets the impression that nobody cares about
the (potentially significant) problem.

In these cases, it is a good idea to submit a bug report, because the
tracking involved in the SF bug database guarantees that the report
stays open until it has been considered.

Of course, bug reports are only the right thing if the poster feels
that Python behaves "bad", where "bad" means either incorrect,
non-intuitive, or in an otherwise undesirable way. For a bug report,
one should also research whether this is a well-known issue.

Regards,
Martin



From michael at stroeder.com  Tue Sep 24 05:47:55 2002
From: michael at stroeder.com (=?ISO-8859-1?Q?Michael_Str=F6der?=)
Date: Tue, 24 Sep 2002 11:47:55 +0200
Subject: Sessionhandling
References: <c5b3549a.0209230650.34a30fca@posting.google.com> <f72dac0d.0209231211.5614b398@posting.google.com> <amp3h2$1ot7$1@ariadne.rz.tu-clausthal.de>
Message-ID: <bvo56-tf1.ln1@junker.stroeder.com>

Lutz Schroeer wrote:
> langpref = os.environ['HTTP_ACCEPT_LANGUAGE']    
> firstpref = 'en-ie'  
> try:
>   langs = string.split(langpref, ',')
>   firstpref = langs[0]
>   secondpref = langs[1]
>   thirdpref = langs[2] 
> except:
>   pass

One might think about a more complete solution handling the weight:

http://stroeder.com/pylib/PyWebLib/pydoc/pyweblib.helper.html#AcceptHeaderDict

Most users do *not* configure their browsers anyway. Therefore it's 
almost useless to use HTTP_ACCEPT_LANGUAGE or similar env vars.

Ciao, Michael.



From rpm1deletethis at nospamfrontiernet.net  Wed Sep  4 06:49:28 2002
From: rpm1deletethis at nospamfrontiernet.net (RPM1)
Date: Wed, 4 Sep 2002 06:49:28 -0400
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com>
Message-ID: <unbp765rbsn809@corp.supernews.com>

"Benjamin" <phncontact.libero.it@> wrote in message
news:uafbnugdli4c3f8ns6qr33mna5urdno78g at 4ax.com...
> Hi guys (and girls)
> I just begun to learn how to program. And I choosed Python for
> doing that. surfing around the web i found an article about this
> language, they said it's the ideal language for beginners to the
> world of programming. and i admit it, the syntax is very clear making
> the code very readable, i learn easily the concepts of programming
> and i soon wrote some nice programs. but since i finished my
> newbie-tutorial, i am wondering about one question: is it worth to go
> deeper into python? i talked to a friend of mine, he told me that i
> should try php..
> what should i do?? ^_^ in other words: what is python good for?
> why should i continue with python?
> thank you!
>

I think you should try coding in C.  It's used everywhere for
everything.  It's faster than about any other language.
Most other languages are written in C.  C is highly marketable.

But best of all...

You'll learn to *really* appreciate other languages like Python,
Java, ...

If you really want to get a feel for a language, read a fairly large
program that somebody else has written in that language, (not
just little snippets).

Patrick





From henrik.motakef at web.de  Sun Sep 29 05:55:14 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 29 Sep 2002 11:55:14 +0200
Subject: What does Python fix?
References: <e15b6fa3.0209271418.14cd5642@posting.google.com> <flaapucbduj6g4ifrdf3oq99hdrepi1o3v@4ax.com> <3d95fb09@news.vip.fi> <87ptuxub5x.fsf@pokey.henrik-motakef.de> <vcmcpusnjqvpj9lavtqlqa7cs01h4i1iij@4ax.com>
Message-ID: <87znu1rugt.fsf@pokey.henrik-motakef.de>

Courageous <jkraska at san.rr.com> writes:

> >Let's not forget that Python's syntax (most of all the significance of
> >whitespace) is considered strange at least by about all novices. In
> >both cases - significant whitespace and lots of irritating superflous
> >parentheses - people don't just get used to it, but tend to like it
> >after a while.
> 
> I'm fairly sure that you meant to say "Lisp" where you said Python,
> right?

No, I'm talking about how newbies tend to react when you tell them
that block structure in Python is determined by indentation. 

> It doesn't help that the number of big commercial Lisp compilers
> can be counted on the fingers of, err, one finger.

You must have funny fingers then, as the people working for Franz,
Xanalys and Corman will love to explain.

BTW, there aren't that many big commercial Python compilers out there,
and that doesn't seem to be too big a problem for the Python users. :)

Regards
Henrik




From haering_postgresql at gmx.de  Thu Sep  5 19:11:34 2002
From: haering_postgresql at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Fri, 6 Sep 2002 01:11:34 +0200
Subject: What database should I use
In-Reply-To: <3D77763F.23080.BFAFA3@localhost>
References: <3D77763F.23080.BFAFA3@localhost>
Message-ID: <20020905231134.GA1229@lilith.ghaering.test>

* A <printers at sendme.cz> [2002-09-05 15:20 +0200]:
> Hello,
> I am going to program a small application (in Python and wxPython) which consists about 
> 2000 - 3000 records . Can you please recommend which database is the best for that. I 
> use Windows.
> Maybe MS Access  or FoxPro or MySQL or better?
> I want to keep the program as small as possible.

PySQLite. It uses an embedded SQL engine (SQLite), which means you don't
have to set up a separate server process. It only weights a hundred KB
or so and has good SQL support (a lot better than MySQL, for example).

http://pysqlite.sf.net/

-- Gerhard



From tjreedy at udel.edu  Fri Sep 20 19:36:46 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 20 Sep 2002 23:36:46 GMT
Subject: e vs exp()?
References: <mailman.1032550470.8080.python-list@python.org>
Message-ID: <iiOi9.173334$5r1.7100346@bin5.nnrp.aus1.giganews.com>

"Tim Peters" <tim.one at comcast.net> wrote in message
news:mailman.1032550470.8080.python-list at python.org...
> [Terry Reedy]
> > ...
> > On the hypothesis that x**y is calculated as exp(y*log(x)),
>
> Bad hypothesis.  The implementation of a decent x**y function is one
of the
> most difficult tasks in a platform's math library....
>I've written such a beast without benefit of an extended hardware
> precision; I wouldn't want to do it again <wink>.

What was your basic approach?  Power series?  Rational function?
Continued fraction? ???
http://functions.wolfram.com/ElementaryFunctions/Power/
gives 343 formulas (many only relevant to complex domain) but doesn't
seem to indicate which Mathematica actually uses (trade secret I
presume).

Terry J. Reedy




From vlindberg at verio.net  Mon Sep 23 19:21:41 2002
From: vlindberg at verio.net (VanL)
Date: Mon, 23 Sep 2002 17:21:41 -0600
Subject: Downloading huge files via urllib
Message-ID: <3D8FA205.8090602@verio.net>

Hello,

For various reasons, I have to use https to download large (20+ MB) text 
files which I then parse.  I set up a basic function to do this using 
urllib:

response = urllib.urlretrieve(serverURL, 'run.log')

However, I then get a MemoryError.  Tracking down the source of the 
error, I see the offending function in httplib:

    def makefile(self, mode, bufsize=None):
        """Return a readable file-like object with data from socket.

        This method offers only partial support for the makefile
        interface of a real socket.  It only supports modes 'r' and
        'rb' and the bufsize argument is ignored.

        The returned object contains *all* of the file data
        """

I think the problem is that bufsize argument that is ignored.  Does 
anyone know if this is correct, and what I can do about it?  I would 
like to automate the process of downloading this file, but is it possible?

Thanks,

VanL




From ckasso at sprynet.com  Sun Sep 29 17:43:22 2002
From: ckasso at sprynet.com (Chris Kassopulo)
Date: Sun, 29 Sep 2002 17:43:22 -0400
Subject: Running with arguments in IDLE?
References: <3D90DB87.4797EA7@quartercase.com>
Message-ID: <20020929174322.7013132b.ckasso@sprynet.com>

On Tue, 24 Sep 2002 21:39:19 GMT
joe sixpack <joe.sixpack at quartercase.com> wrote:

> the arguments.  What I don't know is how to run the script from 
> within IDLE and specify the argument to use when it is run.  In
> Pythonwin, 'RUN'ning a script brings up the dialog box in which
> arguments may be specified.  How is this done in IDLE?  

I just noticed this post.  It's a few days old and I don't see
any replies, so I'll try.  According to the docs you can't run
with arguments directly.  You need a function that calls your
script with arguments.  See "Running a File" in:

    http://www.python.org/idle/doc/idle2.html

Also, I don't think this is what you want but I'll mention it
anyway.  From a commandline you can run a script in idle with:

    python idle.py ascript.py arg1 arg2 ...


Chris Kassopulo


From h_schneider at marketmix.com  Fri Sep 27 07:20:54 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Fri, 27 Sep 2002 13:20:54 +0200
Subject: py2exe Error creating NT Service
Message-ID: <an1epm$rc5$03$1@news.t-online.com>

Hi,

got an error with py2exe when trying to create a Win32 service:

.
changing back to 'C:\WORK\Software\Python\xqmagent'
creating dist\xqmagentwin\xqmagentwin.exe
Traceback (most recent call last):
  File "C:\WORK\Software\Python\xqmagent\setup.py", line 10, in ?
    scripts=['xqmagentwin.py'])
  File "C:\usr\python\lib\distutils\core.py", line 138, in setup
    dist.run_commands()
  File "C:\usr\python\lib\distutils\dist.py", line 899, in run_commands
    self.run_command(cmd)
  File "C:\usr\python\lib\distutils\dist.py", line 919, in run_command
    cmd_obj.run()
  File "C:\usr\python\py2exe\build_exe.py", line 626, in run
    self.create_exe(exe_name, arcname, use_runw)
  File "C:\usr\python\py2exe\build_exe.py", line 823, in create_exe
    svc_name, svc_display_name, svc_deps = self.get_service_names()
  File "C:\usr\python\py2exe\build_exe.py", line 898, in get_service_names
    exec compile(open(self.script).read(), self.script, "exec") in
mod.__dict__
  File "<string>", line 101

SyntaxError: invalid syntax


The py2exe service sample compiles perfect on my machine, but this script
here does not. There are no differences in the class derived from the
service framework.

Any ideas?

All the best,
Harald






From peter at engcorp.com  Tue Sep 17 08:27:26 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 17 Sep 2002 08:27:26 -0400
Subject: problems getting a single item out of a list
References: <20020917071900.GB23629@partagas.as.de.cw.net> <200209170030.52964.shalehperry@attbi.com> <mailman.1032249801.25943.python-list@python.org>
Message-ID: <3d871fa0$1@news.sentex.net>

Manuel Hendel wrote:
> On Tue, Sep 17, 2002 at 12:30:52AM -0700, Sean 'Shaleh' Perry wrote:
> 
>>just use fields[2] (you start counting at zero).
> 
> 
> This worked fine:
> 
> domainfield = fields[3]
> 
> But the next problem is, that after a 
> 
> domainfield = string.split(fields[3])
> 
> domainfield seams to be a list item again.
> 
> print domainfield
> 
> ['142000']

Well, you don't show what fields[3] contained before the
string.split, but in any case split() returns a list by
definition, so what else did you expect?  If you want to
get the first item from the split list, just use
domainfield[0] this time...

-Peter



From loewis at informatik.hu-berlin.de  Thu Sep  5 10:45:32 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 05 Sep 2002 16:45:32 +0200
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <20020905.072837.159259470.2038@cybermesa.com>
Message-ID: <j4y9ag3377.fsf@informatik.hu-berlin.de>

"Jay O'Connor" <joconnor at cybermesa.com> writes:

> Why not simply disallow the using of a name already defined in an
> enclosing scope? Similar to Smalltalk?

That might break compatibility with existing code.

Regards,
Martin


From timr at probo.com  Fri Sep 27 01:16:03 2002
From: timr at probo.com (Tim Roberts)
Date: Thu, 26 Sep 2002 22:16:03 -0700
Subject: Help - GUI programming question
References: <3d93ca8b_4@goliath.newsgroups.com>
Message-ID: <29q7pu86sh1rjs7k2p8js56pot9li42je6@4ax.com>

 wrote:

>Quick question. The following yields the error: "makemenu takes only one
>argument (2 given)". This should be simple - I'm just not seeing it! Can
>anyone explain? The applicable code follows:

I know!  I know!  What's the prize?

>class WINMENU:
>  def makemenu(parent):

Your makemenu DOES only expect one argument.  You need to make room for the
instance:

   def makemenu(self,parent):

>    menubar = Frame(parent)                        # win=top-level window
>    menubar.pack(side=TOP, fill=x)                 # set its menu option
>	(more code to build a menu.....)
>
>if __name__ == '__main__':
>  root = Tk()
>  root.title('GUI Menu')
>  smenu = WINMENU()
>  smenu.makemenu(root)  
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From BPettersen at NAREX.com  Wed Sep 18 13:20:34 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Wed, 18 Sep 2002 11:20:34 -0600
Subject: Python comment stripper
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192012533FE@admin56.narex.com>

> From: Dennis Reinhardt [mailto:DennisR at dair.com] 
> 
> Does anyone have a Python source comment stripper or can 
> anyone recommend a parser that could be adapted for same?
> 
> Why?  ... well there is a better than 2:1 byte size reduction 
> in both raw and compressed .py files if comments are 
> stripped.  Results of analyzing dospath.py in :
> 
>                       raw Kb          compressed Kb
>                     ----------    --------------------
> original              10.44               2.98
> 4 spaces -> 1 space    8.85               2.86
> + strip comments       4.65               1.34

One question: How big is your disk page size?

-- bjorn



From leipold at acm.org  Thu Sep 26 12:39:17 2002
From: leipold at acm.org (Walt Leipold)
Date: 26 Sep 2002 09:39:17 -0700
Subject: Strange change in xmlrpclib behavior
Message-ID: <71f60588.0209260839.537943bb@posting.google.com>

Late last year, I wrote an app (in Python 2.1) that acquires and 
logs serial data.  It's a standalone app on a remote computer, but 
I wanted to send it an occasional command ("reset", "status?"), so
I included an XML-RPC server handler in its main loop, and set its
socket to nonblocking ("self.socket.setblocking(0)") so it wouldn't
slow down data acquisition.  

The code worked for months.  As of early last week, it no longer 
works.  XML-RPC requests from a remote host (except occasionally
for the very first one after the server starts) raises error 10035
in SocketServer.py ("cannot complete without blocking").  Requests
to the server from localhost work normally.

I've reduced the problem to 43 lines of server code and 18 lines 
of client code, attached below.  It's been tested on six different
machines, and all of them now misbehave.  Other socket-based code 
(like the code that logs the serial data to yet another machine) 
still works.  I don't *think* I'm going crazy, but I could swear 
this code used to work.

Any ideas?  (BTW, it's xmlrpclib v0.9.8, Win2000 SP2 and NT4 SP6.)

------------------------------------------------------------------
"No matter how thin you slice                         Walt Leipold
it, it's still baloney."                           leipold at acm.org
------------------------------------------------------------------

aserver.py:
===========
#!/usr/bin/python
import SocketServer
import xmlrpcserver
import time

class BozoServerHandler(xmlrpcserver.RequestHandler):
    def call(self,method,params):
        try:
            server_method = getattr(self,method)
        except:
            raise AttributeError,"No method named %s!" % (method,)
        return server_method(method, params)

    def status(self,method,args):
        global done,statcount,itercount
        now = time.time()
        statcount += 1
        return time.strftime("%H:%M:%S -- ",time.localtime(now)) + \
            "%d requests in %d iterations." % (statcount,itercount)

    def die(self,method,args):
        global done,statcount,itercount
        done = 1
        return "Aaargh!"

class bozoSocketServer(SocketServer.TCPServer):
    """A TCPServer with a non-blocking socket."""
    # Use the 'server_close()' method to close this server.
    def __init__(self,addr,handler):
        SocketServer.TCPServer.__init__(self,addr,handler)
        self.socket.setblocking(0)

server = bozoSocketServer(('',1234),BozoServerHandler)
done = 0
statcount = 0
itercount = 0

while not done:
    time.sleep(0.1) # Between serial i/o requests...
    itercount += 1
    # Handle a single XML-RPC request, if present.
    server.handle_request()
server.server_close()

aclient.py:
===========
#!/usr/bin/python
import sys
import xmlrpclib
import time

if len(sys.argv) != 2:
    host = "localhost"
else:
    host = sys.argv[1]
s = xmlrpclib.Server("http://%s:1234" % (host,))

try:
    for i in xrange(0,100):
        print s.status()
        time.sleep(1.5)
except KeyboardInterrupt:
    pass
print s.die()


From edwardam at interlix.com  Sun Sep 22 17:43:50 2002
From: edwardam at interlix.com (Edward Muller)
Date: 22 Sep 2002 14:43:50 -0700
Subject: Python HTTPSConnection/HTTPResponse Problem
References: <9abedeb.0209212200.5c669a50@posting.google.com> <mailman.1032691649.13632.python-list@python.org>
Message-ID: <9abedeb.0209221343.74d27ad1@posting.google.com>

I thought I had posted this already, sorry if it sorta doubles.

I can't use python 2.2, the latest I can use is python 2.1 because in
the end this will be done with zope and zope currently only supports
python 2.1.3 and earlier.

And yes, my python is SSL enabled. 

Gerhard H?ring <gerhard.haering at gmx.de> wrote in message news:<mailman.1032691649.13632.python-list at python.org>...
> * Edward Muller <edwardam at interlix.com> [2002-09-21 23:00 -0700]:
> > I am trying to use HTTPSConnection and HTTPResponse objects [...]
> 
> Is your Python SSL-enabled?  Check with:
> 
>     import socket
>     hasattr(socket, "ssl")
> 
> Which version of Python are you using? Be sure to use at least version
> 2.2, as it contains many SSL-related bugfixes.
> 
> -- Gerhard


From mertz at gnosis.cx  Mon Sep 30 01:31:27 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Mon, 30 Sep 2002 01:31:27 -0400
Subject: ANN: Gnosis Utils 1.0.4
Message-ID: <vG+l9kKkXUuf092yn@gnosis.cx>                

Subject: [Announce] Gnosis (XML) Utils 1.0.4
Date: Mon 09-30-2002

BACKGROUND:  Gnosis Utilites contains a number of Python libraries, most
(but not all) related to working with XML.  These include:

    gnosis.xml.pickle       (XML pickling of Python objects)
    gnosis.xml.objectify    (Any XML to "native" Python objects)
    gnosis.xml.validity     (Enforce validity constraints)
    gnosis.xml.indexer      (XPATH indexing of XML documents)
    gnosis.indexer          (Full-text indexing/searching)
    [...].convert.txt2html  (Convert ASCII source files to HTML)
    gnosis.util.dtd2sql     (DTD -> SQL 'CREATE TABLE' statements)
    gnosis.util.sql2dtd     (SQL query -> DTD for query results)
    gnosis.util.xml2sql     (XML -> SQL 'INSERT INTO' statements)
    gnosis.util.combinators (Combinatorial higher-order functions)
    gnosis.util.introspect  (Introspect Python objects)

...and so much more! :-)

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

This release contains a variety of minor bugfixes contributed by
users:

 * gnosis.xml.validity accidentally left out of 1.0.3
   distribution.

 * gnosis.introspect has minor improvements/corrections.

 * User contributed, but not really tested, improvements to
   gnosis.indexer.  Let me know if this breaks something (it has
   to do with reindexing working correctly).

 * Cleanup of gnosis.xml.objectify bug introduced in 1.0.3
   (problem was detecting file object rather than string or
   DOM).

 * Probably something else I forgot.

It may be obtained at:
   http://gnosis.cx/download/Gnosis_Utils-1.0.4.tar.gz

The current release is always available as:
   http://gnosis.cx/download/Gnosis_Utils-current.tar.gz

Try it out, have fun, send feedback!

David Mertz (mertz at gnosis.cx)






From amitp at Xenon.Stanford.EDU  Thu Sep  5 12:16:29 2002
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 5 Sep 2002 16:16:29 GMT
Subject: A Free Idea: Search Engine for Webpages
References: <m27kig8qw1.fsf@boiko_linux.demogr.mpg.de> <m23ct3kl3u.fsf@boiko_linux.demogr.mpg.de> <477762c2.0208260615.3790945@posting.google.com> <3D6AB55D.63D36207@engcorp.com>
Message-ID: <al800t$j17$1@usenet.Stanford.EDU>

 Peter Hansen  <peter at engcorp.com> wrote:
| Dave Brueck wrote:
| > 
| > What I find useful is Google's presentation of cached content (the one
| > you get to by clicking a search result's 'Cached' link instead of the
| > main link presented) because it highlights in a different color each
| > word from your search. It gives you an easy way to quickly and
| > visually scan a document searching for relevant content. I just scroll
| > down the page looking for concentrations of pretty colors! ;-)
| 
| I believe there is some kind of add-in for IE and maybe Netscape
| and/or others which puts a little search field in your browser
| toolbar and lets you search any given page in the same way, including
| highlighting the search terms exactly as in the cached version.

The Google Toolbar does this, but it's IE only.  In addition to the
highlighting, after you search, it shows the search words in the
toolbar and you can click on them to jump to them on the page.  There
are other toolbars now that do the same sorts of things.

Someday I'd like to implement this in Python -- not as a web server or
browser client plugin, but as a proxy.  I can parse the page as it's
coming through and insert a navigation box at the top.  It'd be handy
to see the structure of the page (h1 - h6, but in a tree), the
anchors, links to the words I last searched for.  With DOM and CSS,
you can do some fancy UI stuff to make the navigation box almost as
nice as something that's embedded in your browser.  And the most
important feature: add a button to kill the colors on any page that
uses dark blue links on a black background!

     - Amit
-- 
--
Amit J Patel, Computer Science Department, Stanford University
http://www-cs-students.stanford.edu/~amitp/


From robin at jessikat.fsnet.co.uk  Sun Sep  1 04:46:37 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 1 Sep 2002 09:46:37 +0100
Subject: ANN: New PEP Format: reStructuredText
References: <mailman.1030681307.29889.python-list@python.org>
 <mailman.1030799859.29983.python-list@python.org>
 <XpB0NIABFNc9Ewco@jessikat.fsnet.co.uk>
 <mailman.1030831309.22494.python-list@python.org>
Message-ID: <X5K05iAtPdc9Ewdt@jessikat.fsnet.co.uk>

In article <mailman.1030831309.22494.python-list at python.org>, Richard
Jones <rjones at ekit-inc.com> writes
>On Sun, 1 Sep 2002 12:22 am, Robin Becker wrote:
>> .
>> ....
>>
>> >But if it strictly stays limited to PEPs, the cost might be higher than
>> >the benefits.  It ought to aim some wider scope of applicability.
>>
>> Well I think I can say that if anyone brings the reStructuredText -->
>> pythonpoint translator to reasonable fruition then it would appear as a
>> working demo on www.reportlab.com ie paste in your text, get back your
>> pdf slides. That would be a good advert for python and reportlab.
>
>This is done in the docutils sandbox under the richard/pythonpoint directory. 
>I had this working ages ago (back in March, I think) but hadn't updated it 
>when David merged the dps and restructuredtext projects to form the docutils 
>project. That's done now.
>
>
>    Richard
>
>
So does this appear as a CGI somewhere or should I go ahead and make a
working demo?
-- 
Robin Becker


From ben at remove.Stanford.EDU  Tue Sep 10 19:03:40 2002
From: ben at remove.Stanford.EDU (Ben Escoto)
Date: Tue, 10 Sep 2002 16:03:40 -0700
Subject: Large file support and stat vs stat64 in extension modules
Message-ID: <pan.2002.09.10.16.03.39.756929.4320@remove.Stanford.EDU>

Hi, I'm writing an extension module in C and am having trouble
statting files longer than 2GB, at least under linux (Redhat 7.2, intel).
If I simply use the stat() system call, I get an EOVERFLOW error
(OSError: [Errno 75] Value too large for defined data type).  It seems
that I should be using the newer stat64() call, but:

1.  posixmodule.c appears just to use plain old stat().  I cannot figure
    out why I get errors and it works fine.  Quite a mystery.

2.  I don't know what #if defined's to put in to choose between stat and
    stat64 correctly.

Does anyone know how to correctly stat a file within an extension module?
Why does:

    struct stat st;
    res = stat("bigfile", &st)

not work for me, but in posixmodule.c the same thing (apparently) is done
and works fine?


-- 
Ben Escoto


From maurice at akst.tn.tudelft.nl  Wed Sep 18 03:11:06 2002
From: maurice at akst.tn.tudelft.nl (Maurice van de Rijzen)
Date: Wed, 18 Sep 2002 09:11:06 +0200
Subject: Getting data into an array in NumPy from an other object?
Message-ID: <am98u9$1mr$1@news.tudelft.nl>

Dear all,
I've encountered a problem.
In one object I've a function:

class blah:
	...
	...
	def GetData(self):
		return self.data


Now I want to put his data into an instance of Numeric.array.
The broblem is that I cannot find and think of any way to do this.
Does anyone knows a solution for this.

Maurice



From tjreedy at udel.edu  Thu Sep  5 00:30:30 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 05 Sep 2002 04:30:30 GMT
Subject: Python Source Code Documentation
References: <al69st$n1f$1@driftwood.ccs.carleton.ca>
Message-ID: <G5Bd9.427560$m91.16660188@bin5.nnrp.aus1.giganews.com>

"Sean Ross" <sross at connectmail.carleton.ca> wrote in message
news:al69st$n1f$1 at driftwood.ccs.carleton.ca...
> Hi. I'm interested in trying to get an understanding of Python's
source
> code.

If you mean the C source of the Python interpreter, you mostly are on
your own.  You can try groups.google to retrieve past posted
discussions.  Beyond that, the source is its documentation.   You can
probably learn quite a bit by spending an hour or two.

TJR





From hamish_lawson at yahoo.co.uk  Mon Sep 23 12:48:48 2002
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: 23 Sep 2002 09:48:48 -0700
Subject: Will "Python in a Nutshell" be too short?
References: <mailman.1032288261.15607.python-list@python.org> <LMNh9.15079$T4.1903@fe04>
Message-ID: <915a998f.0209230848.60cc1415@posting.google.com>

Steve Holden wrote:

> I wil be very surprised if Alex' book does not become the canonical
> reference for the language, since his treatment is exceptionally thorough
> and (even for him) more than usually precise.

I had no doubts about this - I'm just a greedy so-and-so who wanted to
have Alex's treatment over as wide a range of material as would fit
into a book that could still qualify to be called a nutshell! I'm
therefore delighted to learn from Alex that his book will in fact
cover a good number of the most popular third-party libraries, and
I'll therefore get to throw away a bunch of printed-out notes after
all (which is what I meant when I hoped that "Python in a Nutshell"
would be a one-stop reference). I know Alex wanted to keep the book to
about 400 pages, but for me the 600 pages that the book has grown to
is still nutshell enough and a small price to pay for the breadth of
coverage. I'm looking forward to it.


Hamish


From lutz at lutz-horn.de  Mon Sep 16 07:54:24 2002
From: lutz at lutz-horn.de (Lutz Horn)
Date: 16 Sep 2002 13:54:24 +0200
Subject: Good book choice?
References: <e6c4872c.0209150635.4816ef98@posting.google.com>
Message-ID: <wky9a2jgjz.fsf@ID-32301.user.dfncis.de>

eva_ag at eresmas.com (Eva) writes:

> I am new to Python but an experienced programmer in Fortran95 (not
> too object oriented). I am looking for a suited Python book. I'd
> like it to have clear ideas about OO, as well as an straight
> pointing into the core language.

Take a look at "Dive Into Python - Python for experienced
programmers"[1]. It's Free (as in speech) and goes straight to the
point.

Regards
Lutz

[1] http://diveintopython.org/


From sschwarzer at sschwarzer.net  Fri Sep 27 16:29:31 2002
From: sschwarzer at sschwarzer.net (Stefan Schwarzer)
Date: Fri, 27 Sep 2002 22:29:31 +0200
Subject: How do you pronounce "__" (double underscore)?
References: <525eaba.0209260956.72bf6786@posting.google.com> <3D934C5C.4000603@mxm.dk> <20020926.130938.1789366143.1524@cybermesa.com> <CnJk9.120675$Pf7.4559925@news1.west.cox.net>
Message-ID: <3D94BFAB.5060506@sschwarzer.net>

Tim Hochberg wrote:
  > "Jay O'Connor" <joconnor at cybermesa.com> wrote:
  >>How about "underder"??
  >
  > Or maybe "dunder"?

I think, you shouldn't use two words which sound very similar when spoken. If you
do this, you might run into problems when discussing code via voice (e. g.
telephone). I like "under" vs. "under-under" most: It's short enough and even
someone who hasn't heard it before, probably will guess what it means.

Stefan



From bokr at oz.net  Mon Sep  2 15:07:07 2002
From: bokr at oz.net (Bengt Richter)
Date: 2 Sep 2002 19:07:07 GMT
Subject: Help with mod_python
References: <GO6c9.22678$Lj6.623043@wagner.videotron.net>
Message-ID: <al0csr$las$0@216.39.172.122>

On Sat, 31 Aug 2002 12:54:09 -0400, "Joseph Youssef" <leader730 at hotmail.com> wrote:

>Hello, I've been trying to get python to work on apache, so I downloaded
>mod_python, after I finished the setup process, I tried to test it just like
>they indicated with the same code they showed but everytime I to run any
>python file on the server, it always gives me this error:
>
>the file I ran here was called mptest.py
>
>-----------------------------------------------
>Mod_python error: "PythonHandler mptest"
>
>Traceback (most recent call last):
>
>  File "C:\PROGRAM FILES\PYTHON\lib\mod_python\apache.py", line 181, in
>Dispatch
>    module = import_module(module_name, _req)
>
>  File "C:\PROGRAM FILES\PYTHON\lib\mod_python\apache.py", line 332, in
>import_module
>    f, p, d = imp.find_module(parts[i], path)
>
>ImportError: No module named mptest
>----------------------------------------------
>I don't know how to fix it so if you know how then please help
>
Apache is probably running as a different user than you (e.g., maybe IUSR_XXX, where XXX is your
computer name), so probably you have to move mptest.py to some directory (you may want to create
one in the tree it already knows about for web stuff) where apache can find it, or tell it where
to find stuff in your personal part of the disk (the latter is not a good idea security-wise).

There could be other problems, but that would be my first guess.

Regards,
Bengt Richter


From Roberto at lupi.an.it  Mon Sep 23 02:42:35 2002
From: Roberto at lupi.an.it (Roberto Lupi)
Date: Mon, 23 Sep 2002 06:42:35 GMT
Subject: Universally unpacking tuples
References: <158d3913.0209222233.123a9675@posting.google.com>
Message-ID: <MPG.17f8d64282bb1ac0989680@powernews.libero.it>

In article <158d3913.0209222233.123a9675 at posting.google.com>, 
peter.bittner at gmx.net says...
> The root of the problem is that I am forced to move from a
> list-solution (iterating over the elements of a list) to a
> tuple-solution (iterating over the elements of a tuple). So, what I am
> trying to do with the above code is nothing more than migrating the
> following list-oriented code to a tuple-thing:
> 
> mylist = ['abc', '123', 'qwert', ...]
> for i in range(len(mylist)):
>   elem = mylist[i]
>   print elem
> 
> 
> Does anyone know a simple solution to this problem?

mylist = ('abc', '123', 'qwert', ...)
for i in range(len(mylist)):
  elem = mylist[i]
  print elem

Or better:

for x in mylist:
  print x

-- 
Roberto Lupi


From bada at skumbad.nu  Sun Sep 22 10:36:59 2002
From: bada at skumbad.nu (harald)
Date: Sun, 22 Sep 2002 14:36:59 GMT
Subject: cursor positioning and color
References: <slrnaoqv2q.hn4.TuxTrax@fortress.tuxnet>
Message-ID: <fAkj9.756$c5.225673@newsb.telia.net>

TuxTrax wrote:
> Being new to python, I would like to know how one goes about such 
> tasks as precise cursor positioning and output to screen in color.
> 
> For instance, If I want to print a character or line, at line 10, 
> column 18, in bright white on red, how would I accomplish this?
> 
> Is there a python library specifically for setting up a text screen 
> to look and function the way you want it to, without having to use a GUI?
> 
> The SLRN screen is a good example. Obviously, text positioning and coloring
> commands are being utilized here, as well as such things as clearing
> the screen and controlled scrolling.
> 
> How would these things be done in python, and what reference material is
> available on line? I couldn't find anything in the Python manual, so any
> URL's would be appreciated

well.. there's the curses module. it's a standard module so the docs are 
in the library reference..

or if you just want to do something simple, you can just use ansi escape 
codes directly, so to 'print a character or line, at line 10, column 18, 
in bright white on red', you could write

print "\033[10;18H\033[37;41m" + string

after a quick search on google i found this reference: 
http://www.dee.ufcg.edu.br/~rrbrandt/tools/ansi.html

--
/harald              bada at skumbad.nu             http://bada.skumbad.nu/

makt leder till missbruk.



From phr-n2002b at NOSPAMnightsong.com  Mon Sep 30 00:42:21 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Sep 2002 21:42:21 -0700
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05> <an7bvd$burft$1@ID-125932.news.dfncis.de> <an7p7t$cpg$0@216.39.172.122> <mailman.1033338759.6775.python-list@python.org> <7xd6qwqtgz.fsf@ruckus.brouhaha.com> <mailman.1033352259.30842.python-list@python.org> <7xelbcgo56.fsf@ruckus.brouhaha.com> <mailman.1033357298.16341.python-list@python.org>
Message-ID: <7x65woks0i.fsf@ruckus.brouhaha.com>

"Chris Gonnerman" <chris.gonnerman at newcenturycomputers.net> writes:
> > What do you mean by "right"?  What do you mean by "all 
> > cases"?
> 
> The original subject is *business* arithmetic.  There is no
> choice but to use decimal arithmetic if you want your math
> results to match good-old-fashioned pen and paper math.

No, I'm still not convinced.  Is business arithmetic well-defined
enough that you can never do the same calculation two ways and get
different results?

Example: bunches are three for a dollar.  How much do you pay if you
buy nine bananas?

    Answer #1:
      price_per_banana = 1.00 / 3
      number_of_bananas = 9
      total_price = number_of_bananas * price_per_banana
      print total_price

    Answer #2:
      price_per_bunch = 1.00   # three bananas
      number_of_bunches = 3
      total_price = number_of_bunches * price_per_banana
      print total_price

With floating point and rounding on output, you get the same answer
both ways.

With decimal arithmetic, you get $2.97 the first way and $3.00 the
second way.

For "business arithmetic" to always give the same answer, it has to
forbid at least one of those methods of doing the problem.



From borcis at removethis.infomaniak.andthis.ch  Sat Sep 21 08:40:44 2002
From: borcis at removethis.infomaniak.andthis.ch (Boris)
Date: Sat, 21 Sep 2002 14:40:44 +0200
Subject: filter(lambda W : W not in 'ILLITERATE','BULLSHIT') (Re : python expression)
References: <d8f8142b.0209161913.12cce045@posting.google.com> <vWBh9.26065$Gp1.687517@weber.videotron.net> <3D86FF01.FE0CECC5@geneva-link.ch> <3d87069c.34652637@news.rcn.com> <3D871316.AA1E8FFC@geneva-link.ch> <3d882cd4.9181792@news.rcn.com> <3f74f974.0209180713.27889f74@posting.google.com> <l9ehouci1ga88ub2csdam1qroiik0ogd4r@4ax.com> <3d88c258$0$707$5402220f@news.sunrise.ch> <v4lhouolv1c0ijmihfom94j0lor2lhntli@4ax.com> <3d89be92$0$707$5402220f@news.sunrise.ch> <s6mjous6o5abg3icq12u156v5tijhcnk7h@4ax.com> <3d89ee85$0$712$5402220f@news.sunrise.ch> <B8pi9.149307$5r1.6007886@bin5.nnrp.aus1.giganews.com> <3d8b4498$0$717$5402220f@news.sunrise.ch> <WSLi9.165658$z91.6788612@bin3.nnrp.aus1.giganews.com>
Message-ID: <3d8c6982$0$719$5402220f@news.sunrise.ch>

Terry Reedy <tjreedy at udel.edu> a ?crit dans le message :
WSLi9.165658$z91.6788612 at bin3.nnrp.aus1.giganews.com...
>
> "Boris"
> >
> > filter(lambda W : W not in 'ILLITERATE','BULLSHIT')
>
> This somewhat-well-known and certainly well-worn programmer joke

Well, *I* invented this formula myself soon after Mr Bush revealed to
the world that "USA" is written in "CRUSADE" just like "US" is in
"BUSH". Adding : "Either you are with US, or you are against US".

It's not a joke, it's a comment on being called to participate to a folly
that is given no other justification than allusions to kindergarten
references that are specific to the US as a pretext to eg licking boots
to the - militarily speaking - most powerful man in the world, and
- intellectually speaking - most dangerous man in the world for a
man in that position.

The full form of it I have never published, and it is :

filter(lambda I : I not in 'USA','CRUSADE'
                        )==(
'C' + filter(lambda I : I not in 'U','RUDE')
                       ) and (
filter(lambda W : W not in 'ILLITERATE','BULLSHIT'))

Since I invented it, I know exactly how much I distributed it myself,
and your I-am-expert-on-the-matter tone falls flat. I would rejoice
on the implication that it has been more widely distributed that
seems to stem from your words, except that, well, there is some
ground for me to doubt your good faith on that point.

Noticed BTW how the typhoon that recently befell Corea was
called RUSA ? US, BUSH, BULLSHIT ; USA, RUSA, CRUSADE

Speak of butterflies...

> functional programming expression, borrowed from Lisp, etc, that
> happens to have been written this time with Python syntax in
> preference to the Lisp or other FP syntax that could have been used
> instead.  (A good choice for presentation to non-programmers, but
> that's another issue.)  It has nothing to do with Python in
> particular.

That's about as precise as saying the following other semi-joke of mine
is base-dependent :

"12 ? - 12 is the least integer symbolizing the natural integers
 by itself. Its successor is 123"

You know as well as I do that although it's clear that the computation
implied by filter(lambda W : W not in "ILLITERATE","BULLSHIT")
can be easily translated to another language, its more direct reading
highly depends on a layout and details of sequence behavior that
python provides and for which it is not a priori clear that immediate
equivalents are available in other programming languages.

> One could just as well justify cross-posting to lisp and

That would be true if the equivalent lisp one-liner existed, and *I* know
I never wrote a translation. Why don't you prove what you say by
providing one ?

But in any case you are probably right on one point : the
right way to advertise the thing is to open a competition
for translations into other programming languages. Then
we could actually measure the material ground of your
easy assertions.

> other functional programming groups, but please don't.
>
> > > and should never have been added.
> >
> > That's a dogmatic opinion that I believe lacking in justification.
>
> I base my opinion (which I still hold) on the complete lack of any
> particular relevance to Python, the computer programming language

I notice that you carefully avoided adressing my point on this
matter. What I pointed out in the message you fake responding to
[in a way, if I may say so, itself depicted in spirit by my formula]
is that the irony in filter(lambda W : W not in "ILLITERATE","BULLSHIT")
is much enhanced when having the technical details of python (such as
is responsibility of experts) put aside in the same manner that the detailed
machinery of Prolog (as experts come to intuit it) is put aside when
contemplating the first-order logic reading of the Prolog code.

Now you answer "us" with a batch of non-sensical Lisp context.
Yes, sure, lambda and filter are found in lisp. The real question is
whether lisp allows to write it down in a form that doesn't give
away that the implementation operates from a differently structured
parse tree. Thus with minimal uncamouflaged clutter between
ILLITERATE and BULLSHIT : no best inquisition rule.

> and
> the presence of very un-Pythonic ugliness

Ah, you are an aesthetician of pythonicity - will you deny the title as
ugly,
pythonically speaking ? But then, what's your competence to formulate
such assertions ?

Okay, say we'd look into the virtues and vices of Python as a programming
language to describe cool and simple algorithms fit to form the flesh of the
code in a file of Python featuring ten lines such as :

from battle.devices.russianRoulette import QuantumRussianRoulette

Class DieteticWar(War) :
       def __init__(self,Democracy,Tyrant) :
            assert self.casusBelli.provided(Tyrant), RelevanceException
            assert self.ultimatum.provided(Democracy), CenturyException

What about it ? Perhaps Dylan is better, perhaps not. The real question
is how much time do we have before the time bandits get it their way

<not-python-related>

no it's more clearly that UN has to check grammatical well-formedness of
any ultimatum Mr Bush submits in the name of furthering the UN's
own intentions. Grammatical well formedness of ultimata implies that
simply saying "I will enter war against you in N days" in camouflaged
language doesn't cut it. Such a language is not an ultimatum, it's a war
ruse directed against the more naive public.

</not-python-related>

Since it is in no way the UN's role to promote war ruses, the UN
should simply force US ultimata to obey ISO 9000-type standards
certification procedures.

I-ll-be-back-unless-nazis-kill-me-in-the-meantime-ly yours,

Boris
--
"W is a mass-creating boson. Mass makes gravity. What do superstrings say ?"




From peter at engcorp.com  Fri Sep 13 21:28:32 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 13 Sep 2002 21:28:32 -0400
Subject: transform list of int in a list of string??
References: <mailman.1031945716.2586.python-list@python.org>
Message-ID: <3d8290af$1@news.sentex.net>

jubafre at brturbo.com wrote:
> my program have other modules imported, like re and sre and i put a convertion int to str but doesn?t work
> t=str(x[i])
> Traceback (most recent call last):
>   File "C:\Documents and Settings\jubafre\Desktop\montador\teste.py", line 78, in ?
>     t=str(x[i])
> TypeError: 'list' object is not callable

This error is most likely because you created a list object
bound to the name "str".  Do *not* use "str" or other such
reserved names as variable names if you want to preserve your
sanity. :-)

> Why i can?t transform a int number in string in my program???
> I tryed with import string, and can?t too, 
> but if i start a new python shell it works, why??

Because when you restarted Python the new session did not have
the variable called "str".

>>>>x=[1,2,3,4,]
>>>>s=[]
>>>>for i in range(len(x)):
> 	t=str(x[i])
> 	s.append(t)
>>>>print s
> ['1', '2', '3', '4']

By the way, although this code works, it is not at all what we
would call "Pythonic".  It looks like code from some other
language translated to Python.  For example, in Python you
don't need to (and rarely want to) use index variables to
iterate through a list like that: you can directly iterate
on the list items like this:

 >>> x = [1, 2, 3, 4]
 >>> s = []
 >>> for item in x:
...     s.append(str(item))
...
 >>> print s
['1', '2', '3', '4']

Note though that there are two tools for doing this kind of
thing much more easily, as shown in the postings by Skip Montanaro
or Erik Max Francis.  The "map" method is older but has advantages
in certain circumstances, while the other method is called a
"list comprehension" and can be both more readable and faster,
again depending on circumstances.

In general, there are a lot of Python idioms which work in a
similar fashion, going in ways traditional languages have not.
They're well worth watching for and learning.  (But your code
has one strong point in its favour, for a newbie: it works! :-)

-Peter



From bdesth at nospam.free.fr  Mon Sep 16 22:12:50 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Mon, 16 Sep 2002 22:12:50 -0400
Subject: Book for win32?
References: <am430k$9l3$03$1@news.t-online.com> <uhegq710h.fsf@online.no>
Message-ID: <3D868FA2.4040503@nospam.free.fr>

Syver Enstad wrote:
> Thomas Guettler <zopestoller at thomas-guettler.de> writes:
> 
> 
>>Hi!
>>
>>I am interested in book about Python on win32.
>>
>>I found one from o'reilly: "Python on win32"
>>
>>Are there better books or alternatives?
>>
>>I am *not* looking for a python introduction, I am
>>programming more than one year with python, but up to
>>now mostly on unix.
>>
>>Now I am intersted in win32 stuff:
>>  - Opening a file-dialog
>>  - COM
>>  - Accessing the registry
>>  - Maybe GUI Programming
> 
> 
> Then you can't go wrong with the o'reilly book. It is a most excellent
> tome for precisely the subjects that you outline there. 
> 
> Be aware that the book is written for Python 1.5 so a couple of
> things, like the need to convert unicode strings passed in by COM
> objects are outdated. As you already have been programming python for
> over year this shouldn't be a problem.
>  
> 

I'd just add one advice : if you're not familiar with the w32 API, go 
and get yourself a good reference of it ! (I think you can get the win32 
API ref on borland's site, with the free version of bcc55).

Laotseu



From ajs at ix.netcom.com  Thu Sep 19 14:27:25 2002
From: ajs at ix.netcom.com (Arthur)
Date: Thu, 19 Sep 2002 14:27:25 -0400
Subject: Python sequences by reference - how to make clear
Message-ID: <00d001c2600a$348dc180$9865fea9@arthur>

   >Yes, this is fully in accordance with how the Python language should
behave:
   >    Res += [rec] inserts references to list object rec,
   >    which are further modified... (he should use copy(rec) instead).

Yes, this issue in various guises bit me for some time after I began using
Python (self.-taught, as a fist programming language).

It had always seemed to me that if there were a copy operator, or
(perferably, to me) that the functionality of the copy module were in the
language core as a built_in, that I would have been on clearer notice that
there was something fundamental in this
area to which I needed to pay attention and become attuned.

In looking into it I did find some discussion of the issue on python-list.
Interestingly it was also math/Numeric folk who seemed to be rasing the
question then and advocating something along the lines of a built_in
operator for copying.  I however think that the issue is quite real for
anyone coming to Python without the benefit of either a programming
background or a formal structure within which they are being taught the
language.

I have felt strongly that others have been too quick to advocate changes to
the language to address "special needs" cases. And that the special needs
cases of the novice have been over-emphasized in discussions, in general.

And I have confirmed to myself in the work I have been doing recently, which
is specifcally addressed to the non-programmer, that those "special needs"
can most often be quite adequately addressed by the right interface design,
using Python's out of the box capabilities..

I admittedly have little understanding of the full implications of a
suggestion to make the copy module capabilities built_in.
But I do believe, all things else being equal (which I suspect they are
not), it would be quite helpful.

Art





From gumuz at looze.net  Thu Sep 12 04:54:40 2002
From: gumuz at looze.net (Gumuz)
Date: Thu, 12 Sep 2002 10:54:40 +0200
Subject: class variable won't icrement!
Message-ID: <3d8054ef$0$234$4d4ebb8e@news.nl.uu.net>

Hello everyone,

I have something very strange and i can not understand it. probably, it's
not _that_ strange : )

I have this simple class which registers itself on a class variable in the
following code:

----------------------------------------------------------------------------
------
class test:
    dict = {}
    id = 1
    def __init__(self, name):
        self.dict[name] = self
        self.id = self.id + 1

    def show(self):
        print "class variable dict element count: ", str(len(self.dict))
        print "class variable id: ", self.id
----------------------------------------------------------------------------
--------

now, look at the following:

>>> john = test("john")
>>> frank = test("frank")
>>> bill = test("bill")
>>> bill.show()
class variable dict element count:  3
class variable id:  2
>>> frank.show()
class variable dict element count:  3
class variable id:  2

The dict variable acts fine, maintaining it's content throughout every
instance of the class.
The id variable however, seems to be only incremented within the instance.
In the end this variable should've been 4.

Could someone explain this to me? Is there something about class variables
whta i should know?

many thanx,
Guyon






From gerhard.haering at gmx.de  Fri Sep 13 15:03:52 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 13 Sep 2002 19:03:52 GMT
Subject: Problem with HTTPS request
References: <uo2fveepf88175@corp.supernews.com> <b0f083db.0209130558.2aa60398@posting.google.com> <uo45ara0l62cdc@corp.supernews.com>
Message-ID: <slrnao4dkn.288.gerhard.haering@lilith.my-fqdn.de>

Christian Reyes wrote in comp.lang.python:
> I'm on Win 98.
> i don't have a development env in which to build the python socket module.
> was hopoing someone could point me to somewhere i could find a premade one.
> 
> Thanks though,
> if you could point me to a site or two, that'd be great'

Please don't top-post.
http://www.cs.tut.fi/~jkorpela/usenet/brox.html
http://www.uwasa.fi/~ts/http/quote.html

> Jeremy Hylton <jeremy at alum.mit.edu> wrote in message
> news:b0f083db.0209130558.2aa60398 at posting.google.com...
>> "Christian Reyes" <creyes at ea.com> wrote in message
> news:<uo2fveepf88175 at corp.supernews.com>...
^^

Broken quoting.

>> > [...] Note: HTTPS support is only available if the socket module was compiled
> with
^^

Broken quoting again. Please throw away that OE crapware or fix it
using Morver or whatever.

Now to answer your question: I've compiled a socket.pyd for Python
2.2.x with SSL support. You can download it at my homepage:
http://www.cs.fhm.edu/~ifw00065/pyssl/

-- Gerhard


From gua81 at XXXyahoo.com  Fri Sep 27 04:46:29 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Fri, 27 Sep 2002 18:46:29 +1000
Subject: Making directory in cgi!
References: <an15gg$s3h$1@lust.ihug.co.nz>
Message-ID: <an15s8$slv$1@lust.ihug.co.nz>

and one more thing......

what's going to happen if let say, the directory simpsons already exist
inside homer.
What will happen to the files inside simpsons???

Thanks.

"CheapSkate" <gua81 at XXXyahoo.com> wrote in message
news:an15gg$s3h$1 at lust.ihug.co.nz...
> HI,
> let say my createdir.cgi is in
> /bug/ger/createdir.cgi
>  and doing ls in /bug/ger/ will list out
> .
> ..
> (file)createdir.cgi
> (dir)homer
> how do I create directory inside homer?
> like homer/simpsons/
> ???
> what are the library that I need to import?
> Thanks......
> (I know this is a simple Q, so please excuse me... im a newbiee)




From donn at u.washington.edu  Wed Sep 11 12:04:47 2002
From: donn at u.washington.edu (Donn Cave)
Date: 11 Sep 2002 16:04:47 GMT
Subject: blocking socket dilemma
References: <3d7f046f$0$230$4d4ebb8e@news.nl.uu.net>
Message-ID: <alnpiv$8fc$1@nntp6.u.washington.edu>

Quoth "Gumuz" <gumuz at looze.net>:
| I am a little bit stuck with my design (again!).
|
| I have this server class, which creates a session object for every accepted
| incoming connection. The session object has a RequestHandler function which
| is then run in a new thread.
|
| all fine, but...
|
| This RequestHandler function actually does a infinite loop. The loop wil
| receive data and take some action and it will send data which is piled up in
| this session's queue. The problem that i'm having is that the loop will hang
| on socket.recv if the client is not sending anything. Ofcourse this is very
| obvious and logical, but i can't think of another way of constructing my
| design, so that this problem doesn't occur.

If it helps, you can use select.select() to determine whether the
device has more input ready, immediately or with a timeout at your
discretion.

	Donn Cave, donn at u.washington.edu


From andrewm at object-craft.com.au  Wed Sep 18 01:08:48 2002
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Wed, 18 Sep 2002 15:08:48 +1000
Subject: How to make a C extension module backwards compatible? 
In-Reply-To: Message from martin@v.loewis.de (Martin v. Loewis) 
   of "18 Sep 2002 06:47:43 +0200." <m3sn07hpjk.fsf@mira.informatik.hu-berlin.de> 
References: <mailman.1032319522.8608.python-list@python.org>  <m3sn07hpjk.fsf@mira.informatik.hu-berlin.de> 
Message-ID: <20020918050848.1CB6E3C21B@coffee.object-craft.com.au>

>> Is there any community wisdom on how to make a C extension module that
>> implements a new-style class (type?) backwards compatible (albeit,
>> at the cost of some functionality)?
>> 
>> Put another way, how *should* I make a class implemented in C behave
>> like a new-style class when installed with python 2.2 and above, and
>> like an old style class when installed with older pythons?
>
>What properties of a new-style class are you interested in? If you
>implement a plain type, it will work on all Python versions, no need
>to change anything for Python 2.2.

That's a good question - this is more of a learning exercise than a
project with a real need. I guess the most useful attribute would be
sub-classing.

We have an existing module that implements a fast csv parser[1] - if
it's easy to make it's "parser" class able to be sub-classed, it would
be worth doing it (I guess the socket module is a decent example).

[1] http://www.object-craft.com.au/projects/csv/

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/



From tjreedy at udel.edu  Thu Sep 12 09:36:08 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 12 Sep 2002 13:36:08 GMT
Subject: [ANN] PySQLite 0.3.0 released
References: <slrnao0e6p.23c.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <cL0g9.335846$On.14066007@bin3.nnrp.aus1.giganews.com>

"Gerhard H?ring" <gerhard.haering at opus-gmbh.net> wrote in message
news:slrnao0e6p.23c.gerhard.haering at haering.opus-gmbh.net...
>   A Python-DB API 2.0 compliant extension module that connects to
>    SQLite databases.
...
>     row = cursor.fetchone()
>
>     while row != None:
>             print "%14s, %15s, %19s, %8s, %25s" % tuple(row)
>             row = cursor.fetchone()

Does anyone know whether there are any plans to update the Python-DB
API to the new iterator protocols, so that the above could be written
more simply something like

for row in cursor:
       print "%14s, %15s, %19s, %8s, %25s" % tuple(row)
?

Terry J. Reedy





From kingprad at mail.com  Mon Sep  9 10:52:10 2002
From: kingprad at mail.com (Carl)
Date: 9 Sep 2002 07:52:10 -0700
Subject: web crawler help?
References: <JARd9.6927$yt3.3340577@newssrv26.news.prodigy.com>
Message-ID: <668b76ce.0209090652.4b074a1b@posting.google.com>

I just wanted to say also that a lot of sites have a robots.txt file
in the root directory with a list of pages the crawler shouldn't troll
through. it's polite to honor it if you're grabbing tons of pages from
a server. Probably fine to ignore if you're not using a lot of server
time and only doing a few simple tasks.


From timr at probo.com  Sat Sep 14 01:58:43 2002
From: timr at probo.com (Tim Roberts)
Date: Fri, 13 Sep 2002 22:58:43 -0700
Subject: Question regarding mysqldb module and variable table names
References: <45ac92e5.0209131920.23b1c22a@posting.google.com>
Message-ID: <eqj5ou49igunaavcg75uv7bkd80b1m2cmt@4ax.com>

ezcoder at mailcity.com (Glenn Scott) wrote:
>
>I want to change it so that the "mytable" value in this example is a
>variable rather than a constant.  So the code would look like this:
>
> c.execute("""INSERT INTO %s \
>        SET playerkey=%s, link=%s, snippet=%s, time_created=%s, \
>        last_seen=%s, relevancy=%s, doc_title=%s """, \
>         (infoTable, playerKey, urlLink, snippetText,
>self.getCurrentTime(), self.getCurrentTime(),\
>         noOfMentions, re.escape(str(self.doc_title))) )
>
>But this doesn't work...for whatever reason, using "%s" in an
>assignment SET variable works, but as a table name it doesn't.
>
>I suppose I can use the "%" string formatter at the end of the
>triplq-quotes to fix this problem, but I'll have to jump through other
>hoops to deal with oddness with my data that is taken care of for me
>by using the "," alternative.

Not at all.  This should work:

 c.execute("""INSERT INTO %s \
        SET playerkey=%%s, link=%%s, snippet=%%s, time_created=%%s, \
        last_seen=%%s, relevancy=%%s, doc_title=%%s """ % infoTable, \
         (playerKey, urlLink, snippetText,
self.getCurrentTime(), self.getCurrentTime(),\
         noOfMentions, re.escape(str(self.doc_title))) )

Do you realize you do not need to escape the ends of lines when you use
triple quotes, nor outside the quotes when you are within parentheses?

 c.execute("""INSERT INTO %s 
        SET playerkey=%%s, link=%%s, snippet=%%s, time_created=%%s,
        last_seen=%%s, relevancy=%%s, doc_title=%%s """ % infoTable,
         (playerKey, urlLink, snippetText,
          self.getCurrentTime(), self.getCurrentTime(),
          noOfMentions, re.escape(str(self.doc_title))) )

If ''' are triple quotes, aren't """ sextuple quotes?
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From pedronis at bluewin.ch  Wed Sep 18 03:19:23 2002
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Wed, 18 Sep 2002 09:19:23 +0200
Subject: Calling Python from C# - please help
References: <6de16ac9.0209130508.5d1b56c6@posting.google.com> <j4elbyt3yw.fsf@informatik.hu-berlin.de> <ur8fyxau0.fsf@online.no> <0iCg9.2350$Ee4.6516@news-server.bigpond.net.au> <3D82FA5B.3050003@mxm.dk> <uptvgvn1h.fsf@online.no> <3%Dh9.15217$Ee4.45118@news-server.bigpond.net.au> <3d877db8_5@news.bluewin.ch> <IpSh9.17251$Ee4.52376@news-server.bigpond.net.au>
Message-ID: <3d882a59_4@news.bluewin.ch>

Mark Hammond <mhammond at skippinet.com.au> wrote in message
IpSh9.17251$Ee4.52376 at news-server.bigpond.net.au...
> Samuele Pedroni wrote:
>
> > How were you going to support eval and exec?
>
> The compiler itself is written in CPython.  The intention was to get the
> compiler good enough to compile itself - then we will have a .NET
> compiler able to be built into the Python runtime.  exec and eval would
> leverage this.
>
> > For the moment the unload granularity of .NET is AppDomains only and
only
> > upon explicit request, and types/modules/assemblies are not garbage
> > collected but simply live as much as the AppDomain,
>
> App Domains can be dynamically created tho, which may help.

that's also the workaround MS proposes to the people that encounter the
issue.
But

> > that's OK for static languages, or environments with a clear run/design
time
> > distinction (one can setup an AppDomain and tear it down for each
> > run-cycle), or for running scripts in isolation but not if one wants
IL/JIT
> > speed for a enviroment that allows for redefinitions and eval, ...
> >
> > In Jython we create Java bytecode and dynamically load classes for all
the
> > code, also eval code because in Java classes are elegible for garbage
> > collection.
> >
> > In .NET it seems one needs a pure interpreter that does not compile to
IL
> > for exec and eval support, and that seems what JScript does for jsc
compiled
> > code containing eval, that means the script is compiled but the
evaluated
> > code is only interpreted,
>
> Im not really convinced of that.  Reflection::Emit has IL generation
> capabilities purely for "dynamic" code - the example used in their docs
> is that regex engines could compile down to IL on-the-fly.

Indeed, that's also what they allow for their regex impl, but if one reads
the fine points about regex

http://msdn.microsoft.com/library/default.asp?url=/library/en-us/cpguide/htm
l/cpconcompilationreuse.asp

"However, generated MSIL cannot be unloaded. The only way to unload code is
to unload an entire application domain ..."
... Out of mem danger.

> > otherwise a long running process using eval could go out of memory.
> >
> > That's something people wanting to write a feature complete impl of
Pyhton
> > for .NET should consider or wait MS to implement a more
> > dynamic-environment-friendly unload policy.
>
> Yes, I see your point, but AppDomains are really quite powerful.  I
> believe there are no restrictions between cross domain calls.

but that means Remoting ...

>  Managing
> the lifetime of the dynamically created domains may be an issue though.

I fear yes, also given that the reload unit of Python is a module, one would
need one AppDomain for each module,exec unit
etc produced code and then a way to manage the lifetime of all this
distributed-across-AppDoms stuff.
A lot of headaches for what externally seems just a very shortsighted design
decision on their part.

Otherwise, as I said, there's the interpreter route, which is JScript eval
approach...

regards.




From jb at yahoo.de  Thu Sep 26 00:36:25 2002
From: jb at yahoo.de (JB)
Date: Thu, 26 Sep 2002 06:36:25 +0200
Subject: Python on a Novell server
Message-ID: <3d928cb5_6@news.newsgroups.com>

We have a Novell net at our school and are trying to install 
Python 2.2 on a Novell server. This does not work at the 
moment as we get the error message "python.dll not found" 
at the moment.

Is it possible to install Python a a server or has it to be 
installed locally on each workstation?

-- 
Janos Blazi


-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From loewis at informatik.hu-berlin.de  Sun Sep 15 14:01:53 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 15 Sep 2002 20:01:53 +0200
Subject: Python11
References: <xwMg9.1433$Dy4.464@nwrddc03.gnilink.net> <m3admk47pf.fsf@mira.informatik.hu-berlin.de> <o2Qg9.53$uo5.66284@news.uswest.net>
Message-ID: <j4bs6zrv1q.fsf@informatik.hu-berlin.de>

"Kevin Altis" <altis at semi-retired.com> writes:

> Is Portland, OR the official location, I didn't think that had been decided
> yet? 

Might indeed still be undecided. I found it on the O'Reilly page
(somewhere), so I'd consider it semi-official.

Regards,
Martin


From chrisabraham at yahoo.com  Thu Sep 26 14:19:23 2002
From: chrisabraham at yahoo.com (Chris Abraham)
Date: Thu, 26 Sep 2002 11:19:23 -0700 (PDT)
Subject: Dedicated Server Recommendations?
In-Reply-To: <JHEOKEOOLIGLDHCMAHMOMENEEBAA.marklists@mceahern.com>
Message-ID: <20020926181923.18596.qmail@web20703.mail.yahoo.com>

Hey all:

I would like to ask you all if you could recommend some inexpensive-but-good
dedicated server hosts that I could hire to run
Zope/Pyhton-CGI/PHP4/MySQL/PostgreSQL/IMAP?

I always prefer to ask Open Source people for this kind of thing because in
general, pricepoint is very important.

Thanks in Advance!

Chris

=====
-- 
chris abraham <chrisabraham at yahoo.com> +1 202-452-7442 tel
po box 15163 washington, dc 20003-0163 +1 202-478-0276 fax



From padraig at linux.ie  Sun Sep 15 15:57:27 2002
From: padraig at linux.ie (Padraig Brady)
Date: Sun, 15 Sep 2002 19:57:27 GMT
Subject: python performance
Message-ID: <3D84E66F.6020701@linux.ie>

I was wondering about the performance characteristics
of python and ran a simple test. The 2 programs
below are functionally equivalent and just read
in fields into a list. file.fields contains
720 fields in each of 405 lines of the form of
repeating: oneoneone twotwo "thre e three"

The time to run is shown above each program
from which I've inferred the following:

1. The function call version is (6.3%) faster because
    the cumulative cost of parsing the simpler expressions
    and function call overhead is smaller than parsing the
    1 single complex expression? Or the function is parsed
    only once and doesn't have to be reparsed. This would
    suggest that top level code is parsed for each iteration?

2. Anyway I thought that parsing affects would be removed by doing the
    parsing only once, i.e. compiling the code to .pyc (I used
    py_compile.compile()). However this makes no difference at all?
    Surely compiling is not just for code obfuscation.

Note I did do the test several times and averaged the results.

cheers,
P?draig.

----------------------------
2.514s
----------------------------
#!/usr/bin/env python2.2
import re

reFieldFinder = re.compile('[^ "]+|"[^"]+"') #unquoted|quoted
def getFields(line):
     fields = reFieldFinder.findall(line)
     return [field.replace('"', '') for field in fields]

for line in open("file.fields").readlines():
     listLine = getFields(line[:-1])

----------------------------
2.672s
----------------------------
#!/usr/bin/env python2.2
import re

reFieldFinder = re.compile('[^ "]+|"[^"]+"') #unquoted|quoted
for line in open("file.fields").readlines():
     listLine = [field.replace('"', '') for field in 
reFieldFinder.findall(line[:-1])]



From dsavitsk at e-coli.net  Wed Sep 18 02:53:41 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Wed, 18 Sep 2002 06:53:41 GMT
Subject: Win32COM & Detaching from automation servers
References: <ef3269df.0209171441.2d6f8e59@posting.google.com>
Message-ID: <VpVh9.10348$yt3.4873797@newssrv26.news.prodigy.com>

"Brian" <codeapocalypse at msn.com> wrote in message
news:ef3269df.0209171441.2d6f8e59 at posting.google.com...
> I've searched this newsgroup for an answer to my problem, but haven't
> come up with anything.  (Perhaps the solution is obvious and I'm just
> not seeing the forest for the trees...)
>
> I want to write a Python program on Windows which will use the
> automation interface for Excel.  What I want it to do is launch Excel,
> open a template, stuff data into it (based on command line parameters
> to the Python app), detach from Excel and exit, leaving Excel open.
>
> I know how to do everything except for the detach part.  Any help is
> much appreciated.

I don't know the real answer (or if there is one other than it seems to work
from Word and not from Excel), but one solution might be to save the
generated file, close it, and call WinExec() or its ilk to reopen the file
visibly.

-doug




From peter at engcorp.com  Mon Sep  9 08:01:36 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 09 Sep 2002 08:01:36 -0400
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <bd751532.0209082142.54236307@posting.google.com>
Message-ID: <3d7c8d8a@news.sentex.net>

Brian Slesinsky wrote:
> significant.  Long-term Python programmers tend to forget how weird it
> is to some newcomers to end a function or class just by *stopping*. 

I don't forget how weird it was to me as a newcomer.  I also remember,
however, that about fifteen minutes after writing *real* code with it,
I realized it was not weird at all, but very sensible and readable.

> I also think there would be a certain satisfaction in having a proper
> ending, even for folks who are entirely comfortable with Python's
> indentation.  It's like "Q.E.D." at the end of a proof, or "The End"
> at the end of children's book, or signing a message even though the
> "From" header line makes it entirely redundant.

There's some truth in that.  I often end my files with a line that
reads "# EOF", which helps me not try to scroll down to find more
code, and when I print the file (rare) I can be sure I've found the
last page of code.

On the other hand, I always put "// EOF" in my C++ and Java code,
and /* EOF */ in my C code, so this bit has nothing to do with
block structures.

And I'd regard a strong desire to do something similar after
every single block as an unnatural compulsion.  Therapy would be
a good next step for people with this problem.  <0.87 wink>

-Peter



From mhuening at zedat.fu-berlin.de  Fri Sep 13 02:44:39 2002
From: mhuening at zedat.fu-berlin.de (Matthias Huening)
Date: 13 Sep 2002 06:44:39 GMT
Subject: [Tutor] Paradox  table
References: <mailman.1031848942.4346.python-list@python.org>
Message-ID: <Xns928858F503770mhueningzedatfuberli@130.133.1.4>

montana <sarmxiii at knology.net> wrote in 
news:mailman.1031848942.4346.python-list at python.org:

> Is there a good example on how to use a CSV parser in Python?
> 

You could try one of these modules:
http://tratt.net/laurie/python/asv/
http://www.object-craft.com.au/projects/csv/

Hope this helps, Matthias


From Oschler at earthlink.net  Mon Sep  2 08:24:21 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Mon, 2 Sep 2002 08:24:21 -0400
Subject: Must have Python modules?
References: <_THc9.42153$vY2.1039887@e3500-atl2.usenetserver.com> <slrnan6k4s.eg.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <FCIc9.30407$FJ1.5585669@e3500-atl1.usenetserver.com>

"Gerhard H?ring" <gerhard.haering at opus-gmbh.net> wrote in message
news:slrnan6k4s.eg.gerhard.haering at haering.opus-gmbh.net...
> Robert Oschler wrote:
>
> What do you want to hear? Everyone's list of favourite Python extensions?
> There's likely is more than one for most of us. My favourite is the "this"
> module.
>

Gerhard,

In C++, which has thousands of libraries, snippets, etc. there were only a
few that I personally used religiously.  For example, the Boost Smart
Pointer library.  In Delphi the Jedi libray is a must-have library with a
very broad coverage of functions.  Most programmers I talk to, regardless of
the language, have a few libraries or [beans/components/plug-ins] which they
have fallen in love with and use in almost every, if not every, project.  It
is that uber-list I want to hear about.  I went to freshmeat.net and there
are over 695 Python entries.  Scanning every project and then figuring out
how useful they are would be fun but probably not very productive.  In
python I already have two that I'm very find of, Secret Labs RegExp module
and the xml.dom module.  (And the wonderful WingIDE environment for Python
development!)  I want to know what the other favorites are "out there".

thx






From bruce_deletethis_dawson at cygnus-software.com  Thu Sep 26 02:09:25 2002
From: bruce_deletethis_dawson at cygnus-software.com (Bruce Dawson)
Date: Thu, 26 Sep 2002 06:09:25 GMT
Subject: SET_LINENO
Message-ID: <3D92A46E.2030706@cygnus-software.com>

Executive summary: Has the SET_LINENO removal been tested with the HAP 
debugger, and wouldn't a breakpoint instruction give much larger benefits?


As a partial author of the HAP Python Debugger (available on source 
forge) I read about the imminent removal of SET_LINENO with some worry. 
It *sounds* like it has been done in such way to not cause debugger 
problems, but I'm curious if anyone has tested HAP with 2.3.

Also, it sounds like the reason for removing SET_LINENO was to improve 
the performance of Python scripts. However one of the posts on this 
topic (can't find it now) suggested that the hoped for improvements had 
not actually materialized.

It seems to me (from the sidelines/peanut gallery/wherever) that the 
biggest performance problems in the SET_LINENO area are when debugging, 
not when running free, and that therefore this optimization is targeting 
the lesser problem.

The need to check for breakpoints at each line has always made Python 
debugging performance poor at best. Technically the performance under 
the debugger doesn't matter, and yet sometimes it does. Having just 
helped complete a Python based sports game for PC, where real-time 
performance was critical, I can say that the slowdown in HAP was 
definitely annoying.

So, wouldn't a break-point opcode be a more substantial improvement than 
removing SET_LINENO? Just bringing up that old topic again to see if it 
gets any more traction :-)

Bruce Dawson
Humongous Entertainment

P.S. For reference I include the canonical? link to the SET_LINENO 
discussion.

http://mail.python.org/pipermail/python-dev/2000-July/007652.html

P.P.S. A breakpoint instruction would also be a boon for asserts - I'd 
love to be able to stop on the assert when running under a debugger, to 
see what is wrong before the stack frame unwinds.



From ehagemann at comcast.net  Tue Sep 24 21:47:05 2002
From: ehagemann at comcast.net (Eric Hagemann)
Date: Wed, 25 Sep 2002 01:47:05 GMT
Subject: help with print function in extension
Message-ID: <tA8k9.270440$AR1.11448440@bin2.nnrp.aus1.giganews.com>

I'm handcrafting an extension in C

If have defined both a str() and a repr() function and when I execute either
repr(obj), or str(obj) I get what I expect

however if I execute 'print obj' I get the following error

Traceback (most recent call last):
  File "<stdin>", line 1, in ?
SystemError: error return without exception set

although 'print str(obj)' works like I think it should

Am I mistaken in thinking that  'print str(obj)' should be the same as
'print obj'

Cheers





From cliechti at gmx.net  Fri Sep 27 17:51:14 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 27 Sep 2002 23:51:14 +0200
Subject: Better terminal for windows?
References: <3D94B037.2020202@motorola.com> <mailman.1033158756.25404.python-list@python.org> <an2ivs$8fg$1@peabody.colorado.edu>
Message-ID: <Xns9296F2B5187A6cliechtigmxnet@62.2.16.82>

Fernando P?rez <fperez528 at yahoo.com> wrote in
news:an2ivs$8fg$1 at peabody.colorado.edu: 
> The other problem is that the windows terminal doesn't properly honor
> ANSI color escapes. I tried a lot to get ipython's coloring to work
> under windows and finally gave up in disgust. Why is it that MS can't
> be bothered to _ever_ follow the simplest standards?
> 
> So some of ipython's nicest functionality (full support for tab
> completion for both objects and files,  and color-enhanced
> information, tracebacks and even syntax highlighted source info) are
> gone under Windows. 
> 
> If you know of a way to fix this, patches are welcome :)

well there is ansi.sys. if that driver is loaded the terminal undestands 
ansi escapes. long ago i used it under win95. it's still there in win2k but 
don't know how to use it. it the old days you needed to load it in 
config.sys.

chris
-- 
Chris <cliechti at gmx.net>



From ckasso at sprynet.com  Thu Sep 26 16:46:41 2002
From: ckasso at sprynet.com (Chris Kassopulo)
Date: Thu, 26 Sep 2002 16:46:41 -0400
Subject: How do you pronounce "__" (double underscore)?
References: <525eaba.0209260956.72bf6786@posting.google.com>
Message-ID: <20020926164641.1753ebbb.ckasso@sprynet.com>

On 26 Sep 2002 10:56:45 -0700
pknotz at sandia.gov (Pat Notz) wrote:

> @ = "at"
> ! = "bang"
> . = "dot"
> * = "star"
> 
> Is there a short (one or two syllable) pronounciation for "__" (double
> underscore).  "Underscore underscore" and "double underscore" are kind
> of long.  Somthing like "splat"?  Anyone?
> 

How about:

 _  = "un"
 __ = "dun"
 
Chris Kassopulo


From tjreedy at udel.edu  Fri Sep 20 17:56:53 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 20 Sep 2002 21:56:53 GMT
Subject: e vs exp()?
References: <3a838998.0209191933.1f22e977@posting.google.com> <amflhl$3eb$1@peabody.colorado.edu>
Message-ID: <FQMi9.171268$AR1.7178378@bin2.nnrp.aus1.giganews.com>

"Fernando P?rez" <fperez528 at yahoo.com> wrote in message
news:amflhl$3eb$1 at peabody.colorado.edu...
> Buckshot wrote:
>
> > Can anyone here explain why I'm getting different results from
> > e**2 and exp(2)?  Which is more accurate?
> >
> > Python 2.2 (#1, Mar 27 2002, 14:56:58)
> > [GCC 2.95.3 20010315 (release)] on sunos5
> > Type "help", "copyright", "credits" or "license" for more
information.
> >>>> from math import *
> >>>> e**2
> > 7.3890560989306522
> >>>> exp(2)
> > 7.3890560989306504
>
> As someone already told you, exp(2) gives a more accurate answer.
The reason
> is, I think, the following: e**2 simply squares e,

Thanks.  That seems to solve the particular problem posed:
>>> e*e
7.3890560989306495
>>> e**2
7.3890560989306495
>>> pow(e,2)
7.3890560989306495
>>> exp(2)
7.3890560989306504

Perhaps x**int is special-cased.  But following must have a different
explanation:

>>> e**2.1
8.1661699125676499
>>> pow(e,2.1)
8.1661699125676499
>>> exp(2.1)
8.1661699125676517

At least two of us expected x**y == pow(x,y) == exp(y*log(x)) (a
standard method, and the only one I could find in Abrahamowitz and
Stegun, Handbook of Mathematical Functions) but there must be a more
direct method being used at least in some C libraries now.

Terry J. Reedy




From tjreedy at udel.edu  Mon Sep 23 21:34:58 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 24 Sep 2002 01:34:58 GMT
Subject: e vs exp()? / financial applications
References: <mailman.1032801271.6597.python-list@python.org>
Message-ID: <6jPj9.161531$216.5939317@bin4.nnrp.aus1.giganews.com>

"Andrew Thompson" <andrew.thompson at ashecastle.com> wrote in message
news:mailman.1032801271.6597.python-list at python.org...
> One must be aware of the representational issues and accuracy of
> arithmetic within a datatype when using numerical types.  This is
not so
> much a Python issue as one of representing the real number line with
> float types.
...
> W=0.89 / 4
> 1+w+w+w+w
> w+w+w+w+1
>
> which are both *different* according to the == operator, but for the
> purposes of numerical accuracy are identical.

Since 'W' != 'w' in Python, both give a NameError unless w defined
elsewise 8<).
However...
>>> w=.89/4
>>> w
0.2225
>>> 1+w+w+w+w
1.8899999999999997
>>> w+w+w+w+1
1.8900000000000001

Although w appears to be exactly represented, it must not be.  Cute
example.

Terry J. Reedy




From gvdbogae at vub.ac.be  Tue Sep  3 06:28:59 2002
From: gvdbogae at vub.ac.be (Geert-Jan Van den Bogaerde)
Date: 03 Sep 2002 12:28:59 +0200
Subject: code doesn't work: question
In-Reply-To: <1551859543.20020903130305@is.lg.ua>
References: <1551859543.20020903130305@is.lg.ua>
Message-ID: <1031048940.915.34.camel@gandalf>

On Tue, 2002-09-03 at 12:03, Ruslan Spivak wrote:
> Hello python-list user,
> 
> I want somebody to explain me why next code doesn't work:
> 
> var = 5
> 
> if (tmp = var) == 5:         # this is the error line
>    print "ok"
> else:
>    print "error"

Python doesn't allow assignment inside expressions, the rationale (from
the Python tutorial at http://www.python.org/doc/current/tut/node7.html
- section 5.5 last sentence): 

"""
Note that in Python, unlike C, assignment cannot occur inside
expressions. C programmers may grumble about this, but it avoids a
common class of problems encountered in C programs: typing = in an
expression when == was intended. 
"""

Given the number of times I've personally made this mistake, I'd agree
with Python's rationale for leaving this out.

As to what to use instead:

>>> var = tmp = 5
>>> if tmp  == 5:         
>>>    print "ok"
>>> else:
>>>    print "error"

Best regards,

Geert-Jan Van den Bogaerde






From richie at entrian.com  Fri Sep 13 04:02:11 2002
From: richie at entrian.com (Richie Hindle)
Date: 13 Sep 2002 01:02:11 -0700
Subject: "str"(expr)"str" interpolation syntax
References: <mailman.1031688793.11971.python-list@python.org> <f7930f23.0209121050.5c1870e3@posting.google.com>
Message-ID: <8a6ba1da.0209130002.37ede06a@posting.google.com>

> Maybe something very simple along the lines of:
> 
> subst("A template string - here: % it goes\n", foo,
>       "Also multiple values: %, %\n", bar, quux,
>       "Quoting percent (%) sings probably not needed...\n", '%')
> 
> could be useful...  Yes it smells like perl formats but the
> intelacing of pictures/values is really near optimal IMHO.

Guido's time machine strikes again:

>>> foo, bar, quux = 1, 2, '3'
>>> print ("A template string - here: %s it goes\n" % foo +
...        "Also multiple values: %s, %s\n" % ( bar, quux ) +
...        "Quoting percent (%s) sings probably not needed...\n" % '%')
A template string - here: 1 it goes
Also multiple values: 2, 3
Quoting percent (%) sings probably not needed...


-- 
Richie Hindle
richie at entrian.com


From tg5027 at citlink.net  Wed Sep 18 17:35:32 2002
From: tg5027 at citlink.net (terry)
Date: Wed, 18 Sep 2002 17:35:32 -0400
Subject: monetary applications (et al)
Message-ID: <E17rm1M-0002iZ-00@mail.python.org>

>>>>> Chris wrote:
Well, sure. This is obviously a stupid way to go about...That's what 
object oriented programming is all about. What you need to do is create 
a currency class that behaves like it should, and then use it 
everywhere you want currency.  <snip>
>>>>>

I do admit to being OO inhibited, so maybe you could take it from this 
short piece code and execution results, and, provide a brief 
description of what you would put into this money class that would 
eliminate this error message, while maintaining the code simplicity.

-------------------------------------------------
class money:
	def __init__(self,name):
		self.name = name

unitprice = money(100)
qty = 2
amount = qty * unitprice

terry at terry:~> python money.py
Traceback (most recent call last):
  File "money.py", line 7, in ?
    amount = qty * unitprice
TypeError: unsupported operand type(s) for *: 'int' and 'instance'
-------------------------------------------------

Support for mixed number types in expressions involving this money 
class must come from somewhere - where?  And, it must be capable of 
determing and setting amount to be of class money to be consistent with 
the rest of Python's complexity hierarchy for the handling of numbers.

The rest of the problem is that the above is a trivial calculation.  
What happens to your recommendation when faced with an additional 
number type?

taxrate = 0.06
amount = qty * unitprice + (qty*unitprice)*taxrate

As I see it, if Python had an intrinsic number type of 'money', the 
logic for determing the results of the above calculation would be 
easily determined and consistent with expectations, as in VB.

Thanks for your time,

terry



From jggramlich at yahoo.com  Wed Sep  4 12:23:28 2002
From: jggramlich at yahoo.com (Joshua Gramlich)
Date: 4 Sep 2002 09:23:28 -0700
Subject: PyListObject & C Modules
References: <mailman.1031085786.12546.python-list@python.org> <m3k7m2tf0y.fsf@mira.informatik.hu-berlin.de>
Message-ID: <bdade4c9.0209040823.3624d164@posting.google.com>

Thank you for your replies, here is what I have so far:

############################code block#############################

#include "Python.h"
#include <stdio.h>

PyObject*
get_first_elem(PyObject *unused, PyObject *args)
{
  PyObject *list;
  PyObject *first;
  PyObject *result;
  if(!PyArg_ParseTuple("!0", &list))
    return NULL;

  if(!PyList_Check(list)) {
    PyErr_SetString(PyExc_TypeError, "get_first_elem expects a list");
    return NULL;
  }

  if(PyList_Size(list) == 0){
    PyErr_SetString(PyExc_ValueError, "empty lists not allowed");
    return NULL;
  }

  first = PyList_GetItem(list, 0);
  if (first == NULL) return NULL;
  result = PyList_New(1);
  if (!result) return NULL;
  Py_ADDREF(first);
  if (PyList_SetItem(result, 0, first) != 0) {
    Py_DECREF(first);
    Py_DECREF(result);
    result = NULL;
  }
  return result;
}

####################################################################

Now, when I compile that, I get:

make

gcc -I/fnal/ups/prd/python/v2_1/SunOS+5/include/python2.1/ -c slist.c
slist.c: In function `get_first_elem':
slist.c:10: warning: passing arg 1 of `PyArg_ParseTuple' from
incompatible pointer type
slist.c:10: warning: passing arg 2 of `PyArg_ParseTuple' from
incompatible pointer type
gcc -G slist.o -o slist.so

[d0ora3 ~/work/grid_query/test]$ python hello.py 
Traceback (most recent call last):
  File "hello.py", line 2, in ?
    import slist
ImportError: ld.so.1: python: fatal: relocation error: file
./slist.so: symbol Py_ADDREF: referenced symbol not found
[d0ora3 ~/work/grid_query/test]$ 

Commenting PyADDREF out of the code causes the shared object not to
build at all.  My hello.py looks like this:


import slist

stuff = [2, 3, 8, 23]
print(slist.get_first_elem(stuff))


From warkid at storm.ru  Tue Sep  3 04:53:54 2002
From: warkid at storm.ru (Kerim Borchaev)
Date: Tue, 3 Sep 2002 12:53:54 +0400
Subject: Python to surpass C performance by 2030
In-Reply-To: <ad052e5c.0209022303.32319d0b@posting.google.com>
References: <ad052e5c.0209022303.32319d0b@posting.google.com>
Message-ID: <12537.020903@storm.ru>

Hello Dan,

Tuesday, September 03, 2002, 11:03:14 AM, you wrote:

DB> Erno Kuusela <erno-news at erno.iki.fi> wrote in message news:<kuptw4o65q.fsf at lasipalatsi.fi>...
>> just for fun i ran pystone under python versions since 1.5.2 to
>> current cvs. here are the results:
>>
>> [benchmark results snipped]
>>
>> since 1.5.2 is about 4 years old, the improvement per year is about
>> 13.2% on average. if we assume that 1.5.2 is 50 times slower than
>> C (probably conservative?), then at 13.2% per year since 1998 it
>> will add up to 53x improvement in 32 years.
>> 
>> i wonder if this would be a good application for the time machine...
>> 
>> learn python today, and in 28 years you will have a job writing
>> the speed-critical parts of c programs in python :)

DB> Before that could ever happen, the Python interpreter would have to be
DB> written in something other than C.

Yes. It should be written in Python:
http://psyco.sourceforge.net/plans.html

Best regards,
 Kerim                            mailto:warkid at storm.ru





From hancock at anansispaceworks.com  Thu Sep 26 16:02:01 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Thu, 26 Sep 2002 13:02:01 -0700
Subject: Lambdaizing Python Code
Message-ID: <3D9367B9.844BF599@anansispaceworks.com>

From: gerson.kurz at t-online.de (Gerson Kurz)
> A really usefull study ... NOT. 
> http://p-nand-q.com/e/quicksort_lambda.html
> also known as "making Guido regret he put lambda in THE BEST
> PROGRAMMING LANGUAGE IN THE WORLD(TM)".

AAAAAUUUUURRRRGGGHHHHH!!!!
That was scary. ;-D

Paraphrasing "Ghostbusters" --
"Okay, guys, important safety tip: avoid lambda."

From: Fernando P?rez <fperez528 at yahoo.com>
> Why not just a simple function?
> In [23]: plur=lambda s,n:'%s %s' % (n,s+('','s')[n!=1])
> In [24]: plur('cat',0)
> Out[24]: '0 cats'
> In [25]: plur('cat',1)
> Out[25]: '1 cat'
> In [26]: plur('cat',9)
> Out[26]: '9 cats'
> In [27]: na=1; np=3; nl=1; ng=0
> In [28]: print "Found %s, %s, %s, %s" % (plur('apple',na),plur('plum',np),
>    ....: plur('lemon',nl),plur('grape',ng))
> Found 1 apple, 3 plums, 1 lemon, 0 grapes
> 
> If the list is big the above can easily be automated with a map or a list 
> comprehension.

Not that I want to sign up for the style police or
anything, but ...

Would respectfully like to point out that this could
as well be represented:

def plur(word, number):
    if number > 1:
        word = word + 's'
    return '%d %s' % (number, word)

and would, IMHO, be much more obvious in function to
the reader.  Nothing says you have to use "lambda"
just because you might use a map or list comprehension.

IMHO, if you need to assign lambda to a variable,
you probably want to use a named function anyway. 
(Basically, it's semantically identical to do so,
(I think) -- except that you remove the expression
limitation, which makes for more intuitive code.

It occurs to me that the plur() function is a very
good idea, especially since it is possibly
multi-lingual (e.g. in Russian, Swahili, and Japanese
the pluralization rules are completely different --
Russian typically inflects with an "i" ending, Swahili
uses prefixes (e.g. "mtoto" / "watoto"). Japanese is
potentially more challenging, as the word remains
unchanged, but the number acquires a counter
"empitsu sambon" (and they're normally separated in
the sentence (e.g. "empitsu ga sambon aru" = "there
are three pencils").  I wonder if there are any
other challenge cases to consider, and whether a
more general pluralization scheme is possible.

PUB (http://py-universe.sourceforge.net) has a problem
with this, in that much of the pluralization code is
mixed in, and assumes English (or other languages where
you can make a plural by adding "s").

Mind you, even in english the plur() function will
break down for sheep, tooth, calf, and a host of
other exception words.

Fun concept.

Cheers,
Terry

-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
Anansi Spaceworks                 
http://www.anansispaceworks.com 
P.O. Box 60583                     
Pasadena, CA 91116-6583
------------------------------------------------------



From ataylor at ucsd.edu  Tue Sep 24 21:36:06 2002
From: ataylor at ucsd.edu (Adam Taylor)
Date: Tue, 24 Sep 2002 18:36:06 -0700
Subject: Getting started
Message-ID: <3D911306.7080609@ucsd.edu>

About this:

Alex Martelli wrote:
 > I'm somewhat surprised at seeing you make such assertions at
 > obvious variance with type theory -- I thought you had a solid
 > CS background (stronger than my own -- I'm just an engineer,
 > more interested in pragmatics than in theory).  We _do_ share
 > the fundamentals, I hope?  E.g., my favorite "essential Luca
 > Cardelli bibliography":
 >
 > Cardelli, Wegner, "On understanding types, data abstraction, and
 > polymorphism", ACM Computing Surveys 1985
 >
 > Cardelli, "Type Systems", in "Handbook of Computer Science and
 > Engineering", CRC Press, 1997.

This last is available at:
http://research.microsoft.com/Users/luca/Papers/TypeSystems.pdf

It says, in part:

   A program variable can assume a range of values during the execution
   of a program.  An upper bound of such a range is called a type of the
   variable. For example, a variable x of type Boolean is supposed to
   assume only boolean values during every run of a program. If x has
   type Boolean, then the boolean expression not(x) has a sensible
   meaning in every run of the program. Languages where variables can be
   given (nontrivial) types are called typed languages.

   Languages that do not restrict the range of variables are called
   untyped languages: they do not have types or, equivalently, have a
   single universal type that contains all values. In these languages,
   operations may be applied to inappropriate arguments: the result may
   be a fixed arbitrary value, a fault, an exception, or an unspecified
   effect. The pure \lambda-calculus (see Chapter 139) is an extreme case
   of an untyped language where no fault ever occurs: the only operation
   is function application and, since all values are functions that
   operation never fails.

Based on this definition, it seems to me that Python _is_ untyped,
since the type of a variable can change arbitrarily in mid-program.

He later discusses the distinction between safety and typed-ness, and
shows this table:

          Typed   Untyped
Safe     ML      LISP
Unsafe   C       Assembler

Python is safe (at least pure Python, I think), so it should go in the
same square as Lisp, correct?

Cardelli doesn't use the term "weakly typed" himself, but on page 5 he
seems to be say that "weakly typed" is the commonly used term for what
he calls "weakly checked".  C is weakly checked, in his terminology, and
that's why it's unsafe.

But people on this thread seem to be using "weakly typed" to mean
roughly the same thing as Cardelli's "untyped".  Which is fine, and
pretty close to Cardelli's terminology.  But clearly the terms "untyped"
(=="weakly typed" on this thread) and "weakly checked" (=="weakly typed"
in common usage) are very different things.

Adam

-- 
Adam Taylor
ataylor at ucsd.edu
858-534-4763




From chris.gonnerman at newcenturycomputers.net  Thu Sep 26 01:10:31 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Thu, 26 Sep 2002 00:10:31 -0500
Subject: Tabs vs. Spaces, VIM vs. EMACS (Was Re: Larry Wall's comment on python...)
References: <F7DB8D13DB61D511B6FF00B0D0F06233025D48F5@mail.hebmex.com> <3D9266B0.7090906@cascade-sys.com>
Message-ID: <006101c2651b$0df0c340$0101010a@local>

----- Original Message ----- 
From: "James J. Besemer" <jb at cascade-sys.com>
 
> I myself still prefer tabs as I find "indenting" and "out 
> denting" sections of code to be a common operation (in Python 
> and in other languages).  Inserting or removing an "if" or 
> changing inline code to a function makes indenting and out 
> denting a common operation.  In the environments I use, this 
> works well with tabs and not so well or at all with spaces.  

At the risk of starting another sort of flamewar, I'll weigh
in here.  I am a user of VIM, and with the "smartindent" and
"expandtab" options activated, spaces are just as convenient
as tabs... I use the tab *key* to get my four-character
indents, and can backspace to cancel an indent (i.e.
"outdent").  No actual tab characters ever get saved that 
way.

When I receive code that contains tabs, I use a global
replacement to quickly detab any module I need to modify.

Surely emacs can do as well as VIM...  (and I expect a rabid
emacs user to fill me in posthaste).

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From g_will at cyberus.ca  Mon Sep 23 21:09:31 2002
From: g_will at cyberus.ca (Gordon Williams)
Date: Mon, 23 Sep 2002 21:09:31 -0400
Subject: httplib 'getaddrinfo failed'
Message-ID: <001401c26367$09dd13e0$9504e640@wnt20337>

Hi All,

It must have been something that I was doing before, though I tried multiple
times.
Now it is working properly.

Thanks,

Gordon Williams


----- Original Message -----
From: "Gordon Williams" <g_will at cyberus.ca>
To: <python-list at python.org>
Sent: Friday, September 20, 2002 1:50 PM
Subject: httplib 'getaddrinfo failed'


> Hi All,
>
> I recently got a DSL line to connect to the internet and my code is now
> broken.  Could this be due to the DSL being different from a dialup
> connection?
>





From markus.vonehr at ipm.fhg.de  Thu Sep  5 02:54:44 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Thu, 05 Sep 2002 08:54:44 +0200
Subject: tk8.4 in python22 ?
Message-ID: <3D76FFB4.FE5EF48A@ipm.fhg.de>

Hi,

I'd like to use tk8.4 with python22, in fact the windows installer
installs tk8.3.
How can I replace tk8.3 to 8.4? Do I have to compile any python files?

Thanks for any hints,

Markus


From davbucko at deletethis.yahoo.com  Mon Sep  2 04:21:07 2002
From: davbucko at deletethis.yahoo.com (davbucko)
Date: Mon, 2 Sep 2002 09:21:07 +0100
Subject: Keyerror problem
References: <_lyc9.7164$7N.29584@newsfep3-gui.server.ntli.net> <slrnan5f2o.4s9.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <0cFc9.1718$yB6.25319@newsfep1-gui.server.ntli.net>

Thanks!  I wasn't too sure about what the backticks meant, but I thought
that they might be necessary.  Oh well

Dave

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote in message
news:slrnan5f2o.4s9.gerhard.haering at lilith.my-fqdn.de...
> davbucko wrote in comp.lang.python:
> > I have a class Flags as follows:
> >
> > class Flags:
> >     def __init__(self):
> >         self.flags = {}
> >
> >     def getInverseFlags(self):
> >         for i in self.flags.keys():
> >             #switch some keys/values around
> >
> >     def setFlag(self,name,value):
> >         self.flags[`name`] = value
>                      ^^^^^^
>
> Here's the bug. It should read: self.flags[name] = value.
>
> `name` calls repr on name, which is most likely not what you want.
>
> Btw. if you want to invert a dictionary, there are probably hundreds
> of examples out there, here's a hopefully readable one:
>
> def invertdict(orig):
>     new = {}
>     for k, v in orig.items():
>         new[v] = k
>     return new
>
> Gerhard
> --
> mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
> web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
> public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
> reduce(lambda x,y:x+y,map(lambda
x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From scotty at yang.id.au  Thu Sep 26 20:29:09 2002
From: scotty at yang.id.au (Scott Yang)
Date: Fri, 27 Sep 2002 00:29:09 GMT
Subject: Copying Elements using PyXML
References: <mailman.1032866191.25215.python-list@python.org>
Message-ID: <slrnap79ou.c64.scotty@heater.xden.xplantechnology.com>

Assuming you want to have the result DOM as
<xformimport>
    <docID />
    <document />
    <person>
        <name />
        ...
    </person>
</xformimport>

You should be able to do it easily using PyXML.

# Parse the XML document into DOM objects.
from xml.dom.ext.reader.PyExpat import Reader
person = Reader().fromString(doc1)
xformimport = Reader().fromString(doc2)

# Clone person's document node.
person_cloned = person.cloneNode(1, xformimport)

# Append cloned node to xformimport's child nodes.
xformimport.documentElement.appendChild(person_cloned)


In article <mailman.1032866191.25215.python-list at python.org>, Jeremy Rew wrote:
> I have 2 XML documents. I would like to copy the entire structure of one 
> document into a child element of the other, but i cannot copy the root node 
> across:
> I get the following error from Zope:
> Error Type: HierarchyRequestErr
> Error Value: Node manipulation results in invalid parent/child relationship.
> 
> the structures are as follows:
> 
><person>
>          <name/>
>          <address/>
>          <phone/>
></person>
> 
> and would like to copy this into the following format:
> 
><xformimport>
>          <docID col="blah" id="blahblah"/>
>          <document>
>                  <person>
>                          <name/>
>                          <address/>
>                          <phone/>
>                  </person>
>          </document>
></xformimport>
> 
> I have written all the code to build 'xformimport','docID' and 'document', 
> but just need to copy 'person' across.
> I figure i may be missing an easy solution here, so any help/ideas will be 
> appreciated.
> I am using PyXML 0.8.1 and python 2.1.3.
> 
> thanks
> Jeremy
> 
> 


-- 
Scott Yang :: scotty at yang.id.au :: http://scott.yang.id.au/ :: PGP:0xF9DCFA8C


From bokr at oz.net  Thu Sep  5 16:04:08 2002
From: bokr at oz.net (Bengt Richter)
Date: 5 Sep 2002 20:04:08 GMT
Subject: Gratitude due to Red Hat? More work for the PBF?
References: <Iyn89.11545$aC5.4453@atlpnn01.usenetserver.com> <mailman.1030142061.12228.python-list@python.org>  <xuyptw3knfx.fsf@halden.devel.redhat.com> <mailman.1031250971.7837.python-list@python.org>
Message-ID: <al8dbo$408$0@216.39.172.122>

On 05 Sep 2002 11:28:14 -0700, Cliff Wells <LogiplexSoftware at earthlink.net> wrote:
[...]
>
>The aggravating thing is that this issue would be a snap for RedHat to
>fix.  Instead of preempting /usr/bin/python for their own use, they
>could maintain a /usr/bin/python-rh and leave python for user-installed
>versions (instead of requiring the user version to be called "python2").
>
>Problem solved, everyone happy.
>
So why don't they do this? Do they need help?

Regards,
Bengt Richter


From loewis at informatik.hu-berlin.de  Mon Sep  2 09:42:55 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 02 Sep 2002 15:42:55 +0200
Subject: Does IDLE handle unicode?
References: <QdIc9.2412$Ak.1132837@kent.svc.tds.net>
Message-ID: <j4u1l8jyn4.fsf@informatik.hu-berlin.de>

"Edward K. Ream" <edream at tds.net> writes:

> IDLE took an exception writing a .py file containing the copyright
> character.  This happened a while ago; IIRC, I lost the file.
> 
> Is this a known problem?  I see nothing about this in the FAQ.

It will be fixed in Python 2.3.

Regards,

Martin


From mbonig at hotmail.com  Sat Sep 14 04:28:53 2002
From: mbonig at hotmail.com (Matthew Bonig)
Date: Sat, 14 Sep 2002 08:28:53 GMT
Subject: Tkinter listbox clicking problem
References: <sGog9.307183$aA.53945@sccrnsc02> <altfrk$1eva$1@nntp6.u.washington.edu> <2D7E3697CF5D1521.5D9711540E657ED8.7EF0451C598438FC@lp.airnews.net>
Message-ID: <9rCg9.262546$kp.888557@rwcrnsc52.ops.asp.att.net>

I'm confused... I looked at these examples but I didn't see anything pop out
that answered my question... or was this mostly for the other post?

Thanks
Matthew Bonig


"Cameron Laird" <claird at starbase.neosoft.com> wrote in message
news:2D7E3697CF5D1521.5D9711540E657ED8.7EF0451C598438FC at lp.airnews.net...
> In article <altfrk$1eva$1 at nntp6.u.washington.edu>,
> Russell E. Owen <owen at nospam.invalid> wrote:
> >In article <sGog9.307183$aA.53945 at sccrnsc02>,
> > "Matthew Bonig" <mbonig at hotmail.com> wrote:
> >
> >>Right now I have it bound so when a Tkinter listbox gets a <Button-1>
> >>(lb.bind("<Button-1>", someFunc)) it calls function someFunc. Now,
> >>someFunc() simply prints the curselection(). The problem is it's
lagged...
> >>if I first select list item 1 (index 0), and then 2, and then 3 (array
> >>indexes 0,1,2) it prints out themout incorrectly.
> >>(I have someFunc actually do some try, except stuff so I don't get
uncaught
> >>exceptions) Instead of printing 0,1,2 it prints "nothing selected", 0,
1.
> >>The function get's called before the listbox updates it's selection
markers!
> >>
> >>Does anybody know how I could force listbox to update it's selection
markers
> >>so instead of printing "nothing selected, 0,1  it prints 0,1,2 like it
> >>should?
> >
> >I'm hoping somebody has better ideas, but here are some to try:
> >
> >Call update_idletasks before you ask for the selection. I have no idea
> >if this will work, but it might.
> >
> >An ugly solution that would probably work:
> >- extract the y coordinate from the event callback
> >- call nearest(y_coord) to get the item closest to the mouse click
> >- explicitly select that index item. You could try skipping this step,
> >but you risk a possible error between what you think they clicked on and
> >what is hightlighted. Unfortunately, this step will may be a real pain
> >for a list that supports multiple selected items.
> >
> >Poll for selection changes -- thus it may take a bit of time before your
> >code notices a change, but it will always be right once it gets there.
> >
> >-- Russell
>
> Whoa!  There's a far easier solution.  Read up on 'bindtags'
> in <URL: http://web.pydoc.org/2.2/Tkinter.html > and elsewhere.
> <URL: http://www.cwi.nl/~jacob/textui.py.html > has a small
> example that almost is pertinent.
> --
>
> Cameron Laird <Cameron at Lairds.com>
> Business:  http://www.Phaseit.net
> Personal:  http://phaseit.net/claird/home.html




From aleax at aleax.it  Fri Sep 27 04:07:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 27 Sep 2002 08:07:24 GMT
Subject: comparing all values of a list to regex
References: <mailman.1032947731.3297.python-list@python.org> <Nbgk9.140934$ub2.3077596@news1.tin.it> <mailman.1033039492.1772.python-list@python.org>
Message-ID: <0lUk9.149240$ub2.3291527@news1.tin.it>

<posted & mailed>

Manuel Hendel wrote:
        ...
> This is the input, it is "|" seperated textfile with about 15000 lines
> of pop3 accounts:
> 
|Number|String|Number|String(Domain)|String(Account)|String(Login/Email)|String(Password)|String|String|
> 
> These are the fields I care about.
> 
> String(Domain): This is the domain of the Account
> String(Account): This is the part in front of the @doamin. This can
> also be a * for a catchall account.
> String(Login/Email): This is the local pop3-account or a emailaddress,
> or both, comma seperated.
> String(Password): This is the password if String(Login/Email) is a
> pop3-account.
> 
> This should be the output:
> 
> Three text files. One with only forwardings (only emailaddresses in
> the String(Login/Email). One with only pop3-accounts in the
> String(Login/Email), and one mixed, where String(Login/Email) has
> pop3-accounts and emailaddresses.

OK, for each line you want fields line.split('|')[4:8], each field
space-stripped (at least, you strip them in the code you posted,
though that's not in the specs you give above), _plus_ you need
to know the set of domains (uniquely) and classify per field 4
(domain) -- you build this in the code you posted, althogh, again,
I can't see any trace of that requirement in the specs.

So we'll want an auxiliary function to tell us which of the three
"bins" to put an entry to depending on the login/email field, e.g.:

def classify(login_email):
    l_e = login_email.split(',')
    assert 1 <= len(l_e) <= 2, "More than one comma in (%s)" % login_email
    if len(l_e)==2: return 2         # both
    elif l_e.find('@'): return 1     # email, I guess
    else return 0                    # other case (local account, I guess)


Now the rest, net of imports and file open/close ops:

classified = [ [] for i in range(3) ] # 3 separate empty lists
per_domanin = {}                      # initially-empty dict

for line in inputfile:
    fields = [ field.split() for field in line.split('|')[4:8] ]
    per_domain.setdefault(fields[0],[]).append(fields)
    classified[classify(field[2])].append(fields)


Now you only need the output -- presumaby each line must be
output in the same way, e.g. with another auxiliary function:

def outline(fileobj, fields):
    fileobj.write('|'.join(fields))
    fileobj.write('\n')

so you only need to loop on each of the three lists of lists
in 'classified', and on the keys of dictionary 'per_domain'
(sort them too, if you wish, of course), in order to emit
the results to appropriate files.  Presumably your exact
specs are not quite as I tried to guess them from a mix of
what you wrote and what you coded, but I hope this outline
can still be useful to you.


Alex



From Oschler at earthlink.net  Sun Sep  1 17:19:36 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sun, 1 Sep 2002 17:19:36 -0400
Subject: Add methods to a class at runtime?
References: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com> <slrnan4dl3.18h.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <wmvc9.21757$FJ1.3181934@e3500-atl1.usenetserver.com>

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote in message
news:slrnan4dl3.18h.gerhard.haering at lilith.my-fqdn.de...
> >>> class Foo: pass
> ...
> >>> foo = Foo()
> >>> foo.bar()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> AttributeError: Foo instance has no attribute 'bar'
> >>> def bar(self): print "bar"
> ...
> >>> Foo.bar = bar
> >>> foo.bar()
> bar

Ok bear with me guys (Gerhard and Peter).  I'm not lazy, the reason I don't
test certain things I should is because Python is capable of such amazing
introspective dynamism that, having a _very_ long experience bank in
strongly typed "static" languages like C++ and Pascal, I find I don't even
consider yet even trying certain possibilies that Python is capable of.
Some interesting tidbits I found while trying out your suggestions,  as you
will see in the two ".py" files that follow, I discovered that to create a
method you can dynamically add to a class _instance_, you need to omit the
'self' parameter in the method declaration.  For a method you add
dynamically to a class definition itself, then you must include the 'self'
parameter in the method declaration.  If not, you will get an argument count
mismatch complaint from the interpreter.  Fascinating stuff.

I'm running 2.1.1, anybody know if 2.2.2 is going "disable" either of these
abilities? (class or class instance dynamic method addition).  I saw a note
about such activities requiring a "dynamic" keyword, which is fine.  I'd
really miss either of these abilities.  Code follows below,

Thanks again.

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


// The test program, test_main.py.

import test1
a = test1.Account("Guido", 1000)
a.MethodToAdd = test1.MethodToAddToInstance
a.MethodToAdd()

test1.Account.MethodToAdd = test1.MethodToAddToClass
c = test1.Account("Fred", 2000)
c.MethodToAdd()

// The module test1.py.

class Account:
  "A simple class"
  account_type = "basic"
  def __init__(self, name, balance):
    "Initialize a new account balance"
    self.name = name
    self.balance = balance
  def deposit(self, amt):
    "Add to the balance"
    self.balance = self.balance + amt
  def withdraw(self, amt):
    "Subtract from the balance"
    self.balance = self.balance - amt
  def inquiry(self):
    "Return the current balance"
    return self.balance

def AddAttr(Acct):
  Acct.number = 7

def MethodToAddToInstance():
  "I want to add this method dynamically to an instance of a class."
  return 10

def MethodToAddToClass(self):
  "I want to add this method dynamically to a class."
  return 11








From emile at fenx.com  Thu Sep 26 19:33:40 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 26 Sep 2002 16:33:40 -0700
Subject: Who knows somefunction?
References: <F7DB8D13DB61D511B6FF00B0D0F06233026399A9@mail.hebmex.com>
Message-ID: <an05eg$9ht$1@main.gmane.org>

gus:
> I haven't compiled it yet; what happens
> when you don't "del x" before you declare x1?


Python 2.3a0 (#29, Sep 26 2002, 16:04:25) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> from pointer import pointer
>>> x = range(3)
>>> y = id(x)
>>> x1 = pointer(y)
>>> x1
[0, 1, 2]
>>> del x
>>> x1
[0, 1, 2]
>>> del x1
>>> x1 = pointer(y)
>>> x1
[0, 1, 2]
>>>

Anybody know anything you can do with a hunk of rope?  ;-)

--

Emile van Sebille
emile at fenx.com

---------






From mnations at airmail.net  Fri Sep 20 11:49:31 2002
From: mnations at airmail.net (Marc)
Date: 20 Sep 2002 08:49:31 -0700
Subject: Strange Telnet error
References: <4378fa6f.0209181245.7b06124d@posting.google.com> <qj8louc5615jicn6covqe5b7df9sp8amdv@4ax.com>
Message-ID: <4378fa6f.0209200749.1b873c4e@posting.google.com>

> Your code is multithreaded.  Could another thread have closed the socket on
> you?

Yeah, the timing was off between one thread and another queue which
was closing it prematurely (DOH!) Next time I will know that the
"...sendall socket..." implies a socket that has already been closed.
An error that says something about trying to use a closed socket would
have been helpful, but when I first read the error it sounded like I
was sending bad or corrupt information or something.

Thanks for the help,
Marc


From gerhard.haering at gmx.de  Thu Sep 26 20:53:51 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 27 Sep 2002 00:53:51 GMT
Subject: POP3 and Multipart Mime
References: <mailman.1033084656.24584.python-list@python.org>
Message-ID: <slrnap7b0v.22h.gerhard.haering@lilith.my-fqdn.de>

David McInnis wrote in comp.lang.python:
> I need to log into a POP3 account (which I have been able to do using
> poplib).
>  
> I need to be able to read and decode the message.  The problem is that
> most of our users send word docs as attachments (1 or more).  I need to
> take these files and save them to disk.

Check out the email module in Python 2.2 and later.

-- Gerhard


From anton at vredegoor.doge.nl  Fri Sep 20 21:00:31 2002
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Sat, 21 Sep 2002 03:00:31 +0200
Subject: average?
References: <mailman.1032541648.27654.python-list@python.org> <amfscs$paa$1@solaris.cc.vt.edu>
Message-ID: <amggcu$195$1@news.hccnet.nl>

On Fri, 20 Sep 2002 15:10:47 -0400, Carl Banks <imbosol at vt.edu> wrote:

>Mark McEahern wrote:
>>  def average(seq):
>>    return reduce(operator.add(seq) / len(seq))

>> 1.  Is there anything else dumb about the above?
>
>Well, yes; it seems to me you meant this:
>
>    def average(seq):
>        return reduce(operator.add,seq)/len(seq)

>> 2.  Am I missing a builtin cousin of min, max that would do this for me?
>
>In pure Python, I don't think so.  I would be surprised if the above
>function (when typed correctly) isn't the fastest way in pure Python.

Just a few more posters and we will get this function right, I am sure
...

 >>> import operator
 >>> def average(seq):
         return reduce(operator.add,seq)/len(seq)

 >>> average([1,2,3,4])
 2
 >>> def average(seq):
         return reduce(operator.add,seq)/float(len(seq))

 >>> average([1,2,3,4])
 2.5
 >>> 

Anton.


From david at no.westcontrol.spam.com  Wed Sep 18 10:23:25 2002
From: david at no.westcontrol.spam.com (David Brown)
Date: Wed, 18 Sep 2002 16:23:25 +0200
Subject: BibTeX module?
References: <Xns928D9F1FBAA22mhueningzedatfuberli@130.133.1.4>
Message-ID: <ama1sn$rpf$1@news.netpower.no>

"Matthias Huening" <mhuening at zedat.fu-berlin.de> wrote in message
news:Xns928D9F1FBAA22mhueningzedatfuberli at 130.133.1.4...
> Hi,
>
> I am looking for a simple BibTeX module, allowing me to read and
> manipulate/convert a bibliography file.
>
> Thanks, Matthias

Are you sure you are posting to the right newsgroup here?  comp.text.tex
might be a better bet, unless by chance someone has written a BibTeX module
in Python - unlikely, since most TeX related programs are written in C,
Pascal or perl.





From diltonm at pacbell.net  Sun Sep  1 18:58:21 2002
From: diltonm at pacbell.net (Dilton McGowan II)
Date: Sun, 01 Sep 2002 22:58:21 GMT
Subject: urllib bug?
References: <mailman.1030848662.20745.python-list@python.org>
Message-ID: <hYwc9.1745$De5.69654800@newssvr21.news.prodigy.com>

"Steve Clift" <sgclift at earthlink.net> wrote in message
news:mailman.1030848662.20745.python-list at python.org...
>
> Is this *really* a urllib bug, or am I missing something blindingly
obvious?
>
>   Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
>   Type "copyright", "credits" or "license" for more information.
>   IDLE 0.8 -- press F1 for help
>   >>> import urllib
>   >>> f = urllib.urlopen('http://www.python.org')
>   Traceback (most recent call last):
>     File "<pyshell#4>", line 1, in ?
>       urllib.urlopen('http://www.python.org')
>     File "C:\PROGRAM FILES\PYTHON22\lib\urllib.py", line 73, in urlopen
>       return _urlopener.open(url)
>     File "C:\PROGRAM FILES\PYTHON22\lib\urllib.py", line 178, in open
>       return getattr(self, name)(url)
>     File "C:\PROGRAM FILES\PYTHON22\lib\urllib.py", line 276, in open_http
>       if not host: raise IOError, ('http error', 'no host given')
>   IOError: [Errno http error] no host given
>
> Fails the same way with 2.1. Works OK with 1.5.2 and (IIRC) 2.0.
>
> -Steve
>

Works on Windows 2000 using Idle, PythonWin and Cygwin Python all version
2.2.




From claird at starbase.neosoft.com  Fri Sep 13 08:58:37 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 13 Sep 2002 07:58:37 -0500
Subject: concurrent programming
References: <fe31142f.0209130052.33174fcc@posting.google.com>
Message-ID: <B7A63460808730BA.82D6BAEAFA5E4E78.4731C570A5FC8A2B@lp.airnews.net>

In article <fe31142f.0209130052.33174fcc at posting.google.com>,
Chris <cs26 at tecco.at> wrote:
>Hi!
>
>I have a problem in my project I startet a few weeks ago. The
>requirement is to program a Task Scheduler who gets a Python script
>with the tasks spezified. So I have to write some extensions in C++
>which provide the functionality. So far so good, now comes the
>problem. It should be possible to write something like:
><code>
>PARBEGIN
>    if ....
>        exeute some shell scripts on distributed computers
>        some method calls from the extension module, e.g. write in a
>database
>    do something else
>    while ...
>    ...
>PAREND
></code>
>
>This means that the code bracketed in PARBEGIN PAREND should be
>executed parallel. The contrary would be SEQBEGIN SEQEND for
>sequential execution.
>
>The Task Scheduler is for testing some programs, so the scripts should
>be easy to write and not too complicated. My idea would be to handle
>this with a call of a function from the extension module which gets a
>function pointer to a method with the code inside PARBEGIN PAREND as a
>parameter, so that I can do a callback from the C++ module e.g.:
><code>
>def f():
>    if ....
>        exeute some shell scripts on distributed computers
>        some method calls from the extension module, e.g. write in a
>database
>    do something else
>    while ...
>    ...
>
>exec_par(f())
></code>
>
>But how can this be executed concurrent? Is there another possibility
>in Python?
>
>thx

There are several possibilities for styling concurrent
processing in Python.

Why are we talking about C++, though?  Would a pure-Python
solution not meet your requirements?
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From tim.one at comcast.net  Mon Sep 16 12:18:33 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 16 Sep 2002 12:18:33 -0400
Subject: Are generator functions thread-safe?
In-Reply-To: <eppstein-C9F3CA.08581716092002@news.service.uci.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEHEBEAB.tim.one@comcast.net>

[David Eppstein]
> Also, if a call to the generator object is running from one thread, you
> will get an exception if you attempt to simultaneously call it (e.g.
> from another thread).  I don't know how thread-safe is the code that
> prevents the generator from being called twice simultaneously, though.

A generator-iterator has an internal "am I currently running?" flag that's
inspected and mutated under protection of the GIL, so it's thread-safe.  The
flag is exposed as the gi_running attribute of a generator-iterator object,
although I haven't found a use for that yet <wink>:

>>> def g():
...     print i.gi_running
...     yield 42
...
>>> i = g()
>>> i.gi_running
0
>>> x = i.next()
1
>>> x
42
>>> i.next()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
StopIteration
>>>




From loewis at informatik.hu-berlin.de  Wed Sep 11 16:30:04 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 11 Sep 2002 22:30:04 +0200
Subject: Global Interpreter Lock
References: <3822aca1.0209111211.8a1538f@posting.google.com>
Message-ID: <j4wups2rsj.fsf@informatik.hu-berlin.de>

codic at stasis.org (Ian Crawford) writes:

>     I'm working on a project with an embedded Python
> interpreter which I'd like to be capable of parallel
> execution on a multi-processor machine.  I'm still
> trying to figure out one thing in particular.  Is
> there a global interpreter lock per interpreter (i.e.
> if I call Py_NewInterpreter, does that interpreter
> have it's own lock), or will I have to embed seperate
> interpreters in their own processes to have parallel
> execution?

The latter: even if you have multiple interpreter objects, the Python
runtime still needs to protected with a single lock.

Notice that you can have multiprocessor parallelism in Python if the
threads invoke long-running C code: if the extension module releases
the interpreter lock before the compute-intensive C code starts, this
computation can be performed on a separate processor.

Regards,
Martin



From KBASS1 at nc.rr.com  Thu Sep 12 06:04:37 2002
From: KBASS1 at nc.rr.com (Kevin Bass)
Date: Thu, 12 Sep 2002 10:04:37 GMT
Subject: Newby Question: os.rename
References: <AcGc9.68214$Xa.3523883@twister.southeast.rr.com> <tLGc9.68215$Xa.3526150@twister.southeast.rr.com> <fDdd9.6312$6i4.491988@newsread1.prod.itd.earthlink.net> <e81be8b2.0209082039.67f33ab4@posting.google.com>
Message-ID: <VEZf9.13406$jF4.1265334@twister.southeast.rr.com>

"Dean Goodmanson" <ponderor at lycos.com> wrote in message
news:e81be8b2.0209082039.67f33ab4 at posting.google.com...
> > >
> > > "Kevin Bass" <KBASS1 at nc.rr.com> wrote in message
> > > news:AcGc9.68214$Xa.3523883 at twister.southeast.rr.com...
> ...
> > >
> > > I have figured it out. Thanks!
> > >
> > >
> > >
>
> "CotaBas" <cotabas at earthlink.net> wrote in message
news:<fDdd9.6312$6i4.491988 at newsread1.prod.itd.earthlink.net>...
>
> >
> > LOL if you wait long enough people solve there own problems....
>
> If you're skimming or Google'ing for a similar problem...you'll find
> the problem listed 3 times and no answer in the follow-up!

Good point!!  Let me place the solution to my problem. :)

Here is the code that I used to solve this problem. I basically (1) give the
program the directory that I wanted to create the archived directoy off of,
(2) created a timestamp within a variable, (3) joined the directory with the
timestamp to form a new varaible housing the new directory, (4) created the
new directory with timestamp, (5) use glob to get a list of files, (6)
separated the directory name from the files names and placed them into
varaibles, (7) joined the new directory to the files that I glob'd, (8) then
rename the files which placed the files from one directory to another
directory.

I know that you probably did not want to know all of this but walking
through the steps helps me to undetsand any problems that I may encounter.
Like I just noticed that I have to place a try clause in the function so
that if the newly created directory is already there (the program was
executed more than once) then the program should not attempt to create the
directory again. Or I should check to see if the directory if already
present before attempting to create the directory.

def archive_data(archive):

   if archive == 'yes':
     datadir = '/home/oracle/data/'

     if os.path.isdir(datadir):
       the_date = strftime("%Y%m%d")
       archdir = os.path.join(datadir, the_date)
       os.mkdir(archdir)

       # Retrieve IBSS data from a previous inventory and archive it.
       for file in glob.glob('/home/oracle/data/*.XP'):
          dir, filename = os.path.split(file)
          oldfile = os.path.join(datadir, filename)
          newfile = os.path.join(archdir, filename)
           try:
              os.rename(oldfile, newfile)
          except:
            print "*** Unable to rename %s" %oldfile

       # Retrieve manual data from a previous inventory and archive it.
       for file in glob.glob('/home/oracle/data/*.csv'):
          dir, filename = os.path.split(file)
          oldfile = os.path.join(datadir, filename)
          newfile = os.path.join(archdir, filename)
          try:
            os.rename(oldfile, newfile)
          except:
            print "*** Unable to rename %s" %oldfile
   else:
      pass




From ajs at ix.netcom.com  Mon Sep 16 18:55:28 2002
From: ajs at ix.netcom.com (Arthur)
Date: Mon, 16 Sep 2002 18:55:28 -0400
Subject: Not sure how to phrase this question
Message-ID: <000501c25dd4$38711460$9865fea9@arthur>

Hoping I can get some help digesting the following
behavior.
************************************
class returnClassic(object):
    def __new__(self):
        return Classic()
class callClassic(object):
    def __new__(self):
       Classic()
class returnNewStyle(object):
    def __new__(self):
        return NewStyle()
class callNewStyle(object):
    def __new__(self):
       NewStyle()
class returnNewStyle(object):
    def __new__(self):
       return NewStyle()

class Classic:
    def __init__(self):
        print "Classic class init"
class NewStyle(object):
    def __init__(self):
        print "NewStyle class init"


if __name__ == "__main__":
   print "call to callClassic()" 
   cc=callClassic()
   print cc.__class__
   print "*************" 
   print "call to returnClassic()" 
   rc=returnClassic()
   print rc.__class__
   print "*************" 
   print "call to callNewStyle()" 
   cns=callNewStyle()
   print cns.__class__
   print "*************" 
   print "call to returnNewStyle()" 
   rns=returnNewStyle()
   print rns.__class__
   
Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help
>>> 
call to callClassic()
Classic class init
<type 'NoneType'>
*************
call to returnClassic()
Classic class init
__main__.Classic
*************
call to callNewStyle()
NewStyle class init
<type 'NoneType'>
*************
call to returnNewStyle()
NewStyle class init
NewStyle class init
<class '__main__.NewStyle'

***********************************************8
Problem being for what I am after the "returnClassic" seems
to be the simplest alternative.  But would  I be relying on what is
essentially a side-effect.

The "returnNewStyle" seems to actually create 2 instances, while
the "callNewStyle" seems to create a class in some netherworld
between existence and non-existence.

Would hate to have to resort to truely understanding
this stuff to get to where I am hoping to get. 


Art





From fperez528 at yahoo.com  Fri Sep 20 13:23:01 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Fri, 20 Sep 2002 11:23:01 -0600
Subject: e vs exp()?
References: <3a838998.0209191933.1f22e977@posting.google.com>
Message-ID: <amflhl$3eb$1@peabody.colorado.edu>

Buckshot wrote:

> Can anyone here explain why I'm getting different results from
> e**2 and exp(2)?  Which is more accurate?
> 
> Python 2.2 (#1, Mar 27 2002, 14:56:58)
> [GCC 2.95.3 20010315 (release)] on sunos5
> Type "help", "copyright", "credits" or "license" for more information.
>>>> from math import *
>>>> e**2
> 7.3890560989306522
>>>> exp(2)
> 7.3890560989306504

As someone already told you, exp(2) gives a more accurate answer. The reason 
is, I think, the following: e**2 simply squares e, and this amplifies the 
error in the value for e (which is stored at finite precision, as a C double 
inside). OTOH, exp(2) computes e**2 to a given accuracy, so it will not be 
propagating the finite-precision error inherent in the 'e' value. As long as 
the exp() function is correctly implemented in the system's math library, it 
should be more accurate than hand-calculations done with 'e'. In most cases 
the difference shouldn't matter, but if you are doing something very 
sensitive to numerical roundoff or require delicately balancing 
cancellations, you may need to pay attention to this.

Cheers,

f.


From sholden at holdenweb.com  Sun Sep 29 06:29:02 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 29 Sep 2002 06:29:02 -0400
Subject: List/Tuple bug or feature?
References: <8765wv6o0d.fsf@nospam.eml.cc>
Message-ID: <4DAl9.4938$pD4.1871@fe05>

"Mahesh Padmanabhan" <news at nospam.eml.cc> wrote ...
>
> Hi,
>
> I spent a lot of frustrating hours tracking this down and I am
> confused as to why I see this behavior.
>
> >>> x = ('a' 'b')
> >>> x
> 'ab'
>
The two adjacent string constants are concatenated. You don't get a tuple
because there is no comma folowing the singleton element, so technically the
RHS in the assignment is simply a parenthesised expression.

> >>> x = ['a' 'b']
> >>> x
> ['ab']
>
Here no commas are needed because there is no ambiguity. Again, the strings
are concatenated by Python. This allows you to more easily specify strings
in chunks.

> As you might have guessed, I missed a comma between two elements in
> a fairly large list.
>
> If this is a concatenation feature, how can I prevent something like
> this happening in the future ?
>
Don't miss out the commas! It *is* a feature, and it *will* bite you if you
use incorrect syntax.

> This is with Python 2.1.3 on Debian Linux.
>
But is valid for everything since 1.5.2 in all known implementations, I
would guess.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From markus.vonehr at ipm.fhg.de  Tue Sep 10 06:00:48 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Tue, 10 Sep 2002 12:00:48 +0200
Subject: compile error on SWIG file
References: <3D7C55EE.D9278B1C@ipm.fhg.de> <3D7D97A6.A63610AF@ipm.fhg.de>
Message-ID: <3D7DC2D0.8D74E062@ipm.fhg.de>

Hi,

I succeeded, in the SWIG generated file there's a line:

#define SWIG_init    init_mvelib

I changed it to:

#define SWIG_init    initmvelib

Now it's linked succesfully. Haven't tried it yet, let's see if it
works...

Markus


Markus von Ehr schrieb:
> 
> Hi Lyle,
> 
> thanks for your answer and your efforts!
> I changed everything to .cpp
> The first error has gone
> Now I have an error (translated from german):
> 
> mvelib.obj: error LNK2001: unresolved external symbol initmvelib
> build\temp.win32-2.1\Release\mvelib.lib : fatal error LNK1120: 1
> unresolved reference
> LINK : fatal error LNK1141: Creation error of export file
> error: command 'c:\Programme\VC98\BIN\link.exe' failed with exit status
> 1141
> 
> Markus
> 
> Markus von Ehr schrieb:
> >
> > Hi,
> >
> > compiling my c-file generated from SWIG
> > (swig -c -python -o mvelib.c mvelib.i)
> > with a setup-file, I receive the error:
> >
> > c:\programme\vc98\include\eh.h(32) : fatal error C1189: #Error: "eh.h
> > is only for C++!"
> >
> > Anybody knows what I have to change???
> >
> > Thanks Markus
> >
> > ****************************
> > setup.py
> > ****************************
> > from distutils.core import setup, Extension
> >
> > setup (name = "mvelib",
> >         version = "1.0",
> >         maintainer = "Markus von Ehr",
> >         maintainer_email = "markus.vonehr at ipm.fhg.de",
> >         description = "mvelib Python module",
> >
> >         ext_modules = [Extension('mvelib',sources=['mvelib.c'])]
> >  )


From emile at fenx.com  Sun Sep  8 14:35:32 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 8 Sep 2002 11:35:32 -0700
Subject: transform a list of lists in a lit of strtings???
References: <-1588066870.1031507786228.JavaMail.nobody@webmail2.brturbo.com>
Message-ID: <alg56j$akv$1@main.gmane.org>

<jubafre at brturbo.com>
> I have a list of lists like this:
>
> labels=[['D1'], ['D2'], ['D3']]
>
> and i want to transform ia a list of strings
>
> labels2==[ 'D1' ,  'D2' ,  'D3' ] <<<<<<<----------- how i can get
it????
>

You'll want a flatten function.  Look into these threads:


http://groups.google.com/groups?as_q=flatten&as_ugroup=comp.lang.python

--

Emile van Sebille
emile at fenx.com

---------






From ark at research.att.com  Wed Sep 25 11:43:15 2002
From: ark at research.att.com (Andrew Koenig)
Date: Wed, 25 Sep 2002 15:43:15 GMT
Subject: Newbie References Question -> Exemple
References: <3d91c72f$0$8558$79c14f64@nan-newsreader-02.noos.net> <3d91d2cf$0$1484$79c14f64@nan-newsreader-01.noos.net>
Message-ID: <yu99hegejcrw.fsf@europa.research.att.com>

Guy> But now, if I change the coordinate of one point:
Guy> p[0] = [5,5]
Guy> I will have:
>>>> p
Guy> [[5, 5], [0, 1], [1, 1], [1, 0]]
Guy> but:
>>>> quad
Guy> [[0, 0], [0, 1], [1, 1], [1, 0]]

That is because you are not changing the coordinates of the point; you
are causing p[0]--the 0th element of p--to refer to a completely new
point.

If you wanted to change the coordinates of the point, you should
have written this:

        p[0][0] = 5
        p[0][1] = 5

or, more succinctly:

        p[0][:] = [5, 5]

If you try this, I think you will get the behavior you want.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From jdavis at empires.org  Tue Sep 24 05:56:20 2002
From: jdavis at empires.org (Jeff Davis)
Date: Tue, 24 Sep 2002 09:56:20 GMT
Subject: Strange values in Accept-Language header?
References: <f72dac0d.0209231303.12349f2e@posting.google.com>
Message-ID: <8FWj9.39597$V7.10466682@twister.socal.rr.com>

> 
> Accept-Language: en-us,x-ns1rDGeT4e2FpA,x-ns2f67971hgDw1
> Accept-Language: x-ns13V8cN8S1Xz9,x-ns2T329gfxKa7d
> 
> I have no idea what these could mean (I've changed the actual values
> BTW, on the remote off-chance that somebody's highly personal
> information was coded in there)

If I had highly personal information, I probably wouldn't send it out with 
all of my HTTP requests ;-)

> 
> 
> D: Some form of cracking attempt involving a buffer-overrun
> vulnerability?
> 

That seems the most likely of your suggestions. 

I really don't know what those extra characters mean. Do you think it might 
have to do with some related preference like fonts or the screen or 
something? Perhaps if you posted the secret info someone else might notice 
something.

Regards,
        Jeff


From imbosol at vt.edu  Wed Sep 25 14:28:45 2002
From: imbosol at vt.edu (Carl Banks)
Date: Wed, 25 Sep 2002 14:28:45 -0400
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com> <3D8E6353.7030303@something.invalid> <96zj9.150553$pX1.5399868@news2.tin.it> <3D911CFA.4060405@something.invalid> <Wefk9.163287$pX1.5841874@news2.tin.it>
Message-ID: <amsvhl$cql$3@solaris.cc.vt.edu>

Alex Martelli wrote:
> Greg Ewing wrote:
>> 
>> I agree there are good reasons to prefer spaces-only
>> for code that will be transmitted or displayed by
>> dumb software.
> 
> To put it differently: use tabs only if you ARE willing to
> bet a very substantial amount that the source file in question
> will NEVER be transmitted, displayed, nor otherwise processed
> in any way, shape, or form, by software you consider "dumb"
> (take it up with KDE KNode's maintainers whether their news
> reader is "dumb"... I like it, though it doesn't show tabs:-).


Dumb?  I would call it broken.


-- 
CARL BANKS
http://www.aerojockey.com


From quatrelle at gmx.ch  Sat Sep  7 19:09:52 2002
From: quatrelle at gmx.ch (Colin Thefleau)
Date: Sun, 08 Sep 2002 01:09:52 +0200
Subject: serve_forever() and non-blocking connection
Message-ID: <ale100$bj4$1@nets3.rz.RWTH-Aachen.DE>

Hi

I have a problem programing a small proxy with non-blocking sockets. I
guess asyncore would be better, but I have no idea how to implement it
and the application is runing now (just too slow).
So I would like to use threads. First here a small example of what I use
for the connection (inside  a function):

http_proxy = HTTPServer(configuration.listen, RequestHandler)
http_proxy.serve_forever()

Now this serve_forever() process one request at a time.
Ok this is working but too slow. I tried to to use handle_request()
instead of serve_forever() in a while loop and put some

thread.start_new_thread(http_proxy.handle_request, ())

in there. But this looks too dirty and apparently does not help much.

So here my newbee question: what would be the (better) right approach to
be able so serve many connections at the same time? Knowing that each
process can be quite CPU hungry.

Any advice is greatly aprecieted
Thanks
Colin



From peter at engcorp.com  Thu Sep 19 19:30:44 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 19 Sep 2002 19:30:44 -0400
Subject: Nokia 3310 SMS Application, done:connecting to serial port(linux/win),
 done:reading some hex like \x00 . . ., problem:write/send command! noresponse
References: <52facf1d.0209190533.3048e62c@posting.google.com> <kFli9.146453$5r1.5856420@bin5.nnrp.aus1.giganews.com>
Message-ID: <3d8a5e19@news.sentex.net>

Terry Reedy wrote:
> "Polerio Babao Jr.II" <ptbjrii at hotmail.com> wrote in message
> news:52facf1d.0209190533.3048e62c at posting.google.com...
> 
>>Greetings: Python Masters
>>
>>I wanted to create an SMS application using Nokia 3310 with
>>DAU-9P(data cable),
> 
> <snip>
> 
>>When I connect to the port for the first time I get the ser.read()
>>output like this \x00\x0c I think. But that's two response is what I
>>get on my first connect. After that when I envoke ser.read() again,
> 
> I
> 
>>get nothing.
> 
> <snip>
> 
> I know nothing about Nokia but have had experience (and battle scars)
> programming serial i/o.  Two suggestions.
> 
> First: make *sure* you have serial protocol set right --  baud rate,
> data bits, stop bits, parity, handshaking?  Also make sure you have
> proper pins active and properly connected.  It is easy to be off.
> 
> Second: if you do not have one, get a breakout box that lets you see
> activity at the port.  Its the equivalent of having a
> volt-ohm-ammeter.  You plug cable into box and run another cable from
> box to port.  A simple one just has lights that flash when a signal
> passes thru.  More complicated one allows some signal patching.  I got
> mine long ago from Inmac.  Would check google for source now.

Third: almost anything you can do from Python you could do using
something like Hyperterminal (ugh) under Windows, or miniterm
under Linux.  If you can't get it working that way, you're probably
dealing with a non-Python issue.  If you *can*, then duplicating the
configuration carefully as Terry advises above in "First:" will
almost certainly leading to better results.


-Peter



From spendius at muchomail.com  Tue Sep  3 04:09:33 2002
From: spendius at muchomail.com (Spendius)
Date: 3 Sep 2002 01:09:33 -0700
Subject: Newbie's unsurmountable pb with IO stream trick
References: <aba30b75.0209020633.26ba4c70@posting.google.com> <3D738EE2.3A8756D@engcorp.com>
Message-ID: <aba30b75.0209030009.3ccb0207@posting.google.com>

Peter Hansen <peter at engcorp.com> wrote in message...

OK I forgot to say I'm trying to develop my stuff under Windoze.

> at once here.  For one thing, why are you trying to use 
> global variables at all when the original code was in
> Java, which doesn't have such things?
Do you mean Java doesn't have global variables or I'm mistaken ?!?
 
> And what does not declaring variables have to do with
> the "global keyword issue" (whatever that is)?
I kept getting the error message
# UnboundLocalError: local variable 'cpt' referenced before assignment
where I really didn't expect to get such an error, and noticed in this
forum a lot of guys who know Python complain and talk themselves about
a "mess" (my 'cpt' var. is a counter incremented every time a method
gets called and I must sometimes re-initialize it to 0 inside an 'if'
statement: merely IMPOSSIBLE). 
 
> If you have an SQL database to access, why not use one of the
> standard DBAPI modules to get to it?  You might be doing a lot
> of extra work when you don't have to.
And why not ? Anyway SQL*Plus contains plenty of built-in commands
that allow you to format your output how you like: I'd like to
keep that functionality. And let's say that I want to do that for
the sole purpose of learning the handling of IO streams in Python !
 
> If you're just learning Python, you've taken on a fairly
> complicated and large task right at the beginning.  Why not
I admit it. But I did *exactly* the same with Java (writing
the app I'm trying to port) but never got stuck into problems
that blocking...
 
Thanks.


From paul at boddie.net  Thu Sep 19 03:29:01 2002
From: paul at boddie.net (Paul Boddie)
Date: 19 Sep 2002 00:29:01 -0700
Subject: linux python ideas
References: <3D88E578.6010001@diespammerdieuselesspython.com> <amapge$22r$1@peabody.colorado.edu>
Message-ID: <23891c90.0209182329.52b55550@posting.google.com>

Fernando P?rez <fperez528 at yahoo.com> wrote in message news:<amapge$22r$1 at peabody.colorado.edu>...
> Rob Andrews wrote:
> 
> > I'll be giving a presentation on Python to my local LUG
> > (http://lugoj.org) in a few weeks, and would like to point out at least
> > a few things that would be of particular interest to linux users.
> 
> <blatant plug>
> Check out ipython at http://www-hep.colorado.edu/~fperez/ipython

I'd very much recommend a demonstration which uses the interactive
mode of Python. That was one of the things that impressed me about
Python in the beginning: the ability to do things like connecting to
servers interactively and with barely any effort, and I amused myself
by perusing the Python newsgroups at the Python prompt and thinking
how straightforward it would be to write a newsreader in Python
instead of C, for example.

Of course, stuff like "network programming" is fairly easy with other
languages, but unless Perl has spawned an interactive mode, for
example, you'll be showing people things they haven't seen before if
you show some experimentation at the command prompt. You might also
get bonus points by then demonstrating the same things again in
Jython. :-)

Paul


From max at alcyone.com  Mon Sep 30 22:28:36 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 30 Sep 2002 19:28:36 -0700
Subject: ANN: empy 2.0 -- Embed Python in template text as markup
Message-ID: <mailman.1033470243.12282.clpa-moderators@python.org>

Summary

    empy is a system for embedding Python expressions and statements
    in template text; it takes an empy source file, processes it, and
    produces output.  This is accomplished via expansions, which are
    special signals to the empy system and are set off by a special
    prefix (by default the at sign, '@').  empy can expand arbitrary
    Python expressions and statements in this way, as well as a
    variety of special forms.  Textual data not explicitly delimited
    in this way is sent unaffected to the output, allowing Python to
    be used in effect as a markup language.  Also supported are "hook"
    callbacks, recording and playback via diversions, and dynamic,
    chainable filters.  The system is highly configurable via command
    line options and embedded commands.

    Expressions are embedded in text with the '@(...)' notation;
    variations include conditional expressions with '@(...?...:...)'
    and the ability to handle thrown exceptions with '@(...$...)'.  As
    a shortcut, simple variables and expressions can be abbreviated as
    '@variable', '@object.attribute', '@function(arguments)',
    '@sequence[index]', and combinations.  Full-fledged statements
    are embedded with '@{...}'.  Forms of conditional and repeated
    expansion are available via '@[...]'.  A '@' followed by a
    whitespace character (including a newline) expands to nothing,
    allowing string concatenations and line continuations.  Comments
    are indicated with '@#' and consume the rest of the line, up to
    and including the trailing newline.  '@%' indicate
    "significators," which are special forms of variable assignment
    intended to specify per-file identification information in a
    format which is easy to parse externally.  C-style escape
    sequences can be specified with '@\...', and finally a '@@'
    sequence expands to a single literal at sign.


Getting the software

    The current version of empy is 2.0.

    The latest version of the software is available in a tarball here:
    "http://www.alcyone.com/pyos/empy/empy-latest.tar.gz",
    http://www.alcyone.com/pyos/empy/empy-latest.tar.gz.

    The official URL for this Web site is
    "http://www.alcyone.com/pyos/empy/",
    http://www.alcyone.com/pyos/empy/.


Requirements

    empy should work with any version of Python from 1.5.x onward.


License

    This code is released under the "GPL",
    http://www.gnu.org/copyleft/gpl.html.

...

Release history [since 1.x]

    - 2.0; 2002 Sep 30.  Parsing system completely revamped and
      simplified, eliminating a whole class of context-related bugs;
      builtin support for buffered filters; support for registering
      hooks; support for command line arguments; interactive mode with
      -i; significator value extended to be any valid Python
      expression.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ To be adult is to be alone.
\__/ Jean Rostand
    Crank Dot Net / http://www.crank.net/
 Cranks, crackpots, kooks, & loons on the Net.



From jepler at unpythonic.net  Mon Sep 30 14:34:04 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Mon, 30 Sep 2002 13:34:04 -0500
Subject: conditional expressions
In-Reply-To: <oOZl9.162526$ub2.3640508@news1.tin.it>
References: <33b491f.0209181820.17b6ea70@posting.google.com> <ofLk9.304267$z91.13145351@bin3.nnrp.aus1.giganews.com> <dITk9.175519$pX1.6209886@news2.tin.it> <t_3l9.334924$AR1.14880141@bin2.nnrp.aus1.giganews.com> <oOZl9.162526$ub2.3640508@news1.tin.it>
Message-ID: <20020930183403.GB21522@unpythonic.net>

On Mon, Sep 30, 2002 at 03:09:08PM +0000, Alex Martelli wrote:
> I've shown this i18n'd with gettext's convention _( ... ) , but
> of course not all languages can pluralize by a constant suffix
> (English can't -- if ever a number of leaves or fishes was
> among the found things, we'd be in trouble...!-), so a better
> approach might be, e.g., for English:
[...]

It's even worse than this.  For instance, a language where the form for
1 is unique, numbers ending with 2, 3, 4 (except for 12, 13, and 14 --
I'm not sure about 111, 1211, etc) get a second form, and the rest of
the numbers get a third form.  I quote from the gnu gettext manual:

 Additional functions for plural forms
 -------------------------------------

    The functions of the `gettext' family described so far (and all the
 `catgets' functions as well) have one problem in the real world which
 have been neglected completely in all existing approaches.  What is
 meant here is the handling of plural forms.

    Looking through Unix source code before the time anybody thought
 about internationalization (and, sadly, even afterwards) one can often
 find code similar to the following:

         printf ("%d file%s deleted", n, n == 1 ? "" : "s");

 After the first complaints from people internationalizing the code
 people either completely avoided formulations like this or used
 strings like `"file(s)"'.  Both look unnatural and should be avoided.
 First tries to solve the problem correctly looked like this:

         if (n == 1)
           printf ("%d file deleted", n);
         else
           printf ("%d files deleted", n);

    But this does not solve the problem.  It helps languages where the
 plural form of a noun is not simply constructed by adding an `s'
 but that is all.  Once again people fell into the trap of believing
 the rules their language is using are universal.  But the handling
 of plural forms differs widely between the language families.
 For example, Rafal Maszkowski `<rzm at mat.uni.torun.pl>' reports:

      In Polish we use e.g. plik (file) this way:
           1 plik
           2,3,4 pliki
           5-21 pliko'w
           22-24 pliki
           25-31 pliko'w
      and so on (o' means 8859-2 oacute which should be rather okreska,
      similar to aogonek).
[.... about 250 more lines describing how gettext attempts to address
these situations]



From thomas.heller at ion-tof.com  Thu Sep 26 02:59:43 2002
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: 26 Sep 2002 08:59:43 +0200
Subject: SET_LINENO
References: <3D92A46E.2030706@cygnus-software.com>
Message-ID: <vg4tp76o.fsf@ion-tof.com>

Bruce Dawson <bruce_deletethis_dawson at cygnus-software.com> writes:

 
> P.P.S. A breakpoint instruction would also be a boon for asserts - I'd
> love to be able to stop on the assert when running under a debugger,
> to see what is wrong before the stack frame unwinds.

Have you seen this?
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/65287

Thomas


From henrik.motakef at web.de  Tue Sep 24 13:39:33 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 24 Sep 2002 19:39:33 +0200
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com> <amet8v$e50$1@scotsman.ed.ac.uk> <L6Gi9.119475$ub2.2533804@news1.tin.it> <3D8B3250.BA964BCF@ObjectWay.it> <amq6e5$85gpo$1@ID-139865.news.dfncis.de>
Message-ID: <87n0q7ffsa.fsf@pokey.henrik-motakef.de>

Dave Kuhlman <dkuhlman at rexx.com> writes:

> And, a large part of the book describes how to program without variables and 
> state.  That's weird and interesting, just like the Elephant Man was weird 
> and interesting.  But, after that, then what?  Am I supposed to try to 
> program that way?

Read "The Koan of Side Effect" at
<http://www.bagley.org/~doug/ocaml/Notes/okoans.shtml>. Become enlightened.

> It did not teach me good Python programming style.  I could not hardly read 
> my own code with all those many layers of parentheses.
> 
> It did not teach me how to use Python data structures.  I found myself doing 
> too many screwy things with lists.

As the book's title indicates, it's about how computer programs work,
not about using Python data structures.

Regards
Henrik


From mhammond at skippinet.com.au  Wed Sep 11 18:54:40 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 11 Sep 2002 22:54:40 GMT
Subject: untrusted dynamic module: win32api
References: <unk63h7q26lefe@news.supernews.com> <OAwe9.5047$_l1.12416@news-server.bigpond.net.au> <unmai51quucsde@news.supernews.com>
Message-ID: <QQPf9.22825$_l1.60831@news-server.bigpond.net.au>

Guy Rabiller wrote:
> [Mark]
> 
>>It appears that SoftImage is configuring ActiveScripting for "untrusted"
>>scripts - ie, SoftImage is requesting the scripting engine switch into
>>untrusted mode.  Thus, you really need to ask SoftImage is the
>>application can be configured to not request restricted execution mode
>>of the language
> 
> 
> After a closer look, it seems that this has nothing to do with Softimage,
> but with MSIE.
> 
> Even if I include ( for exemple ):
> 
> <SCRIPT Language="Python">
> import win32com.client
> w = win32com.client.Dispatch("Photoshop.Application")
> </SCRIPT>
> 
> in a simple html file, then launching it in MSIE, totaly outside from
> Softimage, it doesn't work, I still have the same error message.
> ( it works in VBScript and JScript )
> 
> I think this has to do with restricted execution mode, but I have no clue
> how to customize it for allowing the loading of 'untrusted dynamic modules'.
> 
> Any help on this would be greatly apreciated.

IE certainly does force restricted execution mode.  There is no standard 
way to remove this restiction so facilities like you request can be 
performed on any installation.  If necessary, you could modify the 
engine source code to remove this restriction - mail me for details.

Mark.



From jeremy at alum.mit.edu  Fri Sep 13 09:58:39 2002
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 13 Sep 2002 06:58:39 -0700
Subject: Problem with HTTPS request
References: <uo2fveepf88175@corp.supernews.com>
Message-ID: <b0f083db.0209130558.2aa60398@posting.google.com>

"Christian Reyes" <creyes at ea.com> wrote in message news:<uo2fveepf88175 at corp.supernews.com>...
> I got the following response when trying to make an HTTPS request.
> 
> <h1>SSL required to access this page</H1>
> 
> Then I read in the documentation the following
> 
> Note: HTTPS support is only available if the socket module was compiled with
> SSL support.
> 
> Does anyone know where i can get my hands on a socket module that was
> compiles with SSL support?

What platform are you running?  Many people compile such a socket
module themselves.  If that isn't an option, you'll have to look at
binary packages for your platform.

Jeremy


From somebody at nojunk.invalid  Sun Sep 29 18:15:21 2002
From: somebody at nojunk.invalid (Russell E. Owen)
Date: Sun, 29 Sep 2002 16:15:21 -0600
Subject: [Mac] Editor on Macintosh?
References: <1fj7za2.1erbqps1uvzpj8N%news@agapow.net>
Message-ID: <an7u1s$33de$1@nntp6.u.washington.edu>

In article <1fj7za2.1erbqps1uvzpj8N%news at agapow.net>, news at agapow.net 
(Paul-Michael Agapow) wrote:

>Sadly, it seems like the excellent editor Pepper is going away. (See
><http://www.hekkelman.com>.) I found it the best editor for Python on
>the Macintosh, having never been really happy with the alternatives. 
>
>So what are other MacPython using? There are some Java-based editors,
>but I'm not persuaded that that they'll perform adequately. There's are
>numerous X11 based options, but who wants to swap into X everytime they
>need to edit a file? The once excellent editor Alpha seems to have
>disappeared permanently into testing for its next release.
>
>Ideas? Anyone? Bueller?

BBEdit is the standard. On the whole it works well--it's polished, 
reliable, fast and programmable. However, I loathe its find/replace 
dialog box:
- To do a "search in selection" you have to UNCHECK "from the top" or 
"wrap" before you are allowed to check "in selection". What sadist 
programmed that?
- Then when you want to return to normal mode, you uncheck "in 
selection" and then check "wrap" or "from the top", and heaven help you 
if you forget to check one of those, because:
- If you leave both "from the top" and "wrap" unchecked, then a replace 
all does not do what it says, but instead it only replaces from the 
current cursor to the end of the document. I've corrupted quite a few 
source code files that way.
- The dialog box is modal. Good luck if you want to enter text by 
copying from or reading from your document and you forget to read/copy 
it before opening the dialog box. (I agree that it's nice to have the 
box go away after pressing "Find", but a non-modal dialog can be 
programmed to do that, too!).
Also it does not have rectangular selection, and you cannot edit files 
from browser windows (such as a find results window).

jEdit (the Java editor, not the Japanese editor) worked pretty well when 
I tried it, but has an unusual document model (a given window may show 
different files at different times). Maybe I'd get used to it, but I 
decided not to try because I had Pepper.

Alpha is quite configurable and programmable but also quite capable of 
getting into strange states (at least it has done so each of the few 
times I have test-driven it) where typing no longer works at all or does 
really weird things (it's been awhile, I don't remember the details). If 
there's a stable version that runs well under X perhaps you should just 
use it and not worry about future versions?

CodeWarrior's IDE is nice, but does not have Python syntax coloring.

Python has the Mac IDE is a bit limited (need an extra plug-in even to 
do syntax coloring) but OK. IDLE is showing promise and runs w/out 
xfree86 with Tcl/Tk 8.4 (which is now Aqua native). But Tk bugs may play 
havoc with key accelerators and generally cross-platform code doesn't 
seem to feel very Mac-like.

-- Russell

P.S. I wish Nisus Software hadn't stopped supporting Qued. It was a 
superb editor, but I think they couldn't maintain it and Nisus at the 
same time. It had first-rate multiple undo from the beginning, 
rectangular selections and even multiple selections (surprisingly handy).

Pepper is the best I've found and is what I use. It has some rough 
edges, but very clean design. Unfortunately, the author really decided 
to pick up his toys and go home. You cannot buy a license anymore, you 
cannot get a free license, you cannot get source code.


From eddie at holyrood.ed.ac.uk  Fri Sep 20 06:28:47 2002
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Fri, 20 Sep 2002 10:28:47 +0000 (UTC)
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com>
Message-ID: <amet8v$e50$1@scotsman.ed.ac.uk>

If you want to learn "how it (computer languages) works" I suggest reading
"Structure and Interpretation of Computer Languages" by Abelson and Sussman.
This is about understanding the principles rather than how to do things in a
particular language and is probably the best book on the subject.

Eddie



From Janssen at rz.uni-frankfurt.de  Sun Sep 22 07:02:41 2002
From: Janssen at rz.uni-frankfurt.de (Michael Janssen)
Date: Sun, 22 Sep 2002 13:02:41 +0200
Subject: Python about file I/O
References: <amgnnl$5qc9n$1@ID-49758.news.dfncis.de>
Message-ID: <3d8da3df@nntp.server.uni-frankfurt.de>

Ken wrote:

> Hi, I have 2 questions on IO, not necessary related to one another:
> 
> 1.) How do you list the filenames within a given directory so user can
> choose which file to open?
> 
> 2.) How do you ensure the program doesn't crash if I specify a
> filename to open but the file doesn't exist? (I.E. how to check if
> file exist before opening the file?)
> 
> Thanks

existence-check is:
os.access(file,os.F_OK)

Michael


From dsavitsk at e-coli.net  Thu Sep 26 01:31:22 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Thu, 26 Sep 2002 05:31:22 GMT
Subject: Python and Scite
References: <mailman.1033005453.8052.python-list@python.org>
Message-ID: <KYwk9.4$XX3.8191@newssrv26.news.prodigy.com>

PythonWin is my favorite editor, and it is in the win32all (python for
windows extensions) package at http://starship.python.net/crew/mhammond/

-d


"Sperry Russ" <sperry at newcreations.info> wrote in message
news:mailman.1033005453.8052.python-list at python.org...
Chris,

Thanks for your help.  The problem was in the path  which I did not
understand at first. Is th PythonWin your favorite editor?

Sperry


----------------------------------------------------------------------------
-----------------------------
> I am new to programming.  I have downloaded Python 2.2.1 and Scite
> 1.48 to learn how to program.  I am trying to configure my Scite so
> that it will run the python files by using go.  When I try to do this
> i get a message in the second window at the bottom -- unable to find
> file.  I have tried configuring the global options and have failed can
> anyone help me? I need a step by step example if you can because I
> cannot understand it according to the documentation that came with
> Scite

works out of the box for me. if you're on windows, make sure that python is
found in the PATH. i.e. edit PATH that it contains the python directory.
and you could look at PythonWin, it uses the editing component used in
scite and has some additional features for the python programmer.
additionaly it comes also with useful, de-facto standard, windows specific
extensions.








From aleax at aleax.it  Mon Sep 30 11:09:08 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 30 Sep 2002 15:09:08 GMT
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <ofLk9.304267$z91.13145351@bin3.nnrp.aus1.giganews.com> <dITk9.175519$pX1.6209886@news2.tin.it> <t_3l9.334924$AR1.14880141@bin2.nnrp.aus1.giganews.com>
Message-ID: <oOZl9.162526$ub2.3640508@news1.tin.it>

Terry Reedy wrote:
        ...
>> It's _dangerous_ -- to use it, one always has to be fully aware
>> of the risk that subexpression b might be false.
> 
> If b is a non-null constant, there is no such risk.  Consider

If it's a constant, you don't need short-circuiting!

> na=1; np=3; nl=1; ng=0
> print "Found: %d apple%s, %d plum%s, %d lemon%s, and %d grape%s." %\
> (na, na!=1 and 's' or '', np, np!=1 and 's' or '', nl, nl!=1 and 's'
> or '', ng, ng!=1 and 's' or '')
> 
> Found: 1 apple, 3 plums, 1 lemon, and 0 grapes.
> 
> Alternatives to this 'risky' construct:

def plur(*ns):
    for n in ns:
        yield n
        yield 's'[n!=1:]

print "Found: %d apple%s, %d plum%s, %d lemon%s, and %d grape%s." % tuple(
    plur(na, np, nl, ng))

You can inline plur's logic if you want, of course -- although I
believe it's more readable in this way.  But anyway, the point is
that slicing is handier than and/or in this important special case.
For other cases in which you don't need short-circuiting, indexing
is fine -- " ('ba','glor')[n!=1] " if that's what a certain natural
language wants for singular/plural endings.  But when one accepts
that a function for such pluralization is preferable anyway, then
if-else becomes attractive once again:

def plur(*ns, singular=_(''), plural=_('s')):
    for n in ns:
        yield n
        if n == 1: yield singular
        else: yield plural

I've shown this i18n'd with gettext's convention _( ... ) , but
of course not all languages can pluralize by a constant suffix
(English can't -- if ever a number of leaves or fishes was
among the found things, we'd be in trouble...!-), so a better
approach might be, e.g., for English:

def plur(*numbers_and_nouns):
    for i in range(len(numbers_and_nouns)):
        number, noun = numbers_and_nouns[i:i+2]
        yield "%d %s" % (number, pluralize(number, noun))

where:

plural_exceptions = {'man':'men', 'child':'children', # etc etc
        'leaf':'leafes', 'fish':'fishes',             # etc etc
    }
def pluralize(number, noun):
    if number == 1:
        return noun
    else:
        return plural_exceptions.get(noun, noun+'s')


You could factor this up in different ways, but please DO ensure
your software never output "plurals" such as "mans", "childs",
"leafs", or "fishs".

> 1. forgo the nicety of correct English output

That's basically what you're doing when you choose to just
conditionally append s's.  If you do take such a shortcut
(because you know that none of the nouns you handle can ever
be one of the many exceptions in English pluralization),
then slicing is one option -- but overall I prefer a short
function to wrap the pluralization task, I think that the
ability to give the function a name helps clarity a LOT,
whether you ARE willing to "just stick an s at the end" OR
waht to ensure "the nicety of correct English output".

> 2. nest if/else four levels deep to choose correct variant of 16 print
> statements (ugh)
> 3. precalculate 4 suffixes and store in 4 temp vars with 4 if/else's
> (8 or 16 lines);
>    though clearly superiour to 2), this is still tedious with some
> risk of typo error.

"Risk of typo error" is quite present even in your repetition of
the and/or idiom (and to a lesser extent in a similarly inlined
repetition of the slicing or indexing idioms).  Better to wrap
things into a function: NO repeated code...!


Alex



From sscherte at mailbox.epfl.ch  Tue Sep 10 04:25:36 2002
From: sscherte at mailbox.epfl.ch (Sebastien Schertenleib)
Date: Tue, 10 Sep 2002 10:25:36 +0200
Subject: Embedding python with a C++ multithread apps -> pause and resume python script
Message-ID: <001601c258a3$a3e19a20$4c4eb280@epfl.ch>

Hi,

I am developping a C++ application and I want to embedd python. I have succeeded to export my C++ classes and data using SWIG then I was able to create multithreading call to python and now I have my app running several script at the same time (~). However, I am stack as I would like to be able to pause and resume my python script from C++ side. 

If we imagine a looping script, my C++ thread in still in the Py_Run_SimpleString() method, and I would like to pause it inside python.

I have try to go directly inside the Interpreter state with all the thread queue but it seem that I don't understand how python access to data since I have set to NULL some thread state and my script continue to run as it.

If someone can give me a hint how to block a python script directly from C++, it would help me a lot.

A second problem is that when I try to redirect the sys.stderr and sys.stdout from python (create a python class which has a write method, and then which called inside my C++ method), It's working only until a certain time then it crash, how can I do a redirection in case of a c++ multithreading app? (I have synchronized the C++ method as well))

Thanks a lot for any help!
 

So, here is my thread run method:

/* Wait for the initialization thread if there is one.*/
if (_initScript != _PythonScriptHandle(NULL))

    _initScript->_thread->join();

PyEval_AcquireLock();

PyInterpreterState *interp = _service->_mainTstate->interp;

// create new PyThreadState only if needed

JTCThread * currentThread = JTCThread::currentThread();

PyThreadState * tstate;

vhdPythonServiceBody::_Impl::ThreadInfoMap::iterator it = _Impl::threadInfoMap.find(currentThread);

if (it == _Impl::threadInfoMap.end())

{

    vhdPythonServiceBody::_Impl::ThreadInfo threadInfo;

    threadInfo.pyThreadState = PyThreadState_New(interp);

    threadInfo.outputBuffer = "";

    it = vhdPythonServiceBody::_Impl::threadInfoMap.insert    (vhdPythonServiceBody::_Impl::ThreadInfoMap::value_type( currentThread, threadInfo)).first;

}

tstate = (*it).second.pyThreadState;

PyThreadState_Swap(tstate);


/* Execute the given script.*/

switch (PyRun_SimpleString((char*) _scriptText->cstr()))

{

case 0:

    _executionResult=RES_OK;

    break;

case -1:

    _executionResult=RES_ERROR;

    break;

}

PyThreadState_Swap(NULL);

// only for internal threads created by vhdPythonService (by using send for execution methods)if (_bServiceInternalThread)

{

    // Reset the thread state structure.

    PyThreadState_Clear(tstate);

    // Nullify the current thread state, release the global 

    // interpreter lock, delete the thread state structure.

    PyThreadState_Delete(tstate);

    // removal of ThreadInfo that was used for internal thread

    // we will not resues it as it was only for the internal thread

     _Impl::threadInfoMap.erase(currentThread);

}


PyEval_ReleaseLock();


/* Notify the service about the termination of the thread.

*/

_service->_notifyTermination(this);

}




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020910/d9561698/attachment.html>

From chris.gonnerman at newcenturycomputers.net  Sun Sep 29 18:33:57 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sun, 29 Sep 2002 17:33:57 -0500
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05> <an7bvd$burft$1@ID-125932.news.dfncis.de> <an7p7t$cpg$0@216.39.172.122>
Message-ID: <006701c26808$5385cd20$ba01010a@local>

----- Original Message ----- 
From: "Bengt Richter" <bokr at oz.net>


> On 29 Sep 2002 17:06:53 GMT, Christopher Browne <cbbrowne at acm.org> wrote:
> 
> >The problem with FP is that FP involves the use of binary
> >approximations, which commonly cannot exactly represent 
> >decimal values.
> >
> >Notably, the decimal fraction 0.1, which is /exactly/ 1/10, 
> >does not have any exact encoding in binary.  Its 
> >representation in binary is a repeating binary fraction.
>
> How is that a significantly different problem from repeating 
> decimal fractions? (E.g., 1/3. or 1/7. or wxyz/9999. which 
> will give you 0.wxyzwxzy..?)

Humans think in decimal.  To a human, 1.0/10.0 == 0.1, not
0.10000000000000001.  OK, so this *rounds* to 0.10 at two
decimal places, but what about:

>>> .70 * .05
0.034999999999999996

Rounded to two decimals, that's 0.03; but done in proper
decimal numbers, the answer should be 0.035, rounded to 0.04
using either "classic" or "banker's" decimal rounding.

If you are calculating sales tax, for instance, this is 
unacceptable.

(I borrowed this example from memory from a previous poster,
and if I could remember who it was I'd give credit...)

You ask how it is a different problem.  Simple.  A human,
training in decimal math in grade school, knows that 

    1/3 = 0.3333333...

and would *expect* it to round to 0.33 (still in two decimal
places).  The problem with *binary* floating point is the
"surprise factor."

So, Bengt, how do I do this right using only floats?  I can't.
I have to have decimal arithmetic to get the right answer.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net





From juliana at pdi.inpe.br  Tue Sep 24 14:58:48 2002
From: juliana at pdi.inpe.br (Juliana Braga)
Date: Tue, 24 Sep 2002 15:58:48 -0300
Subject: help
Message-ID: <amqclg$6uu$1@server.ltid.inpe.br>

I need help.
For example:
x = ?name1?
How can I get the value of variable x (in this case the value of variable x
is ?name1?). After , I?d like to do this:
value of variable x = ?teste?.





From joconnor at cybermesa.com  Wed Sep  4 20:28:39 2002
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Thu, 05 Sep 2002 07:28:39 +0700
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de>
Message-ID: <20020905.072837.159259470.2038@cybermesa.com>

In article <j4d6rt61nd.fsf at informatik.hu-berlin.de>, "Martin v. L?wis"
<loewis at informatik.hu-berlin.de> wrote:

> Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:
> 
>> >Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:
>> >
>> >> So why hasn't this been extended when nested scopes were introduced,
>> >> e.g. by reusing the global directive or some other one?
>> >
>> >I think the main reason is that nobody could propose an acceptable
>> >syntax. Reusing the global directive does not work.
>> 
>> Can you explain why it does not work?
> 
> Consider
> 
> x = 1
> 
> def foo():
>   x = 2
>   def bar():
>     x = 3
>     def baz():
>       global x
>       x = 4
> 
> If you invoke baz(), which of the variables is changed? How do you
> denote that you want to change the others?
> 
> Regards,
> Martin


Why not simply disallow the using of a name already defined in an
enclosing scope? Similar to Smalltalk?

Consider the following:

| y | 
aBlock := [
	| x | 
	y := [	
		| x |
		x := 5
		x
	] value

] value .
^y

Here I'm defining x twice, once inside the outer most block and once
inside the inner most block.  Attempting to execute this (actually,
attempting to compile it) raises an error "x is already defined (perhaps
in an outer block)"

changing the inner x to 'x1' or something solves the problem

This works for instance variables in methods of a class, too.  If an
object has an instance variable named foo...you can't have a method with
a temporary variable named foo also, the compiler will error out with
the above error about foo already being defined

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.r4h.org/r4hsoftware


From mike at bindkey.com  Fri Sep 27 18:24:04 2002
From: mike at bindkey.com (Mike Rovner)
Date: Fri, 27 Sep 2002 15:24:04 -0700
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <amvcev$cf6$1@s1.read.news.oleane.net> <7nHk9.172208$pX1.6117207@news2.tin.it> <ofLk9.304267$z91.13145351@bin3.nnrp.aus1.giganews.com> <dITk9.175519$pX1.6209886@news2.tin.it> <t_3l9.334924$AR1.14880141@bin2.nnrp.aus1.giganews.com>
Message-ID: <an2lte$amv$1@main.gmane.org>

"Terry Reedy" <tjreedy at udel.edu> wrote in message
news:t_3l9.334924$AR1.14880141 at bin2.nnrp.aus1.giganews.com...
> na=1; np=3; nl=1; ng=0
> print "Found: %d apple%s, %d plum%s, %d lemon%s, and %d grape%s." %\
> (na, na!=1 and 's' or '', np, np!=1 and 's' or '', nl, nl!=1 and 's'
> or '', ng, ng!=1 and 's' or '')
>
> Found: 1 apple, 3 plums, 1 lemon, and 0 grapes.
>
> Alternatives to this 'risky' construct:
> 1. forgo the nicety of correct English output
> 2. nest if/else four levels deep to choose correct variant of 16 print
> statements (ugh)
> 3. precalculate 4 suffixes and store in 4 temp vars with 4 if/else's
> (8 or 16 lines);
>    though clearly superiour to 2), this is still tedious with some
> risk of typo error.
>
4. Try:
use_s=('','s')
print "Found: %d apple%s, %d plum%s, %d lemon%s, and %d grape%s." %\
(na, use_s[na!=1], np, use_s[np!=1], nl, use_s[nl!=1], ng, use_s[ng!=1])







From mstenner at phy.duke.edu  Wed Sep  4 13:29:51 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Wed, 4 Sep 2002 13:29:51 -0400
Subject: simple xml pickle
In-Reply-To: <QEjd9kKkXEbH092yn@gnosis.cx>; from mertz@gnosis.cx on Wed, Sep 04, 2002 at 12:13:04PM -0400
References: <mailman.1031104505.31870.python-list@python.org> <QEjd9kKkXEbH092yn@gnosis.cx>
Message-ID: <20020904132951.A7928@phy.duke.edu>

On Wed, Sep 04, 2002 at 12:13:04PM -0400, David Mertz, Ph.D. wrote:
> Michael Stenner <mstenner at phy.duke.edu> wrote previously:
> |  the "old" xml_pickle.py, which appears to not work at all with
> |  python2.2, but is otherwise appealing....
> |  the "new" xml.pickle as it appears in the Gnosis tools.  This is
> |  less appealing because it's very large and distributed over several
> |  files...
> 
> I think there are some good reasons why the "new" gnosis.xml.pickle is
> organized into a package namespace.  And there are some new capabilities
> and modularity that go along with that.

You're quite right.  I should have said "less appealing as a solution
in my current task" or something similarly qualified.

> However, the "old" xml_pickle (use the latest such one, labelled 0.51)
> can be fixed up for Python 2.2 by adjusting two uses of dir().  That
> built-in function changed meaning in Python 2.2--or at least it changed
> behavior.  Just replace occurrences of 'dir(obj)' with
> 'obj.__dict__.keys()', and I'm pretty sure all will be well.

That's the version that I was playing with last night.  I'll take
another look at it.  I think there were some api changes in the DOM
stuff too, though.  I'm not sure of the specifics.

The xml marshalling does almost exactly what I want, and the
modifications to make it "standalone" for Python 2.2 were very easy
(patch forthcoming) so I may just stick with that.  Nonetheless, I'll
look at xml_pickle again and see what I can see.

> Yours, David (the xml_pickle guy)...

Yeah, I remembered the name... when you a google search on "xml
pickle" you see "David Mertz, Ph.D" popping up everywhere :)

					-Michael
-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From steve at ferg.org  Fri Sep 13 09:06:40 2002
From: steve at ferg.org (Stephen Ferg)
Date: 13 Sep 2002 06:06:40 -0700
Subject: Newbie Looking for direction
References: <uo23o95eb8tg87@corp.supernews.com>
Message-ID: <b16e4ef7.0209130506.457b1db3@posting.google.com>

> So far it seems like I'm a long way away from being able to be productive
> with Python...

This might help in getting you productive quickly:  

http://www.ferg.org/easygui/


From jnana4 at DELETEhotmailCAPS.com  Sun Sep  1 16:56:34 2002
From: jnana4 at DELETEhotmailCAPS.com (jano)
Date: Sun, 01 Sep 2002 20:56:34 GMT
Subject: EOFError using cPickle.load
Message-ID: <6avc9.264595$me6.34329@sccrnsc01>

Hi,

I am getting an EOFError when i try to load a cPickled object that was
cPickled in binary format.

>>> import cPickle
>>> cPickle.dump(keywordDict, open('pickle.p', 'w'), 1)
>>> somedict = cPickle.load(open('pickle.p', 'r'))
Traceback (most recent call last):
  File "<pyshell#60>", line 1, in ?
    somedict = cPickle.load(open('pickle.p', 'r'))
EOFError
>>>

It works fine with small test dictionaries, but it doesn't seem to like my
keywordDict dictionary, which is a very long inverted index, an excerpt of
which follows:

{'cellularli': ['4620'], 'non-cooper': ['2350'], 'liar': ['5800', '5800'],
'consider': ['880', '1270', '1690', '3420'], 'founder': ['3420'], 'illustr':
['3180'], 'fly-bottl': ['6090'], 'hatr': ['1200', '3490'], 'instinctu':
['2260'], 'four': ['20', '1640', '1800', '3420', '3700'], 'presupposit':
['4400'], 'sleep': ['2300'], 'phenomenologist': ['3350', '3350', '3420'],
'conclus': ['470', '1830', '2990'], 'follow': ['390', '390', '1070', '2030',
'3700', '3880'], 'children': ['1990', '5030'], 'abid': ['650'], 'skeleton':
['1110'], 'accus': ['3490'], 'incap': ['1060'], 'tu': ['3960'], 'swai':
['1600'], 'articl': ['4230', '4350'], 'self-piti': ['1200'], 'program':
['440'], 'invari': ['1250'], 'ti': ['4530', '5280'], 'hermann': ['3420'],
'smile': ['1070'], 'pride': ['5180'], 'worth': ['1220', '2480', '2490',
'2790', '3010', '4750', '5890'], 'volit': ['1690', '1710'], 'case': ['1270',
'3700'], 'lure': ['3710'], 'meritrend': ['700'], 'everi': ['390', '500',
'610', '820', '840', '1090', '1150', '1190', '1290', '1530', '1620', '1830',
'1830', '2050', '2060', '2400', '2410', '2620', '2630', '2660', '2670',
'2890', '3240', '3320', '3330', '3460', '3670', '3670', '3670', '4110',
'4120', '4160', '4440', '4520', '4520', '4800', '4900'], 'parcel': ['390'],
'physician': ['5670'], 'fals': ['570', '1180'], 'void': ['1020', '1250',
'2560'], 'rise': ['3620', '4000'], 'lurk': ['3610'], 'voic': ['3710'],
'util': ['390'], 'lumin': ['1190'], 'candid': ['2180'], 'mechan': ['300'],
'lowland': ['140'], 'fall': ['0', '1160', '1250', '1740', '3710', '3710',
'3710', '3790'], 'veri': ['450', '1190', '1320', '1490', '1640', '1900',
'2470', '2680', '2680', '2680', '3460', '3490', '3520', '3680', '4590',
'4620', '4670', '4670', '4670', '5470'], 'affect': ['250'], 'puffi':
['250'], 'strang': ['700', '1480', '3160'], 'vast': ['2870'], 'school':
['1360', '1370', '2370'], 'heaven': ['710', '5260'], 'affirm': ['3310',
'3310'], 'previou': ['3330'], 'sand': ['4060'], 'subhuman': ['1270'],
'tereza': ['3700'], 'small': ['330', '3630', '4620'], 'customari': ['2830'],
'everth': ['1380'], 'disabus': ['1020'], 'triumph': ['1470'], 'fritter':
['5560', '5590'], 'enjoy': ['4570'], 'chew': ['270'], 'abil': ['5870',
'6180'], 'direct': ['390', '390', '1780', '4340'], 'dogmat': ['3220'],
'past': ['20', '430', '430', '430', '440', '1270', '1320', '1960', '1960',
'2060', '3330', '3530', '3530'], 'consequ': ['1680', '1830', '2550'],
'second': ['110', '3490', '3700', '3700', '3700', '5140'], 'street': ['800',
'870', '3700'], 'design': ['4430', '4560'], 'pass': ['2890', '3260',
'5140'], 'link': ['3470'], 'illumin': ['3600', '4620'], 'theory': [],
'excit': ['1020', '1210'], 'aim': ['380', '390', '1420', '5110', '6090'],
'hide': ['2700', '3500', '3730', '4340'], 'barrier': ['3750'], 'microphon':
['3620', '3700'], 'solemn': ['250'], 'sun': ['4340', '4340'], 'sum':
['1280']}

It works fine in text mode, but is too slow, so i tried it in binary mode,
but i get the EOFError everytime.

Anybody know what is going on here or how I can get this to work in binary
mode?

thanks,

jano




From fperez528 at yahoo.com  Mon Sep 23 16:04:49 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 23 Sep 2002 14:04:49 -0600
Subject: Non-interpreted Python?
References: <3D8F65A4.38124945@engcorp.com> <y7Kj9.400$UC2.36205@read2.inet.fi>
Message-ID: <amns51$abt$1@peabody.colorado.edu>

Tapio Kelloniemi wrote:

>>There are other approaches, but there's no point describing them until
>>you explain why you want to do this.  What are your requirements?
> I'm using Linux and searching for a good high-level language to use in
> my projects. I don't have any special ideas, it depends on language.

Early optimization is the root of all evil. --- D. Knuth.

_After_ you:

1. Know your code is correct.
2. Know you _do_ have a problematic bottleneck.
3. Know that bottleneck is due to python's execution speed (and not choice of 
a poor algorithm, I/O or network bottlenecks, etc).

_Then_ you consider optimizing it away in a compiled language. At that point, 
python offers a plethora of solutions: Numeric (for number crunching), pyrex, 
weave, boost.python, SWIG, CXX, SCXX, ...  Some easier than others, some more 
tailored to some problems than others. 

But don't worry, chances are you'll find that 9 times out of 10 python is 
fine, and for the cases when there is indeed a critical bottleneck (which of 
course do exist), there are plenty of very convenient ways of addressing that 
problem in python.

Cheers,

f.

ps. Anyone care to put a summary of this issue into a FAQ, since it seems to 
show up around here about once a week? Something reasonably detailed and with 
links to all the current projects (I mentioned the ones I know about above, 
I'm sure there's more).


From whisper at oz.net  Tue Sep 17 06:20:18 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 17 Sep 2002 03:20:18 -0700
Subject: More on Protecting Source Code
In-Reply-To: <20020917062412.GA85457@hishome.net>
Message-ID: <GCEDKONBLEFPPADDJCOEKEEKFAAA.whisper@oz.net>

> The state has given you the 'right' to use the power of its court and
> police to forcibly restrict people that use information you have created
> in violation of copy-restriction laws.

I don't know what country you're in, but in mine, at least in theory, we ARE
the state! We elected those who make the laws we live by and, again at least
in theory, we can vote them out when they get out of line (like the current
bums, but I digress). We have an independent judiciary to protect us against
the more egregious misventures of our legislation and executive branches
too.

> But that is not enough for you. You also want the 'right' to
> complain about
> something that was given to you freely and expect that righteous rants
> about your view of intellectual property laws will somehow compel people
> to modify it to your specification without compensation.
>
> 	Oren
>

Please explain to me where I was complaining? I pointed out what I see as a
limitation of Python. As I just said elsewhere, Python didn't get to where
it is by impuning and misrepresting the character of those who offer
criticisms.

Vote Democrat in November!

Dave LeBlanc
Seattle, WA USA




From duncan at NOSPAMrcp.co.uk  Tue Sep 24 09:00:27 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 24 Sep 2002 13:00:27 +0000 (UTC)
Subject: Guido Van Rossum to speak at UK Python Conference
Message-ID: <Xns92938DB0E89C8duncanrcpcouk@127.0.0.1>

This is a repost of the CFP: if you missed it the first time it isn't
too late to put yourself forward as a speaker at the UK Python
Conference. No previous experience necessary, just tell me what you want
to talk about or suggest a range of topics if you aren't sure.

The provisional list of speakers for this conference includes Guido
van Rossum, David Abrahams (Boost Python), Andrew Koenig, Marc Andr?
Lemburg, and many more. See http://www.accuconference.co.uk/ for more
about the conference.


                    *** Call for Participation ***
                      UK Python Conference 2003

                            April 2-3 2003
                         Holiday Inn, Oxford

The UK Python Conference 2003 is being held in conjunction with the
larger ACCU conference. The conference will include two days of papers
arranged in 90 minute speaking slots with either one or two tracks.

The audience is highly technical and will include experts in many
fields of computing, some of whom may know little or nothing about
Python. Talks will be welcomed on any aspect of Python, but we are
especially interested in the following:

        - Commercial applications for Python
        - Cross-language talks

Speakers are welcome from the UK or the rest of the world, however our
budget for overseas speakers may be limited. Speakers get reasonable
travel and accomodation costs paid as well as free attendance at the
Python conference and the associated ACCU conference.

To submit a proposal, send an email message to duncan at rcp.co.uk by 2nd
October 2002. In your message, please include the following:

        - Speaker name and affiliation
        - Contact address
        - Presentation title
        - A one paragraph abstract



-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From ch_knievel at gmx.de  Wed Sep 25 12:01:07 2002
From: ch_knievel at gmx.de (christopher knievel)
Date: Wed, 25 Sep 2002 18:01:07 +0200
Subject: [newbie] mod_python apache w32
References: <ami5rr$des$04$1@news.t-online.com> <amrvgp$lfk$1@newsreader2.netcologne.de>
Message-ID: <amsmkd$cjq$05$1@news.t-online.com>

Hi Donald,
i still didn't find a working solution, but i am trying now with a
standalone apache server installation
and hey there are some other error messages....but i am still trying :)
regards,
christopher knievel

>"Donald Lessau" <nc-lessaudo at netcologne.de> schrieb im Newsbeitrag
news:amrvgp$lfk$1 at newsreader2.netcologne.de...
> hi christopher,
>
> i've got exactly the same problem... Did you find a solution in the
> meantime?
>
> thanks
> Donald
>
> "christopher knievel" <ch_knievel at gmx.de> schrieb im Newsbeitrag
> news:ami5rr$des$04$1 at news.t-online.com...
> > Hi all,
> > i got some problems running mod_python with apache under windows.
> > I am using python 2.2 and apache 1.3.26.
> > As written in the manual i configured httpd.conf like this:
> >
> > LoadModule python_module modules/mod_python.dll
> >
> > AddModule mod_python.c
> >
> > <Directory "C:/Programme/Apache Group/Apache/python">
> >   AddHandler python-program .py
> >   PythonHandler mptest
> >   PythonDebug on
> > </Directory>
> >
> > I copied the mod_python.dll to C:/Python22/lib/mod_python.
> >
> > The Apache server is starting with "Apache/1.3.26 (Win32)
mod_python/2.7.6
> > Python/2.2.1 running..."
> > So i guess it should be working....
> >
> > The test script looks like this:
> >
> > #! C:/Python22/python.exe
> >
> > from mod_python import apache
> >
> > def handler(req):
> >     req.send_http_header()
> >     req.write("<html><body>Hello World!</body></html>")
> >     return apache.OK
> >
> > my error.log (last entry):
> > [Sat Sep 21 18:00:13 2002] [error] [client 192.168.1.99] python_handler:
> > make_obcallback returned no obCallBack!
> >
> > i've read several newsgroups and googled a lot but i didn't found a
> working
> > solution
> > Thanks a lot in advance...
> > christopher knievel
> >
> >
>
>




From thoa0025 at mail.usyd.edu.au  Mon Sep 16 04:36:23 2002
From: thoa0025 at mail.usyd.edu.au (:B nerdy)
Date: Mon, 16 Sep 2002 08:36:23 GMT
Subject: [xml] convert funny chars to char entites?
References: <vNah9.10101$Ee4.23987@news-server.bigpond.net.au> <m3elbuze67.fsf@mira.informatik.hu-berlin.de>
Message-ID: <bKgh9.10807$Ee4.29949@news-server.bigpond.net.au>

now when i do filehandle.write(str)

on text that has the funny chars in it. it crashes with an error like this

Traceback (most recent call last):
  File "./batch-maid.py", line 85, in ?
    generate_browse_options()
  File "./batch-maid.py", line 70, in generate_browse_options
    f.write(options_composer)
UnicodeError: ASCII encoding error: ordinal not in range(128)

whats going on?
how can i fix it?

cheeers


"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m3elbuze67.fsf at mira.informatik.hu-berlin.de...
> ":B nerdy" <thoa0025 at mail.usyd.edu.au> writes:
>
> > ive got a string "Il Est N?" and i want to put it into a xml document
> >
> > is there a quick and easy way to conver the funny characters to char
> > entities so i can store them?
>
> I recommend you properly declare the encoding of the document, e.g.
>
> <?xml version="1.0" encoding="iso-8859-1"?>
>
> (assuming the ? is meant to be encoded in Latin-1).
>
> XML does not define external entities for non-ASCII characters, so you
> would have to create a DTD first if you want them.
>
> Regards,
> Martin
>




From cce at clarkevans.com  Wed Sep 25 00:50:50 2002
From: cce at clarkevans.com (Clark C. Evans)
Date: Wed, 25 Sep 2002 04:50:50 +0000
Subject: streaming popen to ftp.storbinary 
Message-ID: <20020925045050.GA67481@doublegemini.com>

Hello.  I'm trying to pipe the output of a process 
created by popen3 through FTP to a file on a remote box.
Ideally, the process would block (so that a temporary
file isn't needed).

I have the following...

  (sin,sout,serr) = os.popen3( my_command, 'b' )  # binary stream
  ftp.storbinary("STOR %s " % filename, sout)

Anyway, if I go through an intermediate file, say by using 
os.system and piping to a temp file, and then using open() 
in the storbinary call it works.  

Also, this works under Win32 using ActiveState's python 2.2.1
but fails under version 2.2 compiled for FreeBSD.  Could this
be a version problem? 

Best,

Clark



From thorsten at thorstenkampe.de  Tue Sep 24 04:08:31 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Tue, 24 Sep 2002 10:08:31 +0200
Subject: list problem
References: <mailman.1032818129.8126.python-list@python.org> <3D8F9F5E.DC48B13F@pop.ntlworld.com>
 <mailman.1032828726.20071.python-list@python.org>
Message-ID: <amp6i1$7n9bi$1@ID-77524.news.dfncis.de>

* Anthony Tuininga
> If you have list comprehensions available, use
>
> [b for a in x for b in a]

What about:
reduce(lambda x, y: x + y, a) ?

Thorsten


From kathuw at swbell.net  Fri Sep 27 05:03:11 2002
From: kathuw at swbell.net (Kat & Huw)
Date: Fri, 27 Sep 2002 04:03:11 -0500
Subject: How do I set-up Multiple socket connections
Message-ID: <002a01c26604$b515c3c0$0100a8c0@mshome.net>

Ok, I want to iterate through a server IP list and establish
multiple socket connections to the various IP address' in the list.

I can do the iterations and I can make single connections, but I can't
work out how I can connect to several address at once.

I understand that a socket is closed, or closes after its done getting the ACk or
whatever after sending a request, but how on earth do I connect to like 5
addresses at once?

I've used non-blocking sockets and the sockettimeout module to skip a host
on certain exceptions, but I want to try the exact same thing on several hosts at once.

Any help would really be appreciated.

Yours
opn
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020927/c415dbf0/attachment.html>

From jwold at min.midco.net  Tue Sep 24 00:55:53 2002
From: jwold at min.midco.net (Jarrett Wold)
Date: Mon, 23 Sep 2002 23:55:53 -0500
Subject: PyXML Building on OS X?
References: <B9B3CED8.3B9%jwold@min.midco.net> <j4k7ldghtp.fsf@informatik.hu-berlin.de>
Message-ID: <B9B55A89.6D8%jwold@min.midco.net>

On 9/23/02 4:45 AM, in article j4k7ldghtp.fsf at informatik.hu-berlin.de,
"Martin v. L?wis" <loewis at informatik.hu-berlin.de> wrote:

...

extensions/expat/lib/xmlparse.c:3056: warning: `regparm' attribute directive
ignored
extensions/expat/lib/xmlparse.c:3093: warning: `stdcall' attribute directive
ignored
extensions/expat/lib/xmlparse.c:3093: warning: `regparm' attribute directive
ignored
extensions/expat/lib/xmlparse.c:3107: warning: `stdcall' attribute directive
ignored
extensions/expat/lib/xmlparse.c:3107: warning: `regparm' attribute directive
ignored
...

cc -L/sw/lib -bundle -flat_namespace -undefined suppress
build/temp.darwin-5.5-Power Macintosh-2.2/pyexpat.o
build/temp.darwin-5.5-Power Macintosh-2.2/xmlparse.o
build/temp.darwin-5.5-Power Macintosh-2.2/xmlrole.o
build/temp.darwin-5.5-Power Macintosh-2.2/xmltok.o -o
build/lib.darwin-5.5-Power Macintosh-2.2/_xmlplus/parsers/pyexpat.so
-flat_namespace
/usr/bin/ld: -undefined: unknown argument: -lbundle1.o
error: command 'cc' failed with exit status 1


> Jarrett Wold <jwold at min.midco.net> writes:
> 
>>   If you want some sample output of the build I can certainly rerun
>> it and post it.
> 
> Please do. No need to report the entire output, though - just the
> error messages and a little context is needed, such as the compiler
> line that caused the error message.
> 
> Regards,
> Martin
> 



From lists at gregfortune.com  Wed Sep 18 15:06:56 2002
From: lists at gregfortune.com (Greg Fortune)
Date: Wed, 18 Sep 2002 12:06:56 -0700
Subject: Loading modules from several directories
References: <a1247c5b.0209171610.243f3c02@posting.google.com> <33803989.0209172247.2e6afc1f@posting.google.com>
Message-ID: <3d88ccbb$0$310$39cecf19@nnrp1.twtelecom.net>

If it is indeed a package, you don't need to append both paths.  Just use:

sys.path.append('/somedir/foo')
import foo.foostuff
import foo.project_a.projectstuff

Furthermore, any module in project_a can access modules in foo by using
import foo.foostuff and the package will know how to resolve the location.


Greg Fortune


Miki Tebeka wrote:

> Hello Eugene,
>> Trying to reuse the code I found a problem (with Python or me) where
>> if I have for example, a foo package in two separate directories
>> (where foo/ contains common files and foo/project_a/ contains
>> project-specific files) I can't load from both.
>> 
>> If I put the common path first in the sys.path it tries to load the
>> project-specific stuff from there too. If I try putting the
>> project-specific one first it can't find the common stuff.
>> 
>> Anyone know of a good solution for this problem?
> Doesn't
> 
> sys.path.append('/somedir/foo')
> sys.path.append('/somedir/foo/project_a')
> import foostuff
> import projectstuff
> 
> works?
> 
> Miki



From mhuening at zedat.fu-berlin.de  Wed Sep 18 09:38:32 2002
From: mhuening at zedat.fu-berlin.de (Matthias Huening)
Date: 18 Sep 2002 13:38:32 GMT
Subject: BibTeX module?
Message-ID: <Xns928D9F1FBAA22mhueningzedatfuberli@130.133.1.4>

Hi,

I am looking for a simple BibTeX module, allowing me to read and 
manipulate/convert a bibliography file. 

Thanks, Matthias


From max at alcyone.com  Mon Sep 23 20:45:00 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 23 Sep 2002 17:45:00 -0700
Subject: Python quote in cgi
References: <amoabi$sjh$1@lust.ihug.co.nz> <amoaf8$smq$1@lust.ihug.co.nz>
Message-ID: <3D8FB58C.53EF8CBE@alcyone.com>

CheapSkate wrote:

>  Hi, this might have been discussed b4,
>  can someone clarify for me, when writing cgi script in python,
>  does ' = """
>  eg
> 
> print """<a href="index.html">Main</a>"""
> 
> print '<a href="index.html">Main</a>'
> 
> are they both the same?

Yes.  Triple quotes allow you to embed either form of single-character
quote in them, and additionally triple quotes can span multiple lines
(though in your example they don't).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Ride / Ride this wave of mine
\__/ Res
    Erik Max Francis' bookmarks / http://www.alcyone.com/max/links/
 A highly categorized list of Web links.


From thehaas at binary.net  Thu Sep 26 13:02:01 2002
From: thehaas at binary.net (thehaas at binary.net)
Date: Thu, 26 Sep 2002 17:02:01 GMT
Subject: shell script in distutils
Message-ID: <d4Hk9.104621$TX5.4307361@news1.east.cox.net>

I'm taking over someone else open-source app which is python-based, but
is invoked by a shell script.

What the shell script does is cd to the install directory and run "python
scriptname.py". It's not a bad solution at all.

However, they a Makefile to control installation and to make the rpm's.
I would like to change it to use distutils.  I have it already
recognizing the scripts, but how to I get distutils to put the shell
script in a different directory than the Python scripts??

Am I making any sense here??  if not, let me know and I will clarify . . 

-- mikeh


-- 
Mike Hostetler          
thehaas at binary.net 
http://www.binary.net/thehaas 
GnuPG key: http://www.binary.net/thehaas/mikeh.gpg 


From aleax at aleax.it  Wed Sep 18 17:02:32 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 18 Sep 2002 21:02:32 GMT
Subject: Is there a list comprehension to do this?
References: <H2nJ0I.36A@news.boeing.com>
Message-ID: <IR5i9.111499$ub2.2370766@news1.tin.it>

Joe Sixpack wrote:

> Is there a list comprehension or other 'cleaner' or 'prettier' way
> to do the following:
> 
>>>> xx = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10)
>>>> ww = []
>>>> for i in range(0, len(xx), 2):
>  ww.append((xx[i+1],xx[i]))
> 
>>>> ww
> [(2, 1), (4, 3), (6, 5), (8, 7), (10, 9)]

Sure, you can always use a list comprehension in lieu of a loop
that builds a list by calling .append -- a mechanical transcription:

ww = [(xx[i+1],xx[x]) for i in range(0, len(xx), 2)]

In Python 2.3, where lists finally support generalized slicing:

ww = zip(xx[1::2],x[::2])

but that doesn't work in 2.2 and earlier, alas.


Alex



From jepler at unpythonic.net  Mon Sep 30 23:03:57 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Mon, 30 Sep 2002 22:03:57 -0500
Subject: Reloading a module
In-Reply-To: <bondpaper-06D4B1.20204130092002@nnrp02.earthlink.net>
References: <bondpaper-648404.19153230092002@nnrp04.earthlink.net> <anastb$h2s$1@peabody.colorado.edu> <bondpaper-06D4B1.20204130092002@nnrp02.earthlink.net>
Message-ID: <20020930220356.A14365@unpythonic.net>

On Tue, Oct 01, 2002 at 02:16:15AM +0000, Tom wrote:
> I wish it were that simple. I've used reload(module) many times, each 
> time hoping that it would do what the documentation says it will do. 
> I've tested it by adding a simple variable to my module, watching to see 
> if it shows up in the new import, but it never does. I noticed also that 
> sometimes it reloads from the .pyc file, and if the reload happens 
> before the .pyc file is updated with the new changes, the changes won't 
> be available. Even if I force a reload from the updated .py file by 
> deleting .pyc file, it still retains the old code. I'm at a loss to 
> understand what's going on.

# silly.py
print "That's silly -- and a bit suspect, I think"

    Python 2.3a0 (#1, Jul 13 2002, 09:27:09) 
    [GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
    Type "help", "copyright", "credits" or "license" for more information.
    >>> import silly
    That's silly -- and a bit suspect, I think
    >>>

(editing the file:
# silly.py
print "spam spam spam spam spam"
)
    >>> reload(silly)
    spam spam spam spam spam
    <module 'silly' from 'silly.py'>
    >>>

As for .pyc vs .py, Python follows these steps when importing a module:
    if a .pyc file exists, read the magic number, version, and the
	stored timestamp.  Compare the version to this Python version,
	and the stored timestamp to the .py file's timestamp.  If these
	both match, or the version matches and the .py file doesn't exist,
	use the .pyc file.
    otherwise, compile the .py file and rebuild the .pyc file along the way

The only circumstance that would interfere with this is NFS or another
network filesystem making the check of the timestamp on the .py file
unreliable---i.e., python can fail to detect that the .py file has been
updated, because the OS gives it stale data.  Using a filesystem that has
no timestamp at all (or a very granular timestamp, so that two revisions of
"silly.py" could share the same timestamp) could cause the same problem,
though I don't know of any such filesystems offhand.

(That said, I rarely actually notice problems of this type over NFS)

It's possible that you've discovered some sort of new problem with the way
.pyc caching or reloading work, but the idea is so simple that I can't see it
failing...

Jeff



From sholden at holdenweb.com  Fri Sep 27 09:59:49 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 27 Sep 2002 09:59:49 -0400
Subject: How to upload a file using http protocol
References: <mailman.1032552637.6156.python-list@python.org> <F_Ni9.284613$eK6.8277278@twister.austin.rr.com>
Message-ID: <GsZk9.1802$6g7.1555@fe03>

> "A" <export at hope.cz> wrote in message
> news:mailman.1032552637.6156.python-list at python.org...
> > Hi,
> > Does anyone know how to upload a file through HTTP protocol?Is that a
> possible?
> > Thank you for help.
> > ladislav
> >

"DeepBleu" <DeepBleu at DeepBleu.org> top-posted ...
> Yes it is.  Assume that the user has to enter the file name (with the
path)
> on a form.  You use File type form element which allows the user to browse
> for the file.

Technically, what you need is an <INPUT> element with the TYPE="FILE"
attribute. You may also need to make sure that you have the correct MIME
type (Content-Type: multipart/form-data). You may want to read

    http://www.ietf.org/rfc/rfc2388.txt

to get the details absuolutely right, but it might be easier just to try a
few options until it works...

> Using CGI module, and a Submit button (or some javascript with a submit
> function) this form element has a KEY.  The VALUE of this key is the Byte
> content of the file, NOT the file name.  Thus, by using the VALUE KEY, the
> server is actually receiving the byte content.  From there on, the sky is
> the limit.
> DeepBleu
>
Read the CGI module's documentation to determin how to read the content.

regards
Steve





From martin at v.loewis.de  Mon Sep 16 17:50:40 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Sep 2002 23:50:40 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org>
Message-ID: <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de>

Ian Bicking <ianb at colorstudy.com> writes:

> I don't think this represents the current state of Stackless --
> Christian has tried to get it into the standard Python, and it is not
> for lack of desire or effort that that did not happen.  I think it could
> be considered a difference of opinion between him and Guido -- in part,
> Stackless introduces more complexity which Guido did not want to
> maintain, and in part because Stackless is not portable to forms of
> Python (not to Jython in particular, and certainly any .NET Python --
> probably a Parrot Python as well).
> 
> So the issue is primarily technical/political, and not a matter of
> resources.  

It's definitely not political, and certainly not the "current state".

I looked at the old Stackless patches once, and spotted significant
amounts of unexplicable code - code that did not serve any obvious
purpose, was not documented, or clearly should have been removed.  So
the old Stackless patch was incomplete for integration into Python -
it would have taken efforts to complete it, and to work with Python
maintainers to integrate it. Nobody went through this effort (and
again, I don't blame anybody for not doing that).

>From a shallow inspection, the code merely looked complex - but it
really was complicated, and complex is better than complicated.

However, this is *not* the current state. The current implementation
(which I have not reviewed, yet) supposedly is much simpler, so past
experience with trying to integrate it does not apply.

Regards,
Martin


From fgeiger at datec.at  Thu Sep  5 01:55:50 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Thu, 5 Sep 2002 07:55:50 +0200
Subject: Does BlackAdder really suck?
References: <slrnanci6i.1s8.gerhard.haering@haering.opus-gmbh.net> <al5ivg$e8v$1@rex.ip-plus.net> <slrnancl3o.2fo.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <al6rjq$i24$1@rex.ip-plus.net>

> The NT box? Wise decision

LOL.
No. Yet I can't be w/o NT. But w/o BlackAdder I can be very well.

> Yes.  It seems to support wxSizers only, which IMO is relatively easy to
do
> with just Vim + Python already.

So, what's your problem?

Regards
Franz GEIGER





From mcfletch at rogers.com  Thu Sep 26 21:13:08 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Thu, 26 Sep 2002 21:13:08 -0400
Subject: Help! This is driving me insane!
References: <FYMk9.12763$wH.1166@sccrnsc01>
Message-ID: <3D93B0A4.6060103@rogers.com>

You do two prints:
	[0, 0]
	[0, 0]

Then somehow or another you're getting the result of the whole 
method-invocation printed (likely by the interactive interpreter), which is:

	[[1,1]]

After the method has returned, your "grouptotalcounter" code has been 
called, so there's no mystery there for Python peoples.  The list 
grouptotal's values were updated, just as your method said to do, and 
that list is the first item in the errorlog list you returned.

I _think_ the problem you're having is that you're thinking by appending 
the list object to self.errorlog you are somehow "fixing" the values in 
the list object as those when appended (which is something you would see 
in, for example C, where you were assigning to arrays).

i.e.

	a = []
	b = [1,2]
	a.append(b)
	b[0] = 3

would end with a == [[3,2]] and b == [3,2].

Because a[0] _is_ b, _not a copy_ of b.  That is, a[0] is the same 
_object_ as pointed to by b, so any changes to b after adding it to a 
will show up in a (since it's just pointing to same list b).

Try using a.append( b[:] ) if you want to append a copy of b, rather 
than the original list.  Note, however, that if you're doing nested lists:

	a = [[],[],[]]

You'll want to use the copy module explicitly for creating copies 
(otherwise only the top-level list is copied, the lower-level lists are 
still just referenced in the new list).

HTH,
Mike

kara.klosterman wrote:
...
> http://www.geocities.com/dbasch/insane.txt
> 
> I would expect to get:
> [0, 0]
> [0, 0]
> [[0, 0]]
> 
> but I am getting:
> [0, 0]
> [0, 0]
> [[1, 1]]
> 
> I dont understand how the 'grouptotals' list is
> getting the incremented values prior to the
> 'grouptotalcounter' function being called.
> furthermore, I can place a 'print grouptotals' right
> after the errorlog append and it isnt the incremented
> values! Please tell me this isnt a normal thing!
> Thanks,
> Derek Basch
> 
> 
> 

-- 
_______________________________________
   Mike C. Fletcher
   Designer, VR Plumber, Coder
   http://members.rogers.com/mcfletch/





From drifty at bigfoot.com  Tue Sep 17 15:50:05 2002
From: drifty at bigfoot.com (Brett C.)
Date: 17 Sep 2002 12:50:05 -0700
Subject: Python-dev summary for 2002-09-01 to 2002-09-15
Message-ID: <mailman.1032439083.17023.clpa-moderators@python.org>

This is a summary of traffic on the `python-dev mailing list`_ between
September 01, 2002 and September 15, 2002 (exclusive).  It is intended
to inform the wider Python community of ongoing developments on the
list.  To comment on anything mentioned here, just post to
python-list at python.org or comp.lang.python in the usual way. Give your
posting a meaningful subject line, and if it's about a PEP, include
the PEP number (e.g. Subject: PEP 201 - Lockstep iteration) All
python-dev members are interested in seeing ideas discussed by the
community, so don't hesitate to take a stance on a PEP (or anything
else for that matter) if you have an opinion.

This is the second summary written by Brett Cannon (hopefully my
sophomoric performance will be better then most sophomore music
albums).

Summaries by me (2002-09-15 to ... when I burn out) are archived at:
    http://www.ocf.berkeley.edu/~bac/python-dev/summaries/index.php
You can find summaries by Michael Hudson (2002-02-01 to 2001-07-05)
at:
    http://starship.python.net/crew/mwh/summaries/index.html
Summaries by A.M. Kuchling (2000-12-01 to 2001-01-31) are at:
    http://www.amk.ca/python/dev/

Please note that this summary is written using reStructuredText_ which
can be found at http://docutils.sourceforge.net/rst.html .  Any
unfamiliar punctuation is probably markup for reST; you can safely
ignore it (although I suggest learning reST; its nice and is accepted
for PEP markup).  Also, because of the wonders of reformatting thanks
to whatever you are using to read this, I cannot guarantee you will be
able to run this text through DocUtils as-is.  If you want to do that,
get the original text from the archive.

I am considering keeping a list of names that people are often
referred to in emails.  This would serve a dual purpose: allows people
who read emails from the list to have a reference to be able to figure
out who is who and makes the summaries easier for me because I can
then make reference to people by the names I know them by.  =)  Any
comments on this idea are appreciated.

.. _python-dev mailing list:
http://mail.python.org/mailman/listinfo/python-dev
.. _reST:
.. _reStructuredText: http://docutils.sf.net/

============================
`To commit or not commit`_
============================

Walter Dorwald asked if there were "any objections against committing
the patch" for implementing `PEP 293`_ (Codec Error Handling
Callbacks).  Guido asked what Martin V. Lowis and M.A. Lemburg had to
say about it.  MAL responded that he was +1 on the patch.  Martin was
"concerned about the massive amounts of C code, most of which could be
expressed way more compact in Python code", but "Walter convinced
[MvL] that this does have a real performance impact for real data" so
he would live with it.  In the end he gave it his vote.

Walter said he would check it in (and he has).  The PEP has now been
moved to the finished PEP list.

.. _To commit or not commit:
http://mail.python.org/pipermail/python-dev/2002-September/028502.html
.. _PEP 293: http://www.python.org/peps/pep-0293.html

=======================================
`Proposed Mixins for Wide Interfaces`_
=======================================

Raymond Hettinger suggested adding mixin classes that automatically
implement magic methods when certain basic magic methods were already
implemented (e.g., "given an __eq__ method in a subclass, adds a
__ne__ method").  David Abrahams said that he thought "these are a
great idea, *in the context of* an understanding of what we want
interfaces to be, say, and do."  Guido brought up some points about
the initial suggestions Raymond made.  He then said that he thought
that there wasn't "enough here to warrant putting this into the
standard library"; the issue will be revisited when a standard type or
interface hierarchy is added to Python (not in 2.3).

.. _Proposed Mixins for Wide Interfaces:
http://mail.python.org/pipermail/python-dev/2002-September/028543.html

===================================
`mysterious hangs in socket code`_
===================================

Jeremy Hylton wrote some threaded code to fetch some web pages that
hung when performing a slow DNS operation.  Apparently, in Python 2.1
"it produces a steady stream of output -- urls and the time it took to
load them".  In Python 2.2 and 2.3, though, "it produces little bursts
of output, then pauses for a long time, then repeats".  Jeremy guessed
that it *might* have something to do with Linux's getaddrinfo() being
thread-safe by allowing only a single lookup at a time.  Aahz said
that "gethostbyname() IIRC has frequently been non-reentrant".

.. _mysterious hangs in socket code:
http://mail.python.org/pipermail/python-dev/2002-September/028555.html

========================================
`Two random and nearly unrelated ideas`_
========================================

Skip Montanaro had two ideas; one was to make the info in `Misc/NEWS`_
(which is a summary of what has been changed in Python for each
release) a web page and the other was "to get rid of the ticker
altogether in systems with proper signal support" (see the `2002-08-16
- 2002-09-01 summary`_ for an explanation of what the ticker is). 
That would get rid of the polling of the ticker and thus reduce the
overhead on threads.

For the first idea, Guido asked Skip to try seeing what it would look
like with reST_ markup and what the resulting page would look like.

In response to the second idea, Oren Tirosh said it couldn't be done
until "all Python I/O calls are converted to be EINTR-safe"
(EINTER-safe means to be able to handle the EINTER signal which what
is raised "When an I/O operation is interrupted by an unmasked
signal").  That "requires a lot of work in some of the hairiest places
in the Python codebase."  Fredrik Lundh said that this "sounds like a
good topic for a "here's what I learned when trying to fix this
problem" PEP.  This is most likely in reference to Skip writing the
patch to make the ticker global instead of a per-thread issue.  Guido
said, in terms of signals, to "just say no"; "it is impossible to
write correct code in the presense of signals".  Guido, in a later
email, gave this whole idea a vote of -1,000,000; so it ain't ever
going to happen.  Some discussion on signals ensued, but Guido never
budged from his position.

Oren pointed out that if some C code used signals and people didn't
handle it in their Python code by checking if IOError was caused by
EINTER (as shown below by Oren's code)::

    while 1:
        try:
            <code>
        except IOError, exc:
            if exc.errno == errno.EINTR:
                continue
            else:
                raise

, it would not restart properly even though there was no reason for it
to have stopped.  Oren said that Python could add the loop in the C
code of the core where EINTR might be raised ("Only low-level
functions like os.read_ and os.write_ that map directly to stdio
functions should ever return EINTR").  The proposed idea was to wrap
functions that might raise this that can be re-entered safely.

.. _Two random and nearly unrelated ideas:
http://mail.python.org/pipermail/python-dev/2002-September/028555.html
.. _Misc/NEWS: http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/dist/src/Misc/NEWS
.. _2002-08-16 - 2002-09-01 summary:
http://www.ocf.berkeley.edu/~bac/python-dev/summaries/2002-08-16--2002-09-01.html
.. _os.read:
.. _os.write: http://www.python.org/dev/doc/devel/lib/os-fd-ops.html

================================================
`Should KeyError use repr() on its arguments?`_
================================================

Originally, when an exception was raised and you passed in an optional
object to act as a description of why the exception was raised (such
as ``KeyError("there is no spoon")`` where ``there is no spoon`` is
the optional argument bound to ``<exception>.args``), it just returned
what args was bound to when you called; ``str(<exception>) ==
<exception>.args``.  Now it calls repr() on what args is bound to;
``str(<exception>) == str(<exception>.args)``.  Much better. =)

.. _Should KeyError use repr() on its arguments?:
http://mail.python.org/pipermail/python-dev/2002-September/028545.html

==========================================
`New 'spambayes' project on SourceForge`_
==========================================

Thanks to great work done by Tim Peters and several other
contributors, Barry Warsaw started an SF project to host the spambayes
code.  It can be found at http://sf.net/projects/spambayes .  There
are two mailing lists:
http://mail.python.org/mailman-21/listinfo/spambayes and
http://mail.python.org/mailman-21/listinfo/spambaye-checkins (yes,
that is Mailman 2.1, and yes, you will "help be a guinea pig for
Mailman 2.1").

.. _New 'spambayes' project on SourceForge:
http://mail.python.org/pipermail/python-dev/2002-September/028626.html

=========================
`Subsecond time stamps`_
=========================

Martin V. Lowis wanted to introduce subsecond timestamps on platforms
that supported it.  He suggested adding another field to stat, create
a new type, or make st_mtime a floating point.  The first one option
is easy, the second has the usual problems of defining a new type, and
the third does not guarantee enough accuracy.

Paul Svensson and Guido said that the last option (turning st_mtime
into a float) was the most Pythonic.  MvL agreed, but worried about
breaking code that expected an int.  Guido then suggested that maybe
the new field is the way to go; define something like st_mtimef that
will contain the float if available or contain an int otherwise.  Tim
Peters also weighed in with his `IEEE 754`_ voodoo about how a float
can hold enough info to be accurate up to 100 nanoseconds if you only
span a 33 years.  That causes an issue starting in 2003 since that is
33 years past the epoch (1970).

But then MvL discovered that st_mtime was already a float on the Mac;
had that caused  issues?  Jack Jansen of course chimed in on this by
saying that it caused him a headache about once a year in the form of
a failing test (other issues caused by timestamps is the Classic Macs
having the epoch at 1904 and not using UTC time).  He said he would
prefer to see the timestamp as a cookie that was passed into a
function that spit out "something guaranteed to be of your liking".

To address the other issues that Jack mentioned, Guido suggested that
all timestamps be converted to UTC time with the epoch at 1970.

MvL has `SF patch 606592`_ up on SF that has already been closed that
makes all the relevant changes to have timestamps return floats.

.. _Subsecond time stamps:
http://mail.python.org/pipermail/python-dev/2002-September/028648.html
.. _IEEE 754: http://grouper.ieee.org/groups/754/
.. _SF patch 606592: http://www.python.org/sf/606592

=================================
`64-bit process optimization 1`_
=================================

Bob Ledwith posted a simple patch for `Include/object.h`_ that changed
the order of certain parts of the PyObject_HEAD macros, affecting
PyObject and PyVarObject.  This was for a 64-bit platform performance
boost (40% for large data sets according to Bob).  The reordering
eliminated some padding in the struct and allows more Python objects
to fit in the L2 cache, or at least that is what Bob thinks is going
on.

Guido pointed out that this would save 8 bytes per object; he thought
all of this was "Interesting!".  But alas, using this patch would
break binary compatibility.  Guido was not sure, though, whether it
had been broken yet between Python 2.2 and 2.3 and thus he might be
"being too conservative here" in terms of saying that it should be
held back for now.

A problem Guido pointed out for 64-bit systems, is that theoretically
the reference count for an object could go negative with enough
references as things stand now.  Guido then suggested that perhaps
refcnt (struct item that holds the reference count) should be a
``long``.  And while dealing with that, Guido suggested that anything
that stores a length should store that number in a ``long``.

Chime in Tim Peters.  He pointed out that it was agreed upon years ago
to move refcnt to ``long`` but no one had bothered to do it.  Heck,
even Guido thought for a long time that it was a long when it wasn't;
it required Tim to "beat that out of [Guido] <wink>" to stop him from
saying that it was a ``long``.  He then pointed out that Win64 was
still only 4 bytes for a ``long``; what was really desired was for it
to be ``Py_intptr_t`` which is the Python way for spelling the C99
type that we wanted.  Apparently C99 has a way to specify that things
be a specific byte length (now if everyone just had a C99 compiler we
wouldn't need these macros; oh, to dream...).

Tim also pointed out that what we wanted for the type that held a
length argument to be size_t since that is what strlen() and malloc()
are restricted by.  He said that he writes all of his "string-slinging
code as using size_t vars now".

Tim pointed out that the issue then became "Whether it's worth the
pain to change this stuff" which "depends on whether we think 64-bit
boxes are just another passing fad like the Internet <wink>".  =)

Martin V. Lowis agreed with the changing of refcnt to a long but had
reservations about using size_t for the length field (ob_size).  He
pointed out that some objects put negative values into that field.

Frederik suggested that the proposed changes be default on 64-bit
systems since the chances that they are willing to recompile is higher
then people on 32-bit systems.  He also suggested making it a compiler
option.  Guido thought it was a good idea.  But then Mats Wichmann
discovered that the switch to long killed the performance boost.  So
Guido re-iterated that he thinks it should be a compiler option only
on 64-bit systems; have "compat", "optimal", and "right" compiler
options.

As of yet nothing has done about this.

.. _64-bit process optimization 1:
http://mail.python.org/pipermail/python-dev/2002-September/028677.html
.. _Include/object.h:
http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/dist/src/Include/object.h

==========================================
`Weeding out obsolete modules and Demos`_
==========================================

Jack Jansen noticed that there demos for some of the SGI-specific
modules that use severely outdated systems and hardware (stuff
discontinued 8 to 12 years ago).  Guido gave the go-ahead to yank them
from CVS.

This has yet to be done.

.. _Weeding out obsolete modules and Demos:
http://mail.python.org/pipermail/python-dev/2002-September/028718.html

==============
`utf8 issue`_
==============

(This thread actually started in August)  There was a bug in Python
2.2 that raised a UnicodeError when trying to decode a lone surrogate
(explanation of surrogates to follow this summary).  This caused
issues in importing .pyc files that contained a lone surrogate because
marshal_ (which is what is used to create .pyc files) encodes Unicode_
literals in UTF-8.  This has all been fixed in Python 2.3, but Guido
was wondering how to backport this for Python 2.2.2.

The option of bumping the magic number for .pyc files was raised and
instantly thrown out by Guido; "Bumping MAGIC is a no-no between dot
releases".  So M.A. Lemburg suggested to either fix the Unicode
encoder or change the Unicode decoder to handle the malformed Unicode.
 MAL wasn't sure, though, if some security issue would be raised by
the latter option.

Guido said go for the latter and didn't see any possible security
issue since "If someone you don't trust can write your .pyc files,
they can cause your interpreter to crash by inserting bogus bytecode".

Explanation of lone surrogates:

    In Unicode, a surrogate pair is when you create the representation
of a character by using two values. So, for instance, UTF-32 can cover
the entire Unicode space (since Unicode is 20 bits, although MvL says
it is really more like 21 bits), but UTF-16 can't.  To solve the issue
for an encoding that cannot cover all possible characters in a single
value a character can be represented as a pair of UTF-16 values.  The
high surrogate cover the high 10 bits while the low surrogate cover
the lower 10 bits.  High and low surrogates can never be the same
since they are defined by a range of possible values and those ranges
do not overlap.  So with the proper high and low surrogate paired
together you can make any possible Unicode character.

    The problem in Python 2.2.1 is that when there is only a lone
surrogate (instead of there being a pair of surrogates), the encoder
for UTF-8 messes up and leaves off a UTF-8 value.  The following line
is an example:

	>>> u'\ud800'.encode('utf-8')
	'\xa0\x80'  #In Python 2.2.1
	'\xed\xa0\x80'  #In Python 2.3a0
	
    Notice how in Python 2.3a0 the extra value is inserted so as to
make the representation a complete Unicode character instead of only
encoding the half of the surrogate pair that the encode was given.

    You can read http://216.239.37.100/search?q=cache:Dk12BZNt6skC:uk.geocities.com/BabelStone1357/Software/surrogates.html
for more info.  Thanks goes to Frederik for the link and Guido for
some clarification.

.. _utf8 issue: http://mail.python.org/pipermail/python-dev/2002-August/028254.html
.. _marshal: http://www.python.org/dev/doc/devel/lib/module-marshal.html
.. _Unicode: http://www.unicode.org/

=====================================
`Documentation inconsistency in re`_
=====================================

Christopher Craig noticed that the docs for the re_ module for the
``\b`` metacharacter was incorrect; it says that "the end of a word is
indicated by whitespace or a non-alphanumeric character".  That would
indicate that an underscore would be the end of a word, which turns
out to be false.  Frederik said that "\b is defined in terms of \w and
\W" and thus allows underscore to be a alphanumeric character.  The
documentaiton has been fixed.

.. _Documentation inconsistency in re:
http://mail.python.org/pipermail/python-dev/2002-September/028644.html
.. _re: http://www.python.org/dev/doc/devel/lib/module-re.html

=======================
`Codecs lookup order`_
=======================

Francois Pinard discovered that for the codecs_ module "one should be
careful about **not** [altered emphasis] naming a module after the
encoding name, when closely following the documentation in the Library
Reference manual".  This is because the codecs module first searches
the registry of codecs, then searches for a module with the same name
and use that module.  The issue comes up when the module does not
contain a function named getregentry(); "\`encodings.lookup()` expects
a \`getregentry` function in that module, does not find it, and raises
a CodecRegistryError, not leaving a chance to subsequent codec search
functions to be used".

M.A. Lemburg said that this has been fixed in Python 2.3 and will be
in 2.2.2 by having encodings.lookup() return None if getregentry() is
not found and thus allowing the search to continue.

.. _Codecs lookup order:
http://mail.python.org/pipermail/python-dev/2002-September/028676.html
.. _codecs: http://www.python.org/dev/doc/devel/lib/module-codecs.html

=================================
`raw headers in rfc822.Message`_
=================================

John Spurling provided a two-line hack to keep the raw headers in an
rfc822.Message_ .  Barry responded that email.Message.Message_ keeps
the raw headers around.

But the reason I am summarizing this is what this thread quickly
changed to is how to properly generate a patch.  Patches should be
generated using UNIX diff, either the -c or -u option with preference
for -c (using cvs diff -c is even better; puts the version of the file
you are diffing with in the output); Mac folk can send MPW diffs, but
UNIX diff is the definitely preference.  Always put the order of the
files ``diff -c OLD_FILE NEW_FILE`` .  And always post the patches_ to
SourceForge_!  Getting random patches, no matter how small, on the
list is annoying (at least to me) because the point of the list is to
discuss the design and implementation of Python, not to patch Python. 
SF is used so that Python-dev does not need to be bothered with
mundame problems like applying patches (and to annoy Aahz with SF's UI
sucking in Lynx_ =).  So please, for my sake and everyone else on
Python-dev, use SF!

For a funny email from Raymond Hettinger about developing for Python
read http://mail.python.org/pipermail/python-dev/2002-September/028725.html
.

.. _raw headers in rfc822.Message:
http://mail.python.org/pipermail/python-dev/2002-September/028682.html
.. _rfc822.Message: http://www.python.org/dev/doc/devel/lib/message-objects.html
.. _email.Message.Message:
http://www.python.org/dev/doc/devel/lib/module-email.Message.html
.. _patches: http://sourceforge.net/patch/?group_id=5470
.. _SourceForge: http://www.sourceforge.net/
.. _Lynx: http://lynx.browser.org/

===================
`type categories`_
===================

Yes, the `same thread`_ from the `last summary`_ is back.  This thread
has become the bane of my summarizing existence.  =)

Aahz asked "why wouldn't we simply use attributes to hold" interfaces
that a class implemented (think of __slots__).  David Abrahams then
brought up the idea of just adding interfaces to the __class__
attribute.

Guido then chimed in on the attributes idea.  He pointed out that this
is how Zope does it, using the __inherits__ attribute.  The limitation
is that "it isn't automatically merged properly on multiple
inheritance, and adding one new interface to it means you have to copy
or reference the base class __inherits__ attribute".  And as for
David's idea of just adding to __class__, that doesn't work because
there is no way to limit the interface; you need "Something like
private inheritance" for when an interface is broken by some inherited
class.  David subsequently added the issue of being able to disinherit
when an interface is not valid but is inherited by default as another
problem for using inheritence for interfaces.

David then brought up the issue of having Python being so dynamic that
you could inject an interface if you used __class__ like he suggested
through black magic code.  If the injected interface didn't work
because of the inheritence chain, then you have a problem.

Barry Warsaw brought in his objections.  He tried playing Devil's
Advocate by saying that Guido had said that inheritance would not be
the only way to handle interfaces, but that it would be the
predominent way.  But this duality would complicate any
conformsto()-like function since it would have to handle two different
ways for a class to get an interface.  Barry then brought up the
objection that he didn't like the idea of using straight inheritence
because he wanted a syntactic way to separate out interfaces.

As a side note, Guido pointed out that __slots__ is provisional; nicer
syntax will eventually surface when Guido gets over his "fear of
adding new keywords".

.. _type categories:
http://mail.python.org/pipermail/python-dev/2002-September/028738.html
.. _same thread: http://www.ocf.berkeley.edu/~bac/python-dev/summaries/2002-08-16--2002-09-01.html#type-categories
.. _last summary: http://www.ocf.berkeley.edu/~bac/python-dev/summaries/2002-08-16--2002-09-01.html

=======================================
`flextype.c  -- extended type system`_
=======================================

Christian Tismer has come up with a replacement for the etype which is
"a hidden structure that extends types when they are allocated on the
heap" (you can find it in `Objects/typeobject.c`_ in the CVS_).  There
is a limitation with the etype where it could not be extended by
metatypes.  Well, Chris worked his magic and came up with a new
flextype that allows overriding of methods.  So with Christian's code
you would be able to override methods in a type without having to hack
something together to handle the overriding correctly; it would be
handled automatically.

Through some clarification from Christian and Guido, it was pointed
out to me (as of this moment I am the only one to make any noise on
this thread, and it was for this summary) that this simplifies an
esoteric issue; note the use of the words "metatype" above.  This is
type/metatype black magic hacking.  Spiffy, but something most of us
"normal" folk will not have to worry about.

.. _flextype.c  -- extended type system:
http://mail.python.org/pipermail/python-dev/2002-September/028736.html
.. _Objects/typeobject.c:
http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/dist/src/Objects/typeobject.c
.. _CVS: http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/#dirlist



From rajarshi at presidency.com  Mon Sep  9 16:10:39 2002
From: rajarshi at presidency.com (Rajarshi Guha)
Date: Mon, 09 Sep 2002 16:10:39 -0400
Subject: a re question
Message-ID: <pan.2002.09.09.16.10.38.482129.20940@presidency.com>

Hi,
  I have a file with lines of the format:

001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23
001 Xya FGh   143557789   7 100 09/05/2002 20:23:23

I am trying to extract the 9 digit field and the single digit field
immediatley after that.

When I use Visual Regexp to try out the regexp 

(\d{9,} {3,}\d)

it highlights the 2 fields exactly. 

But when I use the following Python code I get None:

>> s='001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23'
>> p = re.compile(r'(\d{9,} {3,}\d)')
>> print p.match(s)
>> None

Could anybody point out where I'm going wrong?

Thanks,


From wanner.dx at mellon.com  Wed Sep 11 08:57:18 2002
From: wanner.dx at mellon.com (dxterra)
Date: 11 Sep 2002 05:57:18 -0700
Subject: file upload using msvcrt
Message-ID: <bb2665bf.0209110457.62d78bfe@posting.google.com>

I am trying to upload msword docs to my webserver. Here is the code
that I am using:

#!I:\Python22\Python -u

# $Header: $

# $Log: $

import cgi, sys, string, time, types, os
import posixpath, dospath, macpath
import msvcrt
msvcrt.setmode(0, os.O_BINARY)
msvcrt.setmode(1,os.O_BINARY)

from pso.service import ServiceHandler

# allows for more verbose feedback when a cgi exception occurs
import cgitb
cgitb.enable()

sys.stderr = sys.stdout

UPLOAD_DIR = "o:\\upload"


def splitpath( orig_path ):
	for pathmodule in [posixpath, dospath, macpath]:
		basename = pathmodule.split(orig_path)[1]
		if basename != orig_path:
			return basename
	return orig_path

def save_file(fileinfo):

	basename = splitpath( fileinfo.filename)
	srvrname = os.path.join(UPLOAD_DIR, basename )
	msvcrt.setmode(0, os.O_BINARY)
	msvcrt.setmode(1,os.O_BINARY)
	fp = file( "s:\\cgi-bin\\project\\log.lis", "a+" )
	fp.write( "saving file '%s' as '%s'\n" % (fileinfo.filename,
srvrname) )
	fp.write( "file contents %s\n" % fileinfo.value )
	fp.write( "file size as read from stdin '%s'\n" % (
len(fileinfo.value) ) )
	fp.write( "file mode in '%s'  file mode out '%s'\n" %
(sys.stdin.mode, sys.stdout.mode ) )
	fp.close()

	srvrfile = file(srvrname, "wb" )
	
	while 1:
		data = fileinfo.file.read( )
		if not data: break
		srvrfile.write(data)

	return fileinfo.filename


def upload_file():

	form = cgi.FieldStorage()
	msgs = []

	fileinfo = None

	for field in form.keys():
		if field == "zzz":
			if not form[field].filename:
				msgs.append( "ERROR: filename is missing" )
			else:
				fileinfo = form[field]
				fp = file( "s:\\cgi-bin\\project\\log.lis", "a+" )
				fp.write( "attachment filename is '%s'\n" % (fileinfo.filename) )
				fp.close()

	if fileinfo is not None:
		fp = file( "s:\\cgi-bin\\project\\log.lis", "a+" )
		fp.write( "preparing to save attachment '%s'\n" %
(fileinfo.filename) )
		fp.close()
		srvrname = save_file(fileinfo)

	# process the form data

	for msg in msgs:
		print '<center><font class="RedBold">%s</font></center>' % (msg)


if __name__ == '__main__':
	upload_file()

I cannot get the binary file to write. It get to about 8k and then
stops.
When I query the mode it always says 'r' for stdin and 'w' for stdout.
Can anyone help me?


From syver-en+usenet at online.no  Mon Sep 16 11:06:58 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Mon, 16 Sep 2002 15:06:58 GMT
Subject: python performance
References: <3D84E66F.6020701@linux.ie> <zm7h9.51182$AR1.2015178@bin2.nnrp.aus1.giganews.com> <IThh9.4238$cP3.10401@news.iol.ie> <j4r8fumezm.fsf@informatik.hu-berlin.de>
Message-ID: <uk7lmm0sj.fsf@online.no>

loewis at informatik.hu-berlin.de (Martin v. L?wis) writes:

> Padraig Brady <Padraig at Linux.ie> writes:
> 
> > > Calling functions adds overhead.  Running code within functions
> > > reduces overhead for object access.
> > 
> > Ah OK, this because it has to search a smaller namespace?
> 
> No. Inside a function, local variables are accessed with integer
> indices, instead of by name. This is possible because there is static
> knowledge about the local variables in a function.
> 
> > Wouldn't compiling overcome this? 
> 
> Not if you want to preserve Python semantics; the compiler would need
> to emit code that performs the same actions that the interpreter
> currently does.

What about Psyco?

-- 

Vennlig hilsen 

Syver Enstad


From jeremy at jdyallop.freeserve.co.uk  Tue Sep  3 04:03:49 2002
From: jeremy at jdyallop.freeserve.co.uk (Jeremy Yallop)
Date: 3 Sep 2002 08:03:49 GMT
Subject: getch in python
References: <75ca4e69.0209022326.4cde72e8@posting.google.com>
Message-ID: <al1qd5$1iahu9$1@ID-114079.news.dfncis.de>

* Venkat Venkataraju
| Is there any function that behaves like getch() function in C? 

I'd say that getch() is probably the closest.  It's a method of the
window class in the curses module:

    http://www.python.org/doc/current/lib/curses-window-objects.html

There are some demos that use it in the Demo/curses directory of the
Python distribution.

Jeremy.


From tjreedy at udel.edu  Sat Sep 21 14:50:08 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 21 Sep 2002 18:50:08 GMT
Subject: filter(lambda W : W not in 'ILLITERATE','BULLSHIT') (Re : python expression)
References: <d8f8142b.0209161913.12cce045@posting.google.com> <vWBh9.26065$Gp1.687517@weber.videotron.net> <3D86FF01.FE0CECC5@geneva-link.ch> <3d87069c.34652637@news.rcn.com> <3D871316.AA1E8FFC@geneva-link.ch> <3d882cd4.9181792@news.rcn.com> <3f74f974.0209180713.27889f74@posting.google.com> <l9ehouci1ga88ub2csdam1qroiik0ogd4r@4ax.com> <3d88c258$0$707$5402220f@news.sunrise.ch> <v4lhouolv1c0ijmihfom94j0lor2lhntli@4ax.com> <3d89be92$0$707$5402220f@news.sunrise.ch> <s6mjous6o5abg3icq12u156v5tijhcnk7h@4ax.com> <3d89ee85$0$712$5402220f@news.sunrise.ch> <B8pi9.149307$5r1.6007886@bin5.nnrp.aus1.giganews.com> <3d8b4498$0$717$5402220f@news.sunrise.ch> <WSLi9.165658$z91.6788612@bin3.nnrp.aus1.giganews.com> <3d8c6982$0$719$5402220f@news.sunrise.ch>
Message-ID: <Ab3j9.91527$216.3470808@bin4.nnrp.aus1.giganews.com>

<other newsgroups snipped>

"Boris" <borcis at removethis.infomaniak.andthis.ch> wrote in message
news:3d8c6982$0$719$5402220f at news.sunrise.ch...
>
> Terry Reedy <tjreedy at udel.edu> a ?crit dans le message :
> WSLi9.165658$z91.6788612 at bin3.nnrp.aus1.giganews.com...
> > "Boris"
> > >
> > > filter(lambda W : W not in 'ILLITERATE','BULLSHIT')
> >
> > This somewhat-well-known and certainly well-worn programmer joke
>
> Well, *I* invented this formula myself soon after Mr Bush revealed
to
> the world that "USA" is written in "CRUSADE" just like "US" is in
> "BUSH". Adding : "Either you are with US, or you are against US".

Congratulations on your cleverness.  You have used it as your
signature over 20 times on this mail-list/newsgroup so it is somewhat
familiar to me and, I presume, to other regular readers  According to
google, you seem to have used it at least once in at least 10 other
newsgroups(and a few more with this posting), but pardon me for
overestimating its well-knownness outside the confines of Pythonland.
----
>> and the presence of very un-Pythonic ugliness

> Ah, you are an aesthetician of pythonicity

Yes.  To quote from the post where you added c.l.p to the newgroup
list:
"""
Malev : [after accusing me of an opaque vice on the basis of an
              incompetent interpretation of the headers of his copy
             of my post]
>
> >by this channel. This really has nothing to do with me.  You should
> >abstain from reading technical small print if it's beyond your
ability
> >to interpret it adequately.
>
> Pompous patronizing fool.

I prefer that to being the unrepenting "innocent" liar that you are,
"""

Again, this sort of personal attack and counterattack has nothing to
do with Python and is contrary to the generally friendly spirit here.
I consider it ugly and make no apology for stating so.

> But then, what's your competence to formulate such assertions ?

My humanity and at least 5 1/2 years of participation here.

Terry J. Reedy




From jkraska at san.rr.com  Mon Sep 23 02:15:31 2002
From: jkraska at san.rr.com (Courageous)
Date: Mon, 23 Sep 2002 06:15:31 GMT
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <mailman.1031377930.4225.python-list@python.org> <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com> <3D8E6353.7030303@something.invalid>
Message-ID: <dvbtouciidaufjneja3s4t89ds0r7jd76k@4ax.com>

>Using *tabs* only is the only way to get truly
>preference-independent indentation.

But makes impossible certain constructs that involving lining
indentation level up with specific items of text in surrounding
lines, which actually matters. This is one of the reasons people
often use mixed tabs and spaces. Here is an example:

void MyFunction ( int a,
                  char c,
                  float f );

void MyFunction ( int a,
					char c,
					float f );


void MyFunction (	int a,
					char c,
					float f );

Furthermore, when you use only spaces, you can guarntee that
all those reading will see it formatted as close to correctly
as is possible.

C//



From michael.amrhein at t-online.de  Sat Sep 21 11:45:05 2002
From: michael.amrhein at t-online.de (Michael Amrhein)
Date: Sat, 21 Sep 2002 17:45:05 +0200
Subject: monetary applications [was: Python GUI app to impress the bos
 s?]
References: <mailman.1032459720.3477.python-list@python.org>
Message-ID: <3D8C9401.4000805@t-online.de>

Tim Peters wrote:
 > About http://starship.python.net/crew/aahz/FixedPoint.py
 >
 > If someone is willing and able to do the work (sorry, I can't make
 > time for it), Guido will accept this into the standard library for
 > 2.3.  It's a small, self-contained, pure Python module; it's been in
 > use for a few years with no bugs reported;

There was a short thread in this ng some months ago, where two bugs in
FixedPoint.py were reported (incl. fixes).

 > and is unencumbered by any license (a lawyer has told
 > me that an individual in the USA cannot meaningfully disclaim
 > copyright, so fine, the copyright I didn't want is hereby transferred
 > to the PSF <wink>).
 >
 > What work remains:
 >
 > + Write docs for the Library Reference manual.  I expect the existing
 >  module docstring will be a good start.
 >
 > + Create a test driver for Python's regression suite.
 >
 > + Have fun modernizing it, if you like (for example, "/" should be
 > changed to "//" where integer division is intended; playing with
 > __slots__ may yield some nice savings; whatever).
 >
 > This should be a fun and comparatively easy task -- if you ever
 > wanted to get your name into the ACKS file, here's one way to do it
 > without tithing you income directly to Guido <wink>.
 >
 > although-tithing-saves-me-time-ly y'rs  - tim
 >
 >
I've written a simular class named 'fixed' (see below) with portions
adopted from your FixedPoint, but a slightly different constructor (no
default precision; precision calculated from given value, if no
precision given) and some additional functions. And, yes, it uses
__slots__ and supports the new integer division. So, maybe, it could
serve as the "modernized" version you recommend in your third point.
But, while willing to contribute it to the PSF, I just don't know what
is required to do so. And because I've never written docs for the
Library Reference nor created test drivers I'll need some guidance.
Maybe, Doug can help with this.
Cheers,
Michael

---
##------------------------------------------------------------------------------
## Name:        fixedpoint.py
## Purpose:     Decimal arithmetic with a fixed number of fractional digits
##
## Author:      Michael Amrhein (michael.amrhein at t-online.de)
##
## Created:     2001-12-06
## RCS-ID:      $Id: fixedpoint.py,v 1.2 2002/06/15 12:26:58 am Exp $
## Copyright:   (c) Michael Amrhein
##              Portions adopted from FixedPoint.py written by Tim Peters
## License:     This program is free software; you can redistribute it 
and/or
##              modify it under the terms of the GNU Lesser General Public
##              License as published by the Free Software Foundation; either
##              version 2 of the License, or (at your option) any later 
version.
##              This program is distributed in the hope that it will be 
useful,
##              but WITHOUT ANY WARRANTY; without even the implied 
warranty of
##              MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
##              See the GNU General Public License for more details.
##              You should have received a copy of the license text 
along with
##              this program; if not, get it from 
http://www.gnu.org/licenses,
##              or write to the Free Software Foundation, Inc.,
##              59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
##------------------------------------------------------------------------------
__version__ = 0, 9, 1

from __future__ import division


class fixed(object):

     __slots__ = ['_value', '_precision']

     # initialize fixed point number
     def __init__(self, value=0L, precision=None):
         """Set value adjusted to precision.

         Keyword arguments:
         value       initial value (default: 0L)
         precision   initial number of fractional digits (default: None)

         If precision is given, it must be of type int.

         If value is given, it must be of type fixed, string, float, 
long or int
         (or be convertable to float or long).

         If a string is given as value, it must be a string in one of two
         formats:
             [+|-]<int>[.<frac>][[e|E][+|-]<exp>]
             [+|-].<frac>[[e|E][+|-]<exp>]

         The value is always adjusted to the given precision or the 
precision
         is calculated from the given value, if no precision is given.
         """
         if precision is not None:
             assert isinstance(precision, int), \
                                          "Precision must be of <type 
'int'>."
             assert precision >= 0, "Precision must be >= 0."
         if isinstance(value, fixed):
             vp = value._precision
             if precision is None or precision == vp:
                 self._precision = vp
                 self._value = value._value
             else:
                 self._precision = precision
                 if precision < vp:
                     self._value = _shiftdecimal(value._value, vp - 
precision)
                 else:
                     self._value = value._value * 10L ** (precision - vp)
             return
         if isinstance(value, str):
             i, e = _string2exact(value)
             if i is None:
                 raise TypeError, "Can't convert "+repr(value)+" to fixed."
             else:
                 if precision is None:
                     if e > 0:
                         self._precision = 0
                         self._value = i * 10L ** e
                     else:
                         self._precision = abs(e)
                         self._value = i
                 else:
                     self._precision = precision
                     p = e + precision
                     if p >= 0:
                         self._value = i * 10L ** p
                     else:
                         self._value = _shiftdecimal(i, -p)
             return
         if isinstance(value, float):
             if precision is None:
                 self._precision = vp = len(repr(value).split('.')[1])
                 self._value = long(round(value * 10L ** vp))
                 self.adjust()
             else:
                 self._precision = precision
                 self._value = long(round(value * 10L ** precision))
             return
         if isinstance(value, (int, long)):
             if precision is None:
                 self._precision = 0
                 self._value = long(value)
             else:
                 self._precision = precision
                 self._value = long(value) * 10L ** precision
             return
         # Coerce value to Float?
         ok = 1
         try:
             v = float(value)
         except:
             ok = 0
         if ok:
             self.__init__(v, precision)
             return
         # Coerce value to Long?
         ok = 1
         try:
             v = long(value)
         except:
             ok = 0
         if ok:
             self.__init__(v, precision)
             return
         # unable to create fixed
         raise TypeError, "Can't convert "+repr(value)+" to fixed."

     # copy instance
     def __copy__(self):
         return fixed(self)
     __deepcopy__ = __copy__

     # return v, p so that self == v // 10 ** p and v % 10 != 0
     def __reduce(self):
         sv = self._value
         if sv == 0L:
             return 0L, 0
         sp = self._precision
         while sp and sv % 10L == 0L:
             sp -= 1
             sv = sv // 10L
         return sv, sp

     # return _value adjusted to given precision
     def __get_adjusted_value(self, precision):
         p = precision - self._precision
         if p == 0:
             return self._value
         elif p > 0:
             return self._value * 10L ** p
         else:
             return _shiftdecimal(self._value, -p)

     # adjust precision
     def adjust(self, precision=None):
         """Adjust this fixed to the given precision.

         If the given precision is less than the current precision 
information
         may be lost.
         If no precision is given it is adjusted to the minimum precision
         preserving x == x.adjust().
         """
         if precision is None:
             self._value, self._precision = self.__reduce()
         else:
             assert isinstance(precision, int), "Precision must be of 
type Int."
             assert precision >= 0, "Precision must be >= 0."
             self._value = self.__get_adjusted_value(precision)
             self._precision = precision

     # return precision
     def getPrecision(self):
         """Return precision."""
         return self._precision

     # string representation
     def __repr__(self):
         return "fixed('"+self.__str()+"')"

     # convert to string
     def __str(self):
         sp = self._precision
         if sp == 0:
             return str(self._value)
         else:
             sv = self._value
             s = str(abs(sv))
             n = len(s)
             if n >= sp:
                 return (sv<0)*'-' + s[0:-sp] + '.' + s[-sp:]
             else:
                 return (sv<0)*'-' + '0.' + (sp-n)*'0' + s

     # convert to string (according to current locale)
     def __str__(self):
         from locale import format, localeconv
         sp = self._precision
         if sp == 0:
             return format("%d", self._value, 1)
         else:
             i = self.__long__()
             f = self.__frac()
             s = (i==0 and f<0)*'-'      # -1 < self < 0 => i = 0 and f 
< 0 !!!
             return s + format("%d", i, 1) + localeconv()['decimal_point'] \
                                                 + format("%0*d", (sp, 
abs(f)))

     # compare to fixed or any type that can be converted to a fixed
     def __cmp__(self, other):
         if isinstance(other, fixed):
             p = max(self._precision, other._precision)
             return cmp(self.__get_adjusted_value(p),
 
other.__get_adjusted_value(p))
         elif other is None:
             return 1
         else:
             return self.__cmp__(fixed(other))

     # compute hash index
     def __hash__(self):
         sv, sp = self.__reduce()
         return hash(sv) ^ hash(sp)

     # return 0 or 1 for truth-value testing
     def __nonzero__(self):
         return self._value != 0L

     # return integer portion as long
     def __long__(self):
         sp = self._precision
         if sp == 0:
             return self._value
         else:
             sv = self._value
             result = abs(sv) // 10L ** sp
             if sv < 0:
                 return -result
             else:
                 return result

     # return integer portion as int
     def __int__(self):
         return int(self.__long__())

     # return fractional portion as long
     def __frac(self):
         return self._value - long(self) * 10L ** self._precision

     # return fractional portion as fixed
     def frac(self):
         """Return fractional portion as fixed."""
         return self - long(self)

     # convert to float (may loose precision!)
     def __float__(self):
         sv, sp = self.__reduce()
         return sv / 10L ** sp

     # positiv value
     def __pos__(self):
         return fixed(self)

     # negative value
     def __neg__(self):
         result = fixed(self)
         result._value = -result._value
         return result

     # absolute value
     def __abs__(self):
         result = fixed(self)
         result._value = abs(result._value)
         return result

     # self + other
     def __add__(self, other):
         if isinstance(other, fixed):
             p = self._precision - other._precision
             if p == 0:
                 result = fixed(self)
                 result._value += other._value
             elif p > 0:
                 result = fixed(self)
                 result._value += other._value * 10L ** p
             else:
                 result = fixed(other)
                 result._value += self._value * 10L ** -p
             return result
         else:
             return self.__add__(fixed(other))
     # other + self
     __radd__ = __add__

     # self - other
     def __sub__(self, other):
         if isinstance(other, fixed):
             p = self._precision - other._precision
             if p == 0:
                 result = fixed(self)
                 result._value -= other._value
             elif p > 0:
                 result = fixed(self)
                 result._value -= other._value * 10L ** p
             else:
                 result = fixed(other)
                 result._value -= self._value * 10L ** -p
             return result
         else:
             return self.__sub__(fixed(other))
     # other - self
     def __rsub__(self, other):
         return self.__neg__().__add__(other)

     # self * other
     def __mul__(self, other):
         if isinstance(other, fixed):
             sp, op = self._precision, other._precision
             if sp >= op:
                 result = fixed(self)
                 result._value = _shiftdecimal(self._value * 
other._value, op)
             else:
                 result = fixed(other)
                 result._value = _shiftdecimal(self._value * 
other._value, sp)
             return result
         else:
             return self.__mul__(fixed(other))
     # other * self
     __rmul__ = __mul__

     # self / other
     def __div__(self, other):
         if isinstance(other, fixed):
             if other._value == 0L:
                 raise ZeroDivisionError
             sp, op = self._precision, other._precision
             if sp >= op:
                 result = fixed(self)
                 result._value = _divdecimal(self._value * 10L ** op,
 
other._value)
             else:
                 result = fixed(other)
                 result._value = _divdecimal(self._value * 10L ** (2 * 
op - sp),
 
other._value)
             return result
         else:
             return self.__div__(fixed(other))
     # other / self
     def __rdiv__(self, other):
         return fixed(other).__div__(self)
     # fixed division is true division
     __truediv__ = __div__
     __rtruediv__ = __rdiv__

     # self // other, self % other
     def __divmod__(self, other):
         if isinstance(other, fixed):
             if other._value == 0L:
                 raise ZeroDivisionError
             sp, op = self._precision, other._precision
             if sp >= op:
                 r = fixed(self)
                 sv = self._value
                 ov = other.__get_adjusted_value(sp)
             else:
                 r = fixed(other)
                 sv = self.__get_adjusted_value(op)
                 ov = other._value
             q = sv // ov
             r._value = sv - q * ov
             return fixed(q, r._precision), r
         else:
             return self.__divmod__(fixed(other))
     # other // self, other % self
     def __rdivmod__(self, other):
         return fixed(other).__divmod__(self)

     # self // other
     def __floordiv__(self, other):
         return self.__divmod__(other)[0]
     # other // self
     def __rfloordiv__(self, other):
         return self.__rdivmod__(other)[0]

     # self % other
     def __mod__(self, other):
         return self.__divmod__(other)[1]
     # other % self
     def __rmod__(self, other):
         return self.__rdivmod__(other)[1]

     # self ** other
     def __pow__(self, other):
         if isinstance(other, (int, long)):
             result = fixed(self)
             result._value = _shiftdecimal(self._value ** other,
                                                 self._precision * 
(other - 1))
             return result
         else:
             raise TypeError, "Exponent must be int or long."

     # return percentage of self according to rate
     def percent(self, rate):
         """Return percentage according to rate.

         x.percent(rate) = x * rate / 100

         The result is returned as fixed with the same precision as self.
         """
         if isinstance(rate, fixed):
             result = fixed(self)
             f = 10L ** (rate._precision + 2)
             result._value = _divdecimal(self._value * rate._value, f)
             return result
         else:
             return self.percent(fixed(rate))

     # return net value of self according to rate
     def net(self, rate):
         """Return net value according to rate.

         x.net(rate) = x * 100 / (100 + rate)
         The result is returned as fixed with the same precision as self.
         """
         if isinstance(rate, fixed):
             result = fixed(self)
             f = 10L ** (rate._precision + 2)
             result._value = _divdecimal(f * self._value, f + rate._value)
             return result
         else:
             return self.net(fixed(rate))

     # return gross value of self according to rate
     def gross(self, rate):
         """Return gross value according to rate.

         x.gross(rate) = x * (100 + rate) / 100
         The result is returned as fixed with the same precision as self.
         """
         if isinstance(rate, fixed):
             result = fixed(self)
             f = 10L ** (rate._precision + 2)
             result._value = _divdecimal(self._value * (f + rate._value), f)
             return result
         else:
             return self.gross(fixed(rate))

# functions:

# return percent value according to rate
def percent(rate):
     """Return percent value according to rate.

     percent(rate) = rate / 100
     The result is returned as fixed with precision of rate increased by 2.
     """
     if isinstance(rate, fixed):
         r = rate
     else:
         r = fixed(rate)
     r._precision += 2
     return r

# helper functions:

# divide x by y, return rounded result
def _divdecimal(x, y):
     i, r = divmod(x, y)
     c = cmp(abs(2*r), abs(y))
     if c > 0 or (c == 0 and i >= 0):    # r > y/2 or r = y/2 and i not 
negative
         return i + 1
     else:
         return i

# shift decimal point left, return rounded result
def _shiftdecimal(x, n):
     if n == 0:
         return x
     return _divdecimal(x, 10L ** n)

# parse string
import re
_pattern = r"""
             \s*
             (?P<sign>[+|-])?
             (
                 (?P<int>\d+)(\.(?P<frac>\d*))?
                 |
                 \.(?P<onlyfrac>\d+)
             )
             ([eE](?P<exp>[+|-]?\d+))?
             \s*$
             """
_parse = re.compile(_pattern, re.VERBOSE).match
del re

def _string2exact(s):
     m = _parse(s)
     if m is None:
         return None, None
     ep = m.group('exp')
     if ep:
         e = int(ep)
     else:
         e = 0
     ip = m.group('int')
     if ip:
         fp = m.group('frac')
     else:
         fp = m.group('onlyfrac')
     if ip:
         i = long(ip)
     else:
         i = 0L
     if fp:
         f = long(fp)
         n = len(fp)
     else:
         f = 0L
         n = 0
     i = i * 10 ** n + f
     e -= n
     if m.group('sign') == '-':
         i = -i
     return i, e



From arosado at softhome.net  Mon Sep  2 22:00:03 2002
From: arosado at softhome.net (Andres Rosado)
Date: Mon, 02 Sep 2002 22:00:03 -0400
Subject: Is it worth learning python?
Message-ID: <5.1.0.14.0.20020902220001.00bea188@mail.softhome.net>

At 12:44 AM 8/26/2002 -0400, you wrote:
> > languages. Python is a great first language.
>
>Python is a great last language!

I agree with you, Python is a great language no matter when you learn it. 
But this is my recommendation for the OP, who is learning to program.

P.D. Sorry for the delay, but last week was a tough one.

-----------------------------------
Andres Rosado
Email: andresr at despammed.com
ICQ: 66750646
Homepage: http://andres980.tripod.com/

A word to the wise is enough.
                 -- Miguel de Cervantes




From duncan at NOSPAMrcp.co.uk  Mon Sep  2 05:57:55 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Mon, 2 Sep 2002 09:57:55 +0000 (UTC)
Subject: Why it is so dramatical?
References: <mailman.1030957743.4021.python-list@python.org>
Message-ID: <Xns927D6DBDE418Cduncanrcpcouk@127.0.0.1>

"Bo M. Maryniuck" <b.maryniuk at forbis.lt> wrote in
news:mailman.1030957743.4021.python-list at python.org: 

> The task:
>      Adding a text to a text.
> 
> The problem:
>      It is tragically slow to add string to huge string.
> 
> The question: Why?

Because the text in strings is stored in a single block of memory, so 
extending a string requires making a copy of all the data. Also strings are 
immutable so you cannot extend them 'in-place'. Like all things it is a 
tradeoff. If you make strings mutable then extending them might be a bit 
faster, but many other operations would then require copying strings in 
case they were modified, so you lose out in general more than you gain.

None of this is specific to Python, the same issues arise in almost any 
programming language (although not always the same solutions).

> 
> The other question: 
>      Is it true, that I should through all the Python life 
>      use lists and then join it into single text instead to use +=
>      operator? 
> 
Using lists and joining them should come to hand as the first solution in 
most cases where you are building up a string.

Using cStringIO, array or mmap should also be in your toolbox for those 
cases where they are the most appropriate.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From manuel at hendel.net  Tue Sep 17 04:02:09 2002
From: manuel at hendel.net (Manuel Hendel)
Date: Tue, 17 Sep 2002 10:02:09 +0200
Subject: problems getting a single item out of a list
In-Reply-To: <200209170030.52964.shalehperry@attbi.com>
References: <20020917071900.GB23629@partagas.as.de.cw.net> <200209170030.52964.shalehperry@attbi.com>
Message-ID: <20020917080209.GC23629@partagas.as.de.cw.net>

On Tue, Sep 17, 2002 at 12:30:52AM -0700, Sean 'Shaleh' Perry wrote:
> 
> just use fields[2] (you start counting at zero).

This worked fine:

domainfield = fields[3]

But the next problem is, that after a 

domainfield = string.split(fields[3])

domainfield seams to be a list item again.

print domainfield

['142000']

Manuel

-- 
In Ohio, if you ignore an orator on Decoration day to such an extent as to 
publicly play croquet or pitch horseshoes within one mile of the speaker's 
stand, you can be fined $25.00. 



From martin at v.loewis.de  Wed Sep  4 18:50:43 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 05 Sep 2002 00:50:43 +0200
Subject: Memory fault - core dumped   while compiling 2.2.1 on SCO
References: <3D757EFA.9050401@nospam.yahoo.com> <yu99sn0qqw1r.fsf@europa.research.att.com> <m3bs7etem9.fsf@mira.informatik.hu-berlin.de> <3D765A05.6030004@nospam.yahoo.com> <j48z2h61j9.fsf@informatik.hu-berlin.de> <3D768753.8030302@nospam.yahoo.com>
Message-ID: <m3n0qx8j3w.fsf@mira.informatik.hu-berlin.de>

Anton Graph <aglyportREMOvethispart at nospam.yahoo.com> writes:

> python -E setup.py build in gdb:
> 
> (gdb) where
> #0  0x80016d06 in do_reloc () from /usr/lib/libc.so.1

That sounds like a question to bring to an SCO forum. Apparently, this
crashes when doing relocations while loading the extension
module. This raises a number of questions:

- Why does it do relocations at all? Could it be that you have object
  files that a not position-independent? I.e. is distutils using the
  compiler correctly both for compilation, and for linking?

  It might be correct that it tries to perform relocation, e.g. for
  the global offset table, and the procedure linkage table - but it
  *is* troubling that relocations is where it crashes.

- What is the precise version of the operating system you are using?

- I'm troubled by the fact that in both configurations (with an
  without gcc), the compiler is always referred-to as "cc". Are you
  sure you are not using gcc?

  Please have a look at

  http://gcc.gnu.org/ml/gcc-testresults/1999-04/msg00239.html

  Notice that the failure mode is quite similar... Likewise,

  http://www.bitmechanic.com/mail-archives/dbi-users/Jun1999/0485.html

  reports such a problem when using gcc on SCO.

Regards,
Martin


From chris.gonnerman at newcenturycomputers.net  Sat Sep  7 00:23:36 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Fri, 6 Sep 2002 23:23:36 -0500
Subject: Python from C/SDL
References: <3D793B29.3198AB0E@ihug.co.nz>
Message-ID: <006101c25626$56c9b1c0$0101010a@local>

----- Original Message ----- 
From: "Gib Bogle" <bogle at ihug.co.nz>


> Hi,
> 
> I am a complete beginner with Python.  My interest in it is focussed on
> a great piece of open-source software called pymol, a system for
> visualization of molecules that uses Python and OpenGL.  I'd like to get
> access to pymol's graphical functions from my own program, which is
> built from C, SDL and OpenGL.

Have you looked at PyGame (www.pygame.org)?  It wraps SDL in Python.
Some pretty impressive things can be done there, and in less lines
of code than the equivalent C.

> I'm wondering if (with a bit of surgery) I'd be able to get access to
> the lower-level functionality of pymol, by which I mean the file I/O and
> rendering functions.

Probably not too hard to implement as a C module, if you are already
an accomplished C programmer.  The C API for Python is pretty easy,
once you grasp refcounting.

> I will not want to use the GUI at all.  SDL would
> create the surface(s), and handle the buffer swapping.  My goal is to be
> able to send a request to the pymol API asking for a frustum view to be
> rendered on the provided surface.
> 
> Sorry if my question is one that gets asked and answered frequently.  If
> there is a place where this has been dealt with please direct me.

Never heard this one before...

Good luck!


Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From sholden at holdenweb.com  Tue Sep 17 12:33:25 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 17 Sep 2002 12:33:25 -0400
Subject: don't understand error message
References: <mailman.1032251060.4906.python-list@python.org> <Xns928C66E3994D7duncanrcpcouk@127.0.0.1> <mailman.1032255920.1725.python-list@python.org>
Message-ID: <VYIh9.12001$T4.1018@fe04>

"Manuel Hendel" <manuel at hendel.net> wrote in message
news:mailman.1032255920.1725.python-list at python.org...
> On Tue, Sep 17, 2002 at 09:12:06AM +0000, Duncan Booth wrote:
> > It simply means that at least one line of your file doesn't have 4
fields
> > after you have finished your processing. If the line only has two "|"
> > characters in it, then you will only get three fields which you can
access
> > with indices 0 to 2. Trying to use an index out of range for a list will
> > produce this error.
> >
> > Try sticking a "print fields" line just before the if statement, the
last
> > line printed will show you which line didn't match.
>
> My script already changed a little bit, I tried to fix it by my self,
> but it seams, that I just moved the error. How would you check the
> lines for the number of fields?
> Actually, I think I found that problem. I forgot some "---|--|--" at
> the end.
> But I already got a new one.
>
> [....]
> if domainfield in domain:
>     domainlines.append(line)
> [....]
>
> This should work if the domainfield and domain got the following
> values, shouldn't it?
>
> domainfield = 1234.ab.cdef.de
> damain = 1234.ab.cdef.de
>
> But I get this error message:
>
> TypeError: 'in <string>' requires character as left operand
>

You're just a little bit ahead of reality (have you been using Guido's time
machine without permission?). This should woirk in 2.3, but for now you'll
have to use the .index() or .find() methods of the string objects.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From skip at pobox.com  Thu Sep  5 22:29:06 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 5 Sep 2002 21:29:06 -0500
Subject: pygtk-2 tutorial/doc?
In-Reply-To: <pan.2002.09.06.02.15.03.816977.8282@yahoo.com>
References: <pan.2002.09.06.02.15.03.816977.8282@yahoo.com>
Message-ID: <15736.4850.859179.13616@12-248-11-90.client.attbi.com>

    Matt> I'd like to find some kind of tutorial or at least some
    Matt> introductory documentation on programming with pygtk-2 with python
    Matt> 2.2.

    Matt> Anyone have any ideas on where I should be looking?

I had similar problems when I first used PyGtk.  I had no real Gtk
experience, so it took a little while to get comfortable with its model.  I
think the intention is that the plain old Gtk docs should suffice.  Failing
that, you might look at the examples in the PyGtk distribution.  In any
case, you might get more helpful responses from the pygtk mailing list
(pygtk at daa.com.au).

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From romany at actimize.com  Thu Sep 12 06:18:32 2002
From: romany at actimize.com (Roman Yakovenko)
Date: Thu, 12 Sep 2002 12:18:32 +0200
Subject: class variable won't icrement!
Message-ID: <7647A9F4B298E74DB6793865DA67285004ADF4@exchange.adrembi.com>

Hi it is simple just change 
self.id = self.id + 1 

To 

test.id += 1

Roman

> -----Original Message-----
> From: Gumuz [mailto:gumuz at amoebe.looze.net]
> Sent: Thursday, September 12, 2002 10:55 AM
> To: python-list at python.org
> Subject: class variable won't icrement!
> 
> 
> Hello everyone,
> 
> I have something very strange and i can not understand it. 
> probably, it's
> not _that_ strange : )
> 
> I have this simple class which registers itself on a class 
> variable in the
> following code:
> 
> --------------------------------------------------------------
> --------------
> ------
> class test:
>     dict = {}
>     id = 1
>     def __init__(self, name):
>         self.dict[name] = self
>         self.id = self.id + 1
> 
>     def show(self):
>         print "class variable dict element count: ", 
> str(len(self.dict))
>         print "class variable id: ", self.id
> --------------------------------------------------------------
> --------------
> --------
> 
> now, look at the following:
> 
> >>> john = test("john")
> >>> frank = test("frank")
> >>> bill = test("bill")
> >>> bill.show()
> class variable dict element count:  3
> class variable id:  2
> >>> frank.show()
> class variable dict element count:  3
> class variable id:  2
> 
> The dict variable acts fine, maintaining it's content throughout every
> instance of the class.
> The id variable however, seems to be only incremented within 
> the instance.
> In the end this variable should've been 4.
> 
> Could someone explain this to me? Is there something about 
> class variables
> whta i should know?
> 
> many thanx,
> Guyon
> 
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 



From gerhard.haering at gmx.de  Tue Sep  3 16:36:51 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 3 Sep 2002 20:36:51 GMT
Subject: Servizi Internet
References: <mailman.1031083025.5304.python-list@python.org>
Message-ID: <slrnana7b1.t8.gerhard.haering@lilith.my-fqdn.de>

Powell, Thomas wrote in comp.lang.python:
> Wasn't this spam from an _ISP_ anyway?  An abuse message to the ISP
> might get them to back off their spamming practices, but otherwise,
> what is the only "legal" recourse?  Complain to their backbone
> provider?

Yep. It contains:

"X-Complaints-To: abuse at baymountain.net"

[x] Done.

> -----Original Message-----
> [snip stupid M$ Outhouse quoting, yuck!]

http://www.math.fu-berlin.de/user/guckes/mail/edit.html

-- Gerhard


From marklists at mceahern.com  Mon Sep 23 23:02:38 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 23 Sep 2002 22:02:38 -0500
Subject: confusion about opening files
In-Reply-To: <3d8fd215$0$192$75868355@news.frii.net>
Message-ID: <JHEOKEOOLIGLDHCMAHMOMEOPEAAA.marklists@mceahern.com>

> O. K. Now I am confused.  I thought that when I imported the os module, i
> would have to call its functions like this: os.open()    Why am I 
> just using the function name?

You are assuming open == os.open.  Not so.  Watch...

***

$ python
Python 2.2.1 (#1, Jun 25 2002, 10:55:46)
[GCC 2.95.3-5 (cygwin special)] on cygwin
Type "help", "copyright", "credits" or "license" for more information.
>>> print open.__doc__
file(name[, mode[, buffering]]) -> file object

Open a file.  The mode can be 'r', 'w' or 'a' for reading (default),
writing or appending.  The file will be created if it doesn't exist
when opened for writing or appending; it will be truncated when
opened for writing.  Add a 'b' to the mode for binary files.
Add a '+' to the mode to allow simultaneous reading and writing.
If the buffering argument is given, 0 means unbuffered, 1 means line
buffered, and larger numbers specify the buffer size.
Note:  open() is an alias for file().

>>> import os
>>> print os.open.__doc__
open(filename, flag [, mode=0777]) -> fd
Open a file (for low level IO).
>>> open == os.open
0

***

Cheers,

// m
-




From loewis at informatik.hu-berlin.de  Thu Sep 19 10:21:42 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 19 Sep 2002 16:21:42 +0200
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <mailman.1032329366.6813.python-list@python.org> <j4ptvbaapm.fsf@informatik.hu-berlin.de> <3D890FB0.4030408@something.invalid> <m3admeo4u1.fsf@mira.informatik.hu-berlin.de> <ambso1$4g8vh$1@ID-59885.news.dfncis.de> <j4fzw6tofh.fsf@informatik.hu-berlin.de> <mailman.1032441028.15013.python-list@python.org>
Message-ID: <j4elbqrrex.fsf@informatik.hu-berlin.de>

Christian Tismer <tismer at tismer.com> writes:

> Maybe sufficient for not crashing on the C level.
> But this is not enough for micro-thread-safety.
> See the random module for example, it needs a lock
> or it will give unpredictable results.

Academically, I don't think the results are unpredictable, as
micro-thread switch occur at predictable points in processing.

Regardless, you are saying that the random module is not
thread-safe. A number of Python modules is not thread safe; many of
them won't be 'safe' in the presence of microthreads, either.  With
microthreads, the problem is usually not as bad as with real threads,
since a switch usually cannot occur in the middle of a library routine
(unless the library calls back, and the call-back causes a switch).

So this is something that has been accepted all the years, and should
not cause FUD.

Regards,
Martin


From phr-n2002b at NOSPAMnightsong.com  Thu Sep 26 17:42:30 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 26 Sep 2002 14:42:30 -0700
Subject: Who knows somefunction?
References: <3D937B08.4090704@aon.at> <3D937D4C.2BFDF40A@alcyone.com>
Message-ID: <7xsmzwwhq1.fsf@ruckus.brouhaha.com>

Erik Max Francis <max at alcyone.com> writes:
> > Is there a function <somefunction> in Python
> > for which
> > 
> > something is <somefunction>(id(something))
> > 
> > returns invariably true?
> 
> Not offhand, but there's an easier test.  x is y really just means id(x)
> == id(y), so you could change this test to:
> 
> 	id(something) == id(somethingElse)
> 
> and you probably wouldn't have the need for your hypothetical function.

No I think he's asking for a way to dereference an id of an object.
And the answer is there's no way to do that.


From owen at nospam.invalid  Tue Sep 10 17:11:59 2002
From: owen at nospam.invalid (Russell E. Owen)
Date: Tue, 10 Sep 2002 14:11:59 -0700
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid>
Message-ID: <alln71$1s2s$1@nntp6.u.washington.edu>

In article <3D7C210F.7090708 at something.invalid>,
 Greg Ewing <see_reply_address at something.invalid> wrote:

> "Joseph A. Knapka" <jknapka at earthlink.net> wrote in message news:<3D798461.B3FE0E06 at earthlink.net>...
>>...(text elided)...
>...I hope you're aware that what you're doing in your buttons
>example is NOT equivalent to using the "command" option.
>The latter does a number of extra things: it waits until
>the mouse is released, and triggers the handler only if
>the release is within the button.

(what Joseph was doing was binding the button to the <Button-1> event, 
e.g. mybutton.bind("<Button-1>", afunction))

Alas, too true. Button-1 fires when the mouse button is pressed (very 
nonstandard behavior) and ButtonRelease fires when the mouse button is 
released even if the mouse is no longer over the Button object. I don't 
know of any event binding that does a proper job of emulating the 
standard behavior of pressing a button. Button(command=...) is the way 
to go for this.

If you need to know which button the user pressed (or any other 
information that you know at the time you specify the button command 
callback), define a new callback function that contains the information. 
You certainly don't have to use lambda (and I never do).

Scott David Daniels wrote a very useful class that makes writing 
callback functions that contain stored information a breeze. It is 
available at 
<http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52549>). I make 
heavy use of it and feel it really cleans up my code.

Actually, I use a minor variant I wrote that optimizes the common case 
of no stored keyword arguments. You can get my version (as well a 
Tkinter summary I wrote) at <http://www.astro.washington.edu/owen>; the 
former is called GenericCallback and can be found in util.zip.

-- Russell


From bobx at linuxmail.org  Mon Sep 16 22:35:49 2002
From: bobx at linuxmail.org (Bob X)
Date: Tue, 17 Sep 2002 02:35:49 GMT
Subject: Python equivalent to perldoc?
References: <Olvh9.7898$Lg2.1389903@news2.news.adelphia.net> <mailman.1032226040.2391.python-list@python.org>
Message-ID: <9ywh9.7955$Lg2.1406050@news2.news.adelphia.net>

Skip Montanaro wrote:
>     Bob> In Perl you can run "perldoc warnings" and get the definition of
>     Bob> what "warnings" does. Is there a equivalent in Python?
> 
> Try "pydoc warnings".
> 
What a funny guy!  :-)



From wavefrnt at tampabay.rr.com  Sat Sep 28 06:24:51 2002
From: wavefrnt at tampabay.rr.com (David Edwards)
Date: Sat, 28 Sep 2002 06:24:51 -0400
Subject: print a %
In-Reply-To: <ff8044a.0209280211.345dc898@posting.google.com>
References: <ff8044a.0209280211.345dc898@posting.google.com>
Message-ID: <20020928102451.GA27641@buttercup.3digital.com>

On Sat, Sep 28, 2002 at 03:11:57AM -0700, Claude wrote:
> When I do 
> print "image size=%d%" %(50)
> I've got an error
> 
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> ValueError: incomplete format
> 
> I do not know how to escape the second %
> 
> print "image size=%d%" %(50) 
> image size=50%
> 

You escape the % with another %.

Like this:

print "image size=%d%%" %(50)

-- 
david edwards



From bobx at linuxmail.org  Mon Sep 16 16:46:03 2002
From: bobx at linuxmail.org (Bob X)
Date: Mon, 16 Sep 2002 20:46:03 GMT
Subject: [FEEDBACK] Is this script efficient...is there a better way?
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <L9Rf9.168314$l_4.127784@atlpnn01.usenetserver.com> <4ZRf9.1782$Lg2.248733@news2.news.adelphia.net> <3hSf9.168947$l_4.98525@atlpnn01.usenetserver.com> <8PSf9.1796$Lg2.257378@news2.news.adelphia.net> <_oUf9.170870$l_4.126599@atlpnn01.usenetserver.com> <Am7g9.2344$Lg2.402676@news2.news.adelphia.net> <3pkg9.3313$WV.2782@fe03>
Message-ID: <fqrh9.7075$Lg2.1331983@news2.news.adelphia.net>

Steve Holden wrote:
<snip>

I did a test on the 30MB sized logs and the RE did indeed come out the 
faster, thought not by much. The interesting thing is it caught more as 
well since my script looked for exact matches and the RE based one 
ignored case (as far as my feeble brain understands such things).

I remember why I had the "result" statement. It was because I needed to 
write the keyword found and then the line it was found on to the hits file.

Can't seem to make that work with the RE based one. Then again...I know 
RE as far as the east is from the west.

Bob




From pierpaolo.glave at ciaolab.com  Thu Sep 12 04:11:04 2002
From: pierpaolo.glave at ciaolab.com (Pier Paolo Glave)
Date: 12 Sep 2002 01:11:04 -0700
Subject: Tkinter, updateStringProc for type font
References: <1907a5e4.0209040455.64ae73b1@posting.google.com> <3D7F772C.80205@ActiveState.com>
Message-ID: <1907a5e4.0209120011.7acb1a40@posting.google.com>

Jeffrey Hobbs <JeffH at ActiveState.com> wrote in message news:<3D7F772C.80205 at ActiveState.com>...

> Do you have any other C extensions in your application?  I've not seen this 
> occur myself, except in bugs where other C extensions were abusing Tcl objects 
> internally.

No, I'm not using any C extensions, as far as I know.
Besides the standards distribution, I only installed BLT, but I were
not using it when these errors occurred.

Bye
--
Pier Paolo Glave


From wlfraed at ix.netcom.com  Thu Sep 26 16:06:41 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Thu, 26 Sep 2002 13:06:41 -0700
Subject: Getting last char in string.
References: <mailman.1033027776.19152.python-list@python.org> <amumk5$p52$1@lust.ihug.co.nz> <amuna6$erv$1@news1.wdf.sap-ag.de> <amuq6t$sv9$1@lust.ihug.co.nz>
Message-ID: <hcpvma.2f3.ln@ix.netcom.com>

CheapSkate fed this fish to the penguins on Thursday 26 September 2002 
04:15 am:


> Single character are strings of length 1(as u said) and " is single
> character right?
>
        The line did NOT contain a double-quote (") but rather /two/ 
single-quotes ('') next to each other.

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From max at alcyone.com  Tue Sep 24 20:22:04 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 24 Sep 2002 17:22:04 -0700
Subject: Strange python and cgi!
References: <amquos$rcu$1@lust.ihug.co.nz>
Message-ID: <3D9101AC.842F0513@alcyone.com>

CheapSkate wrote:

> when I go to the webpage, it stops halfway and in "view source: it
> only show
> the code till the open "pre".

It's likely some other error occurred.  Check your Web server error log
for the final word.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ As far as I'm concerned, being any gender is a drag.
\__/ Patti Smith
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.


From 53ab2750 at meowing.net  Wed Sep 25 20:08:14 2002
From: 53ab2750 at meowing.net (itsy bitsy meowbot)
Date: Thu, 26 Sep 2002 00:08:14 GMT
Subject: "which" command?
References: <yibznu5fxvx.fsf@coxeter.math.toronto.edu>
Message-ID: <Odsk9.564$qg5.306@nwrddc02.gnilink.net>

marco wrote:
> 
> Hi,
> 
> A few days ago there was a discussion here about a
> "which" written in python (as opposed to running
> the OS "which" command). Any pointers to that thread?

Are you looking for this? <http://starship.python.net/~tmick/>



From bogle at ihug.co.nz  Sat Sep  7 20:47:22 2002
From: bogle at ihug.co.nz (Gib Bogle)
Date: Sun, 08 Sep 2002 12:47:22 +1200
Subject: Larry Wall's comment on python...
References: <mailman.1031349792.2374.python-list@python.org> <3d796c18@news.sentex.net>
Message-ID: <3D7A9E1A.1171C857@ihug.co.nz>

Peter Hansen wrote:
> 
> mmaddox at hcsmail.com wrote:
> > Although I grant the Python structure and the column-based
> > structures are not completely analogous, they BOTH burden the programmer
> > with arbitrary rules to govern the physical layout of the code. Although
> > this layout may be considered more elegant from a readability standpoint,
> > it's the enforcement of the layout that is irritating to the more
> > Libertarian of us.
> 
> I think you have the wrong meaning for "arbitrary".  From the
> sounds of it, you have exactly reversed which is arbitrary (your
> own "libertarian" coding style) and which is "governed by fixed
> rules" (Python layout).
> 
> > Having read Wall's comments on Slashdot, I think he is really just mildly
> > complaining about ridiculously long page-widths in his editor, not really
> > slashing Python seriously. He's just giving some props to his own baby at
> > the expense of another. (He had to say something, right?) You must agree
> > that Python can lead you to horizontal scrolling in your editor - the bane
> > of user interface design.
> 
> I agree with the comments about Larry "having" to say something like this.
> 
> I definitely don't understand how Python, more or less than any other
> language, leads to more horizontal scrolling.  Pretty much all
> Python code I've ever seen (though I haven't seen yours yet :-) is
> roughly 80 characters max, and mine own has been about that since I
> started writing C code 16 years ago.
> 
> Do you mean there is something special about Python not delimiting
> blocks in the traditional ways which somehow leads you to write
> much wider lines of code?

Maybe he is referring to an effect of default editor tab size.


From a.clarke11 at pop.ntlworld.com  Sat Sep  7 16:23:39 2002
From: a.clarke11 at pop.ntlworld.com (a.clarke11)
Date: Sat, 07 Sep 2002 21:23:39 +0100
Subject: Thinking of joining
References: <3D7970DC.5115BCC8@earthlink.net> <20020907095211.24938.00000006@mb-mt.aol.com>
Message-ID: <3D7A604B.31FE7E73@pop.ntlworld.com>

And its not dead, hasn't expired, gone to Valhalla, isn't kaput or aything like
that. Thanks, teaand bikkie for reminding us what it is all about

TeaAndBikkie wrote:

> IMHO Python has beautiful plumage, isn't nailed to the perch, and won't leave
> you pining for the fields. It also has a good range of cheese, even the ones
> you thought you didn't need, and you won't get stoned by your friends for
> joining.
>
> (Any implied references to Parrot are completely accidental).
>
> I'm too new to Python to give you good technical reasons to use it, but I like
> it :)



From aleax at aleax.it  Thu Sep 26 10:28:11 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 26 Sep 2002 14:28:11 GMT
Subject: Getting started
References: <00a801c26511$4e084750$df7e4e18@brianspiv1700> <mailman.1033021239.13549.python-list@python.org> <iaDk9.145545$ub2.3207724@news1.tin.it> <lm5olw7n.fsf@ion-tof.com>
Message-ID: <%PEk9.146154$ub2.3217944@news1.tin.it>

Thomas Heller wrote:
        ...
> Hidden fields in a union by the discretion of the compiler?
> Don't think that would be allowed by any C standard.

Could you help me out by pointing to chapter and verse that would
make it disallowed, please?  I can't find anything that would
disallow it, and it appears to me that the sentence:
"""
if the value of a member of a union object is used when the most recent 
store to the object was to a different member, the behavior is 
implementation- defined.
"""
(with the specific exception of unions of structs that share
a common initial sequence).

The specs for sizeof seem to go out of their way to NOT constrain
sizeof(aunion) to be the largest sizeof of union members, so as
to avoid prohibiting such internal tags -- it says the size is
*sufficient* to contain the largest member, NOT *is* that member's
size.


So, what _am_ I missing...?


Thanks,

Alex



From fperez528 at yahoo.com  Mon Sep 30 22:58:43 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 30 Sep 2002 20:58:43 -0600
Subject: Reloading a module
References: <bondpaper-648404.19153230092002@nnrp04.earthlink.net> <anastb$h2s$1@peabody.colorado.edu> <bondpaper-06D4B1.20204130092002@nnrp02.earthlink.net>
Message-ID: <anb314$kim$1@peabody.colorado.edu>

Tom wrote:

> I wish it were that simple. I've used reload(module) many times, each
> time hoping that it would do what the documentation says it will do.
> I've tested it by adding a simple variable to my module, watching to see
> if it shows up in the new import, but it never does. I noticed also that
> sometimes it reloads from the .pyc file, and if the reload happens
> before the .pyc file is updated with the new changes, the changes won't
> be available. Even if I force a reload from the updated .py file by
> deleting .pyc file, it still retains the old code. I'm at a loss to
> understand what's going on.

True. I actually don't rely on reload almost ever. I instead tend to use a 
cooked-up version of execfile() for this, or deep_reload. The former is 
available under ipython as @run, the latter as dreload(). Using @run works ok 
only for top-level stuff, of course. dreload() is pretty good but still can 
fail in certain cases.

You may want to give them a try.

Cheers,

f.


From tismer at tismer.com  Thu Sep 19 14:02:52 2002
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 19 Sep 2002 20:02:52 +0200
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <mailman.1032329366.6813.python-list@python.org> <j4ptvbaapm.fsf@informatik.hu-berlin.de> <3D890FB0.4030408@something.invalid> <m3admeo4u1.fsf@mira.informatik.hu-berlin.de> <ambso1$4g8vh$1@ID-59885.news.dfncis.de> <j4fzw6tofh.fsf@informatik.hu-berlin.de> <mailman.1032441028.15013.python-list@python.org> <j4elbqrrex.fsf@informatik.hu-berlin.de>
Message-ID: <3D8A114C.50007@tismer.com>

Martin v. L?wis wrote:
> Christian Tismer <tismer at tismer.com> writes:
> 
> 
>>Maybe sufficient for not crashing on the C level.
>>But this is not enough for micro-thread-safety.
>>See the random module for example, it needs a lock
>>or it will give unpredictable results.
> 
> 
> Academically, I don't think the results are unpredictable, as
> micro-thread switch occur at predictable points in processing.

Yes, right now, but this will change pretty soon.

> Regardless, you are saying that the random module is not
> thread-safe. A number of Python modules is not thread safe; many of
> them won't be 'safe' in the presence of microthreads, either.  With
> microthreads, the problem is usually not as bad as with real threads,
> since a switch usually cannot occur in the middle of a library routine
> (unless the library calls back, and the call-back causes a switch).

I am not saying that the random module is not thread safe (although
I'm not sure if it is). The point is: Microthread safety is more
than thread safety, since microthreads don't care about thread
locks, since they are still *in the same thread*.

It will for sure become a problem, and it is different from
effect, but similar to the C probs by reason: We are switching
context where some code might not expect it. It might feel safe
because it is thread safe, but this is not enough.
I have to do rigid locking of every unknown piece of code,
until somebody explicitly claims some code to be switchable.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From peter at engcorp.com  Fri Sep  6 08:57:53 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 06 Sep 2002 08:57:53 -0400
Subject: Timothy Rue
References: <898662b7d85d2c966c0bd45d563d5d2d@xganon.com>
Message-ID: <3d78a638$1@news.sentex.net>

xganon wrote:
> Does Timothy Rue still Troll here? 
> If not, how did you get rid of him?
> We need him out of our group now,
> maybe we could get him into a Perl
> group for a while and let them deal
> with him?

Have you heard of that nasty "cocktail" treatment that has been
tried in the past for cancer?  It looked a lot like that...

Sorry to hear the infection has spread.

-Peter



From syver-en at online.no  Sat Sep 14 19:57:45 2002
From: syver-en at online.no (syver-en at online.no)
Date: Sat, 14 Sep 2002 23:57:45 GMT
Subject: IDLE
References: <wpbf9.264320$aA.45691@sccrnsc02> <3D7D4F84.318CDB0F@bellatlantic.net> <v4Pg9.428631$me6.49396@sccrnsc01>
Message-ID: <bs70f7ks.fsf@online.no>

"Ali K" <akakakhel at attbi.com> writes:

> Hardware: 386 enhanced, 75MHz, 32MB RAM.
> OS: Win98
> Copied files from other computer to mine.
> It doesn't open when I click on ide.pyw or when I choose to edit a py
> file
> 
> with it.

I suggest you install the python distro from www.python.org.

-- 

Vennlig hilsen 

Syver Enstad


From google at quinceleaf.com  Thu Sep 19 16:38:50 2002
From: google at quinceleaf.com (Brian Ibbotson)
Date: 19 Sep 2002 13:38:50 -0700
Subject: Strange behavior of ClientCookie and site-packages
Message-ID: <6cb82943.0209191238.4b116988@posting.google.com>

I'm running Python 2.2 in RH7.3.

I've been trying to use ClientCookie, and have installed it per its
README instructions (moving it to a ClientCookie dir in
site-packages).

The module is accessible from the 2.2 command line shell, where
sys.path reports

   /usr/lib/python2.2/site-packages/ClientCookie

among the search paths. The command

   import ClientCookie

succeeds without problem.

However, Idle for Python 2.2 cannot do the same. With an identical set
of paths reported by sys.path, including the ClientCookie reference
above, Idle reports:

   >>> import ClientCookie
   Traceback (most recent call last):
     File "<pyshell#0>", line 1, in ?
       import ClientCookie
   ImportError: No module named ClientCookie
   >>> 

Can anyone suggest an explanation for this behavior? Other third-party
modules installed to site-packages (such as ChartDirector and MySQLdb)
have had no problems once they are included in sys.path

- Brian


From lordmacro at hotmail.com  Mon Sep 30 06:01:57 2002
From: lordmacro at hotmail.com (Tom Widrick)
Date: Mon, 30 Sep 2002 05:01:57 -0500
Subject: sizeof PyTypeObject difference
References: <j5Rl9.29807$IO5.4027049@e3500-atl1.usenetserver.com> <m34rc8qamo.fsf@mira.informatik.hu-berlin.de> <mrTl9.30085$IO5.4052423@e3500-atl1.usenetserver.com> <j4adlz3mfs.fsf@informatik.hu-berlin.de>
Message-ID: <JdVl9.85910$5M3.3944098@e3500-atl2.usenetserver.com>

Interestingly enough, it turned out that the easiest way to look at it for
me was to remember that:

class Accumulator:
      __metaclass__ = type
      __value = 0
      value = property(lambda s: s.__value)
      def add(self, v):
        self.value = self.__value + v
      def sub(self, v):
        self.value = self.__value - v

was the same as

tp_dict = {}
def add(self, v):
    self.__value = self.__value + v
tp_dict['add'] = add
def sub(self, v):
     self.__value = self.__value - v
tp_dict['sub'] = sub
tp_dict['__value'] = 0
tp_dict['value'] = property(lambda s: s.__value)
Accumulator = type('Accumulator', (), tp_dict)

I was able to take a "regular" extension type I was working on, defined a
tp_init type member that implemented the (name, bases, dict) interface and
bam, a metatype. So, I guess it could be argued that metatypes in C are
supported by the class/type unification. It's just easier to use the builtin
functions as helpers when implementing the metatype, (like
PyObject_GenericGetAttr). Nearly all of the functionality of the Type object
could be grafted over into my extension type, so I assume that was the
intended way, and not to subtype PyType_Type.

Tom






From kokohh at hotmail.com  Wed Sep  4 23:53:44 2002
From: kokohh at hotmail.com (koko)
Date: Thu, 05 Sep 2002 03:53:44 GMT
Subject: newbie problem: use socke lib to retrieve one web page:
Message-ID: <czAd9.6803$yt3.3242044@newssrv26.news.prodigy.com>

I write this to retrieve one web page using socket lib.

# retrieve webpage
import socket
HOST='www.uic.edu' # remote host
PORT=80     # remort port
s=socket.socket(socket.AF_INET, socket.SOCK_STREAM)
m="""GET /home/events.shtml HTTP/1.0
From: hh at uic.edu
User-Agent: test/1.0
"""
s.connect((HOST,PORT))
s.send(m)
data=s.recv(4096)
print data
s.close()

-----------
But I retrieved nothing and the program stuck.




From sismex01 at hebmex.com  Thu Sep 26 17:33:15 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Thu, 26 Sep 2002 16:33:15 -0500
Subject: How do you pronounce "__" (double underscore)?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F0623302604FB2@mail.hebmex.com>

> 
> In article <525eaba.0209260956.72bf6786 at posting.google.com>, 
> Pat Notz wrote:
> > @ = "at"
> > ! = "bang"
> > . = "dot"
> > * = "star"
> > 
> > Is there a short (one or two syllable) pronounciation for 
> > "__" (double underscore).  "Underscore underscore" and "double 
> > underscore" are kind of long.  Somthing like "splat"?  Anyone?
> 
> I've heard "splat" used for both '#' and '*', so I'd cross it
> off the list...
> 
> -- 
> Grant Edwards                   grante             Yow!  Sign 
> my PETITION.
>                                   at               
>                                visi.com            

Hmmm.. how about a "squish"?


-gus










Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From nhodgson at bigpond.net.au  Mon Sep  9 23:20:36 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Tue, 10 Sep 2002 03:20:36 GMT
Subject: PythonWin bug [very weird]
References: <vU2f9.7489$yt3.3820112@newssrv26.news.prodigy.com> <3D7CC91D.5000607@mxm.dk>
Message-ID: <8ydf9.14146$_l1.38749@news-server.bigpond.net.au>

Max M:

> That is almost as good as when you write a top latin-1 character like
> '?' and you try to delete it via backspace and it only seems to delete
> one of two bytes making it up, and you end up with '?'.

   That is caused by reusing the IDLE automatic indentation smart backspace
code without allowing for characters that use multiple bytes. If you are
interested in fixing it, the code is in
Pythonwin\pywin\idle\AutoIndent.py

   Neil




From 2002 at weholt.org  Thu Sep 19 07:05:00 2002
From: 2002 at weholt.org (Thomas Weholt)
Date: Thu, 19 Sep 2002 11:05:00 GMT
Subject: StructuredText -> HTML module??
Message-ID: <wbii9.19869$0p1.359879@news2.ulv.nextra.no>

Need a module to parse StructuredText/Wiki into HTML. Looked at PyWiki, but
the generated code was not what I expected. Does anybody have module or lib
to do this?

Best regards,
Thomas




From wlfraed at ix.netcom.com  Sun Sep 29 23:19:36 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Sun, 29 Sep 2002 20:19:36 -0700
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com> <3d976fbc$0$23666$91cee783@newsreader02.highway.telekom.at> <8428na.7vj.ln@ix.netcom.com> <3d979184$0$20332$79c14f64@nan-newsreader-03.noos.net>
Message-ID: <8sf8na.86m.ln@ix.netcom.com>

Guy Rabiller fed this fish to the penguins on Sunday 29 September 2002 
04:49 pm:

> 
> yes it works on MS-DOS too.
>
        Ah, good to know...

        Still need -vv to get the details from import though; -v just says 
"not found"

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From aleax at aleax.it  Thu Sep 26 09:32:45 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 26 Sep 2002 13:32:45 GMT
Subject: Getting started
References: <uojinmcfapjf45@corp.supernews.com> <mailman.1032979596.18965.python-list@python.org> <3D920950.323B25B3@alcyone.com> <mailman.1033012423.27861.python-list@python.org>
Message-ID: <10Ek9.170778$pX1.6080993@news2.tin.it>

James J. Besemer wrote:
        ...
>>(I favor the
>>latter definition, since it carries more information about the
>>language.)
> 
> It doesn't really tell you anything.
> 
> Else, in 25 words or less, why does Perl's "+" operator supposedly work
> the one way and Python's "*" operator work the other way?


Perl's + *creates a numeric context*: operands implicitly "become numbers".
Python's x*y calls x.__mul__(y) or else y.__rmul__(x) -- no "context",
nothing "becomes", _objects_ drive.


There -- 24 words by wc's count, so, little space to insert caveats
about implementation quirks &c, such as optimizations which e.g. *
could insert for types it knows about -- conceptually, everything
should work as if operators were implemented by dispatching special
methods, although optimization with invariant program-observable
results is clearly fine; and the like.

I have to agree with you that Perl's overall defined behavior CAN
be implemented by mechanisms akin to Python's (don't know if it
IS -- my knowledge of Perl's internals dates to Perl 4, and back
then, as I recall, it wasn't); but the mental model becomes too
cumbersome that way -- in practice, thinking in terms of various
established-contexts is just about inevitable in Perl (some of
them you can even TEST for, e.g. via wantarray).  Python's own
overall defined behavior OTOH is most simply conceived by seeing
operators as syntax sugar for templates (in the sense of the
"template method" DP -- they aren't methods, but they're templates
in just that way), involving one or more calls to the operands'
special methods.  Stringification does involve "implicitly become"
(via __str__ or __repr__, but that's really no more than "giving
context" as in so many cases in Perl) -- the print statement, and
the % operator with string LHS, being the two cases.  One could
quibble about "boolean contexts" (in statements if and while,
and operators and, or, not) -- arguing that there's a "become"
there, however, is denied by the fact that and and or return one
of their two operands, without having made it become anything,
so I don't think __nonzero__ is best considered a "become" special
method here (but I can see the arguments on the reverse side,
since it IS quite natural to speak of "truth context" and the
like to explain this little part of Python).


Alex



From tarnower at optonline.net  Tue Sep 24 22:10:53 2002
From: tarnower at optonline.net (DT)
Date: 24 Sep 2002 19:10:53 -0700
Subject: Zope Documentation and example sites
References: <503511c2.0209230758.74a64a3d@posting.google.com>
Message-ID: <9bf19f33.0209241810.1fbd28f5@posting.google.com>

Hi Sam,

I am in the same boat as you.  I've just become a Zope user, and I
hope to apply it to an Intranet project I plan to start soon.  I'm
particularly interested in the CMF product.

I also did a search for existing websites.  One I found that seems
pretty well written is: http://spring.parkinsons.org.uk/

I would like to find some working examples that use CMF.  Does anyone
know of any?


sam-nospam- at totallynerd.com (Sam Alexander) wrote in message news:<503511c2.0209230758.74a64a3d at posting.google.com>...
> Hi all,
> 
> I've just installed Zope, and now I'm trying to find reesources to
> learn it.  I have a few questions along these lines...
> 
> 1) What online resources are out there to help a newbie learn Zope? 
> I've downloaded the Zope Book, but it's a bit technical for my taste. 
> I'm looking for something with examples or tutorials showing how to do
> some basic stuff from scratch.  Any suggestions?  Either online or
> books ???
> 
> 2) What sites are using Zope?  I've noticed most if not all the Zope
> sites I've seen don't mention they're using Zope.  Can someone give me
> some examples of sites which use Zope so I can get a better idea of
> the type of layouts I can do with Zope?  I know it's very
> programmable, but other then the Plone site (which is great), I'm not
> sure what other sites use Zope.
> 
> 3) Finally, I've never used Python, but I'm seasoned in Linux and in
> web/app programming (Perl, PHP, C++, Cold Fusion, etc).  How long
> would it take to learn Zope/python well enough to create a production
> site?
> 
> Thanks for your help and suggestions.


From gua81 at XXXyahoo.com  Mon Sep 30 21:00:23 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Tue, 1 Oct 2002 11:00:23 +1000
Subject: PROBLEM with form in cgi
Message-ID: <anas2k$i2b$1@lust.ihug.co.nz>

HI,
I'm having problem trying to figure out what is wrong with my cgi script.

here's the code that's creating the problem in the apple.cgi

print """<form action="delete.html" name="myform" method="post">"""
                             ^
it keeps on saying invalid syntax in the action(as indicated)
and if I try to change it to this
print """<form action="delete.html" name="myform" method="post">"""
                             ^
it gives the error at action.

Please help.




From sabren at manifestation.com  Wed Sep 25 18:21:56 2002
From: sabren at manifestation.com (Michal Wallace)
Date: Wed, 25 Sep 2002 18:21:56 -0400 (EDT)
Subject: a python puzzle
In-Reply-To: <20020926000315.K30315@prim.han.de>
Message-ID: <Pine.LNX.4.44.0209251817040.17059-100000@hydrogen.sabren.com>

On Thu, 26 Sep 2002, holger krekel wrote:

> OK.  something like 
> 
> def decrypt(yourstring):
>     dedict = {'F': 'S', 'I': 'C', 'H': 'I', 'a': 't', 'c': 'f', 'b': 'v', 'e': 'n',
>     'd': 'o', 'f': 's', 'i': 'c', 'h': 'i', 'k': 'p', 'j': 'h', 'm': 'l',
>     'l': 'j', 'o': 'm', 'q': 'u', 'p': 'w', 's': 'b', 'u': 'z', 't': 'd',
>     'w': 'y', 'v': 'x', 'y': 'e', 'x': 'a', 'z': 'g', 'g': 'k'}
>     return "".join([dedict.get(i,i) for i in yourstring])
> returns
...
>     It's time to test your brainpower! I will give
>     six months of free web hosting to the first person
>     to solve this cryptogram. Can you figure it out?


Hi Holger,

Yep, that's it! I figured it would be easy to solve -
especially given the python code, but I didn't expect a
solution THAT quick. Next time I'm using RSA!

> but i don't know if i need free webhosting :-)

Hahaha. Well, if you change your mind, just let me know. :)

Cheers,

- Michal   http://www.sabren.net/   sabren at manifestation.com 
------------------------------------------------------------
Switch to Cornerhost!             http://www.cornerhost.com/
 Low Priced, Reliable Blog Hosting, With a Human Touch. :)
------------------------------------------------------------




From donnal at donnal.net  Mon Sep 23 06:08:25 2002
From: donnal at donnal.net (Donnal Walter)
Date: 23 Sep 2002 03:08:25 -0700
Subject: convention for Pickler file extension
Message-ID: <918bc22f.0209230208.4eaca702@posting.google.com>

When one uses Pickler.dump(object) to serialize an object to a file
associated with the Pickler, is there a Python convention for the file
extension name?

Thanks,
Donnal Walter
Arkansas Children's Hospital


From lbrannma at cablespeed.com  Thu Sep 19 09:03:25 2002
From: lbrannma at cablespeed.com (Lance)
Date: Thu, 19 Sep 2002 06:03:25 -0700
Subject: Getting started
Message-ID: <uojinmcfapjf45@corp.supernews.com>

Hi All,

Last night I attended a Weak typing, C++ Templates, and Python talk by Bruce
Eckel. It was great, my introduction to Python.

I'm sold on Python and weak typing. I want to write a graphics module that
will link to a C application, permitting scatter plots, bar charts, log
scales, etc. I suppose the Python Imaging Library will be used for this
purpose.

Would you be able to suggest an indispensable learning Python book?

Thanks,
Lance

PS: Bruce has Python material on his website, www.bruceeckel.com







From franke at ableton.com  Tue Sep  3 17:15:05 2002
From: franke at ableton.com (Stefan Franke)
Date: Tue, 03 Sep 2002 23:15:05 +0200
Subject: Better tracebacks for exec and execfile
References: <3i99nusjbgpiepva9pkuv3mtl7equ7h88h@4ax.com> <87elcbm1ze.fsf@andreasen.org>
Message-ID: <ed9anuo8i52h2fdq0fug0klabuselpjo90@4ax.com>

On 03 Sep 2002 19:12:21 +0200, "Erwin S. Andreasen" <erwin at andreasen.com> wrote:

>I don't know how to make SyntaxError appear correctly, 

Ah. I didn't understood that until I tried it out. In SyntaxError tracebacks
still "<string>" appears as the filename. This should be considered
a bug in compile(), shoudn't it?




From bogle at ihug.co.nz  Fri Sep  6 22:36:26 2002
From: bogle at ihug.co.nz (Gib Bogle)
Date: Sat, 07 Sep 2002 14:36:26 +1200
Subject: Python from C/SDL
References: <3D793B29.3198AB0E@ihug.co.nz> <slrnanifek.rh.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <3D79662A.402E47B6@ihug.co.nz>

Gerhard H?ring wrote:
> 
> Gib Bogle wrote in comp.lang.python:
> > Hi,
> >
> > I am a complete beginner with Python.  My interest in it is focussed
> > on a great piece of open-source software called pymol, a system for
> > visualization of molecules that uses Python and OpenGL.  I'd like to
> > get access to pymol's graphical functions from my own program, which
> > is built from C, SDL and OpenGL.
> >
> > I'm wondering if (with a bit of surgery) I'd be able to get access
> > to the lower-level functionality of pymol, by which I mean the file
> > I/O and rendering functions.
> 
> You'll have to embed Python
> (http://www.python.org/doc/current/ext/embedding.html), load the pymol
> modules and call functions or classes in it from C.
> 
> This means you'll have to get familiar with the Python/C API. The good
> thing is that this is relatively easy. You should be able to find
> documentation and examples on embedding Python. If you have any
> particular questions on how to use the Python/C API, just ask here.

Thanks Gerhard.  I'll have a look and see if I can grasp it.

Gib


From pinard at iro.umontreal.ca  Fri Sep  6 16:00:14 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Fri, 06 Sep 2002 16:00:14 -0400
Subject: Larry Wall's comment on python...
In-Reply-To: <Pine.LNX.4.44.0209061302050.1523-100000@ns0.tbye.com> (brueckd@tbye.com's
 message of "Fri, 6 Sep 2002 13:03:59 -0700 (PDT)")
References: <Pine.LNX.4.44.0209061302050.1523-100000@ns0.tbye.com>
Message-ID: <oqsn0mnb1t.fsf@carouge.sram.qc.ca>

[brueckd at tbye.com]

> The very idea of Larry Wall lecturing on code readability is hilarious.

While I understand the sentiment, history adds some perspective.

I've been following Perl since version 1, and in the beginnings, despite the
syntax was admittedly eclectic, Perl was much easier to read then, than now.
Perl grew many $X variables where X was a special character, but in the state
of the things at the time, it was still bearable.

The legibility problem developed progressively when adding things to Perl,
while trying to stay compatible to what it was.  Running out of special
characters, $X started to use control characters.  The very notion of
announcing the type of something to a prefix character ($%@&) did not mix well
when references and more complex structures were added.  Larry made an effort
to clean out the syntax a bit when going from Perl 4 to Perl 5, but the damage
was already too intense for the cleanup to be effective, also considered that
the cleanup just could not be too drastic, to protect the existing code base.

I take Larry Wall as a nice and clever guy.  Perl does not marry so well with
evolution, and this is probably the doom of all languages which do not enough
foresee evolution, and overly praise backward compatibility.  On the other
hand, too much rapid or drastic changes might erode the user base or make it
less solid.  Equilibrium is a very difficult art.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From gupsmail at ihatespam.shaw.ca  Wed Sep 25 01:23:21 2002
From: gupsmail at ihatespam.shaw.ca (Grant Hallam)
Date: Wed, 25 Sep 2002 05:23:21 GMT
Subject: problem with dictionaries within lists
Message-ID: <pan.2002.09.25.05.21.07.477231.167@ihatespam.shaw.ca>

the code attached opens a small wxPython frame with menu's to select parsing web survey entries or mailing list 
responses which are then outputted to Excel.  Problem occurs in the parsing part of the code.  After the last 
elif statement I have a list appended with a dictionary that is created my matching keys to the parsed text file. 
 However, the list seems to get populated with the same dictionary repeatedly until the entire file is parsed.  It
 is currently over 305 entries so the resulting list is 305 dictionaries holding the same value instead of unique
 entries.  Any help would be grateful,hopefully it is just a stupid error on my part :)

import string,re
from win32com.client import Dispatch
from wxPython.wx import *
#Global variables holding event ID's for menus
ID_Web = 101
ID_Mail = 102
ID_EXIT = 103
ID_ABOUT = 201
#Compiled regular expressions
#Used in web parser
rcN = re.compile('Name')
rcL = re.compile('Location')
rcT = re.compile('Transit')
rcTN = re.compile('Technical_Knowledge')
rcLS = re.compile('Like_Site')
rcRT = re.compile('Resolve_Tech_Issues')
rcEN = re.compile('Easy_Navigation')
rcPC = re.compile('Publish_Comments')
rcAN = re.compile('Answered_No')
rcFS = re.compile('Find_Support')
rcFSO = re.compile('Find_Support_Other')
rcFI = re.compile('Find_Information')
rcWLS = re.compile('Which_LEOnet_Site')
rcFIO = re.compile('Find_Information_Other')
rcVA = re.compile('Visit_Again')
rcVOS = re.compile('Visited_Old_Site')
rcTS = re.compile('Time_Spent')
rcRTI = re.compile('resolve_tech_issues')
#Used in both web and mail parser
rcP = re.compile('Position')
rcC = re.compile('Comments')
rcB = re.compile('B1')
#Used in mail parser
rcM = re.compile('Membership')
rcEM = re.compile('Email_address')

class MainFrame(wxFrame):
	'''Main window for program, uses menus to call parsing engines that output results to excel'''
	def __init__(self, parent, ID, title):
		wxFrame.__init__(self, parent, ID, title, wxDefaultPosition, wxSize(200, 150))
		#sets up status bar
		self.CreateStatusBar()
		self.SetStatusText("Version 3.1, built by Grant Hallam")
		#sets up file menu
		filemenu = wxMenu()
		filemenu.Append(ID_Web, "&Web Survey", "Go find your file")
		filemenu.Append(ID_Mail, "&Mail Survey", "Go find your file")
		filemenu.AppendSeparator()
		filemenu.Append(ID_EXIT, "&Exit", "Terminates the program")
		#sets up help menu
		helpmenu = wxMenu()
		helpmenu.Append(ID_ABOUT, "&About", "Info about program")
		#adds the menubar
		menuBar = wxMenuBar()
		menuBar.Append(filemenu, "&File")
		menuBar.Append(helpmenu, "Help")
		self.SetMenuBar(menuBar)
		#links events from menus to id and callable def's
		EVT_MENU(self, ID_Web, self.Web)
		EVT_MENU(self, ID_Mail, self.Mail)
		EVT_MENU(self, ID_EXIT, self.Exit)
		EVT_MENU(self, ID_ABOUT, self.About)
	def Web(self, event):
		'''Opens file dialog box to select web data file to parse'''
		dlg = wxFileDialog(self, "Choose a file", ".", "", "*.*", wxOPEN|wxMULTIPLE)
		if dlg.ShowModal() == wxID_OK:
			for path in dlg.GetPaths():
				filename = path
			parsefile = WebParser(filename)
		dlg.Destroy()
	def Mail(self, event):
		'''Opens file dialog box to select mail data file to parse'''
		dlg = wxFileDialog(self, "Choose a file", ".", "", "*.*", wxOPEN|wxMULTIPLE)
		if dlg.ShowModal() == wxID_OK:
			for path in dlg.GetPaths():
				filename = path
			parsefile = MailParser(filename)
		dlg.Destroy()		
	def Exit(self, event):
		'''Closes the program'''
		self.Close(true)
	def About(self, event):
		'''Opens the about dialog box'''
		dlg = wxMessageDialog(self, 'This program parses text files currently coded to parse\n survey & mail responses but future parsing engines\n may be included.', 'About Parsing', wxOK | wxICON_INFORMATION)   
		dlg.ShowModal()
		dlg.Destroy()
class MailParser:
	'''This class parses a text file containing mailing list entries, and outputs the results to excel'''
	def __init__(self, filename):
		'''Gets filename and calls the FileRead method'''
		self.filename = filename
		self.FileRead()
	def FileRead(self):
		'''Opens the text file, parses it, stores the various sections into a dictionary, and then stores that into a list, which is then iterated over to output result in to excel'''
		file = open(self.filename, 'r')
		input = file.read()
		file.close()
		#sets the column title dictionary, the field names dictionary, the pos list for the variable field values of Position, and the maildatabaselist that stores all the field dictionaries
		columntitle = ['Membership','Email_address','Position','Comments']
		dict = {'Membership':'','Email_address':'','Position':'','Comments':'','B1':''}
		pos = []
		maildatabaselist = []
		#This is the guts of the parser, splits the file into blocks,then splits the blocks in to lists of two fields, compares the 1st field to a RE, and adds the 2nd field to the field name dict	
		rawline = string.split(input, '\n')
		for x in rawline:	
			list = string.split(x, '=')
			if rcM.match(list[0]): 
				case = string.capwords(list[1])
				dict['Membership'] = case
			elif rcEM.match(list[0]): 
				dict['Email_address'] = list[1]
			elif rcP.match(list[0]): 
				pos.append(list[1])
			elif rcC.match(list[0]): 
				dict['Comments'] = list[1]
			elif rcB.match(list[0]): 
				dict['B1'] = list[1]
				#joins the various items in pos list into one string that is inserted in the Position field, 
				POS = string.join(pos,',')
				dict['Position'] = POS
				#adds field name dict to maildatabaselist
				maildatabaselist.append(dict)
				fs = []
				fi = []
		#opens Excel, sets to visible, and adds a worksheet
		xlApp = Dispatch('Excel.Application')
		xlApp.Visible = 1
		workbook = xlApp.Workbooks.Add()
		worksheet = workbook.Worksheets.Add()
		#sets the titles of the columns
		count = 0
		for x in columntitle:
			cell = columntitle[count]
			count = count+1
			worksheet.Cells(1,count).Font.Bold=1
			worksheet.Cells(1,count).Value = cell
		#sets up excel cells and inserts maildatabaselist fields into cells
		row = 2
		for x in maildatabaselist:
			row = row +1
			worksheet.Cells(row,1).Value = x['Membership']
			worksheet.Cells(row,2).Value = x['Email_address']
			worksheet.Cells(row,3).Value = x['Position']
			worksheet.Cells(row,4).Value = x['Comments']
		#adds bottom column headers to excel
		count = 0
		titlerow = row +2
		for x in columntitle:
			cell = columntitle[count]
			count = count+1
			worksheet.Cells(titlerow,count).Font.Bold=1
			worksheet.Cells(titlerow,count).Value = cell
class WebParser:
	'''This class parses a text file containing web survey list entries, and outputs the results to excel'''
	def __init__(self, filename):
		'''Gets filename and calls the FileRead method'''
		self.filename = filename
		self.FileRead()
	def FileRead(self):
		'''Opens the text file, parses it, stores the various sections into a dictionary, and then stores that into a list, which is then iterated over to output result in to excel'''
		file = open(self.filename, 'r')
		input = file.read()
		file.close()
		#sets the column title dictionary, the field names dictionary, the fs and fi list for the variable field values of Find support and find information, and the webdatabaselist that stores all the field dictionaries
		columntitle = ['Name','Transit','Location','Position','Knowledge','Like site','Resolve issues','Navigate','Publish','Answer no','Find support','Find support other','Find information','Find information other','Which leonet other','Visit','Old visit','Spent','Comments']
		dict = {'Name': '','Location':'','Transit':'','Position':'','Technical_Knowledge':'','Like_Site':'','Resolve_Tech_Issues':'','Easy_Navigation':'','Publish_Comments':'','Answered_No':'','Find_Support':'','Find_Support_Other':'','Find_Information':'','Which_LEOnet_Site':'','Find_Information_Other':'','Visit_Again':'','Visited_Old_Site':'','Time_Spent':'','Comments':'','B1':''}
		fs = []
		fi = []
		webdatabaselist = []
		#This is the guts of the parser, splits the file into blocks,then splits the blocks in to lists of two fields, compares the 1st field to a RE, and adds the 2nd field to the field name dict	
		rawline = string.split(input, '\n')
		for x in rawline:	
			list = string.split(x, '=')
			if rcN.match(list[0]): 
				case = string.capwords(list[1])
				dict['Name'] = case
			elif rcL.match(list[0]): 
				dict['Location'] = list[1]
			elif rcT.match(list[0]): 
				dict['Transit'] = list[1]
			elif rcP.match(list[0]): 
				dict['Position'] = list[1]
			elif rcTN.match(list[0]): 
				dict['Technical_Knowledge'] = list[1]
			elif rcLS.match(list[0]): 
				dict['Like_Site'] = list[1]
			elif rcRT.match(list[0]): 
				dict['Resolve_Tech_Issues'] = list[1]
			elif rcEN.match(list[0]): 
				dict['Easy_Navigation'] = list[1]	
			elif rcPC.match(list[0]): 
				dict['Publish_Comments'] = list[1]
			elif rcAN.match(list[0]): 
				dict['Answered_No'] = list[1]
			elif rcFIO.match(list[0]): 
				dict['Find_Information_Other'] = list[1]
			elif rcWLS.match(list[0]): 
				dict['Which_LEOnet_Site'] = list[1]
			elif rcFSO.match(list[0]):
				dict['Find_Support_Other'] = list[1]
			elif rcFS.match(list[0]): 
				fs.append(list[1])
			elif rcFI.match(list[0]): 
				fi.append(list[1])
			elif rcVA.match(list[0]): 
				dict['Visit_Again'] = list[1]
			elif rcVOS.match(list[0]): 
				dict['Visited_Old_Site'] = list[1]
			elif rcTS.match(list[0]): 
				dict['Time_Spent'] = list[1]
			elif rcC.match(list[0]): 
				dict['Comments'] = list[1]
			elif rcB.match(list[0]): 
				dict['B1'] = list[1]
				#joins the various items in fs and fi list into one string that is inserted in the Find Support and Find Information fields,
				findS = string.join(fs,',')
				dict['Find_Support'] = findS
				findI = string.join(fi,',')
				dict['Find_Information'] = findI
				#print dict
				#adds field name dict to maildatabaselist
				webdatabaselist.append(dict)
				fs = []
				fi = []
	
		totalnumberrecords = len(webdatabaselist)
		#opens Excel, sets to visible, and adds a worksheet		
		xlApp = Dispatch('Excel.Application')
		xlApp.Visible = 1
		workbook = xlApp.Workbooks.Add()
		worksheet = workbook.Worksheets.Add()
		#sets the titles of the columns
		count = 0
		for x in columntitle:
			cell = columntitle[count]
			count = count+1
			worksheet.Cells(1,count).Font.Bold=1
			worksheet.Cells(1,count).Value = cell
		#sets up excel cells and inserts webdatabaselist fields into cells
		row = 2
		for x in webdatabaselist:
			row += 1
			worksheet.Cells(row,1).Value = x['Name']
			worksheet.Cells(row,2).Value =x['Transit']
			worksheet.Cells(row,3).Value =x['Location']
			worksheet.Cells(row,4).Value =x['Position']
			worksheet.Cells(row,5).Value =x['Technical_Knowledge']
			worksheet.Cells(row,6).Value =x['Like_Site']
			worksheet.Cells(row,7).Value =x['Resolve_Tech_Issues']
			worksheet.Cells(row,8).Value =x['Easy_Navigation']
			worksheet.Cells(row,9).Value =x['Publish_Comments']
			worksheet.Cells(row,10).Value =x['Answered_No']
			worksheet.Cells(row,11).Value =x['Find_Support']
			worksheet.Cells(row,12).Value =x['Find_Support_Other']
			worksheet.Cells(row,13).Value =x['Find_Information']		
			worksheet.Cells(row,14).Value =x['Find_Information_Other']
			worksheet.Cells(row,15).Value =x['Which_LEOnet_Site']
			worksheet.Cells(row,16).Value =x['Visit_Again']
			worksheet.Cells(row,17).Value =x['Visited_Old_Site']
			worksheet.Cells(row,18).Value =x['Time_Spent']
			worksheet.Cells(row,19).Value =x['Comments']

		#adds bottom column headers to excel
		count = 0
		titlerow = row +2
		for x in columntitle:
			cell = columntitle[count]
			count = count+1
			worksheet.Cells(titlerow,count).Font.Bold=1
			worksheet.Cells(titlerow,count).Value = cell
class SurveyParser(wxApp):
	def OnInit(self):
		frame = MainFrame(NULL, -1, "SurveyParser")
		frame.Show(true)
		self.SetTopWindow(frame)
		return true
app = SurveyParser(0)
app.MainLoop()


Sample text file for web parser
Name = Mourad Salji
Location = Western Canada
Transit = 05110
Position = Other
Technical_Knowledge = Expert
Like_Site = Yes
Resolve_Tech_Issues = Yes
Easy_Navigation = Yes
Found_Everything = Yes
Publish_Comments = Yes
Answered_No = 
Find_Support = TSC_Telephone
Find_Support = TSC_Web
Find_Support_Other = (Other - describe here)
Find_Information = TSC_Telephone
Find_Information = Word_Of_Mouth
Which_LEOnet_Site = (Which LEOnet site?)
Find_Information_Other = (Other - describe here)
Visit_Again = Daily
Visited_Old_Site = When_A_Problem_Occurs
Time_Spent = 5_Plus_Minutes
Comments = 
B1 = Submit

Name = Nelson
Location = Eastern Canada
Transit = 08271
Position = Other
Technical_Knowledge = Expert
Like_Site = Yes
Resolve_Tech_Issues = Yes
Easy_Navigation = Yes
Found_Everything = Yes
Publish_Comments = Yes
Answered_No = 
Find_Support = TSC_Telephone
Find_Support_Other = (Other - describe here)
Find_Information = TSC_Telephone
Which_LEOnet_Site = (Which LEOnet site?)
Find_Information_Other = (Other - describe here)
Visit_Again = When_A_Problem_Occurs
Visited_Old_Site = Never
Time_Spent = 5_Plus_Minutes
Comments = great JOB
B1 = Submit

Name = Dina Della Porta
Location = Eastern Canada
Transit = 03511
Position = Other
Technical_Knowledge = Expert
Like_Site = Yes
Resolve_Tech_Issues = Yes
Easy_Navigation = Yes
Found_Everything = Yes
Publish_Comments = Yes
Answered_No = 
Find_Support = Other
Find_Support_Other = My own department
Find_Information = Other
Which_LEOnet_Site = (Which LEOnet site?)
Find_Information_Other = surfing
Visit_Again = When_A_Problem_Occurs
Visited_Old_Site = Never
Time_Spent = Less_Than_2_Minutes
Comments = The site came in handy for a SECAF problem I was working on.  
B1 = Submit

Name = MARYKE GAUVREAU
Location = Western Canada
Transit = 00620
Position = Branch - Other
Technical_Knowledge = Average
Like_Site = Yes
Resolve_Tech_Issues = Yes
Easy_Navigation = Yes
Found_Everything = Yes
Publish_Comments = Yes
Answered_No = 
Find_Support = TSC_Telephone
Find_Support = Word_Of_Mouth
Find_Support = TSC_Web
Find_Support = LEOnet
Find_Support = Folio_HOCircular
Find_Support_Other = (Other - describe here)
Find_Information = TSC_Newsletter
Which_LEOnet_Site = (Which LEOnet site?)
Find_Information_Other = (Other - describe here)
Visit_Again = When_A_Problem_Occurs
Visited_Old_Site = When_A_Problem_Occurs
Time_Spent = 5_Plus_Minutes
Comments = 
B1 = Submit


From h_schneider at marketmix.com  Fri Sep 27 08:58:52 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Fri, 27 Sep 2002 14:58:52 +0200
Subject: SOLVED:  py2exe Error creating NT Service
References: <an1epm$rc5$03$1@news.t-online.com> <j4ptuz4pvh.fsf@informatik.hu-berlin.de>
Message-ID: <an1khb$eoe$06$1@news.t-online.com>

Thanks for the hint!
I solved the prob myself: must have been a tab or whitespace beyond the last
line of the script. That irritated py2exe for some reason. After deleting
everything after the last line it compiled well.

Bye,
Harald




From bokr at oz.net  Wed Sep  4 17:17:24 2002
From: bokr at oz.net (Bengt Richter)
Date: 4 Sep 2002 21:17:24 GMT
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com> <yu99u1l5bkra.fsf@europa.research.att.com>
Message-ID: <al5t94$vse$0@216.39.172.122>

On Wed, 4 Sep 2002 19:46:49 GMT, Andrew Koenig <ark at research.att.com> wrote:

>Gon?alo> Anyway, I've been directed to PEP 227, cheerfully obeyed, and
>Gon?alo> from what I have understood having read-only nested scopes
>Gon?alo> was more of a design decision than something else. And while
>Gon?alo> I can live happily with the decision (some would even say
>Gon?alo> happier) there is something about the "You can look at it but
>Gon?alo> not touch it" attitude that I dislike ;-)
>
>You can do anything to the object that the object support.
>What you can't do is change the binding of the name.
>
OTOH, if "everything is an object," then the namespace of
an enclosing scope could also be an object, and as such
could support rebinding of names therein ;-)

The question would be how to spell such access.

E.g., "global x" is a way to say that x=123 will be short for
the effect globals()['x']=123 within a certain scope.

If you visualized nested scopes as a list of scopes, starting with
locals and ending with globals, you could refer to them by index, e.g.,
scopes[0] for local and scopes[-1] for global, and others in between, if any.

If you treated those spaces like attribute name spaces, you could write
a silly example like

    def foo():
        x=123
        def bar():
           scopes[1].x=456
        bar()
        return x

and expect foo() => 456

Regards,
Bengt Richter


From missive at frontiernet.net  Fri Sep 27 16:41:44 2002
From: missive at frontiernet.net (Lee Harr)
Date: Fri, 27 Sep 2002 20:41:44 -0000
Subject: pygame.event.Event?
References: <r_Lk9.174774$pX1.6183293@news2.tin.it>
Message-ID: <up9gk8asa6g37c@corp.supernews.com>

In article <r_Lk9.174774$pX1.6183293 at news2.tin.it>, Max wrote:
> Hi!
> I can't generate a QUIT event in pygame:
> pygame.event.Event(type, dict, [keyword_args]) -> Event
> 
> What do I have to put in "dict"?
> 


How about an empty dict?

pygame.event.Event(QUIT, {})




From rs at onsitetech.com  Tue Sep  3 11:25:02 2002
From: rs at onsitetech.com (Robb Shecter)
Date: 03 Sep 2002 15:25:02 GMT
Subject: Automatically find the source of a method invocation?
References: <3D6FC3AF.30006@onsitetech.com> <mailman.1030736867.5976.python-list@python.org>
Message-ID: <3D74D43B.1090509@onsitetech.com>

holger krekel wrote:

> Why is this a problem? 

Well, you know; it's just clutter with zero information content since it 
must always be there.

[...]

> 
> It depends on what you really want to do. I wouldn't bother too much ...
> 

Good ideas, thanks.  I can see it'd take a bit of work.  I pretty much 
just want an object receiving a PS message to be able to ignore it if it 
was the sender (for example).



From michael at stroeder.com  Thu Sep 26 11:01:24 2002
From: michael at stroeder.com (=?ISO-8859-1?Q?Michael_Str=F6der?=)
Date: Thu, 26 Sep 2002 17:01:24 +0200
Subject: Page describing http compression with gzip.
References: <f72dac0d.0209260606.4dab583e@posting.google.com>
Message-ID: <53kb6-iu7.ln1@junker.stroeder.com>

Alan Kennedy wrote:
> 
> A few months ago, after spending a couple of days troubleshooting gzip
> compression over HTTP, I put together a page describing how to do HTTP
> compression in python,
> [..]
> http://xhaus.com/alan/python/httpcomp.html

Thanks for writing it down. Things I'd like to note:

Netscape Navigator 4.x (tested with 4.8) does not handle Transfer-Encoding: 
gzip. It displays garbage. Mozilla (tested with 1.1) shows a download select 
box.

Netscape Navigator 4.x (tested with 4.8) does not display images which were 
gzip'ed.

It is possible to deliver dynamic content of unknown length. Personally I've 
implemented my own wrapper class around gzip.GzipFile.

Some people mentioned that it's necessary to add a header line
Vary: Accept-Encoding
Do you have more detailed information about that?

Ciao, Michael.



From vlcus at vlcank.co.jp  Wed Sep 18 21:06:10 2002
From: vlcus at vlcank.co.jp (vlcus at vlcank.co.jp)
Date: 19 Sep 2002 01:06:10 -0000
Subject: Win $ 4,000.00 in cash !
Message-ID: <20020919010610.13444.qmail@www.canausr.com>

Dear Sir/Madam,


This mail has been sent to persons who answered
to the past questionnaire of our company.

Now, we offer you the chance?to take part in our 
'Questionnaire About Digital Camera and Color Printer' 
posted on our web site and to win $ 4,000.00in cash.

This survey is intended for female only.

Please visit now HERE
   http://www.vlcank.com/usa/index.html

We think you will find this web survey very interesting
and your input will help guide manufacturers to improve
their products !

Once you qualify and complete our web survey, you will be
automatically get a chance to win $ 4,000.00 in cash
prizes !

Fill the questionnaire now and become the winner who will
be randomly selected and get $ 4,000.00 in cash !

Sincerely,

Sumio Muramatsu
VLC Inc.

* Please discard this message if you do not have 
  any interest.

* If you'd like to unsubscribe, Send mail to:
   vlcus at vlcank.co.jp



From loewis at informatik.hu-berlin.de  Fri Sep 27 06:37:21 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 27 Sep 2002 12:37:21 +0200
Subject: SET_LINENO
References: <3D92A46E.2030706@cygnus-software.com> <lk8z1nwxjp.fsf@pc150.maths.bris.ac.uk>
Message-ID: <j4u1kb4t26.fsf@informatik.hu-berlin.de>

Michael Hudson <mwh at python.net> writes:

> > So, wouldn't a break-point opcode be a more substantial improvement
> > than removing SET_LINENO? Just bringing up that old topic again to see
> > if it gets any more traction :-)
> 
> How would a breakpoint opcode differ from what SET_LINENO was?

This is easy to answer: a breakpoint instruction could be inserted in
the code at the points where the user wishes to have a
breakpoint. Thus, with a breakpoint instruction, you could disable all
tracing, and wait for the breakpoint callback.

With such a scheme, you can have zero-overhead debugging, for
breakpoints. If your debugger supports watchpoints, you get overhead,
again.

Regards,
Martin



From deltapigz at telocity.com  Tue Sep 10 20:57:57 2002
From: deltapigz at telocity.com (Adonis)
Date: Tue, 10 Sep 2002 20:57:57 -0400
Subject: test (ignore)
References: <3d7e2957$1_2@nopics.sjc> <slrnansspr.6im.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <3d7e950e_2@nopics.sjc>

oh im sorry i was not aware of that, just on another news thread i placed a
post and did not show up for about 20mins later, thought it was a general
issue with my server and i happened to be here and just ran a test. sorry if
it incovenienced you in any way...

Adonis

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote in message
news:slrnansspr.6im.gerhard.haering at lilith.my-fqdn.de...
> Adonis wrote in comp.lang.python:
> > just a test, please ignore.
>
> What exactly is wrong with alt.test or one of the other .test groups?
>
> -- Gerhard




From bokr at oz.net  Tue Sep 10 15:49:42 2002
From: bokr at oz.net (Bengt Richter)
Date: 10 Sep 2002 19:49:42 GMT
Subject: Graph Layout Algorithms
References: <alksio$d2l$1@newsg3.svr.pol.co.uk>
Message-ID: <allicm$cqd$0@216.39.172.122>

On Tue, 10 Sep 2002 14:37:00 +0100, "Duncan Smith" <buzzard at urubu.freeserve.co.uk> wrote:

>Hello,
>         Just wondering if anyone has implemented anything in Python.  I
>need to lay out my graphs (directed acyclic / undirected / trees) in a
>reasonably clear way.  So if anyone has already implemented anything, or if
>anyone has advice on appropriate algorithms, I'd be glad to hear about it.
>Cheers.
>
>Duncan
>
>
IIRC Ka-Ping Yee made some beautiful pdf(?) plots of python data structures, but
I can't find anything right now. If the plotting software is open source and
Python, that would sure be a treat. I think Guido made reference to some of
the plots in some post.

Anyone know where that stuff is?

Regards,
Bengt Richter


From oren-py-l at hishome.net  Tue Sep 10 10:30:13 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Tue, 10 Sep 2002 10:30:13 -0400
Subject: UNIX processes
In-Reply-To: <200209101233.05796.b.maryniuk@forbis.lt>
References: <200209101233.05796.b.maryniuk@forbis.lt>
Message-ID: <20020910143013.GA71272@hishome.net>

On Tue, Sep 10, 2002 at 12:33:05PM +0200, Bo M. Maryniuck wrote:
> Hi, all.
> Somebody know how to list all the processes in RIGHT way on Linux/UNIX?
> I need to find the zombies, stopped processes etc.

Ah, it's tricky. The output of the ps command is not consistent and the 
native API can be anything from variations on a /proc filesystems to 
actually looking at the kernel data structures. It's a lot of work to
support all these different interfaces.

Fortunately, someone already did the hard work - the lsof utility:

http://ftp.cerias.purdue.edu/pub/tools/unix/sysutils/lsof/  (Thanks!)

It lists all open files on the system and as a side-effect also reports
a great deal of information about processes. It also has an output mode 
designed for parsing by other programs. It works for a large number of 
different platforms and is installed by default on many systems.

Opening lsof with popen and decoding the output should be trivial.

	Oren



From phncontact.libero.it at   Wed Sep  4 04:14:02 2002
From: phncontact.libero.it at  (Benjamin)
Date: Wed, 04 Sep 2002 08:14:02 GMT
Subject: Why Python?
Message-ID: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com>

Hi guys (and girls)
I just begun to learn how to program. And I choosed Python for
doing that. surfing around the web i found an article about this
language, they said it's the ideal language for beginners to the
world of programming. and i admit it, the syntax is very clear making
the code very readable, i learn easily the concepts of programming
and i soon wrote some nice programs. but since i finished my
newbie-tutorial, i am wondering about one question: is it worth to go
deeper into python? i talked to a friend of mine, he told me that i
should try php..
what should i do?? ^_^ in other words: what is python good for?
why should i continue with python?
thank you!

_____________________________________________________
I'm going to prove the impossible really exists


From james.kew at btinternet.com  Sun Sep  8 16:18:04 2002
From: james.kew at btinternet.com (James Kew)
Date: Sun, 8 Sep 2002 21:18:04 +0100
Subject: transform a list of lists in a lit of strtings???
References: <mailman.1031507832.24527.python-list@python.org> <algah9$1p0kp3$1@ID-71831.news.dfncis.de>
Message-ID: <algb9t$1q0jg0$1@ID-71831.news.dfncis.de>

"James Kew" <james.kew at btinternet.com> wrote in message
news:algah9$1p0kp3$1 at ID-71831.news.dfncis.de...
> labels1 = [["D1"], ["D2"], ["D3", "D4"]]
> --> labels2 == ["D1", "D2", "D3", "D4"]

> labels2 = reduce(lambda x, y: x+y, labels1)
>
> Can this be written in list comprehension form?

Emile's google link suggests (a) flattening is a hoary old chestnut, and (b)
yes it can:

labels2 = [i for item in labels1 for i in item]

(I find this oddly hard to read.)

James





From merman at snafu.de  Wed Sep  4 10:33:18 2002
From: merman at snafu.de (T. Kaufmann)
Date: Wed, 04 Sep 2002 16:33:18 +0200
Subject: newbie-question: difference between classmethod and staticmethod in Python2.2
Message-ID: <3D7619AE.3050605@snafu.de>

Hi there,

can anybody explain the difference between 'staticmethod' and 'classmethod' in a simple way in 
Python2.2?

When should I use staticmethod or classmethod in classes?

I thought a staticmethod is a classmethod like in C++/ Java?

Thanks for help.

o-o

Thomas



From aleax at aleax.it  Fri Sep 27 02:51:27 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 27 Sep 2002 06:51:27 GMT
Subject: POP3 and Multipart Mime
References: <mailman.1033084656.24584.python-list@python.org>
Message-ID: <PdTk9.148987$ub2.3287189@news1.tin.it>

David McInnis wrote:

> Does anyone know where I can find a tutorial on how to accomplish the
> following?
>  
> I need to log into a POP3 account (which I have been able to do using
> poplib).
>  
> I need to be able to read and decode the message.  The problem is that
> most of our users send word docs as attachments (1 or more).  I need to
> take these files and save them to disk.

The email package, bundled since Python 2.2, makes it easy -- e.g.:

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/86676


Alex



From glingl at aon.at  Thu Sep 26 17:57:03 2002
From: glingl at aon.at (Gregor Lingl)
Date: Thu, 26 Sep 2002 23:57:03 +0200
Subject: Who knows somefunction?
References: <664e469667.69667664e4@icomcast.net>
Message-ID: <3D9382AF.5070800@aon.at>

Tim Peters schrieb:

>[Gregor Lingl]
>  
>
>>Is there a function <somefunction> in Python for which
>>
>>something is <somefunction>(id(something))
>>
>>returns invariably true?
>>    
>>
>
>I'm not sure you're asking the question you intended to ask, but str() 
>is one answer to the question you actually asked.
>  
>
I don't think so:
 >>> something = [1,2,3]
 >>> something is str(id(something))
0

How did you understand my question?
(Errors should never pass silently)
Gregor

>uselessly y'rs  - tim
>
>
>
>  
>




From bbeck at REMOVE-THIS-TO-EMAIL-ME.austin.rr.com  Wed Sep  4 11:59:03 2002
From: bbeck at REMOVE-THIS-TO-EMAIL-ME.austin.rr.com (Brandon Beck)
Date: Wed, 04 Sep 2002 15:59:03 GMT
Subject: Regexp Over Multiple Lines
References: <565e316f.0209040740.36573b56@posting.google.com>
Message-ID: <b5qd9.200380$Yd.8603161@twister.austin.rr.com>

Yes, please see the python online documentation for the re module 
(http://www.python.org/doc/current/lib/module-re.html) the section you 
specifically want is at 
(http://www.python.org/doc/current/lib/node99.html).  You'll want to 
pass the VERBOSE flag to your compilation statement.

pattern = re.compile('''
            <first line>              # comment
            ...
            <last line>               # comment
''', re.VERBOSE);

Additionally you may want to check out the regular expression howto at 
sourceforge (http://py-howto.sourceforge.net/regex/regex.html).




Keith wrote:
> In Python can I have a regular expression that spans multiple lines...
> e.g. In Perl I may define a regular expression like:
> 
> $depends_def =   qr/
>          [\w_\.]+\s*:                     # simobject:
>          \s*
>          [\w_\.]+                         # job name
>          \(\s*(?:[\d]*)?\s*\)             # arg list for job (optional
> num)
>          \s+
>          depends\s+on                     # keywords "depends on"
>          \s+
>          (?:
>           [\w_\.]+\s*:                    # simobject:
>           \s*
>             (?:
>               [\w_\.]+                    # job name
>               \(\s*(?:[\d]*)?\s*\)        # arg list for job (optional
> num)
>               (?:
>                 (?:
>                  \s+
>                  and\s*[\w_\.]*\s*:       # "and simobject:"
>                  \s*
>                  [\w_]*                   # job name
>                  \(\s*(?:[\d]*)?\s*\)     # arg list for job (optional
> num)
>                 )+
>               )?                          # match optional "and" 1 or
> more times
>             )
>          )
>         /sx;
> 
> then later I can use that regular expression in another regular
> expression.
> 
> It helps to span multiple lines for commenting what the heck the ugly
> regular expression is trying to do...
> 
> BTW: I am investigating Python and have no knowledge of it.  Our Perl
> code has just gotten out-of-hand.
> 
> Thanks,
> Keith



From oren-py-l at hishome.net  Sat Sep  7 12:10:37 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Sat, 7 Sep 2002 12:10:37 -0400
Subject: Daemonize?
In-Reply-To: <unjt47pc5r8671@corp.supernews.com>
References: <F7DB8D13DB61D511B6FF00B0D0F06233022B66EC@mail.hebmex.com> <mailman.1031184848.32071.python-list@python.org> <unjt47pc5r8671@corp.supernews.com>
Message-ID: <20020907161037.GA70503@hishome.net>

On Sat, Sep 07, 2002 at 12:43:51PM -0000, A.M. Kuchling wrote:
> Here's the daemonize function I use; I think Greg Ward originally
> gave it to me.
> 
>     def daemonize (self):
>         # Fork once
>         if os.fork() != 0:
>             os._exit(0)
>         os.setsid()                     # Create new session
>         if os.fork() != 0:      
>             os._exit(0)         
>         os.chdir("/")         
>         os.umask(0)
> 
>         os.close(sys.__stdin__.fileno())
>         os.close(sys.__stdout__.fileno())
>         os.close(sys.__stderr__.fileno())

    devnull = os.open('/dev/null', 0)
    os.dup2(devnull, 0)
    os.dup2(devnull, 1)
    os.dup2(devnull, 2)

This is safer because it prevents these file descriptors from being reused.

	Oren



From stencelp at web.de  Thu Sep 12 07:45:10 2002
From: stencelp at web.de (Paulus)
Date: 12 Sep 2002 04:45:10 -0700
Subject: Starting external functions in Python applications
Message-ID: <1d3ffc6b.0209120345.64cf64a1@posting.google.com>

Hi.
How can I link an external dll in a Python programs
and (when possible) start any function from this DLL ? 
Thanks in advance
paulus


From Oschler at earthlink.net  Wed Sep 18 20:07:00 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Thu, 19 Sep 2002 00:07:00 GMT
Subject: PyLog, not require SWI-Prolog right?
Message-ID: <Ey8i9.9524$Lg2.1789572@news2.news.adelphia.net>

PyLog is inpendent of any Prolog interpreter right?  There's PyLog and
PyProlog, the latter I believe is a wrapper and therefore requires
SWI-Prolog (just like bevedere requires GNU-Prolog), but PyLog doesn't, is
that correct?  Just looking for confirmation.

Also, PyLog supports backracking right?

thx





From see_reply_address at something.invalid  Tue Sep 24 22:18:34 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Wed, 25 Sep 2002 14:18:34 +1200
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <mailman.1031377930.4225.python-list@python.org> <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com> <3D8E6353.7030303@something.invalid> <96zj9.150553$pX1.5399868@news2.tin.it>
Message-ID: <3D911CFA.4060405@something.invalid>

Alex Martelli wrote:
 >

> Greg Ewing wrote

>
>>Using *tabs* only is the only way to get truly
>>preference-independent indentation.
>
> ...unless other people use tools that do NOT let them change the
> ways tabs are displayed, and that way is intolerably incompatible
> with your use of tabs.


I agree there are good reasons to prefer spaces-only
for code that will be transmitted or displayed by
dumb software.

I was only pointing out that, whatever arguments there
may be in favour of spaces-only, "it doesn't force the
reader to view it the way you do" is not one of them!

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From tdelaney at avaya.com  Wed Sep  4 18:56:13 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 5 Sep 2002 08:56:13 +1000 
Subject: Privacy and Inheritance
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A572@natasha.auslabs.avaya.com>

> From: Eric Brunel [mailto:eric.brunel at pragmadev.com]
> 
> In languages that have actual "private", "protected" and "public" 
> attributes, private attributes are not inherited. They are 
> visible only in 
> the class that define them, not outside it and not in its 
> sub-classes. What 
> you want is a protected attribute, that is visible in the 
> defining class 
> and in its sub-classes.
> 
> Unfortunately, Python doesn't know anything about protected 
> attributes. 
> AFAIK, there's no simple way to do what you want: in Python, 
> attributes are 
> either private or public; there's nothing in between...

Actually, there is.

Anything beginning with a leading underscore is not imported via a 'from
module import *' (not that anyone should in general use this ;) Thus
anything with a leading underscore is something other than completely
"public".

There is a (strong) convention that anything beginning with a *single*
leading underscore is "protected". No name mangling is performed, but the
name is flagged as "special".

class StandardNormal:

    def letEventZScore(self, eventZScore):
        self._eventZScore = float(eventZScore)

class DerivativeNormal(StandardNormal):

    def getCentral(self):
        if self._eventZScore < 0:

Tim Delaney



From see_reply_address at something.invalid  Tue Sep 24 23:30:11 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Wed, 25 Sep 2002 15:30:11 +1200
Subject: help with print function in extension
References: <tA8k9.270440$AR1.11448440@bin2.nnrp.aus1.giganews.com>
Message-ID: <3D912DC3.7010302@something.invalid>

Eric Hagemann wrote:

> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> SystemError: error return without exception set


Check what your C function implementing str() is
returning. My guess is that you've forgotten a
"return 0" statement, and it's returning some junk
value that in some cases the Python interpreter
thinks is signalling an error.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From markus.vonehr at ipm.fhg.de  Mon Sep  9 04:03:58 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Mon, 09 Sep 2002 10:03:58 +0200
Subject: compile error on SWIG file
Message-ID: <3D7C55EE.D9278B1C@ipm.fhg.de>

Hi,

compiling my c-file generated from SWIG
(swig -c -python -o mvelib.c mvelib.i)
with a setup-file, I receive the error:

c:\programme\vc98\include\eh.h(32) : fatal error C1189: #Error: "eh.h
is only for C++!"

Anybody knows what I have to change???

Thanks Markus


****************************
setup.py
****************************
from distutils.core import setup, Extension

setup (name = "mvelib",
        version = "1.0",
        maintainer = "Markus von Ehr",
        maintainer_email = "markus.vonehr at ipm.fhg.de",
        description = "mvelib Python module",

        ext_modules = [Extension('mvelib',sources=['mvelib.c'])]
 )


From robin at jessikat.fsnet.co.uk  Thu Sep 19 05:31:28 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 19 Sep 2002 10:31:28 +0100
Subject: cgi binary upload
References: <d5t2JzAIBNi9EwtI@jessikat.demon.co.uk> <3D88D492.1080909@mxm.dk>
Message-ID: <3tD8sIAwlZi9EwoD@jessikat.demon.co.uk>

In article <3D88D492.1080909 at mxm.dk>, Max M <maxm at mxm.dk> writes
>Robin Becker wrote:
.....
>That should not be a problem. I have not heard of the cgi module having 
>those kind of problems at least.
>
.......
>
>I don't think your problem is with Python. You are probably having a 
>problem somewhere else in the chain.
>
>At least we should see some code, or you need to be more specific.
>
>regards Max M
>
>
My colleague, who is doing the work reports that

"I've just installed Apache 1.3.26 and the demo with file upload works
completely, but still not with Apache 2.0.40 which I'm running in
parallel."

So that may be something to do with our difficulties.
-- 
Robin Becker


From thomas at xs4all.net  Thu Sep 12 09:44:25 2002
From: thomas at xs4all.net (Thomas Wouters)
Date: Thu, 12 Sep 2002 15:44:25 +0200
Subject: class variable won't icrement!
In-Reply-To: <20020912131422.GB29027@meson.dyndns.org>
References: <7647A9F4B298E74DB6793865DA67285004ADF4@exchange.adrembi.com> <20020912131422.GB29027@meson.dyndns.org>
Message-ID: <20020912134425.GC797@xs4all.nl>

On Thu, Sep 12, 2002 at 09:14:22AM -0400, Jp Calderone wrote:
> On Thu, Sep 12, 2002 at 12:18:32PM +0200, Roman Yakovenko wrote:
> > Hi it is simple just change 
> > self.id = self.id + 1 
> > To 
> > test.id += 1

>   Actually, that turns out not to be the case.  += behaves differently
> depending on its arguments - sometimes rebinding the name, sometimes
> mutating in place.  For integers, it can not mutate in place (ints are
> immutable, of course!) so "self.id = self.id + 1" is exactly the same as
> "self.id += 1".  Were 'id' a mutable type, say a list, then the above
> solution would work (assuming you changed 1 to something iterable).

Notice how Ramon said 'test.id', not 'self.id. He's changing the class
attribute directly, not through the instance. The switch from
addition+assignment to augmented assignment is just a smoke screen :)

-- 
Thomas Wouters <thomas at xs4all.net>

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



From bbrelin at yahoo.com  Wed Sep 18 15:07:45 2002
From: bbrelin at yahoo.com (Braun Brelin)
Date: 18 Sep 2002 12:07:45 -0700
Subject: How do I find out which context I'm passing?
Message-ID: <e0a78451.0209181107.328addf2@posting.google.com>

I'm running into a strange problem.  I've written a small e-mail
sender python
script for Zope/CMF/Plone.  It gets activated from DCWorkflow.  

The script works fine when I test it from the "manage" interface, i.e.
I save
the script and then push the "test" button.  

When I try and run it from my portal, however, the script fails. 

The script looks like this:

----------------------------------------------------------------------------
members = context.portal_membership.listMembers()

try:  
   mailhost = getattr(context,context.superValues('Mail Host')[0].id)
except:
   raise AttributeError,"Can't find a mail host object"

addrs = []
for member in members:
        if 'Member' in member.getRoles():
                email = member.getProperty('email')
                if email:
                        addrs.append(email)

#url = getattr(exprNamespace['object'].absolute_url())
url = 'test url'
to_list = string.join(addrs, ',')
mFrom = 'ciaran at localzope'
mSubj = 'New Content'

mailhost.send(url, to_list, mFrom, mSubj)
#return to_list

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

The script constantly gets a "No mailhost found" error.  Which means,
I presume,
that the getattr(context, context.superValues()) function call fails. 

The only reason that I can think of for this happening is that the
context that's passed when I test it via the manage window is a
different context
than that which gets passed within the users portal window.  

There is definitely a MailHost object defined within the portal
application.

Anybody have any thoughts on how to debug this?

Thanks,

Braun Brelin


From pedronis at bluewin.ch  Thu Sep  5 08:29:54 2002
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Thu, 5 Sep 2002 14:29:54 +0200
Subject: jython memory usage
References: <hVm5ydAY2hd9Eww$@jessikat.demon.co.uk><3d76366a$1_5@news.bluewin.ch> <rnLbuKAv5jd9Ewzz@jessikat.demon.co.uk><3d76494d$1_5@news.bluewin.ch> <YR13DVAUNld9EwXH@jessikat.demon.co.uk><3d765ce6_4@news.bluewin.ch> <zzKd7KAqhnd9EwxT@jessikat.fsnet.co.uk><3d767c4a$1_1@news.bluewin.ch> <YWyHyIAJKyd9EwHA@jessikat.demon.co.uk>
Message-ID: <3d774f40$1_1@news.bluewin.ch>

Ok, I have discovered an undocumented fact and maybe relevant fact.

The JVM ignore the Xmx limit for classes, now classes are garbage collected
if not used, but if that not the case and new classes are continously loaded
the JVM will allocate memory up to exhaustion of total memory, instead of
stopping
at Xmx.

Now one can log class loading, you can edit jython.bat and
put -verbose:class in there:

"..java.exe" -verbose:class "-Dpython.home=..."

then running jython.bat <the-mem-overflowing-code>

you should get an initial burst of [Loaded <classname>]
and then normally nothing or a pattern of

[Loaded ...]
...
[Unloading class ...]

If you get an increasing sequence of [Loaded ...] without corresponding
[Unloading class ...] then we have reduced
the possible suspects.

regards.





From cjw at sympatico.ca  Wed Sep 18 11:05:30 2002
From: cjw at sympatico.ca (Colin J. Williams)
Date: Wed, 18 Sep 2002 11:05:30 -0400
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <20020917.151708.502278611.1586@cybermesa.com> <T5Qh9.479727$q53.16320265@twister.austin.rr.com>
Message-ID: <3D88963A.F61FDD82@sympatico.ca>

You might look at:
      http://www.scopal.com/information06.htm

A digital oscilloscope is controlled by a PC, using wxPython
and Python.

Colin W.

Lee Gray wrote:

> Jay O'Connor wrote:
>
> > In article <96904b50.0209171305.6d65f574 at posting.google.com>, "Lee Gray"
> > <lpgray at uop.com> wrote:
> >
> >>
> >> Can someone point me to a good Python GUI app to demonstrate (other than
> >> IDLE)?
> >
> > Bioreason makes some products in Python.  You can see some screenshots of
> > their Python/Gtk products here: http://www.bioreason.com/lead.html and
> > http://www.bioreason.com/drug.html
> >
>
> Thanks, I'll show those.
>
> Lee



From ledirach at wanadoo.fr  Wed Sep  4 15:39:06 2002
From: ledirach at wanadoo.fr (Pierre-Yves Le Dirach)
Date: 4 Sep 2002 12:39:06 -0700
Subject: Python critical mass /turning point
References: <899f842.0209040617.711664fe@posting.google.com>
Message-ID: <a263a361.0209041139.7cdcc19e@posting.google.com>

X-No-archive:yes

If i dare adding a word, I think you forgot at least one reason among
"Suitability for commercial products" :

2.4) Human-readable and understandable source code (even for
non-techies);
Even if you could open a Perl program in vi (Notepad ?) and read it,
no non-techie could understand it at first sight (sometimes even
techies...;).
But the day my boss (not a programmer, but a marketing guy) had a look
on my Python source, he felt like he understood it. And most
"Microsoft-minded" bosses can't consider selling "understandable"
source to a client.

Closed-source companies... 

Just my 2 cents.

P.


anthony_barker at hotmail.com (Anthony_Barker) wrote in message news:<899f842.0209040617.711664fe at posting.google.com>...
> I have been trying to determine what is holding python back from 
> reaching the critical mass/success that perl/c++/Java have experienced.
> Caveate - I am not a python guru - so some of the technical stuff may be wrong.
> 
> I believe most of the issues are not technical - If technical features were 
> key to critical mass then LISP would be the main programming 
> language.
> 
> 1. Interpreter Available on every desktop/server
> Perl has this on the server
> Java had this in the browser space
> 
> Solution(s)
> Parrot/Perl6 integration - the if python is included
> with the perl interpreter it would be everywhere
> Lobby Sun & IBM to include Python on their Unix CDs
> Lobby Microsoft to get added to their resource kit
> Other options are already available such as freeze....
> 
> 2. Suitability for commercial products
> A couple of the problems I believe are technical and
> reputation.
> 1) Speed issue compared when compared with c/c++
> 2) The ability to enforce interfaces
> Allow enforced type checking and encapsulation - people are often
> surprised when they find out data is hidden in python.
> 3) Access c libraries/apis seamlessly
> 
> Solutions(s)
> Speed issue I believe may be solved by psycho, and time. As python
> matures it seems to be getting faster, while computers are getting
> so fast speed matters less.
> 
> Having a compiler would allow python to be know not only as a 
> scripting language. Unfortunately "scripting language" for some
> is equivalent to - not a real language. Enforced interfaces
> and the ability to enforce data hiding
> 
> SWIG is doing a good job with the last issue.
> 
> 
> 3) Big corporate sponsor(s)
> IBM sponsors Java/Perl/C/C++ & Cobalt 
> by having IDEs and corporate support available for those languages.
> 
> IBM supports PERL?
> Surprise - Tivoli products are glued together with Perl
> Lotus Unix server products use perl under the hood
> Perl is shipped on every AIX box
> Perl courses are available as is support
> etc
> 
> Microsoft Supports
> VB/C/C++/c#/Java(kind of)/Perl etc
> 
> Perl is included in the NT/Win2K resource kits
> Also is part of the tools for Unix
> 
> Sun
> Java/Perl/etc
> 
> Perl is bundled with Solaris
> Sun Solutions has Perl courses available
> 
> 
> Cheers,
> 
> Anthony
> http://xminc.com/anthony/


From mstenner at phy.duke.edu  Thu Sep 12 10:53:05 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Thu, 12 Sep 2002 10:53:05 -0400
Subject: Coding Style: Quotes
In-Reply-To: <bt1g9.17429$Pf7.590136@news1.west.cox.net>; from javajohn@cox.net on Thu, Sep 12, 2002 at 02:25:11PM +0000
References: <bt1g9.17429$Pf7.590136@news1.west.cox.net>
Message-ID: <20020912105305.A6508@phy.duke.edu>

On Thu, Sep 12, 2002 at 02:25:11PM +0000, John Waycott wrote:
> I'm curious if others have adopted any standards for choice of single vs.
> double quotation marks to delimit strings. A look through the standard
> library reveals the choice of one over the other is rather arbitrary. I
> suspect it really makes no difference, but the question has come up during a
> code review.

Since there is no "physical" difference between them, I can't imagine
why it would matter.  The only reason there "needs" to be two is for
nested quoting.  Sure, it would be annoying to alternate for each
string, but beyond that...

Personally, my use tends to be c/perl influenced.  I use single quotes
for short static things and double quotes for longer things and format
strings.

f['key'] = "This is a long string, but it's just _my_ style..."

					-Michael

-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From jkraska at san.rr.com  Thu Sep 26 00:29:35 2002
From: jkraska at san.rr.com (Courageous)
Date: Thu, 26 Sep 2002 04:29:35 GMT
Subject: Larry Wall's comment on python...
References: <F7DB8D13DB61D511B6FF00B0D0F06233025D48F5@mail.hebmex.com> <mailman.1033004794.21467.python-list@python.org>
Message-ID: <mq25pu0bs0159c70pp16rutfoabgh9f8hn@4ax.com>

>Once upon a time, a large class of computers, and their attendant 
>peripherals, OSs and utility programs universally implemented tabs 
>hard-wired at 8 stops.

Indeed; such is this issue that I leave hard tabstops as 8 in my
editor, and set soft tab stops set to 4. What this means, in vim,
is that when tab is struck it "takes the cursor to the next logical
tab stop by inserting enough spaces to get it there." No real tab
exists. In fact, I have to strike ctrl-V[TAB] in order to get a
real one. I also use another setting that displays ascii-187 (?)
wherever a tab is. This is nice for editing those nasty makefiles
(that _REQUIRE_ tabs).

The relevant parts of my .vimrc can roughly be summarized as follows:

set tabstop=8
set softtabstop=4 
set shiftwidth=4 
set et
"ascii-187==?
set lcs=tab:?\ 

If you want to enter ascii-187 manually in vim (and can't cut out
of this message) type ctrl-V187

C//



From Oschler at earthlink.net  Wed Sep 11 22:44:15 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Thu, 12 Sep 2002 02:44:15 GMT
Subject: SpamBayes PIK file?
References: <mailman.1031790134.7684.python-list@python.org>
Message-ID: <3cTf9.1805$Lg2.262739@news2.news.adelphia.net>

"Tim Peters" <tim.one at comcast.net> wrote in message
news:mailman.1031790134.7684.python-list at python.org...
> [Robert Oschler]
>
> It's a binary pickle of a trained instance of class GrahamBayes in the
code
> base.  If what to do next isn't instantly obvious, you should return the
> .pik file for a full refund <wink> -- unless you're a developer on this
> project, you don't stand much chance of getting any good out of it.
>

Thanks Tim, yeah there's two unexpected events that confuse me with Source
Forge, since I've come to expect source code in project.  That's one of them
and the other's an empty project that points to a commercial package that
doesn't even sell the source. :)

thx





From dag at animagicnet.no  Mon Sep 30 05:53:21 2002
From: dag at animagicnet.no (Dag)
Date: 30 Sep 2002 09:53:21 GMT
Subject: is cPickle really this slow?
Message-ID: <slrnapfs9u.8re.dag@omally.animagicnet.no>

I have a script which starts by parsing a large file and turning into
a data stucture (a dictonary of lists) which I then use in the script.  
This operation however takes about 12-15 seconds.  Since the file hardly 
ever changes , and if it changes it's because I changed it, thought I'd 
save some time by pickeling the datastructure and and simply getting that 
at the start of each script.  So I pickle my dictionary, which ends up
being a 3.2MB file and try to run my script again, this time getting 
the pickled dictonary instead of building it at runtime.  This time
however the script takes over 35 seconds to run, close to three times
as long.  Am I doing something wrong is loading and parsing larger
pickled files really very slow.

Dag


From huaiyu at gauss.almadan.ibm.com  Tue Sep  3 13:42:27 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Tue, 3 Sep 2002 17:42:27 +0000 (UTC)
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <mailman.1030907703.26266.python-list@python.org> <akuifb$h0e$2@solaris.cc.vt.edu> <1030962064.628232@ampungk.ozonline.com.au> <al0djh$a33$1@solaris.cc.vt.edu> <1031008966.343893@ampungk.ozonline.com.au> <mailman.1031013186.6118.python-list@python.org>
Message-ID: <slrnan9t42.pb.huaiyu@gauss.almadan.ibm.com>

jepler at unpythonic.net <jepler at unpythonic.net> wrote:
>On Mon, Sep 02, 2002 at 09:23:18PM +1000, John La Rooy wrote:
>> Carl Banks wrote:
>> >
>> >pattern = Group(Any()) + Group(Any()) + Group(Any()) \
>> >          + GroupRef(3) + GroupRef(2) + GroupRef(1)
>> >
>> Err symantically that's exactly the same as the re and my suggestion
>> only the syntax is different. It's still nothing like saying
>> 
>> pattern = "6 character palindrome"
>
>Do you mean something like this?
>
>    def palindrome_re(n):
>        pat = ["(.)" * ((n+1)/2)]
>        for i in range(n/2, 0, -1):
>            pat.append("\\%d" % i)
>        return "".join(pat)
[snip]
>
>I think that building REs in functions is a great approach for more
>complex REs.

It would also be useful if patterns can be built up as structured objects:

    def palindrome_re(n):
        pat = Empty()
        for i in range(n):  pat += Group(Any())
        for i in range(n, 0, -1): pat += GroupRef(i)
        return pat

    class Palindrom:
        def __init__(self, n): self.n = n
        def __call__(self): return palindrome_re(self.n)

    pattern = Palindrom(6)

Once the tree structure is revealed it could be mapped to whatever syntax
that is convenient for the situation.  The perl-like syntax would just be
one of them.

Huaiyu


From astavale at yahoo.co.uk  Fri Sep 13 14:48:34 2002
From: astavale at yahoo.co.uk (Alistair Thomas)
Date: Fri, 13 Sep 2002 18:48:34 GMT
Subject: Signals for GtkEntry widgets
References: <mailman.1031866395.5190.python-list@python.org>
Message-ID: <6qqg9.509135$Aw4.21217664@bin2.nnrp.aus1.giganews.com>

In article <mailman.1031866395.5190.python-list at python.org>, "David
Shifflett" <shifflett at nps.navy.mil> wrote:

> I am trying to find which signal is emitted when the user presses Enter
> while in a GtkEntry widget.

The signal is key_press_event. Which returns a GdkEvent object. The
keyval attribute tells you which key was pressed.

I use Glade and my sample code is based on my own module which uses libglade,
hope this helps:

import gtk, GDK

# My own module based on libglade:
import window 

class text_entry ( window.open ):

	def __init__ ( self, title = 'Enter text:', text = '' ):
		window.open.__init__ ( self, 'text_entry' )
		self.add_handlers ( { 'on_key_press' : self.key_press,
						'closed' : self.closed } )
		self.get_widget ( 'text_entry' ).set_title ( title )
		self.modified = 0
		self.text = text
		a = self.get_widget ( 'text' )
		a.set_text ( text )
		a.grab_focus ()
		gtk.mainloop ()

	def key_press ( self, widget = None, event = None ):
		if event.keyval in [GDK.Return, GDK.KP_Enter] :
			self.text = widget.get_text ()
			self.modified = 1
			self.close ()
			gtk.mainquit ()

	def closed ( self, widget = None ):
		gtk.mainquit ()


--
	Al


From oliver.smith at removethisbit.baesystems.com  Thu Sep  5 06:31:40 2002
From: oliver.smith at removethisbit.baesystems.com (Olly Smith)
Date: Thu, 5 Sep 2002 11:31:40 +0100
Subject: Returning lists from python COM servers
Message-ID: <3d7731f5$1@pull.gecm.com>

Using this snippet on the server side:

==
return (tags,vals)
==

Where tags and vals are equal length lists (actually a dict that maps
strings to strings split up so it can be returned from the COM object)

This snippet on the client side:

==
tags,vals = self.prodcat.getPDF(id)
==

I get this error (mildly snipped cgitb output):

==
 547         tag,val = self.prodcat.getPDF(id)

ValueError: unpack tuple of wrong size
      __doc__ = 'Inappropriate argument value (of correct type).'
      __getitem__ = <bound method ValueError.__getitem__ of
<exceptions.ValueError instance at 0x00849670>>
      __init__ = <bound method ValueError.__init__ of <exceptions.ValueError
instance at 0x00849670>>
      __module__ = 'exceptions'
      __str__ = <bound method ValueError.__str__ of <exceptions.ValueError
instance at 0x00849670>>
      args = ('unpack tuple of wrong size',)
==

Any ideas on where I'm going wrong? I'm pulling my hair out trying to work
this one out.

Thanks

Olly





From SBrunning at trisystems.co.uk  Thu Sep 26 05:19:49 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 26 Sep 2002 10:19:49 +0100
Subject: jython web scripting options
Message-ID: <31575A892FF6D1118F5800600846864DCBD880@intrepid>

> From:	Stefaan Nachtergaele [SMTP:snacht at xs4all.nl]
> I'm prototyping a three tier web application in java. I've looked at a lot
> of java web frameworks but for development speed nothing compares to web
> scripting languages like php, etc. But that's not an option for me cause I
> need to talk with my server side java code (php-java integration is
> dinky).
> My question is, what options are there for jython.
 
Why not stick with J2EE, and code your servlets, tags etc. in Jython? See
<http://aspn.activestate.com/ASPN/Cookbook/Python?kwd=Jython> for some
simple examples.

Oh, and have a look at struts - see
<http://www.brunningonline.net/simon/blog/archives/000357.html>.

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




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



From a12345 at ms1.url.com.tw  Sat Sep 21 15:58:09 2002
From: a12345 at ms1.url.com.tw (Decent)
Date: 21 Sep 2002 12:58:09 -0700
Subject: WxString and python string module?
Message-ID: <36de4780.0209211158.2fc3815@posting.google.com>

Dear all:

I'm trying to rewrite my apps using wxpython,
but how do I convert WxString object to python string?
I have to use python's default string module, not the wxWindow's WxString.

Any Idea?


From ehagemann at comcast.net  Wed Sep 25 17:03:30 2002
From: ehagemann at comcast.net (Eric Hagemann)
Date: Wed, 25 Sep 2002 21:03:30 GMT
Subject: help with print function in extension
References: <tA8k9.270440$AR1.11448440@bin2.nnrp.aus1.giganews.com> <4jfk9.163306$pX1.5842487@news2.tin.it>
Message-ID: <Cwpk9.283583$5r1.12035771@bin5.nnrp.aus1.giganews.com>

Alex, Thanks -- missed that section in the manual.  I blindly assumed that I
could use the repr() as the print func......

ok so for the next trick question. I implemented a function that has the
proto
int print_func(PyObject *obj, FILE *fp,int flags)

when I try to print to fp -- as in --> fprintf(fp,"<fmt>", , )... I crash
the program.

when the print function is used and output is sent to the screen is the fp
set equal to stdout ?

The example in the "Extending and Embedding the Python Interpreter" show a
fprintf to fp as might be expected ?!?

Cheers


"Alex Martelli" <aleax at aleax.it> wrote in message
news:4jfk9.163306$pX1.5842487 at news2.tin.it...
> Eric Hagemann wrote:
>
> > I'm handcrafting an extension in C
> >
> > If have defined both a str() and a repr() function and when I execute
> > either repr(obj), or str(obj) I get what I expect
> >
> > however if I execute 'print obj' I get the following error
> >
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in ?
> > SystemError: error return without exception set
> >
> > although 'print str(obj)' works like I think it should
> >
> > Am I mistaken in thinking that  'print str(obj)' should be the same as
> > 'print obj'
>
> *ALMOST!*  In the PyTypeObject struct, there's a tp_print slot
> that is used by preference, if non-null, when you print something
> and stdout is a real file.  What do you have in tp_print for
> your type?  It's smack in-between tp_dealloc and tp_getattr.
>
>
> Alex
>




From a_salnikov at yahoo.com  Mon Sep  2 03:16:54 2002
From: a_salnikov at yahoo.com (Andy Salnikov)
Date: Mon, 2 Sep 2002 00:16:54 -0700
Subject: Python GUI
References: <3d8d3c0.0208262339.e7bc02c@posting.google.com> <uniomu8pdi7tm169b4g6225d5oqdrkj2go@4ax.com> <akobqe$4en$1@usenet.Stanford.EDU> <3D6FBB3B.5090700@river-bank.demon.co.uk> <r794nugam0su9metpaas0nobn4jsm84j3j@4ax.com>
Message-ID: <akv3b8$7dv$1@usenet.Stanford.EDU>

"Lothar Scholz" <llothar at web.de> wrote in message
news:r794nugam0su9metpaas0nobn4jsm84j3j at 4ax.com...
> On Fri, 30 Aug 2002 19:36:43 +0100, Phil Thompson
> <phil at river-bank.demon.co.uk> wrote:
>
> >Or take the Qt (and PyQt) approach and implement all styles on all
> >platforms and let the user decide on a global or per application basis
> >with a simple command line argument.
>
> Thats completely wrong it the user selects a style from the OS style
> manager. At least for MacOS Users this is very popular and then
> your QT application looks again like an alien. Same is true for WinXP

  Qt on Windows actually uses whatever style is current on the host
(in other words Qt uses native widgets). On X11 they have their own
configuration utility which changes style of Qt-based applications.
Problem with X11 is that there is no "OS style manager" (whatever
this could mean) for X11, or there is no _single_ style manager.

  Andy.




From phil at river-bank.demon.co.uk  Sat Sep 28 04:51:50 2002
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Sat, 28 Sep 2002 09:51:50 +0100
Subject: how to create QLineEdit.focusOutEvents?
In-Reply-To: <an2eos$9v7m1$1@ID-143433.news.dfncis.de>
References: <an2eos$9v7m1$1@ID-143433.news.dfncis.de>
Message-ID: <200209280951.51269.phil@river-bank.demon.co.uk>

On Friday 27 September 2002 10:05 pm, Hans - G. Normann wrote:
> Hello
>
> I'm new in QT and Python. My first project: creating a GUI for Linux tar
> command. My prob:
>
> A QLineEdit contains a directory name. I can create this name by clicking
> (Fileselection maodule) or by type in. On lost focus, I want check the
> Directoryname, but how can I create slots, signals and the event for Line
> Edit?
>
> For creating gui, I'm using QT's designer. It's works fine, except
> focusOutEvent. Is there a (easy) sample?

Sub-class QLineEdit and re-implement focusOutEvent() in your sub-class.

Phil



From dallasm at aiinet.com  Wed Sep 11 14:38:03 2002
From: dallasm at aiinet.com (Mahrt, Dallas)
Date: Wed, 11 Sep 2002 14:38:03 -0400
Subject: Distutils/setup help
Message-ID: <638AA0336D7ED411928700D0B7B0D75B02E1EF96@aimail.aiinet.com>

I am in the process of making modifications to the Python build system to
aid in the cross-compilation of Python. A primary problem I am running into
involves the module detection logic 
In PyBuildExt::detect_modules() of setup.py. It is determining what modules
to build based on the build system. I have modified 'Modules/Setup' to
remove some modules from being compiled, but I would still like to be able
to specify more cleanly which modules to compile and the necessary flags to
use for the hosts system. Is there currently a way to make these
specifications? If not, does anyone have any suggestions for a clean
longer-term solution? Is this the right list to ask these questions?

Thanks.

---
Dallas S. Mahrt               614-923-1194
Software Engineer             Network Management Systems
Applied Innovation Inc.       Dublin, Ohio




From observer at NOSPAM.space.pl  Wed Sep 25 10:17:16 2002
From: observer at NOSPAM.space.pl (Johann)
Date: Wed, 25 Sep 2002 16:17:16 +0200
Subject: Python mixed with HTML like PHP
Message-ID: <o9h3pu4f24ffnupaen8ut8v8dgokacdna6@4ax.com>

I heard about project which can mix python cgi code with html content.
I forgot its name. Who remember?

--
Johann


From al at metroic.no  Mon Sep 23 14:54:44 2002
From: al at metroic.no (Andreas Lydersen)
Date: Mon, 23 Sep 2002 20:54:44 +0200
Subject: Dynamic loading of modules
Message-ID: <PrJj9.12293$Uu.3503@news1.bredband.com>

Hi,
I'd like to load a set of modules from a directory.  I have tried to use the
__import__(...) statement to do so, but with no success.
?
The static approch works (of course), but the dynamic does not!

static:
import plugins.Password

dynamic atttempt:
__import__("plugins.Password")

I have tried 'exec "plugins.Password"', and that works.  However, I need to
get a list of pointers to the modules I import in order to work with them
later.

Bottom line is:  I need to know how to dynamically load modules and get a
pointer to the same module.

Any help is appreciated.

--
Johan Fredrik ?hman




From linuxbrad at rogers.com  Thu Sep 12 14:59:56 2002
From: linuxbrad at rogers.com (Brad Fonseca)
Date: Thu, 12 Sep 2002 18:59:56 GMT
Subject: [newbie] Buying an introductory/reference text
Message-ID: <Mu5g9.1702$U_.1021@news04.bloor.is.net.cable.rogers.com>

Greetings!

I'm looking for a few opinions on the above subject.  I'm a computer 
science student who has had some introductory lectures on Python and I've 
done some very basic mucking around with.  I'm looking for a clear 
reference and/or introductory text.  Any suggestions would be nice.  I have 
looked at http://www.python.org/cgi-bin/moinmoin/IntroductoryBooks but 
there appears to be a number of good books to choose from there.

Regards,

-- 
---
Brad Fonseca


From pinard at iro.umontreal.ca  Tue Sep  3 21:23:27 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Tue, 03 Sep 2002 21:23:27 -0400
Subject: Pyrex 0.4.4
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8BF0A567@natasha.auslabs.avaya.com> ("Delaney,
 Timothy"'s message of "Wed, 4 Sep 2002 10:33:34 +1000")
References: <B43D149A9AB2D411971300B0D03D7E8BF0A567@natasha.auslabs.avaya.com>
Message-ID: <oqwuq2r1io.fsf@titan.progiciels-bpi.ca>

[Delaney, Timothy]

>> I noticed too.  To get going, I got Plex installed separately, and removed
>> the 'Plex' line in `setup.py' for Pyrex.

> I decided against that, since one of the ficed bugs was in Plex.

I wondered a bit as well.  But since the Plex bug never hit me so far that I
know, it would take some bad luck for being hit now.  The fact that Pyrex
traceback got nicely integrated in the overall Python feedback is a nice
feature, which outweigh the fear of the Plex bug.

In one application, I use Pyrex to embed Python within a Bison generated
parser and a Flex scanner, that is, within C code.  This afternoon, I got
Pyrex to generate some code, then cut-and-pasted the traceback tricks it uses
into a new C module, which the parser calls after a Python returns an error.
So even the C code is properly referenced in the traceback.  Wonderful!

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



From kennethg at pd.jaring.my  Sat Sep 21 23:59:41 2002
From: kennethg at pd.jaring.my (Kenneth Gomez)
Date: Sun, 22 Sep 2002 03:59:41 GMT
Subject: Local Python Server Installation
References: <3d8b47c6.277669@news.jaring.my> <slrnaomn6m.39s.gerhard.haering@haering.opus-gmbh.net> <3d8c993c.2190426@news.jaring.my> <mailman.1032625889.13783.python-list@python.org>
Message-ID: <3d8d4009.375344@news.jaring.my>

Chris, you're right. It's an Apache config problem. I'll dig into the
doc's.

Thanks,
Regards,
Kenneth.


On Sat, 21 Sep 2002 11:31:50 -0500, "Chris Gonnerman"
<chris.gonnerman at newcenturycomputers.net> wrote:

>----- Original Message ----- 
>From: "Kenneth Gomez" <kennethg at pd.jaring.my>
>
>
>> I tried your method. But instead of getting a cgi listing of 
>> my environmental listings, all I saw on my browser was the 
>> two input lines ..i.e. import cgi and cgi.test()
>
>Your cgi is being displayed rather than executed.  The 
>configuration options for Apache are so complex, that entire 
>books have been written about configuring it... and Gerhard's
>suggestion only works on Windows platforms (AFAIK).  You didn't 
>name your platform, did you?  Or did I just miss it?
>
>You need to verify that Apache 1) recognizes .py files as CGI 
>scripts, and 2) allows you to run a CGI in the folder you're 
>putting it in.  You probably should put your script in the
>cgi-bin directory to start with.  If you are on a Unix/Linux
>system, you need to make sure that the script is marked 
>executable.
>
>Really, I think you need to spend some time with the Apache 
>docs.  It isn't a Python problem per se.
>
>Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
>http://newcenturycomputers.net
>
>

Cheers,
Kenneth Gomez.


From pknotz at sandia.gov  Tue Sep 24 15:27:53 2002
From: pknotz at sandia.gov (Pat Notz)
Date: 24 Sep 2002 12:27:53 -0700
Subject: Copying Elements using PyXML
References: <mailman.1032866191.25215.python-list@python.org>
Message-ID: <525eaba.0209241127.1ec90a6@posting.google.com>

This is an untested snippet of code that should get you started.  It
looks like you'll want to do you .appendChild on a node other than
.documentElement...

# Merge subsequet DOMs
dom = None
for xmlfile in xmlfiles:
    tmp_dom = xml.dom.minidom.parse(xmlfile)
    if dom == None:
        dom = tmp_dom
    else:
        for node in tmp_dom.documentElement.childNodes:
            e = dom.importNode(node,True)
            dom.documentElement.appendChild(e)


From anthony at computronix.com  Wed Sep 18 00:31:34 2002
From: anthony at computronix.com (Anthony Tuininga)
Date: Tue, 17 Sep 2002 22:31:34 -0600
Subject: can wxpython do this?
In-Reply-To: <3d87fbfc@news.sentex.net>
References: <737aefee.0209151702.3f735252@posting.google.com>
 <spcaou804m5oo6gah31oqaeuqsuj60h0ip@4ax.com> <3d854b44@news.sentex.net>
 <am7hv3$623$1@news01be.so-net.ne.jp> <3d87fbfc@news.sentex.net>
Message-ID: <1032323500.23315.0.camel@localhost>

I have written a set of routines similar to py2exe and I know that
Gordon McMillan's installer does similar things as well. Both of these
options are cross-platform. So which platforms do you need to have this
run on?

On Tue, 2002-09-17 at 22:07, Peter Hansen wrote:
> Erick wrote:
> >     > A better option, although I haven't tried it myself, would be
> >     > to use py2exe to build an executable, then package that up
> > 
> > thanks Peter, but I was talking about something cross-platform. As py2exe's
> > page says:
> > 
> >      py2exe is a distutils extension to convert python scripts into
> >     executable windows programs, which are able to run without
> >     requiring a python installation.
> > 
> > 
> > do you know if it works across platforms?
> 
> I don't believe the current solutions are cross-platform, but
> it could be feasible to build a meta-packager which uses the
> appropriate package to build executables for different platforms
> by calling on the appropriate platform-dependent packager.
> 
> Anyway, do you really want all platforms, or a specific subset?
> 
> -Peter
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list





From markus.vonehr at ipm.fhg.de  Tue Sep 10 02:56:38 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Tue, 10 Sep 2002 08:56:38 +0200
Subject: compile error on SWIG file
References: <3D7C55EE.D9278B1C@ipm.fhg.de>
Message-ID: <3D7D97A6.A63610AF@ipm.fhg.de>

Hi Lyle,

thanks for your answer and your efforts!
I changed everything to .cpp
The first error has gone
Now I have an error (translated from german):

mvelib.obj: error LNK2001: unresolved external symbol initmvelib
build\temp.win32-2.1\Release\mvelib.lib : fatal error LNK1120: 1 
unresolved reference
LINK : fatal error LNK1141: Creation error of export file
error: command 'c:\Programme\VC98\BIN\link.exe' failed with exit status
1141

Markus


Markus von Ehr schrieb:
> 
> Hi,
> 
> compiling my c-file generated from SWIG
> (swig -c -python -o mvelib.c mvelib.i)
> with a setup-file, I receive the error:
> 
> c:\programme\vc98\include\eh.h(32) : fatal error C1189: #Error: "eh.h
> is only for C++!"
> 
> Anybody knows what I have to change???
> 
> Thanks Markus
> 
> ****************************
> setup.py
> ****************************
> from distutils.core import setup, Extension
> 
> setup (name = "mvelib",
>         version = "1.0",
>         maintainer = "Markus von Ehr",
>         maintainer_email = "markus.vonehr at ipm.fhg.de",
>         description = "mvelib Python module",
> 
>         ext_modules = [Extension('mvelib',sources=['mvelib.c'])]
>  )


From jo at johnochiltree.uklinux.net  Mon Sep 16 03:04:42 2002
From: jo at johnochiltree.uklinux.net (John Ochiltree)
Date: Mon, 16 Sep 2002 08:04:42 +0100
Subject: couldnt find __dic__
References: <7153651c.0209151958.4099ac3e@posting.google.com>
Message-ID: <Dofh9.52300$rd2.31724@news-binary.blueyonder.co.uk>

"black" <quiteblack at yahoo.com> wrote in message
news:7153651c.0209151958.4099ac3e at posting.google.com...
> my book told me all dictionaries have __dic__ method but I searched
> each dictionaries and find nothing, where is it ???


I think you'll find your book has a misprint try __dict__ instead.

John




From see_reply_address at something.invalid  Tue Sep 10 01:32:47 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Tue, 10 Sep 2002 17:32:47 +1200
Subject: dynamism
References: <66715c02.0209091923.265fe77e@posting.google.com>
Message-ID: <3D7D83FF.6070202@something.invalid>

Steven Shaw wrote:

> 
> I thought you could just extend an object's attributes at run-time like this:
> 
>>>>p = object()
>>>>p.x = 1


It seems that you can't do this to an instance
of "object" itself, because it's a built-in type.
But you can do it to a subclass of object:

 >>> class C(object):
...  pass
...
 >>> c = C()
 >>> c.x = 1
 >>> print c.x
1
 >>>


> I thought that you could access dictionary values with object syntax:
> 
>>>>x = { 'a': 1, 'b':2 }
>>>>x.a


No, that has never been possible. But instances have
a __dict__ attribute that holds all the other
attributes, so you can do this:

 >>> x.__dict__['a'] = 42
 >>> print x.a
42
 >>>

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From jacek.generowicz at cern.ch  Tue Sep 24 10:55:46 2002
From: jacek.generowicz at cern.ch (Jacek Generowicz)
Date: 24 Sep 2002 16:55:46 +0200
Subject: Three dumb questions (ordered by dumbness descending)
References: <amnvit$7kqdt$1@ID-77524.news.dfncis.de> <pan.2002.09.23.20.43.59.461504.1216@io.com>
Message-ID: <tyfu1kfv3m5.fsf@pcitapi22.cern.ch>

Steven Feil <sfeil at io.com> writes:

> Humm, what is is zip(), a function in the standard Python library or
> is it something you or someone else made up? I did a quick search in
> the Python docs and the only zip I found referred to zip
> compression. If this is something from the standard Python libraries
> please state the library it comes from.

  Python 2.2.1 (#3, Jun  4 2002, 09:56:27) 
  [GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
  Type "help", "copyright", "credits" or "license" for more information.
  >>> zip
  <built-in function zip>
  >>> print zip.__doc__
  zip(seq1 [, seq2 [...]]) -> [(seq1[0], seq2[0] ...), (...)]
  
  Return a list of tuples, where each tuple contains the i-th element
  from each of the argument sequences.  The returned list is truncated
  in length to the length of the shortest argument sequence.
  >>> 


or even


  > pydoc "zip"

Python Library Documentation: built-in function zip

zip(...)
    zip(seq1 [, seq2 [...]]) -> [(seq1[0], seq2[0] ...), (...)]
    
    Return a list of tuples, where each tuple contains the i-th element
    from each of the argument sequences.  The returned list is truncated
    in length to the length of the shortest argument sequence.


From aleax at aleax.it  Thu Sep 19 11:23:36 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 19 Sep 2002 15:23:36 GMT
Subject: Python sequences reference - problem?
References: <mailman.1032443760.5578.python-list@python.org>
Message-ID: <YZli9.114624$ub2.2439313@news1.tin.it>

Michal Kaukic wrote:

> Hi,
>    
>    we like Python - I use it (with Numeric and SciPy modules)
> for teaching of Numerical analysis and my colleague is doing
> some research in discrete optimization in Python. And he often
> surprises me with non-conformal use of language :-)
> 
>    We are mathematicians, not programmers. I suppose, he believes
> that Python "thinks" like Mathematician. Yesterday, he was struggling

I think the concept of "MODIFYING an object" is pretty far from
most areas of maths I've dabbled in (I'm an engineer, not a
mathematician, but do find maths fascinating).  It seems to me
that the only real way to get a language that WILL "think like
a mathematician" is to take your pick among languages that have
no concept of "modifying an object" - which nowadays basically
means, Functional Programming languages.  Personally I find
Haskell the most elegant one -- clean syntax (uses whitespace
much like Python), clean semantics (lazy evaluation the default
everywhere).  Last I checked, the best Haskell compilers were
able to produce very decent code for many problems -- performance
will in most cases be better than Python's.  If you want even
MORE performance, I suspect some less-puritanical but still FP
language (such as O'Caml) can probably deliver it.

If for some reason you don't want to get into FP, then I think
you can forget everything about "thinking like a mathematician".
If data modification IS allowed, then your chosen language needs 
to choose either of two possibilities:

-- COPY-semantics: an assignment such as
        x = y
   inherently copies y, so that any future change to y has no
   more effect on x

-- REFERENCE-semantics: an assignment such as
        x = y
   makes x a synonym of y, so that both refer to the same
   object: any future change to that object affects any
   synonym for the object equally

Many languages choose a MIX of these two kinds of behavior,
but that's the worst of both worlds, as it complicates things
a lot.  Generally, OLD languages (such as Fortran) relied on
copy semantics (references or synonyms were not considered,
and when they could happen, as in argument-passing, they were
prohibited by the language rules, although those rules were
rarely enforced).  Modern languages by and large have moved
to reference semantics, which is generally much cleaner even
though in some cases performance can suffer (optimization is
the main reason some languages complicate things by mixing
reference and copy semantics).  "Generally" is important: LISP,
one of the oldest languages around (second-oldest after
Fortran, among those still widely used today), relied on
reference semantics from the start.  The resulting elegance
has much to do with Lisp's acceptance in the mathematical
community, IMHO.

Python uses reference semantics.  When you want a copy, you
ask for a copy -- it's as simple as this.  If you don't ask
for a copy, you don't get a copy... you get a reference to
the original.  There is no complication: it's _always_ this
way.  (OK, OK, at the margin there are doubtful cases: in
particular, a SLICE is a copy-request for all built-in types,
but a reference, sharing data with the original array, when
you use the Numeric package... Numeric had to move _away_
from copy and onto reference in this case to avoid accidental
copies of huge array slices... take this as the exception
that proves the rule, OK?-).

Python does have some data types that are immutable, which
helps move it a little bit closer to FP -- numbers, strings
and tuples are immutable.  (Even Java has strings as
immutables -- SOME of FP's niceties do start to move into
the mainstream, albeit slowly:-).  But the main style is still
very much NOT FP, rather relying on modifiable data.

> with the code shown below (somewhat simplified):
>     
> 
> L_L=[[1,2,3], [2,3,4], [3,4,5]]
> 
> def do_process(LL):
>     n=len(LL[0])
>     rec = [0]*n
>     Res=[]
>     
>     for row in LL:
>         rec[0] = row[0]
>         # further code - modifying rec
>         # lots of conditional processing and such...
>         # ...
>         Res += [rec]
>                 
>     return Res
> 
> ---------------------------------------------
> 
> After calling do_process(L_L), the "expected" result should be
> 
>    [[1, 0, 0], [2, 0, 0], [3, 0, 0]]
> 
> but the true result is
> 
>    [[3, 0, 0], [3, 0, 0], [3, 0, 0]].
> 
> 
>    Yes, this is fully in accordance with how the Python language should
> behave:
>        Res += [rec] inserts references to list object rec,
>        which are further modified... (he should use copy(rec) instead).

Right in all respects, yes.

 
>    But there is nothing to make this behaviour clearly VISIBLE in code.

*EVERY* assignment in Python uses reference semantics.  Every occurrence
of an assignment, therefore, makes reference behavior "clearly VISIBLE".

Presumably fish don't find water "clearly VISIBLE", being used to that,
much as it is for us and air.  When something "is *EVERYWHERE*", how
can it be "clearly VISIBLE" at the same time?-)


> If I work with pointers in C/C++ I know and see they are pointers.

That's because C uses copy semantics, and uses pointer to emulate
reference semantics when needed.  That's just the reverse of Python,
which uses reference semantics, and explicit calls to copy to
emulate copy semantics when needed.

C++ lets you define "references" which IMPLY reference semantics
rather than copy semantics -- and you don't SEE what they are at
the point of use (only at the point of definition).  Flaming about
that here, one way or another, won't be very useful:-).  But there
are LOTS of precedents: Pascal (!!!) lets you declare arguments
to a procedure as either reference or copy -- then you don't see
the difference at the point of use; Visual Basic emulated that, too
(the default has changed in recent releases of VB, but that's
-another- flame:-); languages with sophisticated macro systems,
such as Dylan or relatively-recent LISPs, let you do all that and
worse.

But C and Python are SIMPLE languages: one semantics is pervasive
(reference for Python, copy for C), you take explicit measures
when you want the OTHER semantics, and those measures are visible
at the point of use (well, _mostly_ -- C does complicate things
a bit by having arrays "decay to" pointers, Python has the
"slice aren't copies in Numeric" issue -- but that's definitely
at the margin).

> You can say - we also know that rec is list object and so be careful
> with it.  Yes, but consider the complex code where the similar constructs
> are very easy to overlook. And debugging of such code can be frustrating.

The issue has nothing to do with rec being a list object, really.
You may see the issue as having to do with the fact that rec is
(any kind of) *MUTABLE* -- indeed, a language without mutable data
DOES shield you from these issues.  If you tried to mutate non-
mutable data, as rec[0] = .... does, you'd get an exception at
runtime (if every kind data was immutable, of course, the language
wouldn't even have any CONSTRUCTS to express mutation).


>    My colleague was in state of despair and made thoughtful remarks
> about FORTRAN and Python similarity. He believes that Python is corrupting
> his (computer) memory...

Fortran (up to F77 -- I lost touch with it afterwards) had reference 
semantics in all parameter passing to subprograms (with aliasing forbidden 
in rather complex ways -- very few compilers ever enforced that, letting
the hard-to-debug cases for the programmers:-)), copy semantics for all 
assignments.  I can't see any parallel between that and Python's
behavior, not even in the "anti-parallel" sense of Python // C.


>    So what is the point? I wrote this message in hope that there are
> more people with similar experience. My question is - how to explain
> to novice non-programmer users (maybe mathematically "infected")
> the quirks of Python sequence objects? Which methodology to use

First of all, somebody with reasonable math bent will prefer a
more general explanation -- so, forget sequence objects, because
that's just one specific cases.  ALL object-mutation works the
same way, whether it be of a sequence or non-sequence.

> in programs so we can clearly see we work with "pointers"?

You're in water.  Water is *everywhere*.  When you think there
is nothing there, there's water.  Just get used to it.  Trying
to muddy up the water in an attempt to make it more visibile
would be counter-productive.

> Or maybe someone can propose the changes to the Language to overcome
> this (psychological) barrier? I feel this as a serious obstacle
> in using Python (my students experienced it too).

It appears, though you never state it, that the behavior you
would consider "natural" would have to do with copy semantics.
I find that weird, absolutely weird -- I've seen it happen to
old died-in-the-wool programmers without solid mathematical
bases, because for 20 or 30 years of their lives they had only
seen copy semantics and could no longer conceive of any other,
but that's exactly the reverse from the "mathematical bent"
you keep mentioning.

Accepting your mention of "mathematical bent" as relevant, I
repeat my suggestion that you look into functional programming.
FP is seriously cool, IF all the programmers involved have a
solid mathematical outlook on life, the universe and everything.

If for whatever reason you want to keep using languages based
on the idea of data modification, you won't find a simpler or
more regular one than Python.  And it would be absurd to pick
old-enough languages that use copy semantics because some
previous experience (NOT with maths itself, surely?!) makes
you feel that's natural and divergences from it should be
made "very VISIBLE" -- besides Python, you'd be cutting yourself
off from Java, Eiffel, C# ... reference semantics IS the wave
of the last decade-plus, and probably of the next one too
(unless FP comes into its own, but I've sort of stopped holding
my breath for THAT -- there just isn't enough of an audience
with a pervasive-enough mathematical turn of mind:-).

Just remembed that EVERYTHING works by reference, and what
could possibly be the problem...?


Alex



From sholden at holdenweb.com  Fri Sep 27 09:16:41 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 27 Sep 2002 09:16:41 -0400
Subject: Local Python Server Installation
References: <3d8b47c6.277669@news.jaring.my> <slrnaomn6m.39s.gerhard.haering@haering.opus-gmbh.net> <3d8c993c.2190426@news.jaring.my> <mailman.1032625889.13783.python-list@python.org> <3d8d4009.375344@news.jaring.my>
Message-ID: <bQYk9.1630$6g7.1090@fe03>

"Kenneth Gomez" <kennethg at pd.jaring.my> wrote in message
news:3d8d4009.375344 at news.jaring.my...
> Chris, you're right. It's an Apache config problem. I'll dig into the
> doc's.


The Wrox book "Professional Apache 2.0" is helpful if you need to understand
how Apache works and, despite its title, it alos covers older versions of
Apache.

If you want to write servers in pure Python you could look at "Python Web
Programming", but you will probably find you can get a lot further than you
thought with simple CGI, then you'll know enough about other directions
you'd like to go.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From peter at engcorp.com  Wed Sep  4 08:42:47 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 04 Sep 2002 08:42:47 -0400
Subject: Matching a string up to a word
References: <slrnanbkbc.huj.dag@omally.animagicnet.no>
Message-ID: <3d75ffaa$1@news.sentex.net>

Dag wrote:
> I have a string that contains an ID number a file path and a date, kind
> of like this:
> 
> 23545 /this/is a/path/to/a file Sun Sep 7 2002.
> 
> I want to extract the file name.  If I could be sure that the file path
> had no spaces or if it would be quited it would be easy.  Unfortunatly 
> many files will have spaces and changing the format of the output I'm
> dealing with isn't possible.  I do however know that the file name starts 
> after the first whitspace and that it's always followed by a three letter 
> day name (like Mon, Tue etc.)  Given this is there an easy way to extract 
> the whole file path from the string.

Not unless you can restrict the possibilities still further.  What
if the file name ended with a space and a three-letter name?

Can you instead guarantee the format of the date that follows the 
filename?  Will it always consist of the day of the week, the month,
the date, and the year?  That would turn this into a relatively
simple problem.

-Peter



From peter at engcorp.com  Tue Sep 10 00:04:11 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 10 Sep 2002 00:04:11 -0400
Subject: Windows/DOS: double clicking a .py file
References: <mailman.1031613313.24086.python-list@python.org>
Message-ID: <3d7d6f26@news.sentex.net>

David LeBlanc wrote:
> cmd.exe will accept a flag telling it to stay open after something is run.
> You can create a shortcut along the lines of:
> 
> cmd /K python myscript.py <args>
[snip rest]

Although, unfortunately for the OP who has users on Win98, cmd.exe does
not exist except in NT/2K/XP.

-Peter



From buzzard at urubu.freeserve.co.uk  Fri Sep  6 20:19:13 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Sat, 7 Sep 2002 01:19:13 +0100
Subject: RPy / Win2k
Message-ID: <albgmp$cgv$1@newsg2.svr.pol.co.uk>

Hello,
         Is there anyone out there who has successfully used RPy under
Win2k?  I get the impression (from Brian Ripley) that it should be possible
to embed R under Win2k, but that it probably isn't straightforward.  Can
anyone help?  Cheers.

Duncan




From kapis at hotmail.com  Sun Sep 15 01:23:22 2002
From: kapis at hotmail.com (Richard)
Date: 14 Sep 2002 22:23:22 -0700
Subject: Problem w/ boost_python_debug.dll
Message-ID: <cb95025b.0209142123.82c016e@posting.google.com>

I'm using Boost.Python (v1) to embed & extend a C++ application. I've
downloaded ActiveState's distribution and built the debug python
library and DLL.

I built the release and python-debug versions of Boost.Python as is
documented. I also define BOOST_DEBUG_PYTHON in my C++ files since I
have the debug python libraries.

Enough with the background, let's get to the problem. For debug
builds, I link with boost_python_pydebug.lib. But when I run the
program, it still tries to load up boost_python.dll, NOT the debug
version! The only way I know to fix the problem is to rename
boost_python_pydebug.dll so that it is loaded instead.

Does anyone know a better fix? Any help would be appreciated.

Thanks.


From jb at cascade-sys.com  Mon Sep  9 21:02:58 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 09 Sep 2002 18:02:58 -0700
Subject: Python primer - comments appreciated!
References: <aljdhe$1q63hj$1@ID-77524.news.dfncis.de>
Message-ID: <3D7D44C2.10901@cascade-sys.com>



Thorsten Kampe wrote:

>"Python, anyone?"... Yep, this is my first one. Comments on
>style/readability and technique/flow appreciated!
>
Looks good overall, certainly you're off to a good start.

I added a few remarks/answers below...

--jb

>
>
>#v+
>def quotient_set(seq, func, partition='nonbool'):
>    """ partition <seq> into equivalence classes
>
>    With the 'bool' parameter 'quotient_set' always returns two lists: the
>    first  containing the 'true' items and the second containing the false
>    ones. """
>    if partition == 'bool':
>    # Yes, I know about 'filter', but this returns just the 'true
>    # elements'...
>    
>        # put this in one line?
>        true_class  = []
>        false_class = []
>

        #### if you are thinking "true_class = []; false_class - []" 
then I'd guess
        #### most here would recommend against bunching up as pointless
        #### and just a tad less readable than two separate statements.

        #### if you are thinking "true_class = false_class = []"
        #### then it would be a BIG NO-NO.
        #### the latter form initializes both variables to the SAME 
(empty) list
        #### and I'm pretty sure you want 2 separate lists.  

        #### 'Mutable' constant objects like [] and {} work differently from
        #### 'Immutable' ones like 1, 0.0 and "abc".  Not sure the 
precise Pythonetics
        #### terminology, but one way to think of it is that mutable 
types are accessed
        #### 'by reference' while immutableones are accessed as if they 
were 'by value'.
        #### 2 + 3 results in a new number.  list.append(...) changes 
the list in place and
        #### affects all 'references' to the list.  

>
>        for item in seq:
>            if func(item):
>                true_class.append(item)
>            else:
>                false_class.append(item)
>        return [true_class, false_class]
>        
>    else:
>        # put this in one line?
>        quotient_set =         []
>        representative_class = []
>
>	for item in seq:
>            representative = func(item)
>
>            # "Try: seq.index(item)" is IMO rather ugly in a loop. But
>            # it's even better than "if item in seq: seq.index(item)",
>            # which goes twice through seq
>            try:
>                quotient_set_index = representative_class.index(representative)
>            except ValueError:
>                quotient_set.append([item])
>

#### teeny tiny nit:  I prefer "quotient_set.extend( item )" to the above.

>
>                representative_class.append(representative)            
>            else:
>                # This somehow "inelegant", because all I want to do is:
>                # "seq[index] = seq[index].append(item)" Suggestions?
>

####  Define your set or sequence types to be a class or classes.  
####  You can then define [] and .append() operations
####  to be whatever you want (e.g., silently exclude duplicates, etc.)

####  Even the most trivial data abstractions usually benefit by being 
made an
####  explicit class/object.

####  if you are 'indexing' you may want to use a dictionary base type 
instead of a list.

####  if you only want to test membership, instead of the 'try' you can 
test:

                  if representative in representative_class:  ....  


>
>                equivalence_class = quotient_set[quotient_set_index]
>                equivalence_class.append(item)
>                quotient_set[quotient_set_index] = equivalence_class
>                
>        return quotient_set
>#v-
>
>To get a feeling for what 'quotient_set' does, try:
>
>>>>f = lambda x: x % 3
>>>>quotient_set([1, 2, 3, 4, 5, 6, 7, 8, 9], f)
>>>>
>[[1, 4, 7], [2, 5, 8], [3, 6, 9]]
>
>>>>quotient_set([1, 2, 3, 4, 5, 6, 7, 8, 9], f, 'bool')
>>>>
>[[1, 2, 4, 5, 7, 8], [3, 6, 9]]
>
>#v+
>def set(seq):
>    """ make <seq> a true set by removing duplicates """
>    # Shortest way to do a 'null function'? 'None' doesn't work...
>    f = lambda x: x  # identity function
>    
>    return [item[0] for item in quotient_set(seq, f)]    
>#v-
>

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020909/60d4e36a/attachment.html>

From ptbjrii at hotmail.com  Thu Sep 19 09:33:07 2002
From: ptbjrii at hotmail.com (Polerio Babao Jr.II)
Date: 19 Sep 2002 06:33:07 -0700
Subject: Nokia 3310 SMS Application, done:connecting to serial port(linux/win), done:reading some hex like \x00 . . ., problem:write/send command! noresponse
Message-ID: <52facf1d.0209190533.3048e62c@posting.google.com>

Greetings: Python Masters

I wanted to create an SMS application using Nokia 3310 with
DAU-9P(data cable), I am now able to get some data from the cell
phone. I'm using Pyserial and Siomodule. I use pyserial in my linux
system. Out of curiosity I've even tested the pyserial on my mouse to
prove that I'm really reading data from the external device connected
on my com2 or /dev/ttyS1 on linux. From this you would perhaps believe
that I've done the first phase of my plan.

The Problem:

I tried almost three weeks writing or sending a command to the
cell-phone but this gives me no response. I'm talking here about the
nokia 3310 connected to my serial port. The command that I send is the
AT command. From my pyserial I made an initialization like this
ser=serial.Serial('/dev/ttyS1',19200) 
ser.readline() 
ser.write()
ser.read()

When I connect to the port for the first time I get the ser.read()
output like this \x00\x0c I think. But that's two response is what I
get on my first connect. After that when I envoke ser.read() again, I 
get nothing.

Now I've tried sending commands using the Nokia AT Command which was
taken from
this link here -> http://3ton.com/besik/ATNOKIA.pdf

The problem is, this still gives no response, when I type
ser.write('ATCGMI') or any other command, I get no response. Please
correct me if I'm wrong for using the Nokia At command on 3310, I
doubt that this At command applies only to some device of nokia
products like 6110 series, 71.. etc.

Thus, ser.read() would not give me anything. And that's my problem. 

I tried installing SMS application that runs on windows
http://www.oxygensoftware.com/download/O2SMS_AX.zip  because I doubt
that my cell phone is not woking on my sockets. But I found that it
works fine. I've tested that software, sending and receiving messages
and that works very ok.

I hope my explanation gives you enough information. I hope you could
help me with this because I plan to integrate this application on my
online/sms borrowing of books project part of my undergraduate
research.

I also tried writing to Mr. Frank Stajano, but it's already 2 weeks
and still I get no response. Perhaps his busy or doesn't want to talk
about this matter. Frank Stajano had already created an application
like this for his research and I think he even asks for help in this
group. His output an SMS python scripts is also not available.

The Thing that I really want to know is what kind of command will I
use to get some response from the nokia 3310? Is it also a hex
commands in which I will envoke \x00,.. etc commands. Or is it a flat
ASCII file also. I hope you could share some idea which regards to
this. Any idea regarding this would be of great help to me, I hope you
could give me one. Thanks

Polerio


From sonnenbu at informatik.uni-bonn.de  Mon Sep  2 07:37:28 2002
From: sonnenbu at informatik.uni-bonn.de (Frank Sonnenburg)
Date: Mon, 2 Sep 2002 13:37:28 +0200
Subject: wie greife ich mit Python auf die c-bibliotheken zu ?
References: <3D7323BB.4070200@uni-koblenz.de>
Message-ID: <akvi7v$13gc$1@f1node01.rhrz.uni-bonn.de>

You can't access directly C++ classes via python. I have made some efforts
to create a python class and instances of it in C, but up to now not really
successfully.

The more easy and so preferrable way is to define a python class in python
and then modify instances of this class in some extension module function,
passing this instance as an argument. To write extension modules, see
extending and embedding python: http://python.org/doc/current/ext/ext.html

Frank



"Oleg Seifert" <misterxx at uni-koblenz.de> schrieb im Newsbeitrag
news:3D7323BB.4070200 at uni-koblenz.de...
> hallo Leute,
>
> k?nnte vielleicht jemand es mir an einem Beispiel zeigen. Ich habe als
> Attachments zwei kleine Anwendungen dazu geschickt.
>
> danke im Vorraus,
>
> Oleg
>


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


>
> /* Dies ist ein Programm in C */
>
> #include <stdio.h>
>
> class person
> {
>   private:
>     char* name;
>     char* geschlecht;
>   public:
>     void set_name(char* n) { name = n; }
>     void set_geschlecht(char* g)  { geschlecht = g; }
>     char* get_name() {return name;}
>     char* get_geschlecht() {return geschlecht;}
> };
>
>
> void main()
> {
> /*
>   person student;
>   student.set_name("Peter");
>   student.set_geschlecht("mannlich");
>   printf("Name : ");
>   printf(student.get_name());
>   printf("\nGeschlecht : ");
>   printf(student.get_geschlecht());
>   printf("\n<<< ende\n");
> */
> }


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


> # dies ist ein Python-Programm
>
>
> # hier instanzierung der Klasse person aus punkt.cpp
>
> print "<<< ende"
>




From david.abrahams at rcn.com  Fri Sep 13 18:48:22 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Fri, 13 Sep 2002 18:48:22 -0400
Subject: Extending a C++ App
References: <996eade8.0209120912.3a653d7@posting.google.com> <n0qnhvfn.fsf@boeing.com> <996eade8.0209122352.45721938@posting.google.com>
Message-ID: <altqiu$fjr$1@bob.news.rcn.net>

"Bwuce_Wee" <bwucew at yahoo.com> wrote in message
news:996eade8.0209122352.45721938 at posting.google.com...

>     I'm busy investigating Boost.Python...

In that case, get Boost.Python v2 using the instructions at

http://mail.python.org/pipermail/c++-sig/2002-May/001100.html

It's nearing release, and v1 isn't long for this world.

--
-----------------------------------------------------------
           David Abrahams * Boost Consulting
dave at boost-consulting.com * http://www.boost-consulting.com





From duncan-news at grisby.org  Fri Sep  6 15:46:46 2002
From: duncan-news at grisby.org (Duncan Grisby)
Date: Fri, 06 Sep 2002 19:46:46 GMT
Subject: Interapp communication under Python[Linux]?
References: <St5e9.25080$w51.7041171@news2.news.adelphia.net>
Message-ID: <GC7e9.116$hZ1.634@news13-win.server.ntlworld.com>

In article <St5e9.25080$w51.7041171 at news2.news.adelphia.net>,
 Robert Oschler <Oschler at earthlink.net> wrote:

>My Python app will talk to 3 or 4 modules written in Java and C/C++.
>Fortunately I have the source for the Java and C/C++ modules and I have
>experience in both languages.  Currently the external modules have no
>provision for being accessed outside of the process they are running as.  So
>I'm wondering what the best protocol would be used to "glue it all
>together".  There a dizzying array of possibilities:

[...]
>6) CORBA? (yikes!)

CORBA has a reputation for being difficult to use, but it's actually
really easy, and it could well solve all your problems for you. Adding
simple CORBA interfaces to your existing code would add around 50
lines to the C++ and Java code, and about 20 to the Python code, plus
maybe another 10 lines of IDL interface definitions. Obviously if you
need more complex interfaces there will be more code involved, but
minimal support really is minimal.

You can see a simple Python CORBA client and server here:

  http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/81254

Cheers,

Duncan.

-- 
 -- Duncan Grisby         --
  -- duncan at grisby.org     --
   -- http://www.grisby.org --


From oren-py-l at hishome.net  Fri Sep 13 17:20:53 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Fri, 13 Sep 2002 17:20:53 -0400
Subject: wrapping yield ?
In-Reply-To: <Xns92888DEA3BB2duncanrcpcouk@127.0.0.1>
References: <alqehl$9qv$1@nntp0.reith.bbc.co.uk> <Xns92885F082A21Cduncanrcpcouk@127.0.0.1> <alsdpt$2nq$1@nntp0.reith.bbc.co.uk> <Xns9288759449799duncanrcpcouk@127.0.0.1> <alsg06$3e2$1@nntp0.reith.bbc.co.uk> <Xns92888DEA3BB2duncanrcpcouk@127.0.0.1>
Message-ID: <20020913212053.GA12991@hishome.net>

On Fri, Sep 13, 2002 at 01:01:15PM +0000, Duncan Booth wrote:
> Don't forget that iterating with a generator is *much* more efficient than 
> calling a Python functions thousands of times.
> 
> Python has a big overhead on each function call, but you only pay that 
> overhead when you first call a generator and the subsequent iterations have 
> virtually no overhead at all. 

In most programming environments function calls are fast and context
switches are slow. Python is quite unusual in having it the other way 
around - a yield statement causes a context switch that is much faster 
than a function call.

	Oren



From ws at mystrobl.de  Sun Sep 22 14:27:43 2002
From: ws at mystrobl.de (Wolfgang Strobl)
Date: Sun, 22 Sep 2002 20:27:43 +0200
Subject: How to get memory usage on Windows?
References: <amilhk$bi8$1@reader1.tiscali.nl> <amkpo5$kur$1@reader1.tiscali.nl>
Message-ID: <f51sou8o4j9pkilhc3g6jevkqmkk7jr3pl@4ax.com>

"GerritM" <gmuller at worldonline.nl>:

>Searching in books , documentation (a.o. win32api) and net did not yet help
>I found this article on the net:
>http://www.developeriq.com/Magazinestories/02jul29memory.php3
>but how to call this function from Python?

I guess thanks to Mark Hammond you already have everything you need.

Have a look at  \Python22\Lib\site-packages\win32\Lib\win32pdhutil.py,
uncomment the test function, run it. 

I snarfed the  following snippet straight from win32pdhutil.py:

import win32pdh

def GetPerformanceAttributes(object, counter, \
 instance = None, inum=-1, 
 format = win32pdh.PDH_FMT_LONG, machine=None):
    path = win32pdh.MakeCounterPath( (machine,object,\
    instance, None, inum,counter) )
    hq = win32pdh.OpenQuery()
    try:
        hc = win32pdh.AddCounter(hq, path)
        try:
            win32pdh.CollectQueryData(hq)
            type, val = win32pdh.GetFormattedCounterValue(hc, format)
            return val
        finally:
            win32pdh.RemoveCounter(hc)
    finally:
        win32pdh.CloseQuery(hq)

counter=r'\Memory\Committed Bytes'

machine, object, instance, parentInstance, index, counterName = \
		win32pdh.ParseCounterPath(counter)
result = GetPerformanceAttributes(object, counterName, instance, \
index, win32pdh.PDH_FMT_DOUBLE, machine)

print result

>>> 264937472.0


-- 
Wir danken f?r die Beachtung aller Sicherheitsbestimmungen


From nessus at mit.edu  Fri Sep  6 17:45:27 2002
From: nessus at mit.edu (Douglas Alan)
Date: 06 Sep 2002 17:45:27 -0400
Subject: Setting the microphone volume
Message-ID: <lcbs7aeqrs.fsf@gaffa.mit.edu>

Hi.  Does anyone know how to set the microphone volume from Python
running on a Windows machine?  (Snack doesn't support doing this under
Windows.)

|>oug


From op73418 at mail.telepac.pt  Wed Sep  4 10:42:08 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Wed, 04 Sep 2002 15:42:08 +0100
Subject: newbie-question: difference between classmethod and staticmethod in Python2.2
References: <3D7619AE.3050605@snafu.de>
Message-ID: <pb6cnusm8jodc6vab0akaqs6316tu39obu@4ax.com>

On Wed, 04 Sep 2002 16:33:18 +0200, "T. Kaufmann" <merman at snafu.de>
wrote:

>Hi there,
>
>can anybody explain the difference between 'staticmethod' and 'classmethod' in a simple way in 
>Python2.2?

Morally, a static method is just a function as a class attribute, while
a class method is a method that is associated with the class, receiving
it implicitly as the first argument (as opposed to intance methods that
receive the instance as the first argument).

class methods are usually used as factory methods (creating and
returning instances of the class) as for static methods I have never
found much use to them since whatever functionality they provide is best
kept at module level. Someone else might have use cases for them.

>
>When should I use staticmethod or classmethod in classes?
>
>I thought a staticmethod is a classmethod like in C++/ Java?
>
>Thanks for help.
>
>o-o
>
>Thomas

Hope this helps somehow, if it was not sufficiently clear, just holler.
Gon?alo Rodrigues


From grabiller at 3dvf.net  Sat Sep 28 11:55:51 2002
From: grabiller at 3dvf.net (Guy Rabiller)
Date: Sat, 28 Sep 2002 17:55:51 +0200
Subject: Keyboard State
Message-ID: <3d95d107$1$6669$79c14f64@nan-newsreader-02.noos.net>

Hi,

sorry if this is a trivial one, but I'm looking for a way to retrieve the 
keyboard state at anytime.

I'm using pyOpenGL and GLUT.

GLUT doesnt allow to call the glutGetModifiers() fonction outside the 
mouse, keyboard and special callbacks, but I need to get the state of the 
SHIFT and CTRL keys at anytime.

Any idea ?

Thanks in advance.

-- 
guy rabiller
3d animator / td
grabiller at 3dvf.net
http://grabiller.3dvf.net


From phil at river-bank.demon.co.uk  Tue Sep 17 09:40:28 2002
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Tue, 17 Sep 2002 14:40:28 +0100
Subject: PyQT License
In-Reply-To: <27E327AB8201B54495D54105FBBBB1485D205A@SRVMSG.GERLAND2.local>
References: <27E327AB8201B54495D54105FBBBB1485D205A@SRVMSG.GERLAND2.local>
Message-ID: <200209171440.28885.phil@river-bank.demon.co.uk>

On Tuesday 17 September 2002 1:56 pm, Maric MICHAUD wrote:
> I did not really understand terms of Qt and PyQt license.
> Is it GPL for both Win32/UNIX platforms ?

No.

There is a commercial version for Win32 and UNIX that allows you to develop 
and ship closed source applications.

There is a GPL version for UNIX.

Phil



From rolf.kalbermatter at citeng.com  Mon Sep  2 04:51:18 2002
From: rolf.kalbermatter at citeng.com (Rolf Kalbermatter)
Date: Mon, 2 Sep 2002 10:51:18 +0200
Subject: version independant object usage counting
Message-ID: <EKEJLGEFHHFGJOBEGJECMEBLCEAA.rolf.kalbermatter@citeng.com>

Hello there,

I'm fairly new in Python and have attempted to do embedding of the Python
core in
another application on Windows NT and succeeded so far. In doing this work I
found
out that Python clients (applications, modules and such) are always linked
with a
version specific import library which even makes a difference for debug and
non-debug.

In using and programming DLLs until now I was always used to attempt to
remove
version dependances as much as possible and so thought about implementing my
own
Python DLL loader to achive full weak linking to Python and possibly allow a
configuration setting to specify where and which Python core DLL to use.
However
although this loader works quite fine now, I found out that I gained little
with
it because the current Python architecture is not really designed to link
dynamically
to it.

Of course more complicated tools may be very difficult or impossible to keep
version
independant but in my case I just call some standard functions from the
Python core
DLL. The problem arises in fact with the refcounting of Python objects. The
macros
Py_INCREF and Py_DECREF access directly structure members in the object
layout and
this layout is different for debugging and non debugging versions and
potentially
different between Python versions as well. By adding two new API functions
PyObject_AddRef() and PyObject_Release() or whatever name seems appropriate
and
exporting them, it would be possible to write clients which import the
necessary
Python API functions dynamically on there own and link to a specific Python
DLL/shared
library much more independant from the used Python core DLL than it is the
case now.

Is there a specific reason that these two API functions are not available
and one
is forced to use the according macros to do proper refcounting, or did I
miss those
functions completely? I do understand that inline code is a little faster
than calling
a function and I do not propose to remove the current macros at all, but
rather would
like to see two additional functions added which allow to do proper refcount
manipulation
over a functional interface instead of direct object structure layout
access.

Any thought about this?

Rolf Kalbermatter
CIT Engineering Nederland BV    tel: +31 (070) 415 9190
Treubstraat 7H                           fax: +31 (070) 415 9191
2288 EG Rijswijk	http://www.citengineering.com
Netherlands		mailto:rolf.kalbermatter at citeng.com





From mjackson at alumni.caltech.edu  Thu Sep 26 15:27:16 2002
From: mjackson at alumni.caltech.edu (Mark Jackson)
Date: 26 Sep 2002 19:27:16 GMT
Subject: How do you pronounce "__" (double underscore)?
References: <525eaba.0209260956.72bf6786@posting.google.com> <3D934C5C.4000603@mxm.dk> <20020926.130938.1789366143.1524@cybermesa.com>
Message-ID: <amvn2k$gd1$1@news.wrc.xerox.com>

"Jay O'Connor" <joconnor at cybermesa.com> writes:
> In article <3D934C5C.4000603 at mxm.dk>, "Max M" <maxm at mxm.dk> wrote:
> 
> > Pat Notz wrote:
> >> @ = "at"
> >> ! = "bang"
> >> . = "dot"
> >> * = "star"
> >> 
> >> Is there a short (one or two syllable) pronounciation for "__" (double
> >> underscore).  "Underscore underscore" and "double underscore" are kind
> >> of long.  Somthing like "splat"?  Anyone?
> > 
> > _ = 'under'
> > 
> > __ = 'under-under'
> 
> 
> How about "underder"??

How about "dunder" (double under)?

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	Yes, but what if it were a parrot?
			- Graham Chapman, to John Cleese




From lpgray at uop.com  Wed Sep 18 15:24:48 2002
From: lpgray at uop.com (Lee Gray)
Date: 18 Sep 2002 12:24:48 -0700
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <3D886F3D.9AE251F9@unice.fr>
Message-ID: <96904b50.0209181124.7eb93f4b@posting.google.com>

Jerome Alet <alet at unice.fr> wrote in message news:<3D886F3D.9AE251F9 at unice.fr>...
 
> Perhaps this is really silly, but why don't you show Zope to your boss ?
> 
> http://www.zope.org/
> 
> not really a GUI, though
> 
> hth
> 
> Jerome Alet

I've heard of, but know nothing about Zope.  I don't think it's really
applicable.  I need a client GUI app to have him get the feel of
what's possible; and the back-end (Oracle DB) isn't going to change.

Thanks though,
Lee


From fb at frank-buss.de  Tue Sep 17 04:53:17 2002
From: fb at frank-buss.de (Frank Buss)
Date: Tue, 17 Sep 2002 08:53:17 +0000 (UTC)
Subject: question about dictionary type..
References: <alt22f$pf8$1@newsreader.mailgate.org> <alv4op$m67$1@newsreader2.netcologne.de> <61Jg9.16334$5r1.665293@bin5.nnrp.aus1.giganews.com> <am0390$cvf$1@newsreader.mailgate.org> <am09qg$78d$1@newsreader2.netcologne.de> <am6k2q$edf$2@newsreader.mailgate.org>
Message-ID: <am6qht$95k$1@newsreader2.netcologne.de>

eugene kim <eugene1977 at hotmail.com> wrote:

> i read a python book while ago, and trying to refamiliaze myself with it 
> recently, your solution looks so cool..

Thanks, I'm not a Python expert, because I wrote my first line of Python 
code one month ago (but I already knew some other languages), so there are 
probably a better solution for your problem using Python. But for me it 
was a good idea to help in this newsgroup for learning a bit more of 
Python :-)

> have to spend some time what u did =)

The add-funtion is obvious. So looking at the main program:

records = re.compile('\((.*?)\)', re.DOTALL).findall(stdin.read())
for record in records:
  fields = map(lambda w: upper(w), re.findall("'(.*)'", record))
  add(tree, (fields[0],fields[2],fields[3]),0)
  add(tree, (fields[1],fields[2],fields[3]),0)

"stdin.read()" reads all from stdin. ".*" matches any chars (zero or more 
occurence). It is written within a group (the '(' and ')'), so findall can 
use the matching result. Because your records are surrounded by '(' and 
')' I wrote '\(' and '\)'. The "?" is needed to avoid greedy matching. 
Without "?" only one record would be matched, starting from the first "(" 
until the last ")" in the whole file. "re.DOTALL" says, that "." should 
match newslines, too.

Next interesting expression is the map-line. "'(.*)'" matches anything 
between two '. It doesn't need to be non-greedy ("?"), because the line 
boundarys are not matched (no re.DOTALL). The result is a list of all 
fields. "map" iterates through all elements of the resulting list, applys 
the function "upper" through the lambda-inline-function to each element 
and returns a new list with the upper-cased elements.

-- 
Frank Bu?, fb at frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de


From rpm at wag.caltech.edu  Wed Sep  4 13:38:01 2002
From: rpm at wag.caltech.edu (Rick Muller)
Date: Wed, 04 Sep 2002 10:38:01 -0700
Subject: Can't exponentiate zero in NumPy
Message-ID: <al5gdq$18g$1@naig.caltech.edu>

I'm hoping someone can help me track down a potential bug that one of my 
users is reporting.

This user claims that exponentiating zero in Numeric raises an 
exception. It isn't clear to me whether the violating code is a single 
value or an array of zeros. That is

from Numeric import *
a = zeros(3,Float)
b = 0
print a**(4./3.)
print b**(4./3.)

I have not been able to get back in contact with this user to double 
check how and when this fails, and everything works on all of the 
machines I have at my disposal here.

Has anyone seen this before? Can anyone verify that it fails? Can anyone 
suggest an elegant workaround for the array case?

Thanks in advance,

Rick



From 2002 at weholt.org  Mon Sep 16 18:14:44 2002
From: 2002 at weholt.org (Thomas Weholt)
Date: Mon, 16 Sep 2002 22:14:44 GMT
Subject: Best way ( easiest ) to generate RSS/RDF on MySQL-data ??
Message-ID: <oJsh9.18870$0p1.344943@news2.ulv.nextra.no>

I got data in a MySQL-database I want to base my RSS/RDF-feeds on. The
columns in the given tables doesn't have fields that correspond to RSS or
RDF tags so I need to do some "mapping" of table columns into RSS/RDF-tags.

Is there a module or piece of code out there that shows a dynamic and
customizable way of mapping these kinds of data to RDF/RSS or even another
XML-based kind of document ?

I've read lots about parsing RSS/RDF but found very few resources on
generating those files, -- in any programming language. It's not that the
format is hard to code doing it the fast and dirty way, but that seem a bit
restricted and by using RDF I want to be able to let users define custom
namespaces etc. too. That's way I need a pretty dynamic method of generating
these feeds.

Any hints or clues on how to generate RSS/RDF-files, not even related to
this problem in particular would be highly appreciated.

Best regards,
Thomas Weholt




From emile at fenx.com  Sun Sep 22 10:42:29 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 22 Sep 2002 07:42:29 -0700
Subject: COM, Word problem
References: <e00ba67.0209211751.1411ad7c@posting.google.com>
Message-ID: <amkkq2$hqr$1@main.gmane.org>

strat_addict:
> com_error: (-2147417846, 'The message filter indicated that the
> application is busy.', None, None)

Google is your friend... perhaps this helps:

http://groups.google.com/groups?selm=MPG.13553ec91f8cb3b98968e%40192.168
.0.28

--

Emile van Sebille
emile at fenx.com

---------






From DennisR at dair.com  Tue Sep 17 10:59:32 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Tue, 17 Sep 2002 14:59:32 GMT
Subject: using python as cgi on apache unix
References: <am6fm301477@enews2.newsguy.com>
Message-ID: <orHh9.167$UJ1.17889587@newssvr14.news.prodigy.com>

> ... want to move this to my unix server.  Unfortunately, no matter what I
> do I can not get apache on my linux box to execute any python scripts.  I
> get the following errors in apache:
> Premature end of script headers: /var/www/html/cgi-bin/test.py

What do you get when you type "test.py" at the shell prompt when logged into
the cgi directory on the RH box at the command line?  Do you get:
---
---
Content-Type: text/html
---
hello

where "---" is taken as a blank line?  I agree with Gerhard that the initial
print "\n\n" is wrong.  But your comment trying the cgi.test suggests that
something else is going on.  The above test may sniff out that something
else.


Dennis Reinhardt

http://www.dair.com




From martin at v.loewis.de  Mon Sep 16 12:42:08 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Sep 2002 18:42:08 +0200
Subject: Py_CompileString and line numbers?
References: <slrnaoc065.9ki.sa@gamegirl.hogia.net>
Message-ID: <m31y7tevj3.fsf@mira.informatik.hu-berlin.de>

Sebastian Andersson <sa at hogia.net> writes:

> I'm parsing a config file and from there I get embedded python code
> which I send to Py_CompileString, with the config file's filename given
> as filename argument to Py_CompileString. Later, when errors are found
> in the python code, the wrong line numbers are reported. Is there
> a way to tell Py_CompileString an offset for the first line that
> I give to it?

No. You'll have to instantiate a tokenizer yourself, and set its
lineno member.

HTH,
Martin



From news.but.NO.SPAM at datatailors.com  Mon Sep 23 07:39:48 2002
From: news.but.NO.SPAM at datatailors.com (Peter van Kampen)
Date: Mon, 23 Sep 2002 13:39:48 +0200
Subject: Cookbook example, DAO error
Message-ID: <slrnaotvc2.4hm.news@datatailors.com>

Hi,

I am trying to run the jet2sql script from the Python cookbook but it
fails. I use win2K with activepython (latest version). I have tried it
on two computers both with the samen result. I have narrowed it down to
a failure to enumerate the Indexes collection. Here's the (relevant part
of) the traceback msg (in pythonwin):

  File "I:\d\Python\jet2sql.py", line 128, in getPrimaryKey
      for idx in currTabl.Indexes:
  File "c:\python22\lib\site-packages\win32com\gen_py\ 
00025E01-0000-0000-C000-000000000046x0x4x0.py" , line 956, in __getitem__
      return self._enum_.__getitem__(index)
  File "C:\Python22\Lib\site-packages\win32com\client\util.py", line 37, in __getitem__
      return self.__GetIndex(index)
  File "C:\Python22\Lib\site-packages\win32com\client\util.py", line 53, in __GetIndex
      result = self._oleobj_.Next(1)
com_error: (-2146825178, 'OLE error 0x800a0c26', None, None)

If I run from the commandline I  get

[same output]
pywintypes.com_error: (-2146825178, 'OLE error 0x800a0c26', None, None)

Has anyone else seem this problem, or does anyone have any idea how
solve it?

TIA,

PterK


From DeepBleu at DeepBleu.org  Thu Sep  5 21:11:59 2002
From: DeepBleu at DeepBleu.org (DeepBleu)
Date: Fri, 06 Sep 2002 01:11:59 GMT
Subject: upgrading python
References: <3D777EA9.F9A96C83@ipm.fhg.de>
Message-ID: <zhTd9.406390$q53.13810121@twister.austin.rr.com>

The problem is how do you get such an installer to work?  Also, how would
the installer know of the different packages?  There are literally hundreds
of packages.  How about the mxODBC package?  Scientifc?  Real?
Having said that - I for one have always run into such issues (please note
that I am using the word issues, not problems).  I just keep track of what
packages I am using and only upgrade when the need arises.  It is much
better to keep the Python 'generic' distribution separate from anything
else.  Then download an duse the packages that you need.
DeepBleu

"Markus von Ehr" <markus.vonehr at ipm.fhg.de> wrote in message
news:3D777EA9.F9A96C83 at ipm.fhg.de...
> Hi,
>
> first of all I'd like programming in python.
> Sometimes you need to grade up to a newer version.
> I have to download everything again and install every single package,
> for example:
>
> -win32all
> -PyOpengl
> -Pyserial
> -Numeric
> -unzip Pmw
> -PIL
> -py2exe
> -PyXML
> etc.
>
> So I just mentioned "only" 8 packages, additionally I have to
> install the newer python version and to uninstall the old packages...
>
> Couldn't it be made easier? Maybe I do something wrong and
> am misunderstanding how to deal with it.
>
> Now I wanted to grade up to tk84. It was quite hard to find
> _tkinter.c, then I downloaded the tcl-sources, tk-sources,
> installed MSVC6, found out that some X11 stuff is missing...
>
> I know that all mentioned packages are different projects but could
> it be made that there's a python distribution with all packages
> up2date and the installer is asking for the desired ones?
>
> Do you have similar experiences?
>
> Markus




From frogger974 at yahoo.com  Thu Sep  5 22:15:04 2002
From: frogger974 at yahoo.com (Matt Rickard)
Date: Thu, 05 Sep 2002 22:15:04 -0400
Subject: pygtk-2 tutorial/doc?
Message-ID: <pan.2002.09.06.02.15.03.816977.8282@yahoo.com>

I'm trying to do some gtk-2 development with python, but I really can't
find a good place to begin.  All the tutorials I've found (python.org)
are based off of old versions of python (1.x) and old versions of gtk
(1.x).  

I'd like to find some kind of tutorial or at least some introductory
documentation on programming with pygtk-2 with python 2.2.

Anyone have any ideas on where I should be looking?

Thanks.


From igetenoughspamalreadythanksjlr at doctor.com  Sun Sep  1 18:21:32 2002
From: igetenoughspamalreadythanksjlr at doctor.com (John La Rooy)
Date: Mon, 02 Sep 2002 08:21:32 +1000
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <mailman.1030907703.26266.python-list@python.org> <akuifb$h0e$2@solaris.cc.vt.edu>
Message-ID: <1030962064.628232@ampungk.ozonline.com.au>

Carl Banks wrote:
> Gerhard H?ring wrote:
> 
>>>which means the real time is not spent in the compile() function, but
>>>in the match or find function. So basically, couldn't one come up with
>>>a *human readable* syntax for re, and compile that instead?
>>
>>That's equally powerful? Most probably not.
> 
> 
> Why not?  It won't be as fast, but it should be able to do anything a
> regexp can do, and would be much more versatile.
> 
>
I think the main problem is that *human readable* doesn't map really 
well onto regular expressions.

What would the equivalent of r"(.)(.)(.)\3\2\1"
This means a "palindrome of 6 characters"
But it is unlikely that the human readable processor would understand 
that (isn't it??)

It would be more likely to look like this (I haven't put too much 
thought into this)
"anything,anything,anything,same_as_3rd,same_as_2nd,same_as_1st"
or would you like to suggest something else?

palindrome_6 = re.compile(r"(.)(.)(.)\3\2\1")
palindrome_6 = 
re.compile("anything,anything,anything,same_as_3rd,same_as_2nd,same_as_1st")

Sure there are some cases where the re is loaded with meta characters...

hmmm
OK is this about writing maintainable code or people not wanting to 
learn all the ins and outs of re's?

John



From Oschler at earthlink.net  Sat Sep 14 23:24:00 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sun, 15 Sep 2002 03:24:00 GMT
Subject: Will xml.dom work with HTML docs?
Message-ID: <k3Tg9.6083$Lg2.908752@news2.news.adelphia.net>

Will the xml document parser module work with plain old HTML docs, if not is
there a good parser module for HTML docs?

thx





From fredrik at pythonware.com  Fri Sep  6 02:29:59 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 06 Sep 2002 06:29:59 GMT
Subject: about parse a link
References: <cVUd9.6962$yt3.3362358@newssrv26.news.prodigy.com>
Message-ID: <HXXd9.9644$HY3.2204710@newsc.telia.net>

"koko" wrote:

> if I have extracted the links on the page:
> e.g:
>
> http://www.uic.edu/index.htm
> on this page: there are
> a.htm
> b.htm
> c.htm
> http://www.uic.edu/home/e.htm
>
> how can I log the a.htm, b.htm, c.htm with the full web address?

    base = "http://www.uic.edu/index.htm"

    url_list = [
        "a.htm",
        "b.htm",
        "c.htm",
        "http://www.uic.edu/home/e.htm"
    ]

    import urlparse

    for url in url_list:
        print urlparse.urljoin(base, url)

prints

    http://www.uic.edu/a.htm
    http://www.uic.edu/b.htm
    http://www.uic.edu/c.htm
    http://www.uic.edu/home/e.htm

</F>

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




From jb at cascade-sys.com  Sun Sep 15 03:31:48 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sun, 15 Sep 2002 00:31:48 -0700
Subject: How to represent sets
References: <alvd8p$1sq$1@news.hccnet.nl>
Message-ID: <3D843764.8030208@cascade-sys.com>

Anton Vredegoor wrote:

>I would like to reopen this discussion.
>
I think BOTH approaches deserve to be in the library.

For dense sets isomorphic with integers, the long integer implementation 
probably will be superior.

For other, more general applications the dictionary approach may be better.

Which is "better" is not a blanket decision we can make in advance.

Include them both and let the user decide.

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From anton at vredegoor.doge.nl  Fri Sep 27 10:12:05 2002
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Fri, 27 Sep 2002 16:12:05 +0200
Subject: watching mutables?
Message-ID: <an1p3c$jrd$1@news.hccnet.nl>

Hello all,

here's a function I would like to have:

<hypothetical code>

>>> m = [1,2]
>>> m
>>> [1,2]
>>> from watcher import watch
>>> watch(m)
>>> m.append(3)
>>> "hey, you changed variable m!, new value is [1,2,3]"

</hypothetical code>

Is this possible? How would this be called?

Anton.


From steve at ferg.org  Fri Sep 13 08:58:00 2002
From: steve at ferg.org (Stephen Ferg)
Date: 13 Sep 2002 05:58:00 -0700
Subject: Newbie Looking for direction
References: <uo23o95eb8tg87@corp.supernews.com>
Message-ID: <b16e4ef7.0209130457.347a8137@posting.google.com>

Some thoughts...

Python is a great general-purpose language.  If your aim is to develop
just database-type applications, using a DBMS that everyone else at
your workplace has, then you are probably better sticking with one of
your database packages.

On the other hand, if you really want to get a taste of a real
general-purpose programming language, one that is not tied to a DBMS,
then Python is an excellent choice.

In my experience, most of the things I do have nothing to do with
databases.  I write a lot of little utilities, one-time quick and
dirty programs, etc.  All the stuff that you'd like to be able to
automate easily.  And for that, you need a programming (or scripting)
language like Python, just the way you (once) needed batch files or
shell scripts or JCL or whatever.  It is something every computer-user
needs.

So I would recommend that you continue to learn and use Python.  But
at the outset, you might want to avoid trying to do database
programming with it.  Play with the language a while, write some
little utilities in it, get a feel for the language.  Once you've done
that, you can decide whether you want to go on and do database
programming with it.  But of course, by that time, you'll also find
that Python has become an invaluable tool in your toolbox, whether or
not you decide to do database programming with it.  And you'll have
found that your concept of what you can do with a programming language
has become much broader than it was when you were programming only
with languages tied to a DBMS.


From bbrelin at yahoo.com  Tue Sep  3 16:31:10 2002
From: bbrelin at yahoo.com (Braun Brelin)
Date: 3 Sep 2002 13:31:10 -0700
Subject: What's this error mean???
Message-ID: <e0a78451.0209031231.1d079c38@posting.google.com>

Hello,

I'm running Zope-2.5.1 on a RedHat Linux 7.3 machine.  I'm trying to
install the ParsedXML Product and I'm getting the following error:

2002-09-03T18:49:49 ERROR(200) Zope Couldn't install ParsedXML
Traceback (innermost last):
  File /usr/local/Zope-2.5.1-linux2-x86/lib/python/OFS/Application.py,
line 637, in install_product
    (Object: Zope)
    (Info: ParsedXML)
  File /usr/local/zope/lib/python/Products/ParsedXML/__init__.py, line
92, in initialize
AttributeError: 'Products.ParsedXML.ParsedXML' module has no attribute
'ParsedXML'

I installed the PyXML product via RPM.  Can anyone tell me what the
error means and how to fix this?

Thanks,

Braun Brelin
bbrelin at yahoo.com


From kwdotbenninkNOSPAM at 12move.nl  Fri Sep 20 09:27:56 2002
From: kwdotbenninkNOSPAM at 12move.nl (Highwaykind)
Date: Fri, 20 Sep 2002 15:27:56 +0200
Subject: Newbia alert .. working my way through Easy Python Montana Edu page, need help..
Message-ID: <MPG.17f540bc75626c66989683@news.worldonline.nl>

I'm asked ( on the website : www.honors.montana.edu\_jjc\easytut
\node10.html ) to write a small programm to calculate rectangle, square 
and circle thingies( forgot the English word .. ) and I came up with 
this code : 

def print_options() :
    print "options:"
    print " 'p' print options"
    print " 'r' Calculate area rectangle"
    print " 's' Calculate area square "
    print " 'c' Calculate area circle "
    print " 'q' Quit"


def rect (width,height):
    return width*height
def square ( width1, height1):
    return width1*height1
def circle (3.14*radius**2):
    return 3.14*radius **2


choice = 'p'
while choice != 'q':
    if choice == 'r':
        a=input ("width")
        b=input ("height")
    print "result",rect (width,height)
    if choice =='s':
        c = input ( "width")
        d = input ( "height")
    print "result",square (width1, height1)
    if choice == 'c ':
        f=input ("radius")
    print "result",circle
    if choice != 'q':
        options()
        choice=raw_input ("option")

But the thing keeps telling me width is not defined, also I can;t use 
the ELIF command .. THink I could do with some help :)

Corien
-- 
For the correct e-mail address : remove NOSPAM and replace dot with .


From op73418 at mail.telepac.pt  Wed Sep  4 10:29:28 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Wed, 04 Sep 2002 15:29:28 +0100
Subject: Why read-only nested scopes?
Message-ID: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com>

Hi

Why is it that we cannot rebind names in outer scopes from within the
inner scope? This is perplexing to me, because you can rebind global
scope variables using the global directive, e.g.

x = whatever

def somefunction(some_args):
    global x
    x = another_whatever

So why hasn't this been extended when nested scopes were introduced,
e.g. by reusing the global directive or some other one? Implementation
issues? Or is this line of development is frowned upon by the language
designers, e.g. the BDFL? I am well aware that this opens up a can of
worms but the global directive does that already, so...

Thanx in advance for any enlightening comments,
Gon?alo Rodrigues


From thoa0025 at mail.usyd.edu.au  Tue Sep 17 00:01:16 2002
From: thoa0025 at mail.usyd.edu.au (:B nerdy)
Date: Tue, 17 Sep 2002 04:01:16 GMT
Subject: raw post data
References: <Wt8h9.9583$Ee4.22928@news-server.bigpond.net.au> <w_fh9.113093$pX1.4054147@news2.tin.it>
Message-ID: <gOxh9.13698$Ee4.40115@news-server.bigpond.net.au>

how could i make it so when i read the stdin it still is there.. cause i use
cgi.FormVariables() also

cheers

"Alex Martelli" <aleax at aleax.it> wrote in message
news:w_fh9.113093$pX1.4054147 at news2.tin.it...
> :B nerdy wrote:
>
> > how would i retrieve the raw post data in cgi python scripts ?
>
> I think something like:
>
> import sys
> all_posted_lines = sys.stdin.readlines()
>
> should do it.  Raw POST'ed data should be no more and no less
> than the contents of the CGI script's standard-input.
>
>
> Alex
>




From loewis at informatik.hu-berlin.de  Tue Sep 17 03:54:24 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 17 Sep 2002 09:54:24 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032245506.8580.python-list@python.org>
Message-ID: <j4n0qhjbkf.fsf@informatik.hu-berlin.de>

Ian Bicking <ianb at colorstudy.com> writes:

> Was there some indication that Stackless would be integrated if the code
> was cleaned up, documented, or otherwise simplified?  I don't believe
> there was, but I may be wrong.  

I certainly meant it that way in

http://groups.google.de/groups?q=g:thl2323460570d&dq=&hl=de&lr=&ie=UTF-8&oe=UTF-8&selm=j4d72xasx0.fsf%40informatik.hu-berlin.de

> Without such an indication, there is no reason to think that mere
> technical changes to Stackless will lead to it being included in
> standard Python.

This is not how free software works. Most contributors don't ask for
an advance blanket incorporation guarantee, and nobody would ever give
one, for any kind of change. Instead, they just contribute the code
they find useful, and then are perhaps told to modify it. If they want
to make sure that the feature itself is undebated, they need to write
a PEP.

There are many features in Python that most maintainers won't care
about (for example, I have no interest in IMAP). That hasn't stopped
people to contribute those features, and hasn't stopped maintainers to
incorporate them.

Contributing needs patience, a skill that contributors sometimes
lack. To give a few examples:

- the IPv6 patches took about 15 months to integrate. For many months,
  the patches were unreviewed, because nobody thought to have the
  expertise to review them. I eventually tought myself the necessary
  background to review the patch, after which about 4 further
  iterations where necessary before the patch was installed, and about
  10 changes after the patch was installed. 

- To build Python as a shared library on Unix, about 4 patches have
  been contributed over several years. Each contributor was told what
  the requirements for such a patch were (must be a compile-time
  option, must work uniformly on many systems, etc), most contributors
  abandoned their patch after learning these requirements. Only the
  most recent such patch could be incorporated, after a few
  iterations.

Both patches were relatively small, compared to the old Stackless
implementation.

Since integrating Stackless will require a significant amount of time
on the side of the Stackless authors, some contributor must step
forward and be willing to guide the entire integration period,
preferably for two Python releases after the patch has been
incorporated. I don't think anybody was willing to contribute that
much time in the past.

Regards,
Martin


From bob at eleaston.com  Sat Sep  7 06:32:24 2002
From: bob at eleaston.com (Bob Easton)
Date: Sat, 7 Sep 2002 06:32:24 -0400
Subject: Interrupting Python
References: <mailman.1031345649.9840.python-list@python.org> <Ef9e9.7309$rU2.218110@news4.srv.hcvlny.cv.net> <albel2$7te$0@216.39.172.122> <R3ce9.9745$rU2.313350@news4.srv.hcvlny.cv.net> <3d796d22@news.sentex.net>
Message-ID: <OAke9.17281$rU2.525682@news4.srv.hcvlny.cv.net>

Yes Peter, it makes sense that PythonWin works that way by design. You are
also right about running from the command prompt after development/debug is
done.

For this particular program, using a file semaphore is ideal since I can
place it at the right point in the program to enable easy restart, something
not easily done with the random ctrl-break.

Thanks to all who helped clear the fog for me.

--
Bob Easton


"Peter Hansen" <peter at engcorp.com> wrote in message
news:3d796d22 at news.sentex.net...
> Bob Easton wrote:
> > Using catchCtlC.py in a command line environment works as you expect.
When I
> > hit Ctrl-Break, Windows displays "^C" and then a command prompt.  If I
hit
> > Ctrl-C, the exception handler catches it and displays the results as
> > specified in catchCtlC/py
> >
> > However, my preferred execution environment is PythonWin, not a command
> > prompt. There, neither Ctrl-break, not Ctrl-C have any effect.  That
could
> > easily be a bug with PythonWin.
>
> Not a bug, I think.  The whole Ctrl-C/Ctrl-Break concept is only
> meaningful for command-line ("console") programs, not GUI apps.
> I don't use PythonWin, but my guess is it does not bother passing
> ctrl-c on to the application, possibly by design.
>
> Can't you just run it under the console?  What advantages do
> you find by running it under PythonWin other than during
> development?
>
> Also, have you looked at the msvcrt module and kbhit() ?
>
> -Peter
>




From pan-newsreader at thomas-guettler.de  Sat Sep 14 06:23:44 2002
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Sat, 14 Sep 2002 12:23:44 +0200
Subject: Skipping Parts of a For Loop
Message-ID: <pan.2002.09.14.12.23.44.103401.723@thomas-guettler.de>

Hi!

Is there a way to skip some parts in a for loop?


for i in range(3):
	print "i before", i
	i=2
	print "i after", i

#Results in:

i before 0
i after 2
i before 1
i after 2
i before 2
i after 2

Is there a way to skip some parts?

"continue" skipps one element. But how can
I skip several?

 thomas


From rolf.kalbermatter at citeng.com  Mon Sep  9 04:32:43 2002
From: rolf.kalbermatter at citeng.com (Rolf Kalbermatter)
Date: Mon, 9 Sep 2002 10:32:43 +0200
Subject: Embedding - Local dictionaries for local scripts
In-Reply-To: <j4k7lvbnz2.fsf@informatik.hu-berlin.de>
Message-ID: <EKEJLGEFHHFGJOBEGJECEEKKCEAA.rolf.kalbermatter@citeng.com>

Maybe your problem is more basic about the actual context to create, to use
such Python functions just as I had in the beginning. Then something like
this
might work for you. I have split the actual compile and execute step because
my original project has some more possibilities than this. But that should
make no difference. Also you should think about proper PyThreadState
implementation and protection if your host application is multithreading and
could possibly call the Python core from different threads.  This has been
left out here for simplicity.

typedef struct tagSESS {
	int32	magic;
	PySession next;
	char *text;
	PyObject *co;
	PyObject *dl;
	PyThreadState *tstate;
} PySessionRec;

int PySetScript(PySession session, char *lpScriptText) {
	if (session) {
		int32 len = strlen(lpScriptText);

		if (session->text)
			free(session->text);
		session->text = (char*)malloc(len + 1);
		MoveBlock(lpScriptText, session->text, len);
		if (session->co) {
			Py_DECREF(session->co);
			session->co = NULL;
		}
		return TRUE;
	}
	return FALSE;
}
int PySetData(PySession session, char *name, void* data, int16*
typedescriptor) {
	if (session) {
		if (!session->dl) {
			/* If no local variable dictionary create new one */
			session->dl = PyDict_New();
		}
		if (session->dl) {
			/* Convert our data into a Python object */
			PyObject *value = PyCreateData(data, typedescriptor);
			if (value) {
				/* Add the input variable to the local variable dictionary */
				PyDict_SetItemString(session->dl, name, value);
				Py_DECREF(value);
				/* Return success/failure */
				success = PyErr_Occurred() ? FALSE : TRUE;
			}
		}
	}
	return FALSE;
}

int PyExecuteScript(PySession session, int *eStart, int *eEnd, char *eText)
{
	if (session) {
		if (!session->co) {
			/* Compile the script */
			session->co = Py_CompileStringFlags(session->text, "<host>",
Py_file_input, 0);
		}
		if (session->co) {
			PyObject *v, *m, *dg;

			/* Extract the global dictionary object */
			if (!(m = PyImport_AddModule("__main__")))
				goto errOut;
			if (!(dg = PyModule_GetDict(m))) {
				goto errOut;
			}
			Py_INCREF(dg);
			if (PyDict_GetItemString(dg, "__builtins__") == NULL) {
				if (PyDict_SetItemString(dg, "__builtins__", PyEval_GetBuiltins()) != 0)
{
					Py_DECREF(dg);
					goto errOut;
				}
			}
			/* Execute the script, passing in our arguments as local variables */
			if (v = PyEval_EvalCode((PyCodeObject *)session->co, dg, session->dl)) {
				Py_DECREF(v);
			}
			Py_DECREF(dg);
		}
		/* Extract eventual error information if any */
		return PyErrorInfo(eStart, eEnd, eText);
	}
errOut:
	return FALSE;
}

int PyGetData(PySession session, char *name, void* data, int16*
typedescriptor) {
	if (session) {
		if (session->dl) {
			PyObject *value;

			/* Retrieve variable value */
			if (value = PyDict_GetItemString(session->dl, name)) {
				/* Value is borrowed reference from dict */
				return PyConvertData(value, data, typedescriptor);
			}
		}
	}
	return FALSE;
}

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Martin v. Lowis
> Sent: Monday, September 09, 2002 9:51 AM
> To: python-list at python.org
> Subject: Re: Embedding - Local dictionaries for local scripts
>
>
> chris.walsh at real-sense.com (Chris Walsh) writes:
>
> >     PyObject* PyRun_String(
> >         char *str,
> >         int start,
> >         PyObject *globals,
> >         PyObject *locals);
> >
> > I have seen some stuff in comp.lang.python about creating a dict and
> > adding the __builtin__ module but I have been searching now for an
> > hour and cannot find it.
> >
> > Can anyone point me to some code which creates this dictionary?
>
> The globals/locals dictionary is not really different from any other
> dictionary; you create them with PyDict_New. Python should
> transparently add an __builtins__ into the globals dictionary if there
> is none, but you can copy the __builtins__ entry just as well yourself
> (via PyDict_SetItemString).
>
> HTH,
> Martin
>
> --
> http://mail.python.org/mailman/listinfo/python-list




From oren-py-l at hishome.net  Wed Sep 18 02:08:23 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Wed, 18 Sep 2002 02:08:23 -0400
Subject: Stackless Python, eventual merge?
In-Reply-To: <m31y7rj49r.fsf@mira.informatik.hu-berlin.de>
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de>
Message-ID: <20020918060823.GB81158@hishome.net>

On Wed, Sep 18, 2002 at 06:44:16AM +0200, Martin v. Loewis wrote:
> Greg Ewing <see_reply_address at something.invalid> writes:
> 
> > Also, I believe that, even on the supported
> > architectures, it plays dangerous tricks with
> > the C stack. I'd prefer to see that fixed
> > before making it a standard part of the
> > distribution.
> 
> They aren't really dangerous, just unconventional. And that cannot be
> fixed, as it is essential to operation of Stackless.

They may be dangerous in some circumstances. IIRC, if a C extension calls
the Python interpreter recursively with pointers to things on its stack
it can cause memory corruption. 

	Oren




From duncan at NOSPAMrcp.co.uk  Tue Sep 24 05:58:14 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 24 Sep 2002 09:58:14 +0000 (UTC)
Subject: instance creation
References: <265e380f.0209240052.4c2988f3@posting.google.com>
Message-ID: <Xns92936D7398E42duncanrcpcouk@127.0.0.1>

paulmg at digitalbrain.com (Paul MG) wrote in 
news:265e380f.0209240052.4c2988f3 at posting.google.com:

> - Having two constructors seems to be illegal - or at least the second
> overwrites the first.
> - Assigning to 'self' in the load method seems to really not work!
> - Putting the create() method outside of the class works fine - but I
> want it to be a class method on Appointment, goddammit!:) Is there no
> way to do this?
> 
> Any help would be appreciated, basically i am looking for the python
> idiom to do this, as my C++ and Smalltalk idiom's clearly aren't in
> the grain of the language!
> 

First off, you want to use a single function and check what arguments were 
passed. For example:

    def __init__(self, name=None, details=None, id=None):
    	 if id is not None:
          assert name==details==None
          i = open("appointment"+str(id))
          tmp = pickle.load(i)
          self.name, self.details = tmp.name, tmp.details
      else:
          self.name = name
          self.details = details

Then you can call the constructor using keyword arguments as appropriate:
    x = Appointment(id='49')
    y = Appointment('Me', 'Something')
    y = Appointment(name='Me', details='Something') # Same as above

You could avoid using keyword arguments and just count the number of 
arguments passed in if you preferred.

If you want to get rid of the extra object and copying of name and details 
you could just pickle the individual fields instead of pickling the whole 
object, or you could use the __new__ method.

__init__ is called when the object already exists to complete its 
initialisation. __new__ is called before the object is created and allows 
you to modify the actual creation of the object. You have to subclass a 
type such as object for __new__ to work:

class Appointment(object):
    def __new__(cls, name=None, details=None, id=None):
      if id is not None:
          assert name==details==None
          i = open("appointment"+str(id))
          return pickle.load(i)
      else:
          self = object.__new__(cls)
          self.name = name
          self.details = details
          return self

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From duncan at NOSPAMrcp.co.uk  Wed Sep 11 04:32:54 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Wed, 11 Sep 2002 08:32:54 +0000 (UTC)
Subject: dynamism
References: <66715c02.0209091923.265fe77e@posting.google.com> <Xns92855D48384F9duncanrcpcouk@127.0.0.1> <3d7e209a_1@news.iprimus.com.au> <Xns92865CE1C916Aduncanrcpcouk@127.0.0.1>
Message-ID: <Xns9286606D44C35duncanrcpcouk@127.0.0.1>

Duncan Booth <duncan at NOSPAMrcp.co.uk> wrote in 
news:Xns92865CE1C916Aduncanrcpcouk at 127.0.0.1:
> The code below gives these timings on my machine:
> 3.84s: 1000000 x New()
> 3.89s: 1000000 x New2()
> 2.39s: 1000000 x Old()
> 5.19s: 10000 x attributes(<class '__main__.New'>)
> 5.81s: 10000 x attributes(<class '__main__.New2'>)
> 4.17s: 10000 x attributes(__main__.Old)

Sorry about the bad form of following up to my own post, but I thought I 
had better add that the above timings are on Python 2.2.1 and I forgot to 
specify -O.

The optimised times for 2.2.1 are:
3.77s: 1000000 x New()
3.81s: 1000000 x New2()
2.33s: 1000000 x Old()
4.94s: 10000 x attributes(<class '__main__.New'>)
5.56s: 10000 x attributes(<class '__main__.New2'>)
3.92s: 10000 x attributes(__main__.Old)

 Using the latest CVS build I get rather different timings:

0.94s: 1000000 x New()
1.05s: 1000000 x New2()
1.39s: 1000000 x Old()
3.52s: 10000 x attributes(<class '__main__.New'>)
4.36s: 10000 x attributes(<class '__main__.New2'>)
3.05s: 10000 x attributes(__main__.Old)

So it appears that creating object-derived classes is now faster than 
creating old-style classes, although both tests are much faster than in 
2.2. Setting attributes is still somewhat slower. Oh, and the -O flag on 
the CVS build has no measurable effect.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From b.maryniuk at forbis.lt  Mon Sep  2 05:07:25 2002
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Mon, 2 Sep 2002 11:07:25 +0200
Subject: Why it is so dramatical?
Message-ID: <200209021107.25761.b.maryniuk@forbis.lt>

The task:
	Adding a text to a text.

The problem:
	It is tragically slow to add string to huge string.

The source:
-----------------8<---------------------------
import time

_body = ''
_data = 'AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA'

START_TIME = time.time()
for a in xrange(0, 0xfff):
    _body += _data
print "'+=' operator:", time.time() - START_TIME, "sek."

_body = []
START_TIME = time.time()
for a in xrange(0, 0xfff):
    _body.append(_data)
print "append() method:", time.time() - START_TIME, "sek."
-----------------8<---------------------------

The usage result:
-----------------8<---------------------------
*bo at oberon:(~/work/spike/benchmarks) python str_vs_list.py 
'+=' operator: 6.96335601807 sek.
append() method: 0.0111969709396 sek.
-----------------8<---------------------------

The question: Why?

The other question: 
	Is it true, that I should through all the Python life 
	use lists and then join it into single text instead to use += operator?

-- 
Regards, Bogdan

It is a mess, pure and simple.  Sucks to be away from Unix, huh?
		-- man perlfaq3




From paul at boddie.net  Thu Sep 19 04:14:07 2002
From: paul at boddie.net (Paul Boddie)
Date: 19 Sep 2002 01:14:07 -0700
Subject: monetary applications [was: Python GUI app to impress the bos s?]
References: <31575A892FF6D1118F5800600846864DCBD817@intrepid> <mailman.1032362843.10509.python-list@python.org>
Message-ID: <23891c90.0209190014.26c5bb9@posting.google.com>

Alex Martelli <aleax at aleax.it> wrote in message news:<mailman.1032362843.10509.python-list at python.org>...
>
> > > http://starship.python.net/crew/aahz/FixedPoint.py

[...]

> To play devil's advocate: DB-API compliant modules have long used
> mxDateTime happily, for date-time values, even though mxDateTime
> was not (and likely will never be) in the core.

Although there's that date/time PEP that will give some kind of
standard interface, thus reducing the significance of that particular
decision.

>                                                  Why is FixedPoint (or
> Decimal or some future Currency -- whatever) different from mxDateTime
> in this regard?  How come DB-API compliant modules seem not to
> even have felt any need to fish around for a type with which to
> represent NUMERIC and DECIMAL, as they did for dates and times?

It could well be that not many people do financial database
programming in Python, bizarre and unlikely as that may seem. Taking a
look at pyPgSQL, as the only database module I have installed that
actually readily documents such things (and that should tell you
something), it would seem that monetary amounts are converted to
floating point numbers - surely the first thing one would want to
avoid in manipulating financial data. Perhaps people use integers, or
perhaps other modules provide better decimal support.

> I don't know the answers, but if "help out the DB-API" is among the
> arguments you plan to muster to convince the BDFL, you might
> want to be prepared in case these questions get raised.  No, sorry,
> I won't volunteer as a paladin for this PEP -- I'll gladly help out with
> code and even drafting the PEP if need be, but this isn't really my
> own crusade...

In that case, what do you think of this module? ;-)

  http://www.boddie.org.uk/python/downloads/ScaledDecimal.py

I'll gladly release this into the public domain, but some commentary
would be nice - I fear that I may not have implemented it as well as
the "numeric gods" might like.

Paul


From notvalid at hotmail.com  Wed Sep 25 22:08:58 2002
From: notvalid at hotmail.com (jano)
Date: Wed, 25 Sep 2002 19:08:58 -0700
Subject: cgi post problem
References: <FVRj9.192186$Jo.71059@rwcrnsc53> <9aWj9.39314$V7.10439005@twister.socal.rr.com> <amq8av$2cum$1@agate.berkeley.edu> <amqgaq$2g05$1@agate.berkeley.edu> <Q14k9.62350$U7.21809935@twister.socal.rr.com>
Message-ID: <amtq7r$fvg$1@agate.berkeley.edu>

"Jeff Davis" <jdavis at empires.org> wrote in message
news:Q14k9.62350$U7.21809935 at twister.socal.rr.com...
>
>
> These two files worked for me (see below). Let me know if you have any
more
> problems.
>
> Regards,
>         Jeff
>
> -------------py2.cgi--------------------------
> #!/usr/bin/python2.2
> import cgi
>
> def main():
>     try:
>         form = cgi.FieldStorage()
>         input = form.keys()
>
>         print "Content-type: text/html"
>         print
>         print("""<?xml version='1.0' encoding='UTF-8'?>
> <!DOCTYPE html PUBLIC '-//W3C//DTD XHTML 1.1//EN'
> 'http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd'>
> <html xmlns='http://www.w3.org/1999/xhtml' xml:lang='en'>
> <head>
> <title>test</title>
> </head>
> <body>""")
>         print "<p>",input,"</p>"
>         print "</body></html>"
>     except:
>         pass
>
> if __name__ == "__main__":
>     main()
>
> ---------------py2.html------------------------------------------
> <html>
> <body>
>
> <form action="py2.cgi" method="post">
>      <input type="text" name="Name" size="20" />
>      <input type="text" name="PartNumber" size="20" />
>      <input type="text" name="Inventory" size="20" />
>      <input type="text" name="Cost" size="20" />
>      <input type="text" name="WholesalePrice" size="20" />
>     <input type="submit" value="Update Records" />
> </form>
>
> </body>
> </html>
>
>
>
> jano wrote:
>
> > There was an error in my prior post.  See ^^^^
> >
> >
> > "jano" <notvalid at hotmail.com> wrote in message
> > news:amq8av$2cum$1 at agate.berkeley.edu...
> >>
> >>
> >> "Jeff Davis" <jdavis at empires.org> wrote in message
> >> news:9aWj9.39314$V7.10439005 at twister.socal.rr.com...
> >> > Hi,
> >> >
> >> > It looks as though you have a lot of code. It might be helpful to try
> >> > to narrow the problem further by removing a lot of the code unrelated
> >> > to
> > your
> >> > problem, which appears to be that "keys" does not contain any items.
> > After
> >> > you work that out, then put the other code back in.
> >> >
> >> > My first impression is that you have some "weird" form element names
> >> > because they contain the "/" character. That might be perfectly
legal,
> > but
> >> > perhaps the cgi.FieldStorage class does not properly handle such
> >> > element names, or perhaps it considers such names erroneous.
> >> >
> >> > Next, it seems that the line "keys = form.keys()" might be out of
> >> alignment
> >> > with the rest of the code in it's block. Perhaps it's just the way it
> >> > looks after you copied your code into the post.
> >> >
> >> > Regards,
> >> >         Jeff
> >> >
> >> >
> >>
> >>
> >>
> >> Yes, the problem is that keys does not contain any items.  The 'weird'
> > names
> >> are not the problem, because i tried it also with regular 1-word names
> >> (no
> >> slashes), with the same results.  Also, the keys=form.keys() line is
> >> properly indented in the code -- it was just a copy/paste problem.
Here
> > is
> >> the code again, with the extraneous stuff stripped out:
> >>
> >> #!/usr/local/bin/python
> >>
> >> import cgitb; cgitb.enable()
> >> import sys, traceback, cgi
> >>
> >> def main():
> >>     try:
> >>         form = cgi.FieldStorage()
> >>         input = form.keys()
> >>
> >>         print "Content-type: text/html\n\n"
> >>         print("""<?xml version='1.0' encoding='UTF-8'?>
> >>         <!DOCTYPE html PUBLIC '-//W3C//DTD XHTML 1.1//EN'
> >> 'http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd'>
> >>         <html xmlns='http://www.w3.org/1999/xhtml' xml:lang='en'>
> >>             <head>
> >>                 <title>test</title>
> >>             </head>
> >>             <body>""")
> >>         print "<p>",keys,"</p>"
> >
> > #                              ^^^^ should be 'input' not 'keys', as it
> > #                              is
> > in my script.
> >
> >>         print "</body></html>"
> >>    except:
> >>         pass
> >>
> >> if __name__ == "__main__":
> >>      main()
> >>
> >> And the form is:
> >>
> >> <form action="http://someurl.com/cgi-bin/updatexml.py" method="post">
> >>      <input type="text" name="Name" size="20" />
> >>      <input type="text" name="PartNumber" size="20" />
> >>      <input type="text" name="Inventory" size="20" />
> >>      <input type="text" name="Cost" size="20" />
> >>      <input type="text" name="WholesalePrice" size="20" />
> >>     <input type="submit" value="Update Records" />
> >> </form>
> >>
> >>
> >> The script, again, is currently just supposed to return the keys, which
> >> I
> > am
> >> expecting to be Name, PartNumber, etc., but it returns only an empty
> >> list. Any ideas.
> >>
> >> thanks,
> >>
> >> jano
> >>
> >>
> >>
>

Jeff, did this work for you just on the command line?  Or over the internet
too?  I get an 'internal server error' when I try to submit from your form,
and I don't have access to the server logs -- it works fine from the command
line, but it seems to be a parsing problem with the form data.  The server
is running python 2.2.1 (apache 1.3.26) on solaris 5.6.  This is a different
problem than my original problem -- keys() returning [] -- since I can't
even get an empty list anymore, but with my original code, i get the
expected list if I use method="submit" (get), and I get an empty list if I
use method="post".

I am about to give up and do what I need to do in Java, as I have deadlines
approaching.

thanks,

jano




From neal at metaslash.com  Thu Sep 19 09:14:42 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Thu, 19 Sep 2002 13:14:42 GMT
Subject: try import except want lineno from Exception
References: <mailman.1032409524.4865.python-list@python.org>
Message-ID: <pan.2002.09.19.13.14.36.452963.1508@metaslash.com>

On Thu, 19 Sep 2002 00:23:39 -0400, Ben Gerblich wrote:

> I am importing module files, and want to catch an error (if any) and
> print the filename and line number of the error. The "Exception ,e" does
> not return, or does not have the attirbutes listed above. It only
> returns e="line() takes exactly 2 arguments (3 given)"

See inspect.getfile() and inspect.getlineno().
You will need a frame which you can get from sys.exc_info()

Neal


From ark at research.att.com  Wed Sep  4 15:46:49 2002
From: ark at research.att.com (Andrew Koenig)
Date: Wed, 4 Sep 2002 19:46:49 GMT
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com>
Message-ID: <yu99u1l5bkra.fsf@europa.research.att.com>

Gon?alo> Anyway, I've been directed to PEP 227, cheerfully obeyed, and
Gon?alo> from what I have understood having read-only nested scopes
Gon?alo> was more of a design decision than something else. And while
Gon?alo> I can live happily with the decision (some would even say
Gon?alo> happier) there is something about the "You can look at it but
Gon?alo> not touch it" attitude that I dislike ;-)

You can do anything to the object that the object support.
What you can't do is change the binding of the name.

    x = []
    def foo(y):
        x.append(y);
    foo(3)
    foo(4)
    foo(5)

After executing this code, x is [3, 4, 5].

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From bobx at linuxmail.org  Wed Sep 11 17:01:41 2002
From: bobx at linuxmail.org (Bob X)
Date: Wed, 11 Sep 2002 21:01:41 GMT
Subject: [FEEDBACK] Is this script efficient...is there a better way?
Message-ID: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net>

I am a newbie at this and this is a script I have come up with. I am 
looking for pointers on any "newbie" gotchas or ways to do it better.

I use this to parse a log file (30MB+) for keywords and write the lines 
that are found into another file.

Any thoughts would be appreciated...  :-)

Bob

#!/usr/local/bin/python -w

# import the needed libs
import sys, string

# make sure the command line arguments are there
if len(sys.argv) < 3:
     print "usage: fread.py [log file] [hit file]"
     sys.exit(1)

# open the files with some error checking
try:
     inFile = open(sys.argv[1],"r")
except IOError:
     print "Cannot open log file!\n"
     sys.exit(1)

try:
     outFile = open(sys.argv[2],"w")
except IOError:
     print "Cannot open hits file!\n"
     sys.exit(1)

# build list of keywords
kw = [ "some", "words" ]

# loop through the list and print the lines to a file
for line in inFile.readlines():
     for badword in kw:
         if line.find(badword) > -1:
             result = '%s %s' % (badword, line)
             print result            # Print the result
             outFile.write(result)   # Write the result

# close the files
inFile.close()
outFile.close()

# let me know when it's done
print "Finished processing file..."



From teaandbikkie at aol.com  Fri Sep  6 10:42:10 2002
From: teaandbikkie at aol.com (TeaAndBikkie)
Date: 06 Sep 2002 14:42:10 GMT
Subject: example for using calldll with python2.2
References: <al921d$1h8$1@news.orst.edu>
Message-ID: <20020906104210.15840.00000001@mb-cv.aol.com>

There is a link to the binary at:

http://www.python.org/ftp/python/binaries-1.4/pythonwin/html/pwmodules.htm
l#calldll



From aleax at aleax.it  Thu Sep 19 14:42:16 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 19 Sep 2002 18:42:16 GMT
Subject: Copy with __slots__
References: <mailman.1032419786.7187.python-list@python.org> <amd082$r5j$1@newsreader.mailgate.org>
Message-ID: <cUoi9.134531$pX1.4784251@news2.tin.it>

Antonio Cuni wrote:

> Griebel, Peer wrote:
> 
>> So my advice now is: Use slots! The program runs faster, uses less
>> memory. And explicitly coding a __copy__ function which copies each
>> attribute isn't a bad thing per se. It makes the things more explicit.
> 
> (sorry for my bad english...)

English excellent, Python less so:-) (...just kidding...)

 
> we can make it easier by providing a metaclass that creates a __copy__
> method for us:
> 
> class autocopy(type):
>     def __new__(cls, name, bases, dic):
>         def __copy__(self):
>             res = self.__class__()
>             for attr in dic['__slots__']:
>                 setattr(res, attr, getattr(self, attr))
>             return res
> 
>         dic['__copy__'] = __copy__
>         return type.__new__(cls, name, (object,), dic)

This doesn't deal with inherited __slots__.  Easy to solve,
mind you -- autocopy.__new__ just needs to look for __slots__
in the bases, too, and accumulate them all into a sequence
on which the nested __copy__ function can loop.  Some care
is needed, though (I can think of several possible anomalies).


Alex



From TOGLIMIcuni at programmazione.it  Thu Sep 19 15:01:50 2002
From: TOGLIMIcuni at programmazione.it (Antonio Cuni)
Date: Thu, 19 Sep 2002 19:01:50 +0000
Subject: Copy with __slots__
References: <mailman.1032419786.7187.python-list@python.org>
Message-ID: <amd082$r5j$1@newsreader.mailgate.org>

Griebel, Peer wrote:

> So my advice now is: Use slots! The program runs faster, uses less
> memory. And explicitly coding a __copy__ function which copies each
> attribute isn't a bad thing per se. It makes the things more explicit.

(sorry for my bad english...)

we can make it easier by providing a metaclass that creates a __copy__ 
method for us:

class autocopy(type):
    def __new__(cls, name, bases, dic):
        def __copy__(self):
            res = self.__class__()
            for attr in dic['__slots__']:
                setattr(res, attr, getattr(self, attr))
            return res

        dic['__copy__'] = __copy__
        return type.__new__(cls, name, (object,), dic)

Python 2.2 (#1, Feb 24 2002, 16:21:58)
[GCC 2.96 20000731 (Mandrake Linux 8.2 2.96-0.76mdk)] on linux-i386
Type "help", "copyright", "credits" or "license" for more information.
>>> import copy, autocopy
>>> class foo(object):
...     __metaclass__ = autocopy.autocopy
...     __slots__ = ['x', 'y']
...
>>> f = foo()
>>> f.x = f.y = 41
>>> bar = copy.copy(f)
>>> print bar.x, bar.y
41 41

Anto
-- 
"Computer science is not about computers any more than astronomy
is about telescopes." -- EW Dijkstra


From gerhard.haering at gmx.de  Sat Sep 28 22:10:02 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 29 Sep 2002 02:10:02 GMT
Subject: Need example for subclassing long in C
Message-ID: <slrnapco7r.617.gerhard.haering@lilith.my-fqdn.de>

I'm trying to add range checks to long and successfully did a Python
implementation. Now I want to translate it to C. Could anybody offer
an example of subclassing a built-in number type at the C level?

Another question I have is how do I do the "long.__add__" call below
at the C-level? I'd like to do it as optimized as possible.
    
    def __add__(self, other):
        val = long.__add__(self, other)
        PgInt8_checkresult(val, "addition")
        return PgInt8(val)

-- Gerhard


From David.Kastrup at t-online.de  Sun Sep 22 11:42:34 2002
From: David.Kastrup at t-online.de (David Kastrup)
Date: 22 Sep 2002 17:42:34 +0200
Subject: Using emacs gud/pdb python debugger under windows
References: <uvg5axb9e.fsf@online.no> <x53cse2dqw.fsf@tupik.goethe.zz> <r8fxwzdw.fsf@online.no> <x5adml216y.fsf@tupik.goethe.zz> <87elbuo46g.fsf@tleepslib.sk.tsukuba.ac.jp> <87n0qiul6j.fsf@computer.localdomain> <x5lm623vtf.fsf@tupik.goethe.zz> <ufzwam04i.fsf@online.no> <x5y9a22b95.fsf@tupik.goethe.zz> <usn09x6br.fsf@online.no> <u1y7t6zld.fsf@hotmail.com> <3cs86d38.fsf_-_@online.no>
Message-ID: <x5fzw2hvyt.fsf@tupik.goethe.zz>

Syver Enstad <syver-en+usenet at online.no> writes:

> Felix <aprogrammer at hotmail.com> writes:
> 
> > u just need to fix a few things to make it recogonize the win32 path.
> > here's how, open gud.el and search for the variable
> > 'gud-pdb-marker-regexp',
[...]
> 
> Ha, ha it works, it works! (Dancing around the room) I want to hug you
> Felix! I didn't think of that, obviously. So it was just pdb failing
> to recognize the output from the windows version of python.

The newest CVS contains this fix: apparently somebody passed it to
the right people.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum
Email: David.Kastrup at t-online.de


From dbasch at yahoo.com  Mon Sep 30 17:26:28 2002
From: dbasch at yahoo.com (Derek Basch)
Date: Mon, 30 Sep 2002 14:26:28 -0700 (PDT)
Subject: Catching user defined exceptions from foreign modules
In-Reply-To: <E17w74V-0000Vv-00@barry.mail.mindspring.net>
Message-ID: <20020930212628.97334.qmail@web20805.mail.yahoo.com>

Thanks for this snippit:
> For class exceptions, in a try statement with an
> except clause that 
> mentions a particular class, that clause also
> handles any exception 
> classes derived from that class (but not exception
> classes from which 
> it is derived). Two exception classes that are not
> related via 
> subclassing are never equivalent, even if they have
> the same name. 

It really helped me understand. I subclassed my
'TestError' exception from a base exception class
'Error' which was derived from the 'Exception' class.
As shown in this example:
http://www.python.org/doc/current/tut/node10.html#SECTION0010500000000000000000
So I could have just caught the imported modules base
exception class and as long as the generated exception
is a subclass of the base class it will be caught.
Cool!! To bad the python book I learned from never
covered this :(. I will have to add this to a bunch
more modules now that I understand how to use it
properly. I think I had better give the base exception
classes unique names as well to avoid import
conflicts.

Thanks for all the help,
Derek Basch

> > Thanks,
> > By 'built in' I meant the python built in module
> exceptions:
> > http://web.pydoc.org/2.2/exceptions.html
> 
> 	I'll have to confess that I haven't kept fully up
> on the changes (like 
> making all exceptions "class-based").
> 
> 	To me, "built in" tends to imply something
> available without explicit 
> imports...
> 
> 	I think the following paragraph from the document
> is applicable...
> 
> For class exceptions, in a try statement with an
> except clause that 
> mentions a particular class, that clause also
> handles any exception 
> classes derived from that class (but not exception
> classes from which 
> it is derived). Two exception classes that are not
> related via 
> subclassing are never equivalent, even if they have
> the same name. 
> 
> 	You created a subclass exception as I recall (I
> deleted the files I'd 
> copied so can't check). The main program could have
> caught that if you 
> had just listed the base exception rather than the
> specific subclass. 
> Without actually testing code (I'm rushing -- have
> an appointment in 45 
> mins)...
> 
> #t1.py
> from t2 import I_Fail
> 
> try:
> 	I_Fail(123)
> except ArithmeticError, e:
> 	print "Arithmetic Error: '%s' Occurred" % e.value
> 
> 
> #t2.py
> 
> class MyFailure(ArithmeticError):
> 	def __init__(self, value):
> 		self.value=value
> 	def __str__(self):
> 		return `self.value`
> 
> def I_Fail(a):
> 	if a is not None:
> 		raise MyFailure("I_Fail received a non-null
> argument")
> 
> 
> 
> >
> > Im assuming these are automatically imported
> anytime the interpreter
> > is fired up. I will use the 'from module import
> alpha, beta,
> > charlie' syntax as this seems to be the best way.
> At least according
> > to the ActivePython documentation/Python HOW
> TO's/Do's and Dont's.
> > They even have your open() idiom in the HOW TO.
> > Thanks for all the help,
> > Derek Basch
> >
> > --- In python-list at y..., Dennis Lee Bieber
> <wlfraed at i...> wrote:
> > > Derek Basch fed this fish to the penguins on
> Monday 30 September
> >
> > 2002
> >
> > > 08:59 am:
> > > > to have to import user defined exceptions? I
> dont
> > > > rememeber having to import any of the built in
> > > > exceptions. So, here is the output of the
> fixed test1:
> > >
> > >         "built-in" -> already there... <G>
> > >
> > > > test2.TestError
> > > > fudge
> > > >
> > > > Am I doing this all wrong?
> > >
> > >         You want to use items defined in an
> imported module. I see
> >
> > three
> >
> > > alternatives...
> > >
> > > import module
> > >
> > >         all uses then become module.item
> > >
> > > from module import item1, item2, ... itemN
> > >
> > >         only the named items are useable, but
> can be used as itemM
> > >
> > > from module import *
> > >
> > >         practically everything within the module
> is available as
> >
> > item
> >
> > > The last is not recommended as you can encounter
> unexpected
> > > redefinition of names... In the last week we had
> one such...
> > >
> > > from os import *
> > >
> > > replaces the Python open() (which I understand
> is being phased out
> >
> > for
> >
> > > file()) with the low-level (C-style) open --
> totally different
> > > arguments needed.
> > >
> > > --
> > >
> > >  >
>
==============================================================
> <
> > >  >   wlfraed at i...  | Wulfraed  Dennis Lee Bieber
>  KD6MOG <
> > >  >      wulfraed at d...     |       Bestiaria
> Support Staff       <
> > >  >
>
==============================================================
> <
> > >  >        Bestiaria Home Page:
> http://www.beastie.dm.net/         <
> > >  >            Home Page:
> http://www.dm.net/~wulfraed/             <
> > >
> > > --
> > >
> http://mail.python.org/mailman/listinfo/python-list
> 
> -- 
> --
>  >
>
==============================================================
> <
>  >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee
> Bieber  KD6MOG <
>  >      wulfraed at dm.net     |       Bestiaria
> Support Staff       <
>  >
>
==============================================================
> <
>  >        Bestiaria Home Page:
> http://www.beastie.dm.net/         <
>  >            Home Page:
> http://www.dm.net/~wulfraed/             <


__________________________________________________
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
http://sbc.yahoo.com



From martin at v.loewis.de  Mon Sep 16 13:14:47 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Sep 2002 19:14:47 +0200
Subject: problems when parsing & with xml.dom.minidom.parseString
References: <a230569c.0209160851.7c6884f@posting.google.com>
Message-ID: <m33cs9dfg8.fsf@mira.informatik.hu-berlin.de>

jcc at ibdosnorte.com (Juan Carlos CORU?A) writes:

> Anyone knows the solution?

>>> node.firstChild.childNodes[0].nodeValue
u'one '
>>> node.firstChild.childNodes[1].nodeValue
u'&'
>>> node.firstChild.childNodes[2].nodeValue
u'two'
>>> node.normalize()
>>> node.firstChild.childNodes[0].nodeValue
u'one &two'

In PyXML 0.8.1, in this case, you will already get normalized data right
after parsing.

HTH,
Martin


From tmohr at s.netic.de  Sat Sep 14 23:38:35 2002
From: tmohr at s.netic.de (Torsten Mohr)
Date: Sun, 15 Sep 2002 05:38:35 +0200
Subject: embedding Python, several instances?
Message-ID: <am0vbr$49p$1@schleim.qwe.de>

Hi,

i'd like to use Python as a scripting language for a C program
and the examples in the documentation work quite well.

When i now run two scripts, the definitions that were done in
the first one are seen in the second one.


How can i run each script in one instance of the python
interpreter?


Best regards,
Torsten.




From tdelaney at avaya.com  Thu Sep 26 02:32:19 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 26 Sep 2002 16:32:19 +1000
Subject: revenge of the python puzzle
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5E8@natasha.auslabs.avaya.com>

> From: David Eppstein [mailto:eppstein at ics.uci.edu]
> 
> I'm not going to spoil this, just point out that my cryptogram applet 
> http://www.ics.uci.edu/~eppstein/cryptogram/ (sorry, not written in 

OK - you *gotta* stop case conversions on that thing ...

Tim Delaney



From mark.charsley at REMOVE_THIS.radioscape.com  Mon Sep 23 11:22:00 2002
From: mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley)
Date: Mon, 23 Sep 2002 16:22 +0100 (BST)
Subject: M$SQL connections from linux
References: <1032528860.599940@news.etel.ru>
Message-ID: <memo.20020923162224.544B@a.a>

In article <1032528860.599940 at news.etel.ru>, black at flamingo.ru (Vasiliy A. 
Chernoivan) wrote:

> Hello all
> Is there some module for python which performs the task?
> Is it possible at allto access data under M$SQL from linux machine?

Not used it myself, but mxODBC should do the job:

http://www.egenix.com/files/python/mxODBC.html

-- 

Mark - personal opinion only, could well be wrong, not representing 
company, don't sue us etc. etc.


From mhammond at skippinet.com.au  Sat Sep 14 04:24:06 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 14 Sep 2002 08:24:06 GMT
Subject: threads or queue for this task
References: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com>
Message-ID: <GmCg9.2354$Ee4.6690@news-server.bigpond.net.au>

robin at execulink.com wrote:
> I am missing something I think with the difference between the
> threading and Queue modules. These could really do with some updated
> docs with examples.
> 
> I wish to have a server continually cycle (until interrupted by
> keyboard, signal, or whatever). It needs to perform several (say 6)
> different tasks, each of which may take significantly different
> lengths of time to perform.
> 
> My first implementation used threading.py. I wrote a dispatcher to
> start each of the threads, and then wait until the thread count was
> back down to 1. This I put in a while loop, and surrounded it with the
> necessary exception code.
> 
> Each cycle through the while has to wait for the slowest process to
> finish before completing. During this time, the other processes just
> sit around doing nothing. What I really want is for each process to
> start up again when they are done, so that there are no more than 6
> going at once. In other words, I want to be sure I don't start a new
> process 3 before the old process 3 is done.
> 
> I have looked at Queue but it doesn't seem to have enough control to
> make this happen.
> 
> I am sure I am overlooking something obvious. Thanks ahead of time for
> any help!

It is common for threading code to look somewhat like:

threads = []
for i in range(number_of_tasks):
   thread.append(thread_for_task)
# let the threads run - just wait for shutdown
for t in threads:
   t.join()

Each thread itself contains a loop which continually look for more work 
to do, or exit.  The queue module is useful for giving threads work to 
do - some other thread can queue new things, and the worker thread 
removes them, blocking while nothing is in the queue.

Look for Aahz's threading tutorial - URL not at hand, but it can't be 
too hard to find.  Indeed, I predict he will tell you exactly where it is ;)

Hope this helps,

Mark.



From a_salnikov at yahoo.com  Mon Sep 16 17:30:06 2002
From: a_salnikov at yahoo.com (Andy Salnikov)
Date: Mon, 16 Sep 2002 14:30:06 -0700
Subject: Anyone can tell me why?
References: <1815200.1031973626@dbforums.com> <1817360.1032087013@dbforums.com>
Message-ID: <am5ik7$ldv$1@usenet.Stanford.EDU>

"hpyhpy" <member at dbforums.com> wrote in message
news:1817360.1032087013 at dbforums.com...
>
> But I try,change the filename as "example.so","_example.so",
> "examplemodule.so",#_#,same error......have other advise?^_^
>
  Hmm, there is no obvious reason why it should not run when
renamed to example.so or examplemodule.so. Here is my list of
possible causes:

1. You run python from directory different from where your .so
  lives in. Python finds everything in the well-defined places, which
  includes local directory. From python prompt type:

>>> import sys
>>> print sys.path

  to see the list of directories where python can find a module. Do not
  put your module outside of any directory in this path.

2. Your python was built without support for dynamic loading. Don't
  know how to check this. Ask the person who installed/configured
  your python.

3. Shared library was not build correctly. I'm not familiar with SCO
  but usually with gcc I do 'gcc -shared object.o -o whatever.so' and that
  works.

4. Whatever else....

  Andy.




From amk at amk.ca  Mon Sep 30 17:42:13 2002
From: amk at amk.ca (A.M. Kuchling)
Date: Mon, 30 Sep 2002 21:42:13 -0000
Subject: curses troubles
References: <mailman.1033018391.15538.python-list@python.org>
Message-ID: <uphh9lf3iu69bb@corp.supernews.com>

In article <mailman.1033018391.15538.python-list at python.org>, Kat & Huw wrote:
> Now, when running this, if I hit either G or P, the applications
>closes as if I'd hit Q 
> any idea what I'm doing wrong here?

Perhaps your application is raising some exception, but the traceback
isn't being displayed because of the terminal mode.  The curses module
contains a wrapper(func, *args) function that will set up the curses
screen and call func(), which should be your main function.  If func()
raises an uncaught exception, wrapper() will restore the terminal mode
and let the traceback be printed.  Maybe that will help.

--amk



From ataylor at ucsd.edu  Fri Sep 27 15:55:51 2002
From: ataylor at ucsd.edu (Adam Taylor)
Date: 27 Sep 2002 12:55:51 -0700
Subject: Getting started
References: <3D911306.7080609@ucsd.edu> <mailman.1032982114.4627.python-list@python.org> <e15b6fa3.0209261128.4edcdfe8@posting.google.com> <mailman.1033119160.31301.python-list@python.org>
Message-ID: <e15b6fa3.0209271155.53ac7fcf@posting.google.com>

>>                      Statically typed   Dynamically typed   Untyped
>>Strongly checked      ML                 LISP
>>Weakly checked        C                                      Assembler
>
>I agree.  So where in this spacewould YOU put Python and Perl?

Python is in the same square as Lisp (which I mentioned).  So is Perl.
At least using this terminology... ;)

> I was translating his internal vocabulary to use the term we started
> with.  He defines "weakly checked" only in terms of "static checking".
> Since they're synonyms, this reference only supports a def of "weakly
> typed" in terms of static checked languages.  So I'm saying the
> reference does not help us nearly as much as I would like.

Okay, that's true, he does define weakly-checked only in terms of
statically checked languages, which I hadn't appreciated:

  In reality, certain statically checked languages do not ensure
  safety. That is, their set of forbidden errors does not include all
  untrapped errors. These languages can be euphemistically called
  _weakly checked_ (or _weakly typed_, in the literature) meaning that
  some unsafe operations are detected statically and some are not
  detected.

But the clear generalization of this (it seems to me) is to define
weakly checked languages as languages in which some unsafe operations
are detected and some are not detected.  In other words, just delete the
word "statically" from the last sentence in the quote above.

Of course, it doesn't sound like any existing languages are both
dynamically typed and weakly checked (and I can't think of any myself):

  Most untyped [read dynamically typed] languages are, by necessity,
  completely safe (e.g., LISP). Otherwise, programming would be too
  frustrating in the absence of both compile time and run time checks to
  protect against corruption.

Adam

-----
"James J. Besemer" <jb at cascade-sys.com> wrote in message news:<mailman.1033119160.31301.python-list at python.org>...
> Adam Taylor wrote:
> 
> >This sounds like the same thing as I said, or damn close.  Am I
> >missing something?
> >
> 
> Not really.  Just different emphasis.
> 
> >Yes.  It seems to me that Table 1 in Cardelli:
> >
> >          Typed   Untyped
> >Safe      ML      LISP
> >Unsafe    C       Assembler
> >
> >Might be better like this:
> >
> >          Statically typed   Dynamically typed    Untyped
> >Safe      ML                 LISP
> >Unsafe    C                                       Assembler
> >
> >(Python would also be in the (Safe,Dynamically typed) position.)
> >
> >Although safety and strongly-checked are not the same thing, this
> >table is also correct:
> >
> >                      Statically typed   Dynamically typed   Untyped
> >Strongly checked      ML                 LISP
> >Weakly checked        C                                      Assembler
> >
> 
> I agree.  So where in this spacewould YOU put Python and Perl?
> 
> >>A literal reading furthermore defines "weakly typed" only in 
> >>terms of static checking.
> >>    
> >>
> >
> >I don't understand what you mean by this.  Cardelli never uses the
> >term "weakly typed" himself, he only uses it that one time, to say
> >that it's the term most people use for the concept that he prefers to
> >call "weakly checked".
> >
> 
> I was translating his internal vocabulary to use the term we started 
> with.  He defines "weakly checked" only in terms of "static checking". 
>  Since they're synonyms, this reference only supports a def of "weakly 
> typed" in terms of static checked languages.  So I'm saying the 
> reference does not help us nearly as much as I would like.
> 
> >Or at
> >least when people start talking about "strongly typed" vs "weakly
> >typed" languages, that first thing to get straight is which axis
> >they're talking about.
> >
> Fair enough.  
> 
> --jb


From Phil.Rittenhouse at dspfactory.com  Wed Sep  4 17:43:23 2002
From: Phil.Rittenhouse at dspfactory.com (Phil Rittenhouse)
Date: Wed, 4 Sep 2002 17:43:23 -0400
Subject: Installing Python 2.2.1 on WinNT
Message-ID: <OFB053F7A9.61AACB6E-ON85256C2A.0062C6B6@dspfactory.com>

Thanks for the help Trent.

I have the lastest NT service pack installed (sp6a).

I'm not sure what DLL could not be loaded, the error message just gives a
bunch of
garbage for the DLL name.   I suspect that it's trying to load a
non-existant DLL for some
reason.

I am using the python.org installer, and I suspect the problem lies in the
WISE code.    To see if
the problem was related to permissions, I tried installing, while logged in
as administrator, and it
still threw up the warning dialog about installing without admin
privileges.   So, clearly, something's
wrong with the detection of administrator privileges.

The ActivePython install worked fine, but we'd rather ship the install from
python.org if possible.

I'd like to hear from any other NT 4.0 users that have tried installing
Python 2.2.   Is it just me?

Thanks!
Phil



                                                                                                                   
                    Trent Mick                                                                                     
                    <trentm at Active       To:     Phil Rittenhouse <phil.rittenhouse at dspfactory.com>                
                    State.com>           cc:     python-list at python.org                                            
                                         Subject:     Re: Installing Python 2.2.1 on WinNT                         
                    30/08/2002                                                                                     
                    07:47 PM                                                                                       
                                                                                                                   
                                                                                                                   





> I've run into a (actually two) problem installing Python 2.2.1 on
> Windows NT 4.0.  At the end of the install, it throws up an error
> "Could not load the DLL library $%#%^$.  The specified module could
> not be found."

Do you have the latest service pack and fixes from MSDN for your NT box?
I have found that getting the latest updates often fixes some "Could not
load the DLL library" and "Page fault in..." and "Invalid entry point
in" errors.

What DLL could not be loaded?

Are you running the python.org installer? You could also try
ActivePython:
    http://www.activestate.com/Products/ActivePython/

Cheers,
Trent


--
Trent Mick
TrentM at ActiveState.com







From umop at psyon.org  Wed Sep 18 17:04:12 2002
From: umop at psyon.org (Eric Arnold)
Date: Wed, 18 Sep 2002 14:04:12 -0700
Subject: Regex question
Message-ID: <amapuf$45t8t$1@ID-144898.news.dfncis.de>

Hi folks,

I'm trying to figure out why this is not working, and wonder if there are
any regexperts out there who can give me a pointer.

The string I'm capturing is as follows:

Name:AAA,BBB,CCC

where I'd like to extract the following data:

<eventName>: Name
<eventParam 1>: AAA
<eventParam 2>: BBB
<eventParam 3>: CCC

this is the regular expression I wrote to do so:

(?<eventName>.*):(?<eventParam>([^,]*(?=,))|((?<=,)[^,]*))+

I tried it with the (?<=,) as another (?=,) too, but no luck.  I imagine
that the problem is that when the (?=) is used once, it can't be used with
the second group, but I don't know a way around that.

Thanks!

Eric




From m2 at plusseven.com  Fri Sep 20 07:42:55 2002
From: m2 at plusseven.com (Alex Polite)
Date: Fri, 20 Sep 2002 13:42:55 +0200
Subject: debugging "Too many open files"
In-Reply-To: <j4znudx52e.fsf@informatik.hu-berlin.de>
References: <mailman.1032454586.3117.python-list@python.org> <j4znudx52e.fsf@informatik.hu-berlin.de>
Message-ID: <20020920114255.GB1710@matijek.plusseven.com>

On Thu, Sep 19, 2002 at 07:27:53PM +0200, Martin v. L?wis wrote:
> Depends on what you mean by "pointing to". It is extremely rare that a
Symlink. My thinks?
> process opens /dev/hdc, since that requires superuser permissions; I
Well. I was running this as root.
> know of no case where a program opens /dev/hdc more than once.
> 

> So by "pointing to", you probably meant something else. Care to
> provide precise data? What did you do, how did the computer respond,
> how do you interpret this response?

Looking closer into the matter I realised that /dev/hdc (my dvdplayer)
was not valid device, since the ide-cd kernel module wasn't loaded. I
loaded the module, restarted the python process and now everything is
working fine. So thanks a lot for the pointers Martin.

No the next step would be to have python check that /dev/hdc is a
valid device and throw an exception if it's not.

Any idea how that should be done?

alex

-- 
Alex Polite
http://plusseven.com/gpg/



From marklists at mceahern.com  Thu Sep 26 12:56:05 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 26 Sep 2002 11:56:05 -0500
Subject: design question
In-Reply-To: <1ha6pukqdqell8f31q1k7rtor2rv5ogjs0@4ax.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOKEMOEBAA.marklists@mceahern.com>

[Gon?alo Rodrigues]
> I know I am being a little vague in my question, but does anyone have
> any ideas? Open-source code where this "same problem" has been tackled?

Have you considered metaclasses?  Do you need a concrete Wrapper for each
class?  Why not a generic wrapper that wraps any given class?

// m




From imbosol at vt.edu  Tue Sep  3 14:39:25 2002
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 3 Sep 2002 14:39:25 -0400
Subject: Design question.... parent/child class
References: <3d74b021$0$234$4d4ebb8e@news.nl.uu.net>
Message-ID: <al30tr$aud$2@solaris.cc.vt.edu>

Gumuz wrote:
> hi all,
> 
> i have a SessionManager class, which contains a list of Session
> objects(instantiated from a Session class). I have a design problem.
> 
> Each session object has a small 'queue' of messages for the specific
> session. How can a session send a message to another session without having
> some kind of parent-reference to the SessionManager object?
> 
> i am a bit puzzled, or is this parent-reference not such a bad idea after
> all?


If you can, make the function that transfers a message a member of
SessionManager class, instead of Session.  (Or, if that's not how your
program is set up, have the function that transfers a message take
your SessionManager instance as an argument, or something like that.
The idea is to access the sender or receiver through the
SessionManager class.)

If you can't, think about how easy it would be to change your code so
that you could.  And if it's not too hard, do it.



Abstractly, you want to operate on two siblings in your hierarchy.
You should want to take advantage of the hierarchy; therfore, your
operation should belong to the parent, which contains both sibling
nodes.  If you want to do this operation without the help of then
parent, then you're circumventing (or at least not taking advantage
of) the hierarchy structure.



-- 
CARL BANKS
http://www.aerojockey.com


From news at NOSPAMmajid.fm  Mon Sep  9 22:36:36 2002
From: news at NOSPAMmajid.fm (Fazal Majid)
Date: Mon, 09 Sep 2002 19:36:36 -0700
Subject: Aristotle and Python
References: <wkr8g7rn4c.fsf@yahoo.com> <mailman.1031360589.11985.python-list@python.org>
Message-ID: <unqmnq2q0leg0c@corp.supernews.com>

James J. Besemer wrote:
>> Larry Wall Wrote:
>>
>> "Python is cool to look at small bits of, but I think the "outline"
>> syntax breaks down with larger chunks of code. I'm with Aristotle on
>> the structure of discourse--a story should have a beginning, and
>> middle, and an end. So should blocks"
> Typically, he substitutes a cute particle of rhetoric for a meaningful 
> response.

That's funny, I hadn't read Larry's article as I am not all that 
interested in Perl (dropped it for Python in 94, haven't looked back 
since), but I set out an Aristotle comparison at about the same time.
One of the good things about Python is that it is non-aristotelian:
http://www.majid.info/radio/2002/09/07.html#a30

-- Fazal Majid



From dwelch91 at nospam.attbi.com  Sat Sep 21 01:31:53 2002
From: dwelch91 at nospam.attbi.com (dwelch)
Date: Sat, 21 Sep 2002 05:31:53 GMT
Subject: newbie seeks advise.
References: <77c84db2.0209192137.51167ebb@posting.google.com> <3D89F605.4030303@nospam.attbi.com> <3D8AB73E.2322DDAC@alcyone.com>
Message-ID: <3D8B441E.304@nospam.attbi.com>

Erik Max Francis wrote:
> dwelch wrote:
> 
> 
>>Simple way, probably not optimal:
>>
>> >>> max_element = max(list(data))
>> >>> max_element
>>74
>>
> 
> max works on any sequence, there's no need to preconvert the tuple to a
> list.
> 
> 

Yes, of course you are correct (that would be inconceivably inconsistent 
for Python). However, when I first looked at using max() with a tuple on 
2.2.1 on WinXP, it didn't work. I got a TypeError stating that max() 
requires a sequence type. Now, I am trying it on Mandrake 8.2 with 
2.2.1, it and it works. Hmmmm.. not sure what is going on here, I will 
have to double check when I am on my XP box again.

Don



From sross at connectmail.carleton.ca  Tue Sep 24 23:36:01 2002
From: sross at connectmail.carleton.ca (Sean Ross)
Date: 25 Sep 2002 03:36:01 GMT
Subject: Removing all occurences of a character from a string.
References: <amr8nh$rd8$1@driftwood.ccs.carleton.ca> <DL9k9.201596$216.7001230@bin4.nnrp.aus1.giganews.com>
Message-ID: <amrav1$4rk$1@driftwood.ccs.carleton.ca>

Thank you Justin.

How about this, then:

I have a list of lists of these strings and I want to apply replace to each
string in the list of lists.

So, say:

attr = [ ["'gill-attachment'", "'a'", "'d'", "'f'",
"'n'"],["'gill-spacing'", "'c'", "'d'", "'w'"], ["'gill-size'", "'b'",
"'n'"]]

I can do the following:
for i in xrange(0, len(attr)):
    for j in xrange(0,len(attr[i])):
         attr[i][j] = attr[i][j].replace("'","")

to get what I want; which is:
[ ['gill-attachment', 'a', 'd', 'f', 'n'],['gill-spacing', 'c', 'd', 'w'],
['gill-size', 'b', 'n']]

Is there a nicer way to do this?
Sean


"Justin Shaw" <wyojustin at hotmail.com> wrote in message
news:DL9k9.201596$216.7001230 at bin4.nnrp.aus1.giganews.com...
> > How can I remove the single quotes so that "'Hello World'" becomes
"Hello
> World"?
>
> Try this one:
>
> >>> s = "'Hello World'"
> >>> print s.replace("'", "")
> Hello World
> >>>
>
> Justin
>
> > I have a string, say "'Hello World'", where Hello World is in single
> > quotes(i.e., 'Hello World'). How can I remove the single quotes so that
> > "'Hello World'" becomes "Hello World"?
> > Thanks
> > Sean Ross
> >
> >
>
>




From aleax at aleax.it  Wed Sep 25 18:13:25 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 22:13:25 GMT
Subject: readlines() and "binary" files
References: <3d90cacf$0$30457$afc38c87@sisyphus.news.be.easynet.net> <9r8k9.46422$V7.11170409@twister.socal.rr.com> <3d91de2d$0$30457$afc38c87@sisyphus.news.be.easynet.net>
Message-ID: <9yqk9.143972$ub2.3161265@news1.tin.it>

Justin wrote:
        ...
> at a time, so that's rather overkill. If anyone can think of a more
> economical way of doing it (short of defining my own iterator), I'd
> be interested.

Pity you ruled out defining your own iterator -- in Python 2.2.1...:

from __future__ import generators

def splitby(fileobj, splitter, bufsize=8192):
    buf = ''

    while True:
        try: 
            item, buf = buf.split(splitter, 1)
        except ValueError:
            more = fileobj.read(bufsize)
            if not more: break
            buf += more
        else:
            yield item + splitter

    if buf:
        yield buf


I think it's more or less the simplest approach.


Alex



From leafnode-usenet at schabi.de  Thu Sep 19 09:32:05 2002
From: leafnode-usenet at schabi.de (Markus Schaber)
Date: Thu, 19 Sep 2002 15:32:05 +0200
Subject: linux python ideas
References: <3D88E578.6010001@diespammerdieuselesspython.com>
Message-ID: <20020919153205.6fd58389.leafnode-usenet@schabi.de>

Hallo,

Rob Andrews <rob at diespammerdieuselesspython.com> schrieb:

> Any suggestions would be appreciated, since my notes are pretty
> generic so far.

I held such a presentation here in Ulm (Germany), you can see the German
data at http://ulm.ccc.de/~schabi/python/index.html - this file is
identical to the presentation, just use Opera 6 or later and press F11
(or use any other CSS presentation mode capable browser)

Gru?,
Markus


From gerhard.haering at gmx.de  Fri Sep 27 12:01:44 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Fri, 27 Sep 2002 18:01:44 +0200
Subject: DBI cursor behaviour with multiple statements?
In-Reply-To: <7e964d1d.0209270137.568a3f3@posting.google.com>
References: <7e964d1d.0209270137.568a3f3@posting.google.com>
Message-ID: <20020927160143.GA1055@lilith.ghaering.test>

Hi Henrik,

* Henrik Weber <Henrik.Weber at sys.aok.de> [2002-09-27 02:37 -0700]:
> I'm trying to write a DBI2 compliant module for Microsoft Access
> databases.

Ok. Then the Database SIG's mailing lists is the appropriate place to
ask. You'll find that almost all module authors and many users are
subscribed there.

> Now I'm not quite sure how to interpret the DBI2 specification when it
> comes to the execution of multiple data returning statements with the
> same cursor.

The behaviour is undefined for more than one DQL statement. And
.nextset() isn't meant for this case, either.

In fact, I asked the same question on the SIG's mailing list some time
ago:

Date: Tue, 28 May 2002 21:36:47 +0200
To: Python DB-SIG Mailing List <db-sig at python.org>
Subject: [DB-SIG] Multiple Statements
Message-ID: <20020528193647.GA902 at lilith.my-fqdn.de>

... and was referred to a discussion about a year ago (then), where the
consensus was what I summarized above.

Here's the SIG's mailing list: http://python.org/sigs/db-sig/

Btw. an updated form of the specification several interesting "optional
features" is available as PEP 0249.

-- Gerhard



From mcherm at destiny.com  Mon Sep 16 12:22:13 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Mon, 16 Sep 2002 12:22:13 -0400
Subject: Skipping Parts of a For Loop
References: <pan.2002.09.14.12.23.44.103401.723@thomas-guettler.de> <uo84kscns3l067@corp.supernews.com>
Message-ID: <am50f3$k8n$1@main.gmane.org>


> Thomas Guettler wrote:
>>Is there a way to skip some parts in a for loop?
>>   [...]
>>"continue" skipps one element. But how can
>>I skip several?

Andrew Henshaw wrote:
> While not really answering your question... In my opinion, it would be 
> better to use a while loop for that purpose.  The code would be much 
> clearer.

I agree with Andrew. I think the cleanest and most readable way to do it 
would be as follows:


i = 0
while i < UPPER_BOUND:
    do_some_processing()
    if condition:
        i += AMOUNT_TO_SKIP
    do_more_processing()
    i += 1

It makes it very explicit that you are skipping several values of i, and 
any programmer could read this code and understand what it's doing. It 
would be even BETTER if you used a meaningful name for i (perhaps 
"line_number", or "record_id"... use something appropriate to what 
you're doing).

-- Michael Chermside






From henrik.motakef at web.de  Sat Sep 28 16:11:38 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 28 Sep 2002 22:11:38 +0200
Subject: What does Python fix?
References: <e15b6fa3.0209271418.14cd5642@posting.google.com> <flaapucbduj6g4ifrdf3oq99hdrepi1o3v@4ax.com> <3d95fb09@news.vip.fi>
Message-ID: <87ptuxub5x.fsf@pokey.henrik-motakef.de>

Pekka Niiranen <krissepu at vip.fi> writes:

>>Shrug. Lisp trades one form of complexity for another. By reducing
>>the solution to a universal grammar, they increase the symbolic load
>>on the user, who is awash in a page full of very similar symbols.
>>This is intimidating in particular to novices, for whatever reason,
>>who are instantly alienated. Without sweeping in novices, you lack
>>a grass roots movement, and without a grass roots, a language is dead.
>>
>>The parts of lisp that really matter to the community have since
>>been adopted in part by many other languages. Lisp is dead. Long live
>>Lisp.

> Sure, Python's syntax looks nice, but to use another language to get
> performance (i.e speed) ? That sucks.

Let's not forget that Python's syntax (most of all the significance of
whitespace) is considered strange at least by about all novices. In
both cases - significant whitespace and lots of irritating superflous
parentheses - people don't just get used to it, but tend to like it
after a while.

Another rather cool feature of lisp that python lacks are macros, btw,
or in general the uniform handling of code and data.

> I have a feeling Python might renew peoples interest to various lisps.

That would be nice. The one thing lisp lacks is a huge community,
manifesting itself in lots of ready-to-use libraries, tutorials etc.

I guess the mere existence and visibility of Python (and Ruby, for
example) is a Good Thing - there's choice again, and when people are
thinking about what language they should use, it's not just the
question "should I use C or Perl" (or, depending on the platform you
are socialized on, "should I use C++ or Visual Basic") any more - it
is more obvious that there are a lot of high-level languages one could
use, and that they differ significantly. That doesn't only help the
highly visible languages like Python, but also the more obscure ones
like Lisp, or Objective Caml for example.

> After studying Python for 2 years I myself have
> recently switched to DrScheme because I have no interest
> relearning C to write extensions to Python.
> 
> What an eye opener it has been to study scheme...

Ack. Learning Lisp (I prefer Common Lisp over Scheme myself, but IMHO
you should know both) is something any programmer should do
sometimes. Of course, learning and using Python isn't a bad thing,
either...

Regards
Henrik


From jpm at papercut.org  Mon Sep  9 20:17:55 2002
From: jpm at papercut.org (Joao Prado Maia)
Date: Mon, 9 Sep 2002 20:17:55 -0400 (EDT)
Subject: [ANN] Papercut 0.9.0 - python nntp server
Message-ID: <Pine.LNX.4.44.0209092014180.7406-100000@mercury.sabren.com>

Papercut is a multi-threaded news server written in Python. Its main
objective is to integrate existing web based message board software
(Phorum [http://phorum.org] is the most supported one) with an Usenet 
front-end.

However, its extensibility enables developers to write their own 
containers for the storage of the Usenet articles (messages). That means 
that the code is extensible enough that _you_ could write new containers 
to integrate the news server with other web message board projects or even
with other ways to store the messages.

The source code is available freely under the BSD license, which means you
can do whatever you like with it. I do welcome any and all contributions
to the project :)

A quick description of the changes on this release:

- Mostly a lot of testing regarding the NTTP authentication code. The 
authentication stuff should be pretty much rock solid and it was tested 
successfully using the following Newsreaders: Mozilla Newsreader, Netscape 
4.79 Newsreader, Outlook Express and Knode (KDE's default newsreader).

More information and download instructions can be found on the project 
site -> http://papercut.org

Cheers,
Joao Prado Maia




From bloke at ii.net  Sun Sep 15 11:15:09 2002
From: bloke at ii.net (Rob Hall)
Date: Sun, 15 Sep 2002 23:15:09 +0800
Subject: confused with class inheritance
References: <3d83fa2c$0$24041@echo-01.iinet.net.au> <mailman.1032061218.4602.python-list@python.org>
Message-ID: <3d84a3ed$0$24051@echo-01.iinet.net.au>

Doh




From phil at river-bank.demon.co.uk  Thu Sep 19 08:27:27 2002
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Thu, 19 Sep 2002 13:27:27 +0100
Subject: Is there a widget or program  for a database browsing
In-Reply-To: <3D89DA60.8878.8D30C74@localhost>
References: <3D89DA60.8878.8D30C74@localhost>
Message-ID: <200209191327.27196.phil@river-bank.demon.co.uk>

On Thursday 19 September 2002 1:08 pm, A wrote:
> Hi,
> In my program users will search  a database and I need to display choosen
> record(s ) in a nice graphical way. Is there a module( program), that I
> could use together with Python, that  could easily  show the database
> record(s) and take care of all thinks during browsing? Thanks for help.
>  Ladislav

PyQt - check out the examples in the examples3/SQL directory.

Phil



From thp at cs.ucr.edu  Sun Sep 22 22:13:55 2002
From: thp at cs.ucr.edu (thp at cs.ucr.edu)
Date: Mon, 23 Sep 2002 02:13:55 +0000 (UTC)
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <%8ci9.138477$5r1.5541622@bin5.nnrp.aus1.giganews.com>
Message-ID: <amltd3$3k$1@glue.ucr.edu>

Terry Reedy <tjreedy at udel.edu> wrote:

+ "daniel w. moore" <dan65536 at hotmail.com> wrote in message
+ news:33b491f.0209181820.17b6ea70 at posting.google.com...
+> It's occurred to me after some brainstorming on the subject that the
+> C/perl construct
+>
+>   a ? b : c
+>
+> can be arranged like this in python:
+>
+>   (a and [b] or [c])[0]

+ If you *know* that b will evaluate as True (perhaps because it is a
+ non-null constant, then this can be simplified to 'a and b or c'.  I
+ first used this several years ago.

But why use grotesque hacks in the first place?  "a ? b : c" is a
well-established syntax for conditional expressions.  Why doesn't
Python simply adopt it?

Tom Payne




From alex at ivascu.comNoSpam  Sat Sep 28 01:01:13 2002
From: alex at ivascu.comNoSpam (Alex Ivascu)
Date: Sat, 28 Sep 2002 05:01:13 GMT
Subject: windows xp command prompt
References: <576c1752.0209271411.6d7bb4c4@posting.google.com>
Message-ID: <tIal9.428915$kp.1219622@rwcrnsc52.ops.asp.att.net>

Did you install perl, etc. while the command prompt window was open?  try
starting another window.... just a thought...

Alex Ivascu


"mike henley" <rsrchstr at hotmail.com> wrote in message
news:576c1752.0209271411.6d7bb4c4 at posting.google.com...
> i have installed the activestate perl, python, tcl, and pragmatic
> programmer's ruby. when i go to the command prompt on windows xp and
> type perl for example... (or python, ruby... etc) i get the following
> message.
>
> 'perl' is not recognized as an internal or external command,
> operable program or batch file.
>
> in the control panel| system | advanced | environment variables | Path
> i have
>
>
C:\ruby\bin;C:\Python22;C:\Perl\bin;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOW
S\System32\Wbem
>
> does anyone know what i should do to fix this problem? (apart from the
> obvious of restarting the machine)




From stefanus.arief at ai.astra.co.id  Mon Sep 23 04:17:22 2002
From: stefanus.arief at ai.astra.co.id (Arief)
Date: Mon, 23 Sep 2002 15:17:22 +0700
Subject: os module and sys module~
In-Reply-To: <20020923080057.955.qmail@web21310.mail.yahoo.com>
Message-ID: <PLELIKCBPOPJEENEJHFBOEOPEBAA.stefanus.arief@ai.astra.co.id>

Based on Python documentation:

sys.path is "A list of strings that specifies the search path for modules.
Initialized from the environment variable PYTHONPATH, or an
installation-dependent default".

os.path is "Common pathname manipulations".

So in simple words, sys.path shows you only the search path where you
imports modules (using "import" statement).
The os.path is collection of functions that manipulate directory, file or
path name.


Mostly os module handles Operating System functions.
-----
Arief

-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]On
Behalf Of black
Sent: Monday, 23 September 2002 3:01 PM
To: python-list at python.org
Subject: os module and sys module~

Hi, all~
    I couldnt tell the difference between them immediately, and some same
method confused me much, for example they both have "path" attribute but
what's the difference ?




From Oschler at earthlink.net  Fri Sep  6 22:06:04 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sat, 07 Sep 2002 02:06:04 GMT
Subject: Multiple Python Apps on Linux
Message-ID: <gade9.25681$w51.7126863@news2.news.adelphia.net>

If I run several single-threaded Python apps simultaneously, each launched
as their own process running on Linux, are they truly independent or will
Python's "global library lock flag" come into play like it does for a single
Python app that is multi-threaded?

thx





From fperez528 at yahoo.com  Fri Sep 27 18:01:06 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Fri, 27 Sep 2002 16:01:06 -0600
Subject: Better terminal for windows?
References: <3D94B037.2020202@motorola.com> <mailman.1033158756.25404.python-list@python.org> <an2ivs$8fg$1@peabody.colorado.edu> <Xns9296F2B5187A6cliechtigmxnet@62.2.16.82>
Message-ID: <an2kf3$9bt$1@peabody.colorado.edu>

Chris Liechti wrote:

>> So some of ipython's nicest functionality (full support for tab
>> completion for both objects and files,  and color-enhanced
>> information, tracebacks and even syntax highlighted source info) are
>> gone under Windows.
>> 
>> If you know of a way to fix this, patches are welcome :)
> 
> well there is ansi.sys. if that driver is loaded the terminal undestands
> ansi escapes. long ago i used it under win95. it's still there in win2k but
> don't know how to use it. it the old days you needed to load it in
> config.sys.

Thanks for the tip. I'll include it in the docs for the next release, for the 
benefit of Win users. Maybe I'll hear from them and get that part fixed (I 
don't use Windows myself).

Cheers,

f.


From nhodgson at bigpond.net.au  Tue Sep 24 21:56:50 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Wed, 25 Sep 2002 01:56:50 GMT
Subject: Pythonwin (win32all build 148) on Python 2.2.1 on Win2K, bogus syntax e
References: <u47k9.46401$V7.11091779@twister.socal.rr.com>
Message-ID: <CJ8k9.16479$6g7.52569@news-server.bigpond.net.au>

Aaron Dimsdale and Alex Moser:

> Pythonwin barfed on the quote marks marked by carets, returning: "Failed
to run script -
> syntax error - invalid syntax".

   Try turning on View | End-of-line markers to see if you have non-standard
line endings such as pure carriage returns, which cause just this sort of
error. On Windows, you want line endings to be carriage return + line feed.

   Neil





From emile at fenx.com  Fri Sep  6 10:32:32 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 6 Sep 2002 07:32:32 -0700
Subject: Privacy and Inheritance
References: <mailman.1031112136.4594.python-list@python.org> <3D776CC2.CA928E38@fluent.com>
Message-ID: <alae3g$5ip$1@main.gmane.org>

"Mark Moales" <mmoales at fluent.com> wrote in message
news:3D776CC2.CA928E38 at fluent.com...
> Dennis,
>
> I think the point of the double underscores is to create "private"
> attributes similar to C++'s or Java's private keyword.  In other
words,
> only the class that defines the attribute can access it.  Subclasses
and
> others can't.  We use a single underscore here to denote "protected"
> attributes, or attributes that are accessible by subclasses.  However,
> the only way to enforce this in Python is by having well behaved
> programmers ;-)  Here's an example:

Just watch out for coincidental reuse of __privateVar in subclasses:

class BaseClass:
    def __init__(self):
        self.publicVar = 'Foo'
        self._protectedVar = 'Bar'
        self.__privateVar = 'Spam'
    def showme(self):
        print self.__privateVar

class SubClass(BaseClass):
    def aMethod(self):
        print self.publicVar
        print self._protectedVar
        # Can't do this because it's private
        # print self.__privateVar
        self.__privateVar = 'Ham'

a = SubClass()
a.showme()

a.aMethod()
a.showme()


--

Emile van Sebille
emile at fenx.com

---------






From loewis at informatik.hu-berlin.de  Fri Sep 27 08:59:39 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 27 Sep 2002 14:59:39 +0200
Subject: Unicode (inc. Japanese) in Python and TK
References: <f0fd5987.0209270334.47cbd9c0@posting.google.com>
Message-ID: <j4lm5n4mh0.fsf@informatik.hu-berlin.de>

gtaylor at lowebroadway.com (Garry Taylor) writes:

> I'm just seeking a little reassurance regarding use of unicode in
> Python, before I write my program, and then later find out I'm going
> to have to re-write in Java or something. Basically, am I going to run
> into any brick walls by doing the following:

It will all work out well if you are willing to adjust your
understanding of character sets, and Unicode, in the process of
developing this application.

> 1) Parsing XML with unicode characters.

XML, per se, does not have 'unicode characters'. XML is typically
stored in files, or other byte sequences. Those byte sequences have an
encoding. If the encoding is known to Python, you can parse the
document. Python will represent the strings as Unicode objects to you.

> 2) Measuring unicode characters using TK

I'm not sure whether this is possible.

> 3) Printing unicode to a file

Same comments as 1): If you know what encoding you want to use in the
file, and if Python supports that encoding, it will work. "unicode" is
not an encoding; "utf-8" and "ucs-2" are.

Regards,
Martin


From calvin at xmission.com  Tue Sep  3 19:01:17 2002
From: calvin at xmission.com (calvin)
Date: 3 Sep 2002 16:01:17 -0700
Subject: Mac os X 10.2 readline
Message-ID: <7bc40a7a.0209031501.77819c51@posting.google.com>

Readline is not built in the default python 2.2 installed with Mac os
X 10.2

Is there an easy way to rebuild it from /usr/lib/python2.2/config ? Or
do I have to download the source and rebuild it, yet again?

thanks

-calvin


From shalehperry at attbi.com  Fri Sep  6 18:08:29 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 6 Sep 2002 15:08:29 -0700
Subject: Larry Wall's comment on python...
In-Reply-To: <200209062206.g86M6UVm092251@mail.hcsmail.com>
References: <200209062206.g86M6UVm092251@mail.hcsmail.com>
Message-ID: <200209061508.29893.shalehperry@attbi.com>

On Friday 06 September 2002 15:06, mmaddox at hcsmail.com wrote:
>
> Having read Wall's comments on Slashdot, I think he is really just mildly
> complaining about ridiculously long page-widths in his editor, not really
> slashing Python seriously. He's just giving some props to his own baby at
> the expense of another. (He had to say something, right?) You must agree
> that Python can lead you to horizontal scrolling in your editor - the bane
> of user interface design.
>

that was the original poster's point.  When you start side scrolling it is 
time to refactor.  I find that I write python in much the same physical style 
as I wrote in C so this has always been how I approached coding.  However not 
everyone has.



From max at alcyone.com  Sat Sep 14 18:16:16 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 14 Sep 2002 15:16:16 -0700
Subject: newbie-question: abstract classes
References: <3D839852.2060405@snafu.de>
Message-ID: <3D83B530.100873B6@alcyone.com>

"T. Kaufmann" wrote:

> I want to create an abstract class but it doesn't work.
> What' s wrong here?
> 
> class Abstract:
> 
>      def __init__(self): return

As others have pointed out, it's strange to have an explicit return
statement in .__init__.  Better would be to use pass when the .__init__
method is truly empty; otherwise it's not necessary.

> class Concrete:

You don't say how it doesn't work, but one obvious issue here is that
Concrete doesn't derive from Abstract.  This should read:

	class Concrete(Abstract):
	    ...

An idiom I use for abstract classes is to have its .__init__ method
check to make sure that it truly is something other than the base class.

	class Abstract:
	    def __init__(self):
	        if self.__class__ is Abstract:
	            raise NotImplementedError

	    def method(self): raise NotImplementedError

	class Concrete(Abstract):
	    def __init__(self):
	        Abstract.__init__(self)

	    def method(self): print "in Concrete.method"

Now even trying to create an instance of the Abstract class will result
in an exception.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ A man that studieth revenge keeps his own wounds green.
\__/ Francis Bacon
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.


From fb at frank-buss.de  Sun Sep  1 04:57:42 2002
From: fb at frank-buss.de (Frank Buss)
Date: Sun, 1 Sep 2002 08:57:42 +0000 (UTC)
Subject: Crossword generator
References: <pan.2002.08.05.23.29.50.205338.718@thomas-guettler.de> <aknmta$h7f$3@newsreader2.netcologne.de> <a873nukqsjots9id3dgbnoomoajs3ieg8v@4ax.com>
Message-ID: <akskq6$8c3$1@newsreader2.netcologne.de>

Tim Roberts <timr at probo.com> wrote:

> These are only pseudo-crosswords.  Like most automatic crossword
> generators, it doesn't worry about symmetry, and it only crosses a few
> words here and there.
> 
> Creating a challenging, interesting, and legitimate crossword puzzle
> is very, very difficult.  Few humans can do it well, and it will be a
> while yet before they can describe their own algorithms in such a way
> that it can be automated.

I don't think so. Perhaps there are no program in Python, but here is an 
example for a program, which looks like it produces crossword puzzles with 
symmetry and lots of words crossing, and really fast online as a CGI 
script:
http://www.wordfit.com/dlframe.html

-- 
Frank Bu?, fb at frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de


From martin at v.loewis.de  Wed Sep 11 18:43:13 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 12 Sep 2002 00:43:13 +0200
Subject: SuSE 7.3 Yast2-able Python 2.2.1?
References: <IjPf9.1729$Lg2.215939@news2.news.adelphia.net>
Message-ID: <m3k7lsgnb2.fsf@mira.informatik.hu-berlin.de>

"Robert Oschler" <Oschler at earthlink.net> writes:

> I have the feeling that the lack of yast2's help, especially when it comes
> to SuSEConfig, is somehow causing other apps not to be able to find the
> version of OpenSLL that definitely _is_ installed on my system (0.9.6)

If you can't satisfy the dependencies, you can try installing building
the source RPM.

I'd recommend to update to SuSE 8.0, which comes with Python 2.2

Regards,
Martin



From aleax at aleax.it  Sun Sep 15 03:33:50 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 15 Sep 2002 07:33:50 GMT
Subject: two or more whitespace regular expression?
References: <am12a3$qo0$1@newsreader.mailgate.org>
Message-ID: <yJWg9.96506$ub2.1934842@news1.tin.it>

eugene kim wrote:

> i used this..
> 
> words = string.split(line, '['+string.whitespace
> +']'+'['string.whitespace+']+')
> 
> which didn't work..

The split method of a string (and the old split
function in module string, which internally just
invokes the split method) takes a string, not a
regular expression.  To use regular expressions,
you must import re first.  Then, for example:

import re
twoormorewhites = re.compile(r'\s\s+')

words = twoormorewhites.split(line)


Alex



From awinston at scn.org  Mon Sep  9 00:59:18 2002
From: awinston at scn.org (Alan Winston)
Date: Sun, 8 Sep 2002 21:59:18 -0700
Subject: Pysol Solitaire for Windows wanted
References: <algahc$e8q$1@news7.svr.pol.co.uk> <2MPe9.47925$Jo.5395@rwcrnsc53> <alh591$sqa$1@brokaw.wa.com>
Message-ID: <alh9r7$vni$1@brokaw.wa.com>

"Alan Winston" <awinston at scn.org> wrote in message
news:alh591$sqa$1 at brokaw.wa.com...
> [From rec.games.playing-cards -- crossposted to comp.lang.python]
>
> PySol requires Python because it is a Python program (hence the "Py" in
> "PySol") and Python is an interpreted language.
>
> I have done the Google exercise several times over recent months, and have
> never gotten a working Windows installable, only the pyc files, which
> briefly open a Python window, display some messages (presumably errors)
then
> close again before I can read the messages.
>
> It hasn't been quite a compelling enough problem for me to have resolved
it,
> since I have assumed that if I plunge in, the Windows install package will
> show up almost simultaneously with completion. but I would be rather
> interested in having the software working.
>
> The PySol card images are gorgeous, and I consider having the working
> software worth having just to play with them, although my computer
gameplay
> is almost exclusively PalmOS.
>
> --
> Alan Winston
> Seattle
>
>
>
> "Tom Sloper" <tomster at sloperama.com> wrote in message
> news:2MPe9.47925$Jo.5395 at rwcrnsc53...
> >
> > "PORK SODA" <Bogeys at Bum.com> wrote...
> > > I only seem to be able to find the macOS version at the moment.
> >
> >
> > Go to google.com and for your keyword search string type "pysol windows
> > download" - you will find several download sites. Looks like you may
also
> > need to download some other software called python for some reason.
> >
> >
> > --
> > Tom Sloper - Game Designer, Producer, Consultant, Author, Speaker.
> > Sloperama Productions. Services for game developers and publishers;
> "Making
> > Games Fun, And Getting Them Done."
http://www.sloperama.com/business.html.
> > 21 web pages of helpful free information and bulletin boards for game
> > industry aspirants; a new article every month.
> > http://www.sloperama.com/advice.html.
> > The Maj Exchange. Over 100 web pages of free information and bulletin
> boards
> > about the games of mah-jongg and hanafuda.
> > http://www.sloperama.com/majexchange/
> > Los Angeles, CA 90066, USA.
> > Tel: (310) 915-9945, Fax: (310) 745-0925, Cell: (310) DIG-SURF
> > tomster at sloperama.com
> >
> >
>
>




From wweexxsseessssaa at telusplanet.net  Sun Sep 29 15:18:30 2002
From: wweexxsseessssaa at telusplanet.net (John Hall)
Date: Sun, 29 Sep 2002 19:18:30 GMT
Subject: Why does QT Windows need Visual Studio?
Message-ID: <3ljepusnje8iuca90im2r9vceja5tfc5l7@4ax.com>

I was about to download the Windows non-commercial version of QT, but
noticed that it requires MS Visual Studio, so didn't.
This requirement surprises me - is there a way around that?

<rant>
I have the Visual Studio CDs and key, but can't install it.
I was wanting to get away from Visual Basic etc for all the usual
reasons, but after a lot of thrashing about trying other tools,
decided I would try VB just once more for a small project. Began the
install, but it hung. Now I can't use VB because it's not installed,
though it appears on the start menu. I can't re-install it because it
IS installed (or perhaps because it's NOT installed - my mind turned
to mush about then). I can't remove it because it's Not installed. I
don't want to reinstall Win 2000 Pro from scratch just to play with
something that is obsolescent.
</rant>

I'm working on an app in Rekall (Windows beta) but need to step
outside the main-stream Rekall stuff for a small but complex portion. 
This will be coded in Python & QT (like Rekall itself), but I thought
I should work on that completely outside of Rekall initially, and
integrate if when (or if) it works.

-- 
John W Hall <wweexxsseessssaa at telusplanet.net>
Calgary, Alberta, Canada.
"Helping People Prosper in the Information Age"


From agenkin at cdf.toronto.edu  Wed Sep  4 00:17:05 2002
From: agenkin at cdf.toronto.edu (Arcady Genkin)
Date: Wed, 4 Sep 2002 04:17:05 GMT
Subject: MD4 and DES for unicode strings (samba passwd file manipulation)
Message-ID: <uoit1mxubi.fsf@bashful.cdf.toronto.edu>

I'd like to manipulate on Samba's password file from Python in order
to set/change samba passwords.  Samba stores the password encrypted
two ways: with DES algorithm, and hashed with MD4[1].  I found and
installed 'pycrypto' module[2], which provides the above algorithms.
And yet I can't create the right crypted strings:

  >>> from Crypto.Hash import MD4
  >>> 
  >>> m = MD4.new()
  >>> m.update( u'111222333' )
  >>> m.hexdigest().upper()
  'A672E5EDFDAD90A15228E0AEC2AA9161'
  >>> # SAMBA-encoded password "111222333":
  >>> samba
  'B66B458E27BD714F99A05EE3C479FBF1'

Perhaps the problem lies with Unicode encoding?  Samba's man page (see
the link below) says that the string is obtained by taking MD4 hash on
the password, represented as a string of 16-bit little-endian unicode
characters.

Any other ideas?  Any help greatly appreciated.

Footnotes: 
[1]  http://www.samba.org/samba/docs/man/smbpasswd.5.html
[2]  http://www.amk.ca/python/code/crypto.html

-- 
Arcady Genkin : CDF sysadmin : http://cdf.toronto.edu/~agenkin/contact.html

guilt is the cause of more disauders
than history's most obscene marorders                       (E.E. Cummings)


From marklists at mceahern.com  Mon Sep 23 22:47:53 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 23 Sep 2002 21:47:53 -0500
Subject: confusion about opening files
In-Reply-To: <JHEOKEOOLIGLDHCMAHMOEEONEAAA.marklists@mceahern.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOAEOOEAAA.marklists@mceahern.com>

[I wrote]
> This is a FAQ.  You should probably find the Python FAQ and read
> it.  Short story:  Don't use 'from x import *' unless you know
> what you're doing.

I should have read the post I was replying to more carefully.  I saw the
"integer required" error and didn't notice that the OP was using os.open.
So the 'from x import *' was a red herring.  As Erik Max Francis pointed
out, the OP should use open instead of os.open.

My apologies.

Cheers,

// mark

-




From a_salnikov at yahoo.com  Wed Sep  4 22:47:57 2002
From: a_salnikov at yahoo.com (Andy Salnikov)
Date: Wed, 4 Sep 2002 19:47:57 -0700
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com>
Message-ID: <al6gn7$40c$1@usenet.Stanford.EDU>

"Gon?alo Rodrigues" <op73418 at mail.telepac.pt> wrote in message
news:m7ncnugbkm0qghf66qcub6e37b64fesdkq at 4ax.com...
> On 04 Sep 2002 20:38:30 +0200, loewis at informatik.hu-berlin.de (Martin v.
> L?wis) wrote:
>
> >Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:
> >
> >> >Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:
> >> >
> >> >> So why hasn't this been extended when nested scopes were introduced,
> >> >> e.g. by reusing the global directive or some other one?
> >> >
> >> >I think the main reason is that nobody could propose an acceptable
> >> >syntax. Reusing the global directive does not work.
> >>
> >> Can you explain why it does not work?
> >
> >Consider
> >
> >x = 1
> >
> >def foo():
> >  x = 2
> >  def bar():
> >    x = 3
> >    def baz():
> >      global x
> >      x = 4
> >
> >If you invoke baz(), which of the variables is changed? How do you
> >denote that you want to change the others?
>
> I think that crawling up the scope hierarchy and stopping at the first x
> that one finds (in the above case, the one defined in bar) is a sensible
> solution.

  This solution would have one big drawback. Suppose you wrote one day

def foo():
    x = 2
    def bar():
        y = 3
        def baz():
            global x
            x = 4    # intention here is to change x in foo()

and then few days later you add x to bar():

def foo():
    x = 2
    def bar():
        x = 3       # added code
        y = 3
        def baz():
            global x
            x = 4    # intention here is to change x in foo()

and suddenly it stopped working. You go into debugger, spend couple
of hours and then "suddenly" realize why it should not be this way :)

  Andy.





From henrik.motakef at web.de  Sun Sep 15 14:25:05 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 15 Sep 2002 20:25:05 +0200
Subject: Will xml.dom work with HTML docs?
References: <k3Tg9.6083$Lg2.908752@news2.news.adelphia.net> <877khntzvt.fsf@pokey.henrik-motakef.de> <j4y9a3qg3j.fsf@informatik.hu-berlin.de>
Message-ID: <871y7vnm9q.fsf@pokey.henrik-motakef.de>

loewis at informatik.hu-berlin.de (Martin v. L?wis) writes:

> Alternatively, see PyXML. It parses an HTML document into a DOM
> tree. It does so by knowing the HTML DTD, ie. it silently inserts
> missing closing tags whereever needed.

Does it do so by using an actual DTD, or is HTML-specific knowledge
hardcoded in it?

I.e. could this feature be used with other document types?

Regards
Henrik


From tebeka at cs.bgu.ac.il  Wed Sep 25 04:26:37 2002
From: tebeka at cs.bgu.ac.il (Miki Tebeka)
Date: 25 Sep 2002 01:26:37 -0700
Subject: Removing all occurences of a character from a string.
References: <amr8nh$rd8$1@driftwood.ccs.carleton.ca>
Message-ID: <33803989.0209250026.719df7b4@posting.google.com>

Hello Sean,


> I have a string, say "'Hello World'", where Hello World is in single
> quotes(i.e., 'Hello World'). How can I remove the single quotes so that
> "'Hello World'" becomes "Hello World"?

import re
re.sub("'", "", "'Hello World'")

HTH.
Miki


From phr-n2002b at NOSPAMnightsong.com  Thu Sep 19 04:02:33 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 19 Sep 2002 01:02:33 -0700
Subject: shared mem & semaphores module
References: <YWbi9.17569$V7.4455639@twister.socal.rr.com> <pan.2002.09.19.08.02.16.550254.723@thomas-guettler.de> <i6ei9.112316$ub2.2394246@news1.tin.it>
Message-ID: <7xr8fqquee.fsf@ruckus.brouhaha.com>

Alex Martelli <aleax at aleax.it> writes:
> Thomas Guettler wrote:
>         ...
> > AFAIK there is no module for shared memory.
> 
> mmap comes reasonably close.

mmap has no provision at all for semaphores.  Try:

  http://gigue.peabody.jhu.edu/~mdboom/omi/source/shm_source/shm.html

I wish the standard library would include this.


From shifflett at nps.navy.mil  Tue Sep 17 14:24:16 2002
From: shifflett at nps.navy.mil (D. Shifflett)
Date: 17 Sep 2002 11:24:16 -0700
Subject: Signals for GtkEntry widgets
References: <mailman.1031866395.5190.python-list@python.org> <6qqg9.509135$Aw4.21217664@bin2.nnrp.aus1.giganews.com>
Message-ID: <da4423a8.0209171024.38ee828a@posting.google.com>

I have found that the 'activate' event/signal occurs
when the Enter key is pressed in a GtkEntry widget.
Here is a snippet:

# field for text from user
usertext = GtkEntry()
usertext.connect("activate", userinput)

Thanks for your help
David Shifflett

"Alistair Thomas" <astavale at yahoo.co.uk> wrote in message news:<6qqg9.509135$Aw4.21217664 at bin2.nnrp.aus1.giganews.com>...
> In article <mailman.1031866395.5190.python-list at python.org>, "David
> Shifflett" <shifflett at nps.navy.mil> wrote:
> 
> > I am trying to find which signal is emitted when the user presses Enter
> > while in a GtkEntry widget.
> 
> The signal is key_press_event. Which returns a GdkEvent object. The
> keyval attribute tells you which key was pressed.
> 
> I use Glade and my sample code is based on my own module which uses libglade,
> hope this helps:
> 
> import gtk, GDK
> 
> # My own module based on libglade:
> import window 
> 
> class text_entry ( window.open ):
> 
> 	def __init__ ( self, title = 'Enter text:', text = '' ):
> 		window.open.__init__ ( self, 'text_entry' )
> 		self.add_handlers ( { 'on_key_press' : self.key_press,
> 						'closed' : self.closed } )
> 		self.get_widget ( 'text_entry' ).set_title ( title )
> 		self.modified = 0
> 		self.text = text
> 		a = self.get_widget ( 'text' )
> 		a.set_text ( text )
> 		a.grab_focus ()
> 		gtk.mainloop ()
> 
> 	def key_press ( self, widget = None, event = None ):
> 		if event.keyval in [GDK.Return, GDK.KP_Enter] :
> 			self.text = widget.get_text ()
> 			self.modified = 1
> 			self.close ()
> 			gtk.mainquit ()
> 
> 	def closed ( self, widget = None ):
> 		gtk.mainquit ()


From grabiller at 3dvf.net  Mon Sep 30 03:50:28 2002
From: grabiller at 3dvf.net (Guy Rabiller)
Date: Mon, 30 Sep 2002 09:50:28 +0200
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com> <874rc8d9c9.fsf@nospam.eml.cc> <mailman.1033319977.7828.python-list@python.org> <3d976fbc$0$23666$91cee783@newsreader02.highway.telekom.at> <8428na.7vj.ln@ix.netcom.com> <3d979184$0$20332$79c14f64@nan-newsreader-03.noos.net> <KAOl9.1635$vX6.52299899@newssvr21.news.prodigy.com>
Message-ID: <3d980247$0$23236$79c14f64@nan-newsreader-03.noos.net>

you are right, thx for correcting me.
I meant it works on w2k console etc..
--
guy rabiller
3d animator / td
grabiller at 3dvf.net
http://grabiller.3dvf.net
"Dennis Reinhardt" <DennisR at dair.com> wrote in message
news:KAOl9.1635$vX6.52299899 at newssvr21.news.prodigy.com...
> > >         python -v script.py 2>output.txt
> > yes it works on MS-DOS too.
>
> Stderr redirect works as shown on Windows 2000 (which is not DOS!), but
does
> not work for Windows 98 and earlier.
>
> --
>
> Dennis Reinhardt
>
> http://www.dair.com
>
>




From bokr at oz.net  Tue Sep  3 13:49:59 2002
From: bokr at oz.net (Bengt Richter)
Date: 3 Sep 2002 17:49:59 GMT
Subject: Design question.... parent/child class
References: <3d74b021$0$234$4d4ebb8e@news.nl.uu.net>
Message-ID: <al2so7$q34$0@216.39.172.122>

On Tue, 3 Sep 2002 14:56:21 +0200, "Gumuz" <gumuz at looze.net> wrote:

>hi all,
>
>i have a SessionManager class, which contains a list of Session
>objects(instantiated from a Session class). I have a design problem.
>
>Each session object has a small 'queue' of messages for the specific
>session. How can a session send a message to another session without having
>some kind of parent-reference to the SessionManager object?
>
>i am a bit puzzled, or is this parent-reference not such a bad idea after
>all?
>

Seems like the Session class is in a good position to keep track of the
list of session objects instantiated from it. I.e., the __init__ method
could add sessions to a directory or list maintained as a class variable
(a variable associated with the class, not each instance, but accessible
from all instances (and via the class name unless you prevent it)).

If you gave names or id numbers to sessions when they were created, you could
then access one session from another by name. E.g., (I'm not recommending this
verbatim. It's just a quick example to give ideas. Maybe you don't need
a separate SessionManager class, depending on what you really want to do
and how you want to factor things.

 >>> class Session:
 ...     sessions = {}
 ...     def __init__(self, name, *etc):
 ...         if name in self.sessions:
 ...             raise KeyError,'Session name "%s" already used.' % name
 ...         self.name = name
 ...         self.sessions[name] = self
 ...         self.etc = etc
 ...         self.msg_queue=[]
 ...     def accept_msg(self, msg):
 ...         print 'I am "%s," and I accepted message "%s."' % (self.name, msg)
 ...     def send_msg(self, dest, msg):
 ...         print 'I am "%s," sending "%s" message "%s"' % (self.name, dest, msg)
 ...         self.sessions[dest].accept_msg(msg)
 ...
 >>> s1 = Session('one', 1,2,'stuff')
 >>> s2 = Session('two')
 >>> s2.send_msg('one','Hi from instance s2')
 I am "two," sending "one" message "Hi from instance s2"
 I am "one," and I accepted message "Hi from instance s2."
 >>> s3 = Session('two')
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "<stdin>", line 5, in __init__
 KeyError: Session name "two" already used.

Note that the s1 and s2 variables weren't really necessary to access the sessions,
since they were recorded by name in the session directory:

 >>> Session.sessions['one'].send_msg('two','Hi from session "one"')
 I am "one," sending "two" message "Hi from session "one""
 I am "two," and I accepted message "Hi from session "one"."

 >>> Session('three')
 <__main__.Session instance at 0x007D5B70>
 >>> Session.sessions['two'].send_msg('three','Hi from 2')
 I am "two," sending "three" message "Hi from 2"
 I am "three," and I accepted message "Hi from 2."

You could define class methods to implement various session management
functions, so you could write e.g. Session.status() and get back or print
a report with status for each session, etc., etc.

The instance data is also available, of course:

 >>> vars(s1)
 {'etc': (1, 2, 'stuff'), 'name': 'one', 'msg_queue': []}
 >>> vars(s2)
 {'etc': (), 'name': 'two', 'msg_queue': []}
 >>> vars(Session.sessions['three'])
 {'etc': (), 'name': 'three', 'msg_queue': []}

and thus as attributes, e.g.,

 >>> s1.etc
 (1, 2, 'stuff')
 >>> s2.etc
 ()
 >>> Session.sessions['one'].etc
 (1, 2, 'stuff')
 >>> Session.sessions['two'].etc
 ()

Regards,
Bengt Richter


From duncan at NOSPAMrcp.co.uk  Fri Sep  6 11:01:31 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Fri, 6 Sep 2002 15:01:31 +0000 (UTC)
Subject: Difference between static method and a function as a class at tribute?
References: <mailman.1031321469.8366.python-list@python.org>
Message-ID: <Xns9281A252E74F2duncanrcpcouk@127.0.0.1>

sismex01 at hebmex.com wrote in news:mailman.1031321469.8366.python-
list at python.org:

> Any "plain" executable found as an attribute of a class
> object, upon creating an instance of that class, is bound
> to the instance as a method.  If not, then calling
> "inst.testmethod()" should display an empty tuple.

That isn't actually what happens. An unbound or bound method is created 
from the function when you actually reference it, not when you create an 
instance of the class. A good way to see this is:

>>> class C: pass

>>> def f(x): pass

>>> C.method = f
>>> inst = C()
>>> p = inst.method
>>> q = inst.method
>>> print p
<bound method C.f of <__main__.C instance at 0x009194E8>>
>>> print q
<bound method C.f of <__main__.C instance at 0x009194E8>>
>>> print id(p), id(q)
9472384 9252640
>>> 

Each access of inst.method actually gets a *different* bound method.

>>> r = C.method
>>> s = C.method
>>> print id(r), id(s)
9252512 9222064

The unbound methods are also created when you access them. So again you get 
a different one each time.

> I suspect that staticmethod(func) and classmethod(func)
> wrap the function in a proxy which first extracts some
> info from the actuall call (say, the instance's class)
> and pass *whose* arguments to the real function.

As of Python 2.2, when you try to get access an object which is stored in 
the class (not in the instance), the object's __get__ method is called. In 
the case of a function, its __get__ method creates a bound or unbound 
method as appropriate. staticmethod and class method wrap the function in 
an object with a different __get__ method:
>>> f.__get__(None, C)
<unbound method C.f>
>>> f.__get__(inst, C)
<bound method C.f of <__main__.C instance at 0x009194E8>>

staticmethod's __get__ simply returns the original function:
>>> staticmethod(f).__get__(None, C)
<function f at 0x00919878>
>>> staticmethod(f).__get__(inst, C)
<function f at 0x00919878>

classmethod's __get__ creates a method bound on the type of C instead of on 
C:
>>> classmethod(f).__get__(None, C)
<bound method class.f of <class __main__.C at 0x009151B8>>
>>> classmethod(f).__get__(inst, C)
<bound method class.f of <class __main__.C at 0x009151B8>>


-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From bloke at ii.net  Wed Sep 18 07:51:16 2002
From: bloke at ii.net (Rob Hall)
Date: Wed, 18 Sep 2002 19:51:16 +0800
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <Xns928CED692541cliechtigmxnet@62.2.16.82> <6_Ph9.479711$q53.16315681@twister.austin.rr.com> <am9jj9$c2m$1@wanadoo.fr>
Message-ID: <3d8868b1$0$28683@echo-01.iinet.net.au>

> Matter of taste, really... I do prefer Tkinter, because I find it far
> simpler than other GUI toolkits. But then, I've been into Tk for some
years
> now, and maybe I'm too used to it.
>

Yes I agree.  It is a matter of taste to some extent.  I evaluated both,
coming from a position of not knowing either, and wx seemed much simpler to
me.

Also, PythonCardPrototype is an excellent system for using wx.  It puts a
lot of the hard work in a separate file, letting you concentrate more on the
application logic.


> The first time I've looked at wxWindows and seen what's needed to open a
> simple window, it just made me run away screaming ;-).

I did that with tkinter.  I ran, and ran.  Then I found wx.  I thought it
wasn't so bad.  Then when I looked at tk again, it didn't seem too bad
either.


I settled on wx for the following reasons...
 - I needed to settle on something, and pretty soon
 - wx has a native look and feel
 - wx is fast - compiled C++, whereas tkinter is a script.  I found it to be
almost unacceptably slow in some instances.
 - PythonCardPrototype would simplify things even further.

Rob






From peter at engcorp.com  Wed Sep 18 23:14:25 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 18 Sep 2002 23:14:25 -0400
Subject: Minimal debug/rep functionality
References: <J%ai9.893$1D.574@nwrddc01.gnilink.net>
Message-ID: <3d894106@news.sentex.net>

Mark wrote:
> Hello all,
> 
> In a nutshell, what I'd like to do is develop a utility that would print out 
> every line in a python program as the program executes (I may even just 
> want to print line numbers).  I can see hints of how to do this in pdb (and 
> more directly in the underlying bdb class) and also in the code module.
> 
> So, 
> 
> 1) Should I try and inherit a class and override some methods or should I 
> just pull out the relevant code and hand spin a small interpreter.
> 
> 2) Anyone have a 20 liner solution for this?

Sorry, I couldn't find a solution that was larger than six lines:

 >>> def trace(frame, event, arg):
...   if event == 'line':
...      print '%s: %s' % (frame.f_code.co_filename, frame.f_lineno)
...   return trace
...
 >>> import sys
 >>> sys.settrace(trace)

Well, if you want to print the actual line of source you could probably
stretch it out to eight or ten lines... although I think you need
only one more line for the simplest approach.  ;-)

-Peter



From gerhard.haering at gmx.de  Sat Sep 21 16:12:21 2002
From: gerhard.haering at gmx.de (Gerhard Haering)
Date: Sat, 21 Sep 2002 22:12:21 +0200
Subject: mySQL and Python 2.2
In-Reply-To: <000801c2619d$64c3aba0$f8ffa8c0@lindstrom>
References: <000801c2619d$64c3aba0$f8ffa8c0@lindstrom>
Message-ID: <20020921201221.GA3576@gargamel.ghaering.test>

* Greg & Janet LINDSTROM <yuba at cyberback.com> [2002-09-21 13:33 -0500]:
> Hello-
> I am trying to convert from gadfly to mySQL for my database needs on 
> my Windows 98 box. I have installed mySQL and it works (I have 
> created my tables and can manipulate them manually), and have 
> downloaded mySQL-python-0.9.1, but when I python setup.py build, I 
> get an error about cl.exe not existing (and, sure enough, it 
> doesn't :-).

At http://sourceforge.net/project/showfiles.php?group_id=22307 download and
install MySQL-python-0.9.1.win32-py2.2.exe instead. I haven't gotten around
to building 0.9.2, yet.

> I'm running Windows 98 with Activestate Python2.2. Any help you 
> could supply (including code snippets of how to get connected to 
> MySQL once I get it running) would be appreciated.

Look in the examples and docs directory of the source distribution that you
downloaded. You should be able to find what you're looking for there.

-- Gerhard



From a-steinhoff at web.de  Sat Sep 21 06:59:41 2002
From: a-steinhoff at web.de (Armin Steinhoff)
Date: 21 Sep 2002 03:59:41 -0700
Subject: PyQt setup on Linux Mandrake 8.2
References: <3D8B4898.20808@nospam.attbi.com>
Message-ID: <ddc19db7.0209210259.7beb52d4@posting.google.com>

dwelch <dwelch91 at nospam.attbi.com> wrote in message news:<3D8B4898.20808 at nospam.attbi.com>...
> I would like to setup PyQt 3.4 on my Mandrake 8.2 box. I am having 
> difficulty, and I am worried that if I do the wrong thing, KDE will go 
> belly-up on me since it is Qt based.
> 
> First, question is: why isn't QTDIR defined by default? 

PyQt works with Qt2.x.x and Qt 3.x.x ... so YOU have to choose the
correct directory.

>Is QTDIR a 
> runtime setting or a development setting?

Both ..

> So I set QTDIR to /usr/lib/qt2 (that seems to be where the libqt2 RPMs 
> files are). Next question is: Since I have libqt2 already on my system, 
> how do I add the Qt development pieces without screwing up Qt/KDE? Is 
> /usr/lib/qt2 even the correct location?

Yes ... if you want to use only the latest version of Qt 2.x.x
 
> When I run SIP 3.4 build.py, it complains about tmake not being 
> available. Is tmake included with Qt2 development files?

It's available at the Trolltech homepage or its mirrors.

> SIP doesn't 
> seem to have any docs, so I'm not sure how to proceed.

Nomally you haven't to care about SIP ..

 
> Also, my RPM manager says that I have libqt2-devel (2.3.1-29mdk) 
> installed. No tmake listed in the files, though. And, PyQt-devel 
> (2.5-1mdk) is also listed as installed, but there aren't any .py files 
> included, only .sip files (I found this curious).

Corious ... there must be examples included.
 
I would suggest to dowload the latest version of Qt from the Trolltech
homepage, just for a simple and clean recompilation.

And do the same with SIP and PyQt from the Riverbank homepage ...

It seems to be better to forget that outdated stuff provided by the 
Mandrake distribution ...

Armin

> 
> Installing software on Linux makes me feel like such an idiot!
> 
> Any help or pointers would be much appreciated.
> 
> Don


From kkuroda at crl.ucsd.edu  Fri Sep 20 02:00:26 2002
From: kkuroda at crl.ucsd.edu (Kow Kuroda)
Date: Thu, 19 Sep 2002 23:00:26 -0700
Subject: Larry Wall's comment on python...
Message-ID: <7E6A03F4-CC5E-11D6-B1E5-00050287C0A6@crl.ucsd.edu>

On Thursday, September 19, 2002, at 09:32 PM, Courageous wrote:

>
>> Well, Python code may look awkward --- if not awk-like ---
>> if you don't know how to customize the tab width in your
>> editor of choice.
>
> Well, there's your problem already. One should use spaces
> and not tabs in Python code. Otherwise, any python file where
> tabs and spaces are mixed is also a choice of which indentation
> level other viewers of the code MUST use in their editors. That's
> heinous.

Well, that's not my job; it's the job of tabnanny.py.

Kow




From loewis at informatik.hu-berlin.de  Tue Sep 24 14:06:00 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 24 Sep 2002 20:06:00 +0200
Subject: Solaris 64 bit compilation
References: <c2595393.0209201844.2ed62265@posting.google.com> <m3sn03k44m.fsf@mira.informatik.hu-berlin.de> <c2595393.0209210627.44f672a8@posting.google.com> <m3d6r7i9o0.fsf@mira.informatik.hu-berlin.de> <c2595393.0209240941.5f3fc565@posting.google.com>
Message-ID: <j4fzvzclfb.fsf@informatik.hu-berlin.de>

phoebe_1 at att.net (Holden Caulfield) writes:

> > > "./Include/pyport.h", line 480: #error: "LONG_BIT definition appears
> > > wrong for platform (bad gcc/glibc config?)."
> > 
> > So what value has SIZEOF_LONG in pyconfig.h?
[...]
> To answer your questions:
>   - pyconfig.h does say 8 as SIZEOF_LONG

That raises the question then why the error in pyport.h is
produced. LONG_BIT comes from limits.h, and, if _LP64 is defined, is
64. So I'm still uncertain what you did to provoke this error.

>   - 'curses' modules does not compile. 
>   - 'dbm' modules compiles fine but core dumps during test.
> So, I disabled these modules in 'setup.py' and built it and ran the
> make test.

curses fails for me as well, since it finds a 32-bit ncurses
installation; dbm also core dumps here.

> test test_time failed -- Traceback (most recent call last):
>   File "./Lib/test/test_time.py", line 46, in test_mktime
>     time.mktime, (999999, 999999, 999999, 999999,
>   File "/apps/python/build/Python-2.2.1/Lib/unittest.py", line 279, in
> failUnlessRaises
>     raise self.failureException, excName
> AssertionError: OverflowError

This is a known bug (#460357), which has been fixed in CVS.

> 2)
> test_gettext
> test test_gettext crashed -- exceptions.OSError: [Errno 20] Not a
> directory: 'xx/LC_MESSAGES/gettext.mo'

I believe this also has been fixed in CVS.

> Are there any plans to support 64-bit python executable/module on
> solaris platforms for future versions?

Building python in 64-bit mode on Solaris works already. Some modules
may still not work; patches are welcome.

> Is anyone even interested in the configure.in patches?

I don't really see a need, since it works fine without the patches.

Regards,
Martin


From fgeiger at datec.at  Wed Sep  4 14:22:20 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Wed, 4 Sep 2002 20:22:20 +0200
Subject: Does BlackAdder really suck?
References: <slrnanci6i.1s8.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <al5ivg$e8v$1@rex.ip-plus.net>

"Gerhard H?ring" <gerhard.haering at opus-gmbh.net> schrieb im Newsbeitrag
news:slrnanci6i.1s8.gerhard.haering at haering.opus-gmbh.net...
> I just checked out the BlackAdder demo and was quite disappointed.
>


BlackAdder crashed my NT box, when I ran it the 1st time. Never fired it up
again...

> I expected that this is a tight integration of Python and PyQt. But it
> looks like it's just Qt Designer embedded in the IDE, while still having
to
> run pyuic manually. I don't know if at least the debugger is any good,
> because the demo was quite limited.
>
> Is there any usable commercial IDE for GUI design? BlackAdder and
> PythonWare Pro seem to be the only candidates (yes, I did check out Boa).

Did you have a look at wxDesigner?

Regards
Franz GEIGER





From gerhard.haering at gmx.de  Tue Sep 24 16:05:15 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 24 Sep 2002 20:05:15 GMT
Subject: Using internet modules to download files
References: <3ab719f7.0209241142.1bf5987@posting.google.com>
Message-ID: <slrnap1hbp.116.gerhard.haering@lilith.my-fqdn.de>

Stuart wrote in comp.lang.python:
> I have a slight problem.  The college which I attend has broadband
> access for everyone, but unfortunately insists upon placing a firewall
> between the pupils and downloads, therefore this fast connection is
> wasted upon retrieving HTML files.

So the "firewall" restricts you to downloading HTML documents only?
Are you not able to view any images in a browser? Can you not download
any .exe, .zip, .tgz files?

> I merely want to use the broadband to download a few more python
> libraries, such as win32all, as my home pc doesn't have internet
> access.
> 
> I have been experimenting with Python's internet protocols, in
> particular with trying to retreive files while bypassing the firewall.
>  I am not sure, however, which modules to use (i am a relative python
> neophyte).  I first tried with urllib, but that unforunately merely
> routes you throught the webbrowser, which of course, promptly blocks
> any attempt.

No, urllib doesn't use any webbrowser. It may be that you'll have to
instruct it to use a proxy server (and it sounds like your "firewall"
has one). Did you try urllib.urlretrieve, yet?

Using it like this:

urllib.urlretrieve("http://python.org/", "pydotorg.html") will
download the http://python.org entry page as the file pydotorg.html.
At least here ;-)

> Are then any modules which will bypass the
> browser/firewall, and allow you to download a file directly through
> the script?  Any advice (especially samples of code) would be greatly
> appreciated.

See the above one-liner.

-- Gerhard


From pieroul at attglobal.net  Mon Sep 30 23:30:42 2002
From: pieroul at attglobal.net (Pierre Rouleau)
Date: Mon, 30 Sep 2002 23:30:42 -0400
Subject: Is there a docutils / reStructuredText system for C++ source
 code available?
References: <3D98D9D9.8000009@attglobal.net> <anarjq$6n3$2@bob.news.rcn.net>
Message-ID: <3D9916E2.5030006@attglobal.net>


David Abrahams wrote:
> "Pierre Rouleau" <pieroul at attglobal.net> wrote in message
> news:3D98D9D9.8000009 at attglobal.net...
> 
>>I have been using docutil and reStructuredText for Python code and it
>>works great. Pydoc does a superb job of helping in setting up an
>>internal web site that provides documentation for our Python code, let
>>alone the ability to run doctest test and show auto testing of python
>>documentation on line!
>>
>>That works great for our Python code.  Now i want to do the same thing
>>for our C++ code.  Since Python has been partly implemented in C / C++
>>i tough there might already exist some tools for parsing C++ and some
>>comments (given some formatting conventions) and generate the same kind
>>of output the html tool of docutil does.
>>
>>Does any one know any thing that resembles that?
> 
> 
> See http://synopsis.sf.net
> 
> --
> -----------------------------------------------------------
>            David Abrahams * Boost Consulting
> dave at boost-consulting.com * http://www.boost-consulting.com
> 
> 
Thanks!
I also found another tool called Doxigen  at 
http://www.stack.nl/~dimitri/doxygen/

Anyone has experience with it?

-- 
         Pierre Rouleau



From pedronis at bluewin.ch  Fri Sep  6 11:43:45 2002
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Fri, 6 Sep 2002 17:43:45 +0200
Subject: Python and Jython inconsistencies when encoding strings
References: <ala00d$dij$1@wanadoo.fr> <j4ofbbclfp.fsf@informatik.hu-berlin.de>
Message-ID: <3d78ce33$1_3@news.bluewin.ch>

Martin v. L?wis <loewis at informatik.hu-berlin.de> wrote in message
j4ofbbclfp.fsf at informatik.hu-berlin.de...
> >>> s
> u"\u0153"
>
> Now, U+0153 is LATIN SMALL LIGATURE OE. It so happens that \x9c (what
> the terminal sends) is U+0153 in CP 1252 (which is the ANSI code page
> on your Windows installation). This might be a bug in Java, which
> assumes that bytes sent by the terminal are in the ANSI code page,
> when they are really in the OEM code page.

no it's more the Jython parser that does that, things can be fixed running
Jython as

jython -Dpython.console.encoding=cp850

on the other hand output seems buggy for:

print s.encode("cp850")

[I have reported that on our SF bug tracker]


 > > Does anybody know what is causing this inconsistency? Is there any way
to
> > avoid it?
>
> Yes. Don't use the console.

sticking to ascii there can avoid some troubles :).

regards




From loewis at informatik.hu-berlin.de  Wed Sep  4 12:47:34 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 04 Sep 2002 18:47:34 +0200
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com>
Message-ID: <j4wuq1d7mh.fsf@informatik.hu-berlin.de>

Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:

> So why hasn't this been extended when nested scopes were introduced,
> e.g. by reusing the global directive or some other one? 

I think the main reason is that nobody could propose an acceptable
syntax. Reusing the global directive does not work.

Regards,
Martin


From b.maryniuk at forbis.lt  Tue Sep 10 09:16:48 2002
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Tue, 10 Sep 2002 15:16:48 +0200
Subject: UNIX processes
In-Reply-To: <20020910083924.A4265@phy.duke.edu>
References: <200209101233.05796.b.maryniuk@forbis.lt> <20020910083924.A4265@phy.duke.edu>
Message-ID: <200209101516.48432.b.maryniuk@forbis.lt>

On Tuesday 10 September 2002 14:39, Michael Stenner wrote:
>   1) use the program "ps", which does the right thing for any given
>      os/kernel and presents a fairly standardized report, or
Yes. Especially, if it is not for Linux only.

>   2) pick the os(s) that you care about, and write your own code to do
>      it directly.  On linux, that can be done pretty nicely through the
>      /proc/ filesystem ("man proc").
Yes, I know this. Seems that no more way to do. :(

-- 
Regards, Bogdan

Apologies for taking up the bandwidth with the apology.  Anything else I
can apologise for ...... er no can't think of anything, sorry about that.
                Andy Hunt (Member of British Olympic Apology Squad)




From fredrik at pythonware.com  Wed Sep  4 03:14:08 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 04 Sep 2002 07:14:08 GMT
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <un50ccotaak86b@corp.supernews.com> <al0e96$oms$0@216.39.172.122>
Message-ID: <4pid9.5883$e5.1033514@newsb.telia.net>

Bengt Richter wrote:

> I agree about new syntax, but I wouldn't mind having a re.help(regexp) function
> for interactive use that would just explain in 'English' what the regexp expression
> stands for.

you can ask SRE to dump the internal parse tree
to stdout:

>>> sre.compile("[a-z]\d*", sre.DEBUG)
in
  range (97, 122)
max_repeat 0 65535
  in
    category category_digit

turning this into 'English' is left as an exercise etc.

</F>




From tmohr at s.netic.de  Sun Sep 15 00:29:34 2002
From: tmohr at s.netic.de (Torsten Mohr)
Date: Sun, 15 Sep 2002 06:29:34 +0200
Subject: Tkinter in PyRun_SimpleString(char* s)
Message-ID: <am12be$4j0$1@schleim.qwe.de>

Hi,

it doesn't seem to be as easy as i thought to load a script
from a C program and execute it, if i import Tkinter, it fails:


Traceback (most recent call last):
  File "<string>", line 31, in ?
  File "<string>", line 27, in __init__
  File "/usr/lib/python2.1/lib-tk/Tkinter.py", line 2249, in __init__
    Widget.__init__(self, master, 'frame', cnf, {}, extra)
  File "/usr/lib/python2.1/lib-tk/Tkinter.py", line 1750, in __init__
    BaseWidget._setup(self, master, cnf)
  File "/usr/lib/python2.1/lib-tk/Tkinter.py", line 1725, in _setup
    _default_root = Tk()
  File "/usr/lib/python2.1/lib-tk/Tkinter.py", line 1476, in __init__
    baseName = os.path.basename(sys.argv[0])
AttributeError: 'sys' module has no attribute 'argv'


The script i try to run is the example from the latest library
documentation of Tkinter.

The C program looks like this, it is also a slightly changed
example.

It seems there are some initialisations missing, what do i need
to do to set it all up properly?


Best regards,
Torsten.


int main(int argc, char** argv) {
        // routine to load a file
        f = slFileLoad("script.py", 0);

        Py_SetProgramName(argv[0]);

        Py_Initialize();

        PyRun_SimpleString(f->address);

        Py_Finalize();

        return EXIT_SUCCESS;
}




From tismer at tismer.com  Fri Sep 20 09:33:08 2002
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 20 Sep 2002 15:33:08 +0200
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <mailman.1032329366.6813.python-list@python.org> <j4ptvbaapm.fsf@informatik.hu-berlin.de> <3D890FB0.4030408@something.invalid> <m3admeo4u1.fsf@mira.informatik.hu-berlin.de> <ambso1$4g8vh$1@ID-59885.news.dfncis.de> <j4fzw6tofh.fsf@informatik.hu-berlin.de> <mailman.1032441028.15013.python-list@python.org> <j4elbqrrex.fsf@informatik.hu-berlin.de> <mailman.1032458669.32270.python-list@python.org> <m37khh93z2.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3D8B2394.4000202@tismer.com>

Martin v. Loewis wrote:
> Christian Tismer <tismer at tismer.com> writes:
> 
> 
>>>Academically, I don't think the results are unpredictable, as
>>>micro-thread switch occur at predictable points in processing.
>>
>>Yes, right now, but this will change pretty soon.
> 
> 
> You mean, tasklet switches will occur at unpredictable times?
> 
> This sounds like a bad idea.

Old Stackless had pre-emptive scheduling, and everybody
was happy with it. Microthreads were the major application.
I need to support this again, but in a much more fine-grained
implementation.
It is (again) the user's decision whether he allows
free scheduling of certain code. Some people love it,
some people hate it. I don't care, but support it all.

...

>>I have to do rigid locking of every unknown piece of code, until
>>somebody explicitly claims some code to be switchable.
> 
> 
> This, again, sounds terrible. I doubt it will help, and users will
> complain.

I don't think it makes sense to continue this
discussion in a public forum. Let's sit together
with a beer at some time, and you will see what
I have in mind. I'm not aiming at complaining
users. Instead, they will get the same safe default
behavior as with the old version, only better.
My fault is to provide a flare gun without a locking
bolt. Will to whatever it needs to fix this.

ciao - chris
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From aleax at aleax.it  Wed Sep 25 18:04:32 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 22:04:32 GMT
Subject: Python Compiling
References: <amfveg$5g64k$1@ID-53055.news.dfncis.de> <80a8af7d.0209250113.28ae4c72@posting.google.com> <pdgk9.140947$ub2.3077676@news1.tin.it> <mailman.1032971832.1946.python-list@python.org>
Message-ID: <Qpqk9.167824$pX1.5991467@news2.tin.it>

Lulu of the Lotus-Eaters wrote:

> Alex Martelli <aleax at aleax.it> wrote previously:
> |Or Haskell, www.haskell.org -=- HUGS inteprets it, the Edinburgh
> |compiler compiles it.  Scheme has _many_ interpreters and compilers
> |too.  Hey, so does Visual Basic... it's quite the thing these days.
> 
> Wrong Scottish city, I'm afraid.  Alex probably means the GLASGOW

Perfectly right, and thanks to you and the many others who mailed
me about it for catching out my silly mistake:-(.

The Edinburgh compiler is for *STANDARD ML*; the Glasgow compiler
is for *HASKELL*.  Whether I mixed up Scottish cities or major
functional programming languages, it was still a silly error to
make, and I apologize to Scotland, FP enthusiasts, and the
readership at large.


Alex



From marklists at mceahern.com  Thu Sep 12 10:48:20 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 12 Sep 2002 09:48:20 -0500
Subject: Is there a portable way to copy files in Python?
In-Reply-To: <3D80A78B.5030206@impathnetworks.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOMECJDOAA.marklists@mceahern.com>

[Pierre Rouleau]
> Subject: Is there a portable way to copy files in Python?

Have you tried shutil?

// m
-




From gerhard.haering at opus-gmbh.net  Tue Sep  3 12:59:40 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 3 Sep 2002 16:59:40 GMT
Subject: Servizi Internet
References: <mailman.1031056925.7383.python-list@python.org> <3bl9nu8924gihfj2nlpog874vfm73qqbh9@4ax.com>
Message-ID: <slrnan9qi1.25g.gerhard.haering@haering.opus-gmbh.net>

Ben Jamin wrote:
> excuse me but aren't commercial promotion and such things forbidden on
> newsgoups?

You can't really forbid things, because you have (almost) no means of
sanctioning people on usenet. The only means you have is complain to their
ISP (abuse@{name_of_ISP}). Wrt. to the post you noticed, it's certainly
against netiquette and the charta of this newsgroup to post off-topic
commercial spam here.

Announcements of Python-related commercial offerings are IMO on topic.

-- Gerhard



From duncan at NOSPAMrcp.co.uk  Mon Sep 16 12:05:37 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Mon, 16 Sep 2002 16:05:37 +0000 (UTC)
Subject: CFP: UK Python Conferece 2003
Message-ID: <Xns928BAD1086D00duncanrcpcouk@127.0.0.1>

                    *** Call for Participation ***
                      UK Python Conference 2003

                            April 2-3 2003
                         Holiday Inn, Oxford

The UK Python Conference 2003 is being held in conjunction with the
larger ACCU conference. The conference will include two days of papers
arranged in 90 minute speaking slots with either one or two tracks.

The audience is highly technical and will include experts in many
fields of computing, some of whom may know little or nothing about
Python. Talks will be welcomed on any aspect of Python, but we are
especially interested in the following:

        - Commercial applications for Python
        - Cross-language talks

Speakers are welcome from the UK or the rest of the world, however our
budget for overseas speakers may be limited.

To submit a proposal, send an email message to duncan at rcp.co.uk by 2nd
October 2003. In your message, please include the following:

        - Speaker name and affiliation
        - Contact address
        - Presentation title
        - A one paragraph abstract



-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?



From gerhard.haering at gmx.de  Wed Sep 18 17:24:25 2002
From: gerhard.haering at gmx.de (Gerhard Haering)
Date: Wed, 18 Sep 2002 23:24:25 +0200
Subject: Python as shared library on Unix
In-Reply-To: <EKEJLGEFHHFGJOBEGJECGEIICFAA.r.kalbermatter@hccnet.nl>
References: <EKEJLGEFHHFGJOBEGJECGEIICFAA.r.kalbermatter@hccnet.nl>
Message-ID: <20020918212425.GA6514@gargamel.ghaering.test>

* Rolf Kalbermatter <r.kalbermatter at hccnet.nl> [2002-09-18 22:48 +0200]:
> Under Linux I see no option to build a python.so.2.2.1 or similar to link my
> application to with the dl* functions.

There isn't any except in Python 2.3-CVS and Python's patched from various
distributors. Debian's Python builds a shared library, for example.

OTOH you don't really need to have a shared library to embed Python. Just link
statically against libpython{major}{minor}.a.

-- Gerhard



From spoermeg at voldelig.com  Tue Sep 17 19:43:42 2002
From: spoermeg at voldelig.com (Terje Johan Abrahamsen)
Date: Tue, 17 Sep 2002 18:43:42 -0500
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <4qLh9.1185$%U1.52435185@newssvr21.news.prodigy.com>
Message-ID: <am8eq3$3ml4d$6@ID-53055.news.dfncis.de>

"Don Bennett" <dpbmisc at pobox.com> wrote in message
news:4qLh9.1185$%U1.52435185 at newssvr21.news.prodigy.com...
> Terje Johan Abrahamsen wrote:
>  > If I would like to write some code that would not be possible for
> others to
>  > read, after I distribute the program, would Python be suitable?
>
> Yes.
>
>  > I know that
>  > there are a few programs like Py2Exe, that creates .exe files. But, I
> also
>  > saw this:
>  >
>  > --------------------
>  >  A zip-compatible archive is built, containing all files from this
>  >   directory as well as your script, and appended to a custom python
>  >   interpreter supplied with py2exe.
>  > --------------------
>  >
>  > Doesn't sound like it is protected very well. Is there some
> alternative to
>  > Py2Exe that can do the job, or should I look for a compiled language
>  > instead?
>  >
> Take a look at Tools/freeze.py.
>
> With this tool, you can compile your python files in to C files, and
> then link the C files with the python interpreter to create a
> stand-alone executable that represents your program.

Thanks... I will check that out...




From gerrit.muller at philips.com  Mon Sep 23 06:19:56 2002
From: gerrit.muller at philips.com (Gerrit Muller)
Date: Mon, 23 Sep 2002 12:19:56 +0200
Subject: How to get memory usage on Windows?
References: <amilhk$bi8$1@reader1.tiscali.nl> <amkpo5$kur$1@reader1.tiscali.nl> <f51sou8o4j9pkilhc3g6jevkqmkk7jr3pl@4ax.com> <aml3th$qu3$1@reader1.tiscali.nl> <ap9sou4crm9qjiilnqcov2m7uu2ktlqiiq@4ax.com>
Message-ID: <3D8EEACC.FBE54A37@philips.com>

<...snip...>
>> Have a look at  \Python22\Lib\site-packages\win32\Lib\win32pdhutil.py,
<...snip...>
> >I will have to try this at work, on a windows2000 machine.
> 
> That's what I'd do.
> 
> Another option would be using calldll
> (see http://www.nightmare.com/software.html). This is more dangerous,
> but will most probably work on any win32 system.
> 
The original routine based on win32pdhutil.py works on windows2000!

May be I have to try the nightmare route at home :-)

thanks Gerrit

-- 
--------------------------------------------------------------------
Gerrit Muller                Philips Research Laboratories Eindhoven
Building:   WDC 2 - 007              Prof. Holstlaan 4
Phone: +31 40 27 45242               5656 AA  Eindhoven
Fax:   +31 40 27 45033               The Netherlands
mailto:gerrit.muller at philips.com
http://www.extra.research.philips.com/natlab/sysarch/index.html


From johnroth at ameritech.net  Tue Sep  3 18:20:57 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 3 Sep 2002 18:20:57 -0400
Subject: Would you like a Snobol pattern matcher? (Re: Making regex suck less)
References: <3d725881.345921@news.t-online.de> <un7n06qrl1gm93@news.supernews.com> <3D740DFC.1040009@something.invalid> <un9bkdrqeh305c@news.supernews.com> <al30tq$aud$1@solaris.cc.vt.edu>
Message-ID: <unad783a7o5oa7@news.supernews.com>

"Carl Banks" <imbosol at vt.edu> wrote in message
news:al30tq$aud$1 at solaris.cc.vt.edu...
> John Roth wrote:
> > I don't know much about Snobol, unfortunately.
> > I think my biggest issue here is that we shouldn't
> > reinvent the wheel unless there is a good reason.
> > In other words, Larry is taking Perl in a specific
> > direction. Assuming we want to make major
> > changes to regex,
>
> I don't think anyone wants to make changes to regular expressions, per
> se.  I think this discussion is about adding a higher level syntax to
> regular expressions, that makes them more readable and versatile, for
> cases where that's important.

One of the things Perl has right now is the ability to use spaces
within regular expressions via a suffix flag. In Perl 6, this will
become
standard, and will include the ability to include comments!

Apocalypse 5 is well worth reading. Larry sets out why he
decided to redesign regular expression syntax. Much of the
reason was exactly what you specify: it's hard to read, hard
to understand, and has picked up lots of cruft over the years.

Making regular expressions "more versatile" is a great
idea. There's lots of things I'd like to do. However,
everything you load onto the poor thing makes it that
much more complex, hard to understand, hard to read
and errorprone. At some point, the mess has to be
redesigned. Since Python's re syntax isn't currently
as complex as Perl's (I think) it might not be there yet.

> > is there any _good_ reason
> > for doing something conceptually different and
> > consequently adding to the cacophony?
>
> I think there is a good reason to have a higher level syntax.  It
> serves a purpose heretofore unserved: that complex regular expressions
> no longer have to be unbearably unreadable.
>
> I could live without it, though.
>
>
> --
> CARL BANKS
> http://www.aerojockey.com




From tismer at tismer.com  Fri Sep 27 20:45:39 2002
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 27 Sep 2002 17:45:39 -0700
Subject: Is Boost Python version 2 ready?
References: <MvZk9.12124$Lg2.3584877@news2.news.adelphia.net> <an2pt9$g5h$1@bob.news.rcn.net>
Message-ID: <3D94FBB3.6060100@tismer.com>

David Abrahams wrote:
> "Robert Oschler" <Oschler at earthlink.net> wrote in message
> news:MvZk9.12124$Lg2.3584877 at news2.news.adelphia.net...
> 
> 
>>Is Boost Python version 2 ready for primetime yet?
> 
> 
> I'd say so, except for docs. We're not making any interface-breaking changes
> at this point, so anything you write will continue to work for some time to
> come. I'm finishing the last feature before release as we speak, and
> everything's been passing all tests on a wide range of platforms for quite a
> while. Docs will be finished over the next week or so.

I have been asked whether Boost will be compatible with
Stackless (or vice versa:).
Who else could answer that better than you?

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From spam at ob_scure.dk  Sun Sep  8 03:42:25 2002
From: spam at ob_scure.dk (Thomas Jensen)
Date: Sun, 08 Sep 2002 09:42:25 +0200
Subject: Using Python to call a DCOM server - "'Class not registered"
References: <3D787D95.9000907@ob_scure.dk> <5%ce9.1272$_l1.2960@news-server.bigpond.net.au> <3D7A1359.4010607@ob_scure.dk> <Ytwe9.5037$_l1.12309@news-server.bigpond.net.au>
Message-ID: <3D7AFF61.7070005@ob_scure.dk>

Mark Hammond wrote:
> Thomas Jensen wrote:
> 
>> Is it safe to say that "ActiveX DLL" = InprocServer32 and "ActiveX 
>> EXE" = LocalServer and that I need the later to use DCOM? (I have some 
>> existing components that I'd rather keep as DLL - don't know why 
>> really? :-)
> 
> 
> Yep.  InProcServer32 means that the COM object is implemented in a DLL, 
> and loaded into the calling process.  LocalServer means it is an 
> executable, and loaded into a separate process and IPC used across 
> boundaries.  For DCOM, there is no way the object can be loaded into the 
> same process, so executable based servers must be used.

Ok, I see. Then I guess the LocalServer approach has much more overhead 
on local calls than InProc32?

Searching for "unregister localserver" I found your book "Python 
Programming on Win32" on O'Reilly :-)
If I can find it in Denmark, I think I'll buy it!

/Thomas


-- 
Best Regards
Thomas Jensen
(remove underscore in email address to mail me)



From irmen at NOSPAMREMOVETHISxs4all.nl  Mon Sep 16 13:18:08 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Mon, 16 Sep 2002 19:18:08 +0200
Subject: pyro & parallel processing
References: <rLOb1LAK5fh9EwRi@jessikat.demon.co.uk> <uy9a1ritm.fsf@online.no>
Message-ID: <am53og$hdl$1@news1.xs4all.nl>

Syver Enstad wrote:
> Robin Becker <robin at jessikat.fsnet.co.uk> writes:
> 
>>Has anyone used pyro to do symmetric parallel processing. The pyro
>>framework is nice with master slave type applications, but the remote
>>objects don't seem to be able to message back to their invoker.
> 
> They can, checkout the callback examples in the Pyro distro. Basically
> what happens is that you pass a Pyro proxy as a parameter to the a
> method on the Pyro object. 

This is a good way of letting the 'other' Pyro object call methods on one
of your 'own' (or 'local') objects. It is a more symmetric way but
you have to pass callback object proxies as parameters.

The other - fully symmetric - way, is to build some Pyro objects
on both sides and register them all with the Name Server.
This way everybody can discover the other's objects by themselves.
You won't need to pass callback object proxies around.
Note that this requires a Pyro server (daemon) on all nodes,
but that is no different from the callback object solution.

Irmen de Jong

PS  the Pyro mailing list is usually a better place to ask questions
about Pyro, see http://lists.sourceforge.net/lists/listinfo/pyro-core



From sonnenbu at informatik.uni-bonn.de  Wed Sep 25 16:29:17 2002
From: sonnenbu at informatik.uni-bonn.de (Frank Sonnenburg)
Date: Wed, 25 Sep 2002 22:29:17 +0200
Subject: Newbie References Question
References: <3d91c72f$0$8558$79c14f64@nan-newsreader-02.noos.net> <amt4t7$f8a$1@f1node01.rhrz.uni-bonn.de>
Message-ID: <amt62m$fb0$1@f1node01.rhrz.uni-bonn.de>

I forgot the most important part of the example:



>>> p=Point(3,4)
>>> f=Figure(p)
>>> f
Figure, def. by (3,4)
>>> p.set(99,-5)
>>> f
Figure, def. by (99,-5)




Frank




From chris.gonnerman at newcenturycomputers.net  Sun Sep 29 22:19:10 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sun, 29 Sep 2002 21:19:10 -0500
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05> <an7bvd$burft$1@ID-125932.news.dfncis.de> <an7p7t$cpg$0@216.39.172.122> <mailman.1033338759.6775.python-list@python.org> <7xd6qwqtgz.fsf@ruckus.brouhaha.com>
Message-ID: <001701c26827$cb21d300$ba01010a@local>

----- Original Message ----- 
From: "Paul Rubin" <phr-n2002b at NOSPAMnightsong.com>


> "Chris Gonnerman" <chris.gonnerman at newcenturycomputers.net> writes:
> > >>> .70 * .05
> > 0.034999999999999996
> > 
> > Rounded to two decimals, that's 0.03; but done in proper
> > decimal numbers, the answer should be 0.035, rounded to 0.04
> > using either "classic" or "banker's" decimal rounding.
> > ...
> > So, Bengt, how do I do this right using only floats?  I can't.
> > I have to have decimal arithmetic to get the right answer.
> 
> How about just adding a small amount before rounding at the end?
> 
> >>> (.70 * .05) + .000000001
> 
> rounds to the right thing.

Cool.  Now PROVE that's right in all cases.

Like I said, decimal arithmetic is STILL the only choice.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From martin at v.loewis.de  Wed Sep  4 03:21:42 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 04 Sep 2002 09:21:42 +0200
Subject: Makefile.pre.in ?? installing postgresql --with-python
References: <al43r1$hk8$1@newsreader.mailgate.org>
Message-ID: <m33csqte2h.fsf@mira.informatik.hu-berlin.de>

eugene kim <eugene1977 at hotmail.com> writes:

> ./configure --with-python gives me this error (postgresql7.1.3 config)
> 
> configure: error: The Python extension makefile was expected at 
> `/usr/local/lib/python2.2/config/Makefile.pre.in\'

The Makefile.pre.in procedure for building extension modules is not
supported anymore. Extension authors should use distutils.

If you don't want to write a setup.py for postgresql, you can probably
work around this by disabling the relevant section of postgresql's
configure.in, and building the extension manually - you should look at
their Setup file to find any additional options they want to use for
the Python extensions.

Regards,
Martin


From nbecker at fred.net  Mon Sep  9 20:53:56 2002
From: nbecker at fred.net (nbecker at fred.net)
Date: 09 Sep 2002 20:53:56 -0400
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <un50ccotaak86b@corp.supernews.com> <al0e96$oms$0@216.39.172.122> <4pid9.5883$e5.1033514@newsb.telia.net> <al5m4u$agv$0@216.39.172.122> <e0Dd9.6122$e5.1077418@newsb.telia.net>
Message-ID: <m33csislzv.fsf@nbecker.fred.net>

IMHO, the thing that I really miss in making regex's understandable,
is that you can't include one regexp into another.  If you could, it
would be much easier to write clear expressions.  Flex, for example,
allows this.


From mertz at gnosis.cx  Thu Sep 19 17:10:44 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Thu, 19 Sep 2002 17:10:44 -0400
Subject: finding last match in a file
Message-ID: <U1ji9kKkXcRS092yn@gnosis.cx>                

|..find the last line which matches a certain experssion.  ...Is there
|some way to read through the file backwards and simply find first match
|starting from the end of the file.

I provide one solution in my _Text Processing in Python_ book
(forthcoming as dead trees, available now as ASCII).  The book itself is
at:

    http://gnosis.cx/TPiP/

In chapter 2, I have a section titled:

    PROBLEM: Reading a file backwards by record, line, or paragraph

I'm sure improvements are possible (it's meant to point to the approach,
not necessarily be the final code).  But what I do should be pretty
efficient (as I mention, for 2.2+, I'd rather use 'yield').

Yours, David...

--
---[ to our friends at TLAs (spread the word) ]--------------------------
Echelon North Korea Nazi cracking spy smuggle Columbia fissionable Stego
White Water strategic Clinton Delta Force militia TEMPEST Libya Mossad
---[ Postmodern Enterprises <mertz at gnosis.cx> ]--------------------------




From cliechti at gmx.net  Sat Sep  7 16:31:25 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 7 Sep 2002 22:31:25 +0200
Subject: Simulating "getopt" in functions (newbie)
References: <aldm7h$1orqou$1@ID-77524.news.dfncis.de>
Message-ID: <Xns9282E51C3554cliechtigmxnet@62.2.16.82>

Thorsten Kampe <thorsten at thorstenkampe.de> wrote in 
news:aldm7h$1orqou$1 at ID-77524.news.dfncis.de:

> Hi,
> 
> I want to simulate a "getopt"-like behaviour within a _function_. This 
> is to avoid having to define two functions: "function" and 
> "slightly_different_function".
> 
> If the function was a standalone program, it would work like this 
> ("v" for "verbose"):
> 
> $ function.py 5
> 1
> 
> $ function.py -v 5
> The result is: 1
> 
> I've thought of this to simulate it in a function:
> def function(a, b=None):
>     if b == "v":
>         print "The result is: 1"
>     else:
>         print "1"
> 
>>>> function(5)
> 1
> 
>>>> function(5, "v")
> The result is: 1
> 
> So: what is the "pythonic", "natural" way to do this in Python?

your idea is right. i would however vary the details:
if you just want to swich something then somthing:
  def function(a, verbose=0):
or you could use the presence of a file object in a param:

  def function(a, output=None):
     if output is not None:
         output.write("The result is:)
     print "1"

and i wouldn't use string parameters directly like you did. i someone makes 
a typo within that string you wont get errors, but if you use a variable as 
indirection you get a name error:
VEROSE="v"
function(23, VEROSE)

and using numbers instead of string would be more efficient.

chris

-- 
Chris <cliechti at gmx.net>



From marklists at mceahern.com  Wed Sep 11 18:10:42 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 11 Sep 2002 17:10:42 -0500
Subject: I must be an idiot, please pity me!
In-Reply-To: <VAOf9.2572$Pa7.155732@newsfep1-gui.server.ntli.net>
Message-ID: <JHEOKEOOLIGLDHCMAHMOAEACDOAA.marklists@mceahern.com>

[newbie]
> But it doesn't seem to work like that.

I don't know about all the mind readers out there, but you really haven't
given us any helpful information.  Are you getting an error?  What are the
symptoms of failure?  What are you expecting to happen?  Why not try to
simplify what it is you're trying to do and share *that* code?

For instance, the functions in each of your modules--do they take a database
connection as a parameter?

Have you tried stuffing everything into one module--just to see if it works?

Etc.

// m

-




From enoch at setim.org  Sat Sep  7 11:17:08 2002
From: enoch at setim.org (Enoch)
Date: Sat, 7 Sep 2002 17:17:08 +0200
Subject: untrusted dynamic module: win32api
Message-ID: <unk63h7q26lefe@news.supernews.com>

Hi,

I'm a Softimage|XSI user, a 3D graphic software that support ActiveX
Scripting on win32 platforms.

This software integrate a 'NetView', an integrated internet explorer.

To be able to communicate between the NetView and XSI I have to create the
XSI.Application object.

Works fine with VBScript and JScript, however I'm running into a problem
trying to make it work with Python.

Basicaly I have to include in my html page the following code:

<SCRIPT Language="Python">
import win32com.client
w = win32com.client.Dispatch("XSI.Application")
</SCRIPT>

But when I load this page into the NetView ( or even in an external
brother ) I have this error message:

'ImportError: untrusted dynamic module: win32api'

Is there any way to 'force' the import ?

Sorry if this has been discussed earlier, I'm relatively new to Python.

Tia.


Guy.


3d animator / td
grabiller at 3dvf.net
http://grabiller.3dvf.net





From imbosol at vt.edu  Wed Sep 11 15:25:09 2002
From: imbosol at vt.edu (Carl Banks)
Date: Wed, 11 Sep 2002 15:25:09 -0400
Subject: function pointer
References: <ffcedbc9.0209110215.644c1788@posting.google.com> <aln5ce$5kt$1@news.netpower.no> <slrnanu7i1.22k.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <alo6ic$sto$1@solaris.cc.vt.edu>

Gerhard H?ring wrote:
> David Brown wrote:
>> Functions are first-order objects (is that the right term?)
> 
> It is.


First-class objects.


-- 
CARL BANKS
http://www.aerojockey.com


From tdelaney at avaya.com  Thu Sep 19 19:47:15 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 20 Sep 2002 09:47:15 +1000
Subject: Minimal debug/rep functionality
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5B7@natasha.auslabs.avaya.com>

> From: Peter Hansen [mailto:peter at engcorp.com]
> 
> Hmmm... I didn't see any implication that speed was even remotely
> a concern for the OP.  After all, he did ask to *print* each line
> as it executed.  What could possibly be slower than that?
> (I doubt he would care to deal with the volume of information
> generated when using this with a large program anyway... needs
> a way to turn it on and off, at least.)

Very true - hence my qualification. The speed comes into effect when tracing
a long-running process - for example, while using the "standard" two
coverage tools (trace.py and coverage.py) performance was unacceptably slow
for coverage of some unit tests here (well, once I'd hacked them to actually
allow unittest to determine what was a test case ...).

> My standard comments about Python speed and Knuth's comments about
> premature optimization apply...

Oh yeah. Unfortunately, tracing becomes a real bottleneck if done wrongly -
I've achieved two orders of magnitude improvement since I first started on
this tool - both in performance and memory use.

> I'm looking forward to seeing your tool though, Timothy. :)

I'd love it to eventually become part of the core - I *expect* I'll be able
to release it (but of course, it's up to my employer ...). It's doing
everything I want now except for determining exactly which zero-count lines
can be safely ignored (comments, multiline expressions). Oh - and I haven't
got an actual "trace as you go" mode (which is the topic of this thread ;)
but that will be easy to add. And I haven't got any path coverage analysis
yet - the data is captured, but no analysis is performed. And it currently
does a relatively primitive text-only coverage dump. And ... 

:)

Tim Delaney



From observer at NOSPAM.space.pl  Thu Sep  5 06:30:13 2002
From: observer at NOSPAM.space.pl (Johann)
Date: Thu, 05 Sep 2002 12:30:13 +0200
Subject: urgent: binary driver for MySQL/Windows
Message-ID: <ivbenucnekgan8tp6vm2351h5pq9s5u488@4ax.com>

1. I am using the latest MySQL ODBC driver.
2. I  created ODBC entry in my Windows2K and I can connect to the
external database (test connection was OK)
3. I CANNOT connect to the same database using odbc.odbc('mydsn')
because it works as ODBC at intranet user not authorized with Windows
Server 2000.

Shit. I have no idea how to fix it so I tried the same with Perl using
DBI and DBD-MySQL and IT WORKS!  But.. I wanted to use Python. 

So, how to make Python work with MySQL? Is there any (other than odbc)
way to connection to MySQL? I found in DB-SIG MySQL module. But it is
only source, binaries are for Linux. I need windows binary driver for
my local Win2K workstation. I tried to compile it with Cygwin, but it
failed. :( I am lookin' for MySQL binary driver for Windows2K (for
Active Python  I am using) 

--
Johan


From howard at eegsoftware.com  Wed Sep  4 18:34:08 2002
From: howard at eegsoftware.com (Howard Lightstone)
Date: Wed, 04 Sep 2002 22:34:08 GMT
Subject: Python parser for cross references
Message-ID: <Xns927F9E6AC5904howardeegsoftwarecom@66.75.162.198>

I would like to use Source Navigator to (re)examine a largish application.
It parses the C, C++,assembly and Python jobs I am working right now and 
has outputs that *statisfy* the naive QA people who think tables and 
diagrams are *documentation*.

The Python parser for it was written by Thomas Heller (thank you very 
much).

However, it does not emit cross reference information.  Of course, I first 
thought "why not just add one?".  I can see why Mr. Heller did not.

A cross reference parser has to track scope....and determine function calls 
within argument lists.....and track just 'name' references which are passed 
around...and on and on.

Is there a (C) parser for Python around which does a "just-good-enough" 
parsing job to be adapted for this?  I think it just needs the scope (name) 
and enough logic to recognize what kind of token it has found.  

Of course, the gui guts of SN are written in Tix, Tcl, and ITcl .... oh 
well, back to those again.

Thanks

Howard Lightstone


From aleax at aleax.it  Thu Sep 19 03:42:37 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 19 Sep 2002 07:42:37 GMT
Subject: [Help]: mailbox classes
References: <jylimd-DAD12F.17305418092002@skates.gsfc.nasa.gov>
Message-ID: <Ndfi9.130668$pX1.4682176@news2.tin.it>

J. Li wrote:

> Hi,
> 
> I am trying to extract a message from a unix mailbox file. Python does
> provide few classes for easy and uniform access. So it says! After
> having read the libiary references, I went nowhere.
> 
> Can anyone give me an example as how to use these mailbox classes and

Say I want to know how many messages from GvR are in my inbox:

[alex at lancelot alex]$ python
Python 2.2.1 (#2, Jul 15 2002, 17:32:26)
[GCC 2.96 20000731 (Mandrake Linux 8.1 2.96-0.62mdk)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> from mailbox import UnixMailbox as MB
>>> inbox = MB(open('/home/alex/Mail/inbox'))
>>> from_guido = [x for x in inbox if x.get('from').find('Guido van 
Rossum')>-1]>>> len(from_guido)
108
>>>

Further, let's remove those such messages that are in the
python-dev mailing list:

>>> gvr_notdev = [x for x in from_guido if 
x.get('sender').find('python-dev')==-1]
>>> len(gvr_notdev)
5
>>>

etc, etc.


> what is it relationship with email.message classes.

Not much, intrinsically -- the items in a mailbox object by
default are rfc822.Message instances.  But you can pass as
the second argument to the constructor a factory with a
compatible interface, which should let you get email.Message
instances instead, if that's what you want.


Alex



From bfarwick at juno.com  Mon Sep 23 14:22:33 2002
From: bfarwick at juno.com (brrrent)
Date: 23 Sep 2002 11:22:33 -0700
Subject: cursor positioning and color
References: <slrnaoqv2q.hn4.TuxTrax@fortress.tuxnet>
Message-ID: <1b6dca50.0209231022.6fae7f6d@posting.google.com>

TuxTrax at fortress.tuxnet.net (TuxTrax) wrote in message news:<slrnaoqv2q.hn4.TuxTrax at fortress.tuxnet>...
> Being new to python, I would like to know how one goes about such 
> tasks as precise cursor positioning and output to screen in color.
> 

Hi Mathew,

Curses is the way to go when using Linux/Unix.

Wconio is the way to go when using Microsoft.
You can find in the Vaults of Parnassus.  If you have any trouble
locating it, let me know, I can email it to you.  Small library, works
well.

All the best,

brrrent


:snip


> 
> Cheers,
> 
> Mathew


From bdesth at nospam.free.fr  Thu Sep 12 00:25:59 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Thu, 12 Sep 2002 00:25:59 -0400
Subject: I must be an idiot, please pity me!
References: <VAOf9.2572$Pa7.155732@newsfep1-gui.server.ntli.net>
Message-ID: <3D801757.1020901@nospam.free.fr>

newbie wrote:
> Hello all. I'm getting very desperate now, so I have to ask for some help.
> I'm working on a small project and have slipped into what must be
> namespace hell.
> 
> I have a few controlling CGI scripts that use functions from other modules
> to build parts of a page. Each function in these other modules contains
> the SQL processing and HTML building code grouped by related data. The
> idea being I can reuse the functions in the other scripts. All works fine,
> providing each modules connects to the database as a stand alone job, or I
> pass a database connection object thing to each and every function. I
> suspect that both of these are wrong, due to the ugliness or nasty system
> overhead of the datbase connections.

 > What I thought I should be able to do is:
 >
 > CGI1.py:
 >
 > import mod1, mod2
 >
 > connect to database
 >
 > create part of page using mod1.functiona
 > create part of page using mod1.functionb
 > create part of page using mod2.functionz
 >
 > close database
 >
 >
 > But it doesn't seem to work like that. I've trawled though google and 
deja
 > for days (plus a couple of new rider books), and no one seems to hit this
 > problem. I must have a HUGE misunderstanding somewhere. I know mod1 and
 > mod2 cannot see CGI.py namespace, but there must be a way of sharing the
 > database connection, mustn't there?
 >
 > Help or a kick up the bottom would be really appreciated here!
 >
 > Sammy.




Creating a db connection in each function for sure not be very 
efficient.But what's wrong with passing the database connection to the 
functions in mod1, mod2, and modXXX ???

Anyway, you can make it simpler, like this :

in each module,
- create a global db_connection variable
- add an init() function, that takes a db_connection as argument and 
assign it to the module's global db_connection variable
- and don't forget to call this function before any other.

Another, more OO way to do this would be to rewrite your modules as 
classes, each getting a db_connection argument in it's __init__(), and 
then, in the main script, create the needed objects.

This second solution is somewhat safer with threading (if each thread 
creates it's own connection and it's own objects...)

HTH
Laotseu




From spoermeg at voldelig.com  Tue Sep 17 19:30:29 2002
From: spoermeg at voldelig.com (Terje Johan Abrahamsen)
Date: Tue, 17 Sep 2002 18:30:29 -0500
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <mailman.1032152058.11043.python-list@python.org>
Message-ID: <am8ept$3ml4d$1@ID-53055.news.dfncis.de>

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote in message
news:mailman.1032152058.11043.python-list at python.org...
> * Terje Johan Abrahamsen <spoermeg at voldelig.com> [2002-09-15 23:31 -0500]:
> > If I would like to write some code that would not be possible for
> > others to read, after I distribute the program, would Python be
> > suitable? I know that there are a few programs like Py2Exe, that
> > creates .exe files. But, I also saw this:
> >
> > --------------------
> >  A zip-compatible archive is built, containing all files from this
> >   directory as well as your script, and appended to a custom python
> >   interpreter supplied with py2exe.
> > --------------------
>
> py2exe contains Python byte code. Python byte code can be transformed
> into Python source code, but all comments are lost.
>
> > Doesn't sound like it is protected very well. Is there some
> > alternative to Py2Exe that can do the job, or should I look for a
> > compiled language instead?
>
> You should perhaps reconsider if and why you need to hide your source
> code.

If I am going to sell my programs, I do not want somebody else to take my
sourcecode, change a few things, and sell it themselves.




From DeepBleu at DeepBleu.org  Fri Sep 20 19:17:03 2002
From: DeepBleu at DeepBleu.org (DeepBleu)
Date: Fri, 20 Sep 2002 23:17:03 GMT
Subject: pyrhon and oracle
References: <5b8834c2.0209201412.50e77861@posting.google.com>
Message-ID: <P%Ni9.284614$eK6.8277262@twister.austin.rr.com>

What Oracle package are you using?
DCOracle2?  Something else?
DeepBleu

"hellprout" <hellprout at yahoo.com> wrote in message
news:5b8834c2.0209201412.50e77861 at posting.google.com...
> hi , i just want to have some detail about python and oracle
> what ca i use to query oracle with python ?
>
> i jsut want a little exemple
>
> thanks




From peter at engcorp.com  Tue Sep 10 15:52:07 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 10 Sep 2002 15:52:07 -0400
Subject: How do i print to a printer using python?
References: <6afb9d2c.0209041044.43c5f5f1@posting.google.com> <al73gj$lbt$1@wanadoo.fr> <6afb9d2c.0209091641.1d9b652e@posting.google.com> <3d7d6fd5@news.sentex.net> <uvg5d6b8b.fsf@online.no>
Message-ID: <3D7E4D67.67E42022@engcorp.com>

Syver Enstad wrote:
> 
> Peter Hansen <peter at engcorp.com> writes:
> 
> > Search for Mark Hammond's "win32all" package and install it, or
> > install the ActiveState distribution of Python which has it already
> > included.  win32ui is one of the modules included in that package.
> 
> Just to confuse further, win32all is not a python package. It's a
> distribution of python modules and packages. What I mean is that
> win32ui is a module that is not in a package, although it is included
> in the win32all distribution. I think I'll quit now, before I make
> things clearer. ;-)

Oops, my apologies for starting the confusion.  I stupidly used
conventional meanings for "module" and "package" instead of 
reserving those for their specific Python meanings.  Substitute
something "extensions" for package, and maybe "DLL wrapper" or 
"library interface" for module in my above message and maybe it 
becomes a little clearer.  Or not.

If not, well, we're already all confused enough that this won't
hurt much. :)

-Peter


From yaipa at aol.com  Thu Sep 19 13:01:23 2002
From: yaipa at aol.com (yaipa)
Date: 19 Sep 2002 10:01:23 -0700
Subject: linking extension modules w/ "ld" on Solaris - to be expected?
References: <mailman.1032393445.324.python-list@python.org>
Message-ID: <8d148763.0209190901.6276b37e@posting.google.com>

Skip Montanaro <skip at pobox.com> wrote in message news:<mailman.1032393445.324.python-list at python.org>...
> On a Solaris machine I configured Python (2.1.3 at the moment, but 2.2.1 is
> in the wings) using:
> 
>     ./configure --with-gcc=mycc --with-cxx=mycc
> 
> where "mycc" is a shell script which decides whether or not to invoke Sun's
> C or C++ compiler.  Some (but not all) 3rd party extension modules need to
> be compiled with C++, but I don't want to force all to be C++-compiled.
> (Gcc is not an option here.)
> 
> I noticed in the configure output that LDSHARED got set to "ld -G".  Is this
> typical for Solaris when building with a non-gcc compiler?

Skip,

I've tried compiling both Python 2.1.1 and 2.2.x on Solaris 8 and 9
with gcc 2.9x and 3.x in my path. I consistently get the same build
error each time, so it is likely that there is some *switch* I am not
setting somewhere.  When trying to compile Python I simply run
configure/make as 'root' without modification.  Configure runs without
error, but the build dies part way through claiming several 'function
declaration isn't a prototype'.

Is there a site which covers building Python in a Solaris environment
or maybe you could post a few pointers?

configure 'head'
----------------------
creating cache ./config.cache
checking MACHDEP... sunos5
checking for --without-gcc... no
checking for --with-cxx=<compiler>... no
checking for c++... c++
checking whether the C++ compiler (c++  ) works... yes
checking whether the C++ compiler (c++  ) is a cross-compiler... yes
checking for gcc... gcc
checking whether the C compiler (gcc  ) works... yes
checking whether the C compiler (gcc  ) is a cross-compiler... no
checking whether we are using GNU C... yes
checking whether gcc accepts -g... yes
checking how to run the C preprocessor... gcc -E
checking for AIX... no
checking for minix/config.h... no
checking for Cygwin environment... no
checking for mingw32 environment... no
checking for executable suffix... no
checking for --with-suffix... 
checking for case-insensitive build directory... no
checking LIBRARY... libpython$(VERSION).a
checking LINKCC... $(PURIFY) $(CC)
checking LDLIBRARY... libpython$(VERSION).a
checking for ranlib... :
checking for ar... no
checking for aal... no
checking for a BSD compatible install... /opt/sfw/bin/ginstall -c
checking for --with-pydebug... no
checking whether gcc accepts -OPT:Olimit=0... no
checking whether gcc accepts -Olimit 1500... no
checking whether pthreads are available without options... no
checking whether gcc accepts -Kpthread... no

make trace follows
------------------
....
# first sign of error in the make.

gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H  -o Python/pythonrun.o Python/pythonrun.c
Python/pythonrun.c: In function `initsigs':
Python/pythonrun.c:1348: warning: function declaration isn't a
prototype
Python/pythonrun.c: In function `PyOS_getsig':
Python/pythonrun.c:1448: warning: function declaration isn't a
prototype
Python/pythonrun.c: In function `PyOS_setsig':
Python/pythonrun.c:1470: warning: function declaration isn't a
prototype
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H  -o Python/structmember.o Python/structmember.c
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H  -o Python/symtable.o Python/symtable.c
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H  -o Python/sysmodule.o Python/sysmodule.c
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H  -o Python/traceback.o Python/traceback.c
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H  -o Python/getopt.o Python/getopt.c
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H  -o Python/dynload_shlib.o Python/dynload_shlib.c
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H  -o Python/thread.o Python/thread.c
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H  -o Modules/config.o Modules/config.c
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H  -DPYTHONPATH='":plat-sunos5:lib-tk"' \
        -DPREFIX='"/usr/local"' \
        -DEXEC_PREFIX='"/usr/local"' \
        -DVERSION='"2.2"' \
        -DVPATH='""' \
        -o Modules/getpath.o ./Modules/getpath.c
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H  -o Modules/main.o Modules/main.c
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H  -o Modules/gcmodule.o Modules/gcmodule.c
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H   -c ./Modules/threadmodule.c -o
Modules/threadmodule.o
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H   -c ./Modules/signalmodule.c -o
Modules/signalmodule.o
./Modules/signalmodule.c:89: warning: function declaration isn't a
prototype
./Modules/signalmodule.c: In function `signal_signal':
./Modules/signalmodule.c:213: warning: function declaration isn't a
prototype
./Modules/signalmodule.c:215: warning: function declaration isn't a
prototype
./Modules/signalmodule.c:226: warning: function declaration isn't a
prototype
./Modules/signalmodule.c: In function `initsignal':
./Modules/signalmodule.c:333: warning: function declaration isn't a
prototype
./Modules/signalmodule.c:337: warning: function declaration isn't a
prototype
./Modules/signalmodule.c:356: warning: function declaration isn't a
prototype
./Modules/signalmodule.c:358: warning: function declaration isn't a
prototype
./Modules/signalmodule.c: In function `finisignal':
./Modules/signalmodule.c:562: warning: function declaration isn't a
prototype
./Modules/signalmodule.c:570: warning: function declaration isn't a
prototype
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H   -c ./Modules/posixmodule.c -o Modules/posixmodule.o
./Modules/posixmodule.c: In function `posix_fstatvfs':
./Modules/posixmodule.c:4389: warning: passing arg 2 of `fstatvfs64'
from incompatible pointer type
./Modules/posixmodule.c: In function `posix_statvfs':
./Modules/posixmodule.c:4416: warning: passing arg 2 of `statvfs64'
from incompatible pointer type
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H   -c ./Modules/_sre.c -o Modules/_sre.o
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H   -c ./Modules/newmodule.c -o Modules/newmodule.o
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H   -c ./Modules/symtablemodule.c -o
Modules/symtablemodule.o
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H   -c ./Modules/xxsubtype.c -o Modules/xxsubtype.o
if test -f buildno; then \
        expr `cat buildno` + 1 >buildno1; \
        mv -f buildno1 buildno; \
else echo 1 >buildno; fi
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H  -DBUILD=`cat buildno` -o Modules/getbuildinfo.o
./Modules/getbuildinfo.c
rm -f libpython2.2.a
ar cr libpython2.2.a Modules/getbuildinfo.o
gmake: ar: Command not found
gmake: *** [libpython2.2.a] Error 127
Viking1# 


Thanks, any help greatly appreciated.

Alan Haffner


From m.faassen at vet.uu.nl  Fri Sep  6 16:54:32 2002
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 6 Sep 2002 20:54:32 GMT
Subject: Timothy Rue
References: <898662b7d85d2c966c0bd45d563d5d2d@xganon.com> <alakj8$20qo$1@nntp6.u.washington.edu>
Message-ID: <alb4m8$ksc$1@newshost.accu.uu.nl>

Donn Cave <donn at u.washington.edu> wrote:
> Quoth xganon <remailer at xganon.com>:
> | Does Timothy Rue still Troll here? 
> | If not, how did you get rid of him?
> | We need him out of our group now,
> | maybe we could get him into a Perl
> | group for a while and let them deal
> | with him?

Which group is that anyway? :) Seems like you and Tim have been posting
on comp.sys.amiga misc. Apparently you were looking for a Tim Rue 
FAQ at the time, and the thread has a link, FAQ posted in '99!

http://groups.google.com/groups?q=tim.rue.faq&start=10&hl=en&lr=&ie=UTF-8&scoring=d&selm=7r5lrp%24ftj%241%40ns2.dialnet.net&rnum=15

Fun, fun..

> If the problem is only that he posts to your group, then your
> group is way ahead of comp.lang.python.  When he was here, a
> number of regulars utterly lost touch with whatever common sense
> they may ordinarily possess, and it went on for what seemed like
> months.  If you don't want to engage in his discussion, then don't.
> I think you'll find that it works as well as anything.

I am a regular and I didn't have any common sense anyway. I had great
fun watching Tim Rue in action, though. Perhaps the Python Secret Underground
got to him, but since it doesn't exist that is somewhat unlikely.

I'm a fan of the ruebot. Did you know it was a rogue timbot? It doesn't
say so in the FAQ, but it's true. Perhaps this is secret PSU information,
in which case they would really not want me to say that in addition he


From Hobbes2176 at yahoo.com  Thu Sep 19 00:16:33 2002
From: Hobbes2176 at yahoo.com (Mark)
Date: Thu, 19 Sep 2002 04:16:33 GMT
Subject: Minimal debug/rep functionality
References: <mailman.1032406044.16132.python-list@python.org>
Message-ID: <Bcci9.8985$t%6.5877@nwrddc02.gnilink.net>

Okay, here's my first stab:

from bdb import *

class Mdb(Bdb):
    def user_call(self, frame, args):
        name = frame.f_code.co_name
        if not name: name = '???'
    def user_line(self, frame):
        import linecache
        name = frame.f_code.co_name
        if not name: name = '???'
        fn = self.canonic(frame.f_code.co_filename)
        line = linecache.getline(fn, frame.f_lineno)
        print '+++', frame.f_lineno, name, ':', line.strip()
    def user_return(self, frame, retval):
        None
    def user_exception(self, frame, exc_stuff):
        self.set_continue()

This is copied out of bdb.py and modified from the Tdb class.

To use it, you simply:

import mdb # contains definition of Mdb
import foo # contains function to trace
m = mdb.Mdb()
m.run('foo.main()') # main() is the method I'd like to trace

and out pops a list like (things are a hair cleaner if the input is done 
from a file instead of interactively):

+++ <line number> <code on that line>

My next wish is to NOT have it step into functions .... this isn't a real 
biggie but it would help eliminate some (soon to be) volumious output.

Regards,
Mark


From marco at sferacarta.com  Fri Sep  6 11:54:30 2002
From: marco at sferacarta.com (Marco Mariani)
Date: Fri, 6 Sep 2002 17:54:30 +0200
Subject: ttySx
In-Reply-To: <20020906152418.GA6937@sferacarta.com>
References: <200209061708.25677.b.maryniuk@forbis.lt> <20020906152418.GA6937@sferacarta.com>
Message-ID: <20020906155430.GA7147@sferacarta.com>

On Fri, Sep 06, 2002 at 05:24:18PM +0200, Marco Mariani wrote:

> Here it is.
> Of course, the code is old and sucks. If you have problems with lock
> files, I don't want to know :-)

I forgot an example.

Hope it helps.




import modem
import sys

if len(sys.argv)==3:
    command = sys.argv[2]
    serial = sys.argv[1]
else:
    sys.exit(1)


mod = modem.Modem(serial)
if not mod.open() and not mod.term_init():
    mod.set_speed(2400)
    if not mod.sendline(command):
        for i in mod.answer:
            print i
        mod.purge()
        sys.exit(0)
mod.purge()

sys.exit(1)




From bloke at ii.net  Fri Sep 27 21:44:55 2002
From: bloke at ii.net (Rob Hall)
Date: Sat, 28 Sep 2002 09:44:55 +0800
Subject: Problem with ClientCookie
References: <3d946d0f$0$10364@echo-01.iinet.net.au> <8b5e42a6.0209271211.4f0c6d9f@posting.google.com>
Message-ID: <3d950984$0$11916@echo-01.iinet.net.au>

Thankyou.  That did the trick.

I had also sent a copy of the message to the author of the package.  I must
confess, John (the author) was very prompt and helpful.  I will send him the
solution also.

Rob




From stasZ at linuxisbeter.nl  Wed Sep 18 08:41:51 2002
From: stasZ at linuxisbeter.nl (stasZ)
Date: Wed, 18 Sep 2002 14:41:51 +0200
Subject: Writing linux term color codes to a file
References: <e3gfouc157qqe6oj2i17g51edv8in2s26g@4ax.com>
Message-ID: <am9s83$p6u$1@reader12.wxs.nl>

On Wed, 18 Sep 2002 03:09:37 +0200, Andrei Kulakov wrote:

> Hello,
> 
> I ran into a bit of a problem here.. I'm making a small app that
> visually creates bash prompts (just for fun of it), and at one point it
> has to write the prompt to .profile, optionally with linux terminal
> color codes, like \[044;\]m or something like that (you know what I'm
> talking about). They don't come out right *in* the file. I tested the
> simple case where I'd have a few such codes in a string in python prompt
> and I'd look at it and it looks fine, but as soon as I write it out to a
> file and look at the file, it's "cooked".
> 
> I hope you know what I mean, if anyone wants I can post direct pastes.
> 
> Looks like write() does some sort of processing on the text; is there
> any way to turn that off?

>>>prompt = r'\[044;\]'
>>> prompt
'\\[044;\\]'

>>> f = open('/home/foo/prompt,'w')
>>> f.write(prompt)
>>> f.close()

$ more prompt
\[044;\]

I think your problem is solved if you use a raw string (r'spam').
With a raw string the backslash is just a backslash.

Stas Z

-- 
A computer scientist is someone who, when told to 'Go to Hell', 
sees the 'go to', rather than the destination, as harmful.


From spam07 at thack.org  Mon Sep 23 15:41:18 2002
From: spam07 at thack.org (Tapio Kelloniemi)
Date: Mon, 23 Sep 2002 19:41:18 GMT
Subject: Non-interpreted Python?
References: <3D8F65A4.38124945@engcorp.com>
Message-ID: <y7Kj9.400$UC2.36205@read2.inet.fi>

                          Re: Non-interpreted Python?
Peter Hansen <peter at engcorp.com> wrote:
>Tapio Kelloniemi wrote:
>>
>> I'm new to Python and would like to know, wheter it is possible to
>> compile Python code to binary executables (and on which platforms)?
>
[--]
>There are other approaches, but there's no point describing them until
>you explain *why* you want to do this.  What are your requirements?
I'm using Linux and searching for a good high-level language to use in
my projects. I don't have any special ideas, it depends on language. I
don't like the idea of using any interpreted language, but have heard so
much good about Python. I have used mostly C before and so it would be
best to give up my speed requirements.

>> Or better, if it is possible to compile code into object format and then
>> link with C/C++/whatever code.
>
>Python has excellent facilities for embedding in applications written
>in C/C++.  Check the documentation for the relevant sections.

I have thought of writing some parts of programs in Python (some kind of
library) and then be able to use it in C++/Ruby/anything else, if it is
possible.

--
Tapio



From mhuening at zedat.fu-berlin.de  Wed Sep 18 14:57:14 2002
From: mhuening at zedat.fu-berlin.de (Matthias Huening)
Date: 18 Sep 2002 18:57:14 GMT
Subject: BibTeX module?
References: <Xns928D9F1FBAA22mhueningzedatfuberli@130.133.1.4> <slrnaoh2i3.25g.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <Xns928DD52938F13mhueningzedatfuberli@130.133.1.4>

Gerhard H?ring <gerhard.haering at opus-gmbh.net> wrote in
news:slrnaoh2i3.25g.gerhard.haering at haering.opus-gmbh.net: 

> Did you try Google? It suggests
> http://canvas.gnome.org:65348/pybliographer/ 
> 

Yes, I did. But Pybliographer is a full-fledged bibliography application, 
which is an overkill for my purpose. I just need a simple Python-module 
to work with some BibTeX and/or Endnote/Refer files.

Thanks, Matthias


From =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?=  Fri Sep 27 14:52:41 2002
From: =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?= (=?iso-8859-5?Q?=B4=D0=DC=F8=D0=DD_=B3=2E?=)
Date: Fri, 27 Sep 2002 20:52:41 +0200
Subject: well-formed xml
References: <mailman.1033085436.9260.python-list@python.org>
Message-ID: <3d94a8ee$1@news.mt.net.mk>

>   from xml.dom import minidom

>   s = "<a href='http://google.com/search?hl=en&q=foobar'>search</a>"
>   #                                             ^
>   #                                             - seems to be the problem
>   #
>   # maybe it thinks I'm trying to reference the &q entity?

In XML the & character must be encoded as &amp;



-- 
??????

               Speak softly and carry a cellular phone.



From jepler at unpythonic.net  Fri Sep 20 15:27:55 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Fri, 20 Sep 2002 14:27:55 -0500
Subject: Python 2.2.1 compatible with Tcl/Tk 8.4?
In-Reply-To: <yu9965x05vkr.fsf@europa.research.att.com>
References: <yu9965x05vkr.fsf@europa.research.att.com>
Message-ID: <20020920192746.GC9734@unpythonic.net>

python2.3a0 (CVS) seems to work fine with Tcl/Tk8.4.  However, I don't
believe any special steps were taken to make it compatible.  Thus, I
suspect 2.2.1 will work fine too.

You will have to recompile Python, though, to make the upgrade.

Jeff



From henrik.motakef at web.de  Mon Sep 23 05:17:07 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 23 Sep 2002 11:17:07 +0200
Subject: os module and sys module~
References: <mailman.1032768177.27371.python-list@python.org>
Message-ID: <87znu9jca4.fsf@pokey.henrik-motakef.de>

black <quiteblack at yahoo.com> writes:

> I couldnt tell the difference between them immediately, and some
> same method confused me much, for example they both have "path"
> attribute but what's the difference ?

The sys module basically refers to the Python system itself, the os
module is about the operating system Python runs on.

Regards
Henrik


From hwlgw at hotmail.com  Sun Sep 29 07:18:07 2002
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 29 Sep 2002 04:18:07 -0700
Subject: Help please with menubars
References: <000f01c266f3$2903dbc0$2502a8c0@emily.ewndsr01.nj.comcast.net> <mailman.1033266279.6222.python-list@python.org>
Message-ID: <cb035744.0209290318.2b1fe7ef@posting.google.com>

[Chad Netzer]
> That was almost exactly what the code I posted did...  Hmmmph; Here is 
> a revision that exits when you click on the "Exit" entry in the 
> menubar, WITHOUT dropping down a menu.  

There is no visible difference between such an ``Exit'' entry (calling
a function directly) and a normal dropdown menu entry.  Do you know
how to display ``Exit'' like a Button?

'''
Give your very best today.
Heaven knows it's little enough.
'''


From skip at pobox.com  Fri Sep 13 15:35:18 2002
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 13 Sep 2002 14:35:18 -0500
Subject: transform list of int in a list of string??
In-Reply-To: <MNqg9.8303$WV.6617@fe03>
References: <-1483717605.1031940885180.JavaMail.nobody@webmail2.brturbo.com>
        <mailman.1031941396.19665.python-list@python.org>
        <MNqg9.8303$WV.6617@fe03>
Message-ID: <15746.15862.413159.408138@12-248-11-90.client.attbi.com>

    Skip> y = [int(n) for n in x]

    Steve> Skip of course meant

    Steve>     y = [str(n) for n in x]

    Steve> in his second example.

Yes he did.  Thanks for catching the bug.

S



From tim.hochberg at ieee.org  Thu Sep 26 15:39:14 2002
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Thu, 26 Sep 2002 19:39:14 GMT
Subject: How do you pronounce "__" (double underscore)?
References: <525eaba.0209260956.72bf6786@posting.google.com> <3D934C5C.4000603@mxm.dk> <20020926.130938.1789366143.1524@cybermesa.com>
Message-ID: <CnJk9.120675$Pf7.4559925@news1.west.cox.net>

"Jay O'Connor" <joconnor at cybermesa.com> wrote:

> How about "underder"??

Or maybe "dunder"?

-tim




From alex.andru at gmx.net  Mon Sep  9 15:08:32 2002
From: alex.andru at gmx.net (Alex Andru)
Date: Mon, 9 Sep 2002 21:08:32 +0200
Subject: ldap compile problem?
Message-ID: <alirhc$803$1@news.rrz.Uni-Koeln.DE>

hallo,

i have downloaded the python ldap driver but have problems to make build on
my mandrake 8.1 linux?
does anybody have an idea what is wrong?
i get the  same err message also on cygwin?

thanks for help,
alex


[alex at io python-ldap-2.0.0pre05]$ python setup.py build
running build
running build_py
warning: build_py: file Lib/ldap.py (for module ldap) not found
not copying Lib/ldap/__init__.py (output up-to-date)
not copying Lib/ldap/async.py (output up-to-date)
not copying Lib/ldap/functions.py (output up-to-date)
not copying Lib/ldap/sasl.py (output up-to-date)
not copying Lib/ldap/schema.py (output up-to-date)
not copying Lib/ldap/ldapobject.py (output up-to-date)
not copying Lib/ldap/modlist.py (output up-to-date)
not copying Lib/ldapurl.py (output up-to-date)
not copying Lib/ldif.py (output up-to-date)
warning: build_py: file Lib/ldap.py (for module ldap) not found
running build_ext
building '_ldap' extension
skipping Modules/LDAPObject.c (build/temp.linux-i686-2.1/LDAPObject.o
up-to-date)
skipping Modules/common.c (build/temp.linux-i686-2.1/common.o up-to-date)
gcc -O3 -fomit-frame-pointer -pipe -mcpu=pentiumpro -march=i586 -ffast-math 
-fno-strength-reduce -fPIC -DHAVE_LDAP_START_TLS_S -DLDAPMODULE_VERSION=2.0.
0pre05 -IModules -I/usr/local/openldap2/include -I/usr/include/python2.1 -c
Modules/constants.c -o build/temp.linux-i686-2.1/constants.o
Modules/constants.c: In function `LDAPinit_constants':
Modules/constants.c:159: `LDAP_SCOPE_DEFAULT' undeclared (first use in this
function)
Modules/constants.c:159: (Each undeclared identifier is reported only once
Modules/constants.c:159: for each function it appears in.)
error: command 'gcc' failed with exit status 1





From fperez528 at yahoo.com  Thu Sep 26 20:05:00 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 26 Sep 2002 18:05:00 -0600
Subject: indexing and searching pdf files
References: <pan.2002.09.26.18.05.37.70342.4621@presidency.com>
Message-ID: <an07bd$m8c$1@peabody.colorado.edu>

Rajarshi Guha wrote:

> My question - is there already something like this with python?
> Another question which is slightly off topic is, does anybody know of any
> articles/pages that talk about indexing text files efficiently - index
> generaion algorithms etc?

It doesn't directly discuss algorithms, but the following page is an 
_excellent_ overview of the problem:

http://freshmeat.net/articles/view/286/

Please post here any findings you may make on the python/indexing front. This 
is a problem I expect to have to deal with in a few months, so having it 
solved for me ahead of time would be most pleasant :) 

Full-text indexing of PostScript would also be nice, with provisions for 
automatic indexing of gzipped and bz2 compressed files. While we are at it, 
it would be nice to index code in at least C, C++, Fortran, Mathematica and 
Python and build tables of classes/functions defined in each file for the 
index database.

Mmmhhh, what else? Ah, if a .ps/.pdf has an associated .lyx/.tex file, that 
should be indexed instead, with the abstract, author, keywords, etc. fields 
properly recognized.

A search interface with a basic webserver would be enough. 

That would be a start. 

Not asking too much, am I?

Cheers,

f.


From fredrik at pythonware.com  Wed Sep  4 03:14:10 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 04 Sep 2002 07:14:10 GMT
Subject: Tkinter is extremely slow in drawing plots.... why???
References: <fa06e058.0208281116.8a6db88@posting.google.com> <3D6E4A62.56DF6302@earthlink.net>
Message-ID: <6pid9.5886$e5.1033593@newsb.telia.net>

Joseph A. Knapka wrote

> You can create the entire plot as a single canvas item,
> however, which might speed things up considerably.
> Assuming "dataset" is just a list of 2-tuples,
> then this should work:
>
> def flatten(l):
>   import operator
>   return reduce(operator.add,map(list,l),[])
> item = canvas.create_line(*flatten(dataset),fill="black")
>
> That code will give you a smooth line connecting
> all of your data points. The "flatten()" is just
> to convert your [(x1,y1),(x2,y2),..] into an
> unstructured [x1,y1,x2,y2], which is what create_line()
> wants.

create_line flattens its arguments all by itself (and in recent
versions, this is done by a really fast C function).

in other words, this should work:

    item = canvas.create_line(dataset, fill="black")

</F>

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




From gmuller at worldonline.nl  Sat Sep 21 16:39:15 2002
From: gmuller at worldonline.nl (GerritM)
Date: Sat, 21 Sep 2002 22:39:15 +0200
Subject: How to get memory usage on Windows?
Message-ID: <amilhk$bi8$1@reader1.tiscali.nl>

I am building an access log analyzer. For the first version I did not worry
about performance or memory use, the first exploration was about what is
practical information to extract and how to present. For current result see
http://www.extra.research.philips.com/natlab/sysarch/GaudiStatistics.html
and
http://www.extra.research.philips.com/natlab/sysarch/weblog/index2002.over.h
tml

For the second version memory use becomes an issue. The brute force version
keeps all information simultaneously in memory, without any size
optimization. This results now in memory use of about 250 MByte (measured
via very coarse windows resource tool), which scales linear with the amount
of (ever increasing) input. I would like to be able to measure more specific
how much memory is used per phase and/or per function. Is there any function
returning this information. I noted that on Unix the resource module gives
access to this kind of information, but I could not find any Windows
alternative.

regards Gerrit

--
www.extra.research.philips.com/natlab/sysarch/




From whisper at oz.net  Sat Sep 28 18:17:05 2002
From: whisper at oz.net (David LeBlanc)
Date: Sat, 28 Sep 2002 15:17:05 -0700
Subject: Getting started
In-Reply-To: <3D956756.7010502@cascade-sys.com>
Message-ID: <GCEDKONBLEFPPADDJCOEKEPPFDAA.whisper@oz.net>

<snip>
> I recall a Lisp implementation from long ago that did not distinguish
> between integers and machine addresses.  When applying a function to an
> arg list, if the function name evaluated to a number the interpreter
> blithely JSRed to that address.  This was a Feature, in that it was how
> system built-ins were linked with the higher level stuff.  User code did
> not normally do this and it was not a big issue in practice.  However it
> WAS a loophole that allowed type errors to occur, as it allowed code to
> jump to any random address in the system.
>
> But I think the generalization holds for most dynamic languages.
>
> Regards
>
> --jb
In fact, the names of some Lisp funtions came from the machine registers of
the same name: car (current address register) and cdr (current data
register) are the only ones I recall offhand.

Dave LeBlanc
Seattle, WA USA




From jetman516 at hotmail.com  Thu Sep 12 16:57:14 2002
From: jetman516 at hotmail.com (The Jetman)
Date: 12 Sep 2002 13:57:14 -0700
Subject: Python for DOS lives ?
Message-ID: <a38d43b0.0209121257.40a7df45@posting.google.com>

I did a search for Python 4 DOS just moments ago, on Google.  No 
joy.  I already knew the last iteration of Python4DOS (Python-DX) went 
belly-up sometime ago.  Then I got the thought to check out DJ Delorie's 
site (the author/maintainer of DJGPP.)  I found the following mail archive
msg:

http://www.delorie.com/archives/browse.cgi?p=djgpp/2002/05/19/15:30:25

The bottom line is:

http://phaseit.net/Python/python-djgpp.221.tar.gz    !!!!

The above mail announcement is from a May of this year, so it may be 
well along.  Haven't had a chance to test it, probably won't for a day or 
so, but I need it, so I'll definitely touch base in the next couple of days,  
if anyone else is interested.  According to the author of the port, it's 
working but isn't quite ready for prime-time....Jet


From roy at panix.com  Thu Sep  5 11:47:52 2002
From: roy at panix.com (Roy Smith)
Date: 5 Sep 2002 11:47:52 -0400
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com> <4f49d232.0209050418.7668c21@posting.google.com>
Message-ID: <al7ub8$crj$1@panix2.panix.com>

Laotseu <bdesth at free.fr> wrote:
> Now you should *also* learn other languages. But you'd better go with
> a lo level languages and a functionnal one - C and Lisp could be good
> choices.

I guess it depends on why you're learning languages.  If you're
learning languages to get a paying job as a programmer, lisp would be
a horrible choice.  If you're learning languages to study computer
science, lisp is looking a lot better.


From ajs at ix.netcom.com  Mon Sep 16 23:44:41 2002
From: ajs at ix.netcom.com (Arthur)
Date: Mon, 16 Sep 2002 23:44:41 -0400
Subject: Not sure how to phrase this question
Message-ID: <001501c25dfc$8e7b6b30$9865fea9@arthur>

Emille writes -

>I don't think you need __new__, or if you do, you'll want to read up on
>the existing documentation and prior discussions. Oh... I couldn't find
>the docs. There are some usage examples in the demo/newmetaclasses, but
>they only seem to exist in cvs. All this probably has a bit to do with
>the lack of response you mention

Yeah, if *I* need __new__ I know there is going to problems.

Seems there is indeed a simple solution which works with the infrastructure
I had already created - the obvious one that dawned on me after having made
my original post. My "factories" work simply as functions.

def Intersection(*args,**kw):
  dowhatigottodo
  if its the right thing to do under the circumstances:
     return PlaneIntersetion(args[0],args[1]
  else:
     do what makes more sense based on args.

class PlaneIntersection(new or old,shouldn't matter):
   def __init__(self,plane,line):
       self.plane = plane
       self.line = line
       takeitfromthere

i=Intersection(planeinstance, lineinstance)
#or Intersection(lineinstance,planeinstance ) the dowhatigottodo routine
handles the ordering issues
print i.__class__
>>class '__main__.PlaneIntersection'

As you suggest, this is probably just one of a number of ways to go, but
seems to work most simply for picking up from where I happen to be right
now.

There is something to be said, though, for sort of discovering this stuff,
rather than learning it in the textbook sense. As Terry posts suggests, this
is probably mundane standard procedure.  But, at least I have a pretty good
sense now of why.

Art






From peter at engcorp.com  Sun Sep 22 12:51:59 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 22 Sep 2002 12:51:59 -0400
Subject: Smalltalk vs. Python and COM was Re: Byte Magazine and Smalltalk
References: <MXFh9.23885$jF4.2037258@twister.southeast.rr.com> <3d87a686@tobjects.> <rd8fougqbgdivimn7vuest3ugoj57lvrr6@4ax.com> <Xns928CA588EF17Dtrippyparcplacecom@64.164.98.7> <3d87dd7c$1@tobjects.> <g9Vh9.313655$kp.1041470@rwcrnsc52.ops.asp.att.net> <Xns928D58DF09B6trippyparcplacecom@64.164.98.6> <20020918.115128.1605894428.1527@cybermesa.com> <amkpk0$or2$1@fe2.cs.interbusiness.it>
Message-ID: <3d8df525$1@news.sentex.net>

(Cross-posting removed to avoid language wars.)

Pino Gargiulo wrote:
> For this project I was considering Smalltalk and Python. The first so I
> could be some real stuff done and Python because apparently it permits 
> quite easy access the the mysterious (to me) m$ APIs. I really would
> want to stay awar from VC++, and I'm looking for some quite turnaround 
> evironment since I'll be learning COM on the job.

Could you please explain why you don't think Python could be used for
"real stuff"?  Do you have some other code already written in Smalltalk
with which you're trying to stay compatible, or do you have a library
in Smalltalk that you must use for this job?  I don't see any reason to
use two languages on a project where one will do the job.

-Peter



From tjreedy at udel.edu  Wed Sep 18 09:24:42 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 18 Sep 2002 13:24:42 GMT
Subject: Earing money with Python software (was: More on Protecting Source Code)
References: <mailman.1032232999.22661.python-list@python.org> <m3elbtcg8l.fsf@mira.informatik.hu-berlin.de> <4jt7ma.fse.ln@lenashee.goebel-consult.de> <8a6ba1da.0209180206.2d47f2bd@posting.google.com>
Message-ID: <u8%h9.10443$216.401412@bin4.nnrp.aus1.giganews.com>

"Richie Hindle" <richie at entrian.com> wrote in message
news:8a6ba1da.0209180206.2d47f2bd at posting.google.com...
> For library code (and tools like bridgekeeper which have a library
> component), my approach is to release a single version under two
> licenses, one free (as in both speech and beer) for use in Open
Source
> projects and one with a fee for use in closed source projects.

Perhaps we should call this a Chameleon license: the software takes on
the 'color' of its use-environment.

TJR




From mgerrans at mindspring.com  Tue Sep 17 02:34:25 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Mon, 16 Sep 2002 23:34:25 -0700
Subject: [FEEDBACK] Is this script efficient...is there a better way?
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <L9Rf9.168314$l_4.127784@atlpnn01.usenetserver.com> <4ZRf9.1782$Lg2.248733@news2.news.adelphia.net> <3hSf9.168947$l_4.98525@atlpnn01.usenetserver.com> <8PSf9.1796$Lg2.257378@news2.news.adelphia.net>
Message-ID: <am6idk$aj$1@slb2.atl.mindspring.net>

Well, since every other aspect of this little script has been beaten to
death, I'll add my 2.0*pow(10,-17) Kongbucks on the error messages: I make a
habit of including as much useful context information as I can, so for
instance, instead of:
...
except:
   print 'Failed to open log file.'

I might have something like:
...
except:
   print 'Failed to open %s for output.' % logFileName

This usually makes me a lot happier when the silly thing does fail (and
especially when someone else is using the script and I discover that they
don't have access to the same network resource I do, or don't have the same
configuration, etc.).   The traceback module is handy if you want to display
even more information from the exception.

- Matt






From skip at pobox.com  Thu Sep 19 10:35:54 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 19 Sep 2002 09:35:54 -0500
Subject: strong/weak - dynamic/static [Was: Getting started]
In-Reply-To: <JHEOKEOOLIGLDHCMAHMOIEOJDPAA.marklists@mceahern.com>
References: <F7DB8D13DB61D511B6FF00B0D0F06233024CC847@mail.hebmex.com>
        <JHEOKEOOLIGLDHCMAHMOIEOJDPAA.marklists@mceahern.com>
Message-ID: <15753.57546.804290.415645@12-248-11-90.client.attbi.com>

    Mark> I don't have any practical use for the following, but it seems to
    Mark> demonstrate that what you said is not true...

    [ snip bit about setting __class__ ]

That doesn't make Python weakly typed.  You have to be explicit about the
change.  An example of weak typing would be 

    a = "3.14159" + 7

Python raises a TypeError exception because it doesn't allow you to (can't
concatenate strings and ints or quietly coerce strings to numbers and
perform addition).  Perl, assigns the floating point value 10.14159 to a.

-- 
Skip Montanaro - skip at pobox.com
The need for gutters to be cleaned is directly proportional to how hard it
happens to be raining at the moment.



From tjreedy at udel.edu  Thu Sep 19 00:12:43 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 19 Sep 2002 04:12:43 GMT
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com>
Message-ID: <%8ci9.138477$5r1.5541622@bin5.nnrp.aus1.giganews.com>

"daniel w. moore" <dan65536 at hotmail.com> wrote in message
news:33b491f.0209181820.17b6ea70 at posting.google.com...
> It's occurred to me after some brainstorming on the subject that the
> C/perl construct
>
>   a ? b : c
>
> can be arranged like this in python:
>
>   (a and [b] or [c])[0]

If you *know* that b will evaluate as True (perhaps because it is a
non-null constant, then this can be simplified to 'a and b or c'.  I
first used this several years ago.

TJR




From jepler at unpythonic.net  Mon Sep 16 16:25:05 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Mon, 16 Sep 2002 15:25:05 -0500
Subject: difflib optimization -- wow (tim peters?)
Message-ID: <20020916152504.A7537@unpythonic.net>

wow, what a speedup!   thanks, whoever it was (tim?)

    $ time python2.3 -O ~/pyunidiff.py foo.c foo2.c > /tmp/pat

    real    0m5.457s
    user    0m4.640s
    sys     0m0.110s
    $ time python2.2 -O ~/pyunidiff.py foo.c foo2.c > /tmp/pat

    real    3m14.042s
    user    3m7.480s
    sys     0m1.180s

foo.c and foo2.c are ~300k long C files from different releases of a
piece of software.  the diff produced is nearly 250k.

pyunidiff is at http://unpythonic.net/~jepler/pyunidiff.py

tim, I only need another factor of 30x and it'll beat diff(1) -u in wall
time (or 5x to beat diff(1) -u -d).  It already gives files about 5%
shorter on the files I've tried (or on par with diff -u -d).  I notice
that the diff produced from 2.2 is marginally smaller, another 1.5% or so.
The price to pay for the optimization?

Jeff



From emile at fenx.com  Thu Sep 26 20:09:21 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 26 Sep 2002 17:09:21 -0700
Subject: This is driving me insane!
References: <20020926235643.37002.qmail@web20806.mail.yahoo.com>
Message-ID: <an07hc$e70$1@main.gmane.org>

Derek Basch
> http://www.geocities.com/dbasch/insane.txt
>
> I would expect to get:
> [0, 0]
> [0, 0]
> [[0, 0]]
>
> but I am getting:
> [0, 0]
> [0, 0]
> [[1, 1]]
>


Please post the code that is using the class.  As it is, there are only
two print statements in the class, so how you're getting three output
lines at all is in question...

--

Emile van Sebille
emile at fenx.com

---------






From phil (at) linux2000.com  Tue Sep 24 18:20:10 2002
From: phil (at) linux2000.com (Phil Edwards)
Date: Tue, 24 Sep 2002 23:20:10 +0100
Subject: Shadow password handling
Message-ID: <3d90e36d$0$123$7b0f0fd3@reader.news.newnet.co.uk>

Hi:

Does anybody know of a Python module which will allow username/password 
credentials to be validated ona Linux box which is configured to use md5 
and/or shadow passwords?

The application is a web portal which requires certain virtual sites on 
a Cobalt RaQ to be accessible only to particular users. Just in case 
anybody mentions .htpasswd files, the end user wants to be able to do 
all the username/password maintenance via the RaQ admin interface... :-(



From gerhard.haering at gmx.de  Thu Sep  5 19:17:11 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Fri, 6 Sep 2002 01:17:11 +0200
Subject: What database should I use
In-Reply-To: <Xns9280F2F7F8D12Latzikatz@139.174.2.24>
References: <mailman.1031257921.29400.python-list@python.org> <Xns9280F2F7F8D12Latzikatz@139.174.2.24>
Message-ID: <20020905231711.GB1229@lilith.ghaering.test>

* Lutz Schroeer <Lutz.Schroeer at kybernetik-manufaktur.de> [2002-09-05 21:46 +0000]:
> "A" <printers at sendme.cz> wrote in
> news:mailman.1031257921.29400.python-list at python.org: 
> 
> > I am going to program a small application (in Python and wxPython)
> > which consists about 2000 - 3000 records . Can you please recommend
> > which database is the best for that. I use Windows.
> > Maybe MS Access  or FoxPro or MySQL or better?
> > I want to keep the program as small as possible.
> 
> From your question I derive that you want to run the program on an MS 
> Windows system.
> 
> Access seems to be the best solution.

I disagree. Access requires a lot of setup (win32 extensions if you want
to use it via ADO, mxODBC if you want to use ODBC and have a maintained
module at the same time.) For databases this small, I'd even consider
using shelve or something else from the Python standard library.

PySQLite OTOH is Open Source, and the underlying SQL engine is
reasonably fast and has good support for SQL.

> [with Access] But keep in mind, that you're stuck to Billy's system.

Yes, and for no good reason, if you go this way. Better alternatives
exist, even several Open Source ones.

-- Gerhard




From sismex01 at hebmex.com  Thu Sep 12 13:07:49 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Thu, 12 Sep 2002 12:07:49 -0500
Subject: [ANN] PySQLite 0.3.0 released
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233023FF87B@mail.hebmex.com>

> 
> "Gerhard H?ring" <gerhard.haering at opus-gmbh.net> wrote in message
> news:slrnao0e6p.23c.gerhard.haering at haering.opus-gmbh.net...
> >   A Python-DB API 2.0 compliant extension module that connects to
> >    SQLite databases.
> ...
> >     row = cursor.fetchone()
> >
> >     while row != None:
> >             print "%14s, %15s, %19s, %8s, %25s" % tuple(row)
> >             row = cursor.fetchone()
> 
> Does anyone know whether there are any plans to update the Python-DB
> API to the new iterator protocols, so that the above could be written
> more simply something like
> 
> for row in cursor:
>        print "%14s, %15s, %19s, %8s, %25s" % tuple(row)
> ?
> 
> Terry J. Reedy
> 

You can already do it, using iter():

for row in iter(cursor, None):
	print ....


-gus

Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From fgranger at alussinan.org  Fri Sep 20 13:27:11 2002
From: fgranger at alussinan.org (=?ISO-8859-1?Q?Fran=E7ois_Granger?=)
Date: Fri, 20 Sep 2002 19:27:11 +0200
Subject: [Help]: mailbox classes
References: <jylimd-DAD12F.17305418092002@skates.gsfc.nasa.gov> <Ndfi9.130668$pX1.4682176@news2.tin.it> <1fitpvg.gi8n0y1iem93tN%fgranger@alussinan.org>
Message-ID: <1fitpz7.1h4nj4m19gs2x5N%fgranger@alussinan.org>

Fran?ois Granger <fgranger at alussinan.org> wrote:

> Alex Martelli <aleax at aleax.it> wrote:
> 
> > J. Li wrote:
> > 
> > > Hi,
> > > 
> > > I am trying to extract a message from a unix mailbox file.
> ...
> > 
> > > what is it relationship with email.message classes.
> > 
> > Not much, intrinsically -- the items in a mailbox object by
> > default are rfc822.Message instances.  But you can pass as
> > the second argument to the constructor a factory with a
> > compatible interface, which should let you get email.Message
> > instances instead, if that's what you want.
> 
> email does not handle the unix mailbox properly. It sees only the first
> message.
> 
> I was going the way you suggest by first using mailbox.UnixMailbox, then
> i'd like to pass each message given by the next() method to email for
> further analysis. But this is way over my current level of understanding
> of Python objects and classes. ....

after some thought and some trial and errors, I came with this solution.
I don't like it, it look bad.

def main():
    fp = file(folderfile)
    folder = mailbox.UnixMailbox(fp, factory=email.message_from_file)
    while 1:
        mail = email.message_from_file(fp)
        print mail
        if not folder.next():
            break
        #mail = email.message_from_file(fp)

if __name__ == '__main__':
    main()


-- 
"Ce qui est gratuit ne vaut pas cher" - inconnu


From tjreedy at udel.edu  Sun Sep 15 00:35:03 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Sun, 15 Sep 2002 04:35:03 GMT
Subject: two or more whitespace regular expression?
References: <am12a3$qo0$1@newsreader.mailgate.org>
Message-ID: <X5Ug9.40826$gf6.1295831@bin4.nnrp.aus1.giganews.com>

"eugene kim" <eugene1977 at hotmail.com> wrote in message
news:am12a3$qo0$1 at newsreader.mailgate.org...
> i used this..
>
> words = string.split(line, '['+string.whitespace
> +']'+'['string.whitespace+']+')
>
> which didn't work..
> thanks

What weres 'line' and 'words'?






From see_reply_address at something.invalid  Wed Sep 25 22:08:27 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Thu, 26 Sep 2002 14:08:27 +1200
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <3D7958DB.50903@nospam.free.fr> <3D87402E.6020700@thinkware.se> <amsfib$cq1$1@nyheter.chalmers.se>
Message-ID: <3D926C1B.7080204@something.invalid>

Jacob Hallen wrote:

> Please note that I didn't say anything about block
> delimiters or parentheses in this. I think they were put into programming
> languages to make it easy to write fast compilers.


I'm not sure you should lump parentheses in with
block delimiters in that statement. They've been
in use by mathematicians for centuries before
compilers were invented, and, used in moderation,
I don't think they have any adverse effect on
readability.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From dsavitsk at e-coli.net  Mon Sep  9 12:51:11 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Mon, 09 Sep 2002 16:51:11 GMT
Subject: random.shuffle question
Message-ID: <3k4f9.7500$yt3.3831288@newssrv26.news.prodigy.com>

why does this not work to shuffle part of a list?

>>> l = [1,2,3,4,5,6,7,8,9]
>>> random.shuffle(l[4:8])
>>> print l
[1, 2, 3, 4, 5, 6, 7, 8, 9]

and more importantly, is there a better way to do this?

-d




From shalehperry at attbi.com  Sat Sep  7 17:29:26 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Sat, 7 Sep 2002 14:29:26 -0700
Subject: Larry Wall's comment on python...
In-Reply-To: <1549197.0PZNOebDTH@cartman>
References: <mailman.1031377930.4225.python-list@python.org> <1549197.0PZNOebDTH@cartman>
Message-ID: <200209071429.26401.shalehperry@attbi.com>

On Saturday 07 September 2002 13:46, yka wrote:
> Kow Kuroda wrote:
> > Well, Python code may look awkward --- if not awk-like ---
> > if you don't know how to customize the tab width in your
> > editor of choice. For example, vi and emacs indent by 8
> > characters by default, which I suspect is too wide for Python.
> > Some code with deep nesting wouldn't look very nice with
> > this default.
> >
> > Kow
>
> Well at vim :set ts=4 makes more fitting

actually emacs defaults to 4 space indenting and does not bother with tabs.  I 
find that code which plays games with how large a tab is always looks ugly 
unless viewed in the same editor it was written in.



From sross at connectmail.carleton.ca  Tue Sep 17 19:59:54 2002
From: sross at connectmail.carleton.ca (Sean Ross)
Date: 17 Sep 2002 23:59:54 GMT
Subject: Splitting a list of strings
Message-ID: <am8flq$23i$1@driftwood.ccs.carleton.ca>

Hi.
I need to read a list of strings from a file, weed out the comments and
separate "attribute" descriptions from the data. I have a working
implementation, but I'm interested in finding out if there's a better way to
do it.

Here's the code:

def getDataAndAttributes(filename):
    "Filters *.arff file into lists of data and attribute strings"
    lines = open(filename).readlines()

    # All lines beginning with "@'" are attributes, except the last one.
    attributes = [str for str in lines if str[0] == "@"]
    # Remove attributes from lines, so we don't process them again to find
data.
    lines = [str for str in lines if not str in attributes]
    attributes = attributes[:-1]

    # Filter comments out to get data.
    data = [str for str in lines if not str[0]=="%" ]

    return (data, attributes)

What I'm interested in knowing is, can I extract all of the "attribute"
strings from lines in such a way that I get my list of "attribute" strings,
but all of those strings have been removed from lines as well, all at the
same time? i.e., without filtering the list a second time. As I've stated,
the code above works, but I'm concerned that it may be sub-optimal.

Thanks, in advance, for your suggestions,
Sean Ross




From gregory.p.green at boeing.com  Mon Sep 23 15:43:37 2002
From: gregory.p.green at boeing.com (Greg Green)
Date: Mon, 23 Sep 2002 19:43:37 GMT
Subject: Python COM and variable number of arguments
References: <mailman.1032529407.31552.python-list@python.org> <sn04lhyb.fsf@boeing.com> <TOPi9.1071$6g7.2034@news-server.bigpond.net.au>
Message-ID: <ofao4hli.fsf@boeing.com>

Mark Hammond <mhammond at skippinet.com.au> writes:

> You could also try passing with pythoncom.Missing or pythoncom.Emtpy
> in place of the positional arguments.  You may need to experiment to
> see what one works - it depends on the COM object.

Mark,

Thanks. pythoncom.Empty did the trick. I didn't know about either one of those.

-- 

Greg Green


From peter at engcorp.com  Mon Sep 16 22:07:14 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 16 Sep 2002 22:07:14 -0400
Subject: stdin on Win32
References: <3d865d34$0$22177$afc38c87@news.optusnet.com.au>
Message-ID: <3d868e43$1@news.sentex.net>

nobody wrote:
> This works on Linux, but not Win32.  Win32 just gives me EOF..
> 
> # echo foo|test.py
> 
> where test.py is:
> 
> import sys
> 
> print sys.stdin.readline()

Which flavour of Win32?  Win98 needs you to explicitly
invoke Python and pass the test.py name as a command line
option.  Also, I believe there are problems with non-Win98
Win32 with respect to redirection and the feature you
are trying to invoke.  Checking the archives for this
might give some detail.

-Peter



From pan-newsreader at thomas-guettler.de  Tue Sep 17 13:46:59 2002
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Tue, 17 Sep 2002 19:46:59 +0200
Subject: File upload script
References: <3D871F7E.C313ADD2@student.uib.no>
Message-ID: <pan.2002.09.17.19.46.58.827604.737@thomas-guettler.de>

On Tue, 17 Sep 2002 14:26:38 +0200, Steinar Dragsnes wrote:

> Hi!
> 
> I'm a complete newbe in python and have never written anything more
> advanced then the classical Hallo World example. I have already
> programmed an applet where users should be able to upload images and or
> other files from their local pc. This is of course not possible due to
> the applet security model (sandbox) in Java. But I can start a
> python-script from the running java applet which can do the actual file
> upload for me and save the file to a specific directory on the apache
> server.
> 
> Since the applet that needs the information about what file has been
> uploaded into what node, I need a way to send back information to the
> applet calling the script. I wonder if this is possible at all? Usually
> scripts generate html-pages and call applets from there on they fly,
> with the right paranemer tag values. Is this possible when the applet is
> already running? Would it not destroy the applet context?

I am not a java applet expert, but one thing should be possible:
the Java applet can open an URL on the webserver. This server-side
generated page contains information which file was uploaded to the server


> My main problem is that I do not know how I can pass the parameters of
> type String; sessionID, userID and nodeID, from the appleten to the
> called python-script. I need to know how to write a function that will
> receive these parameters. The python script below is written by a
> colleague of mine, and he is now on a sick-leave....and I need to figure
> this out soon....
> 
> In the java code I call the upload.py script like this:
>      java.net.URL url=new
> java.net.URL("http://"+ip+"/cgi-bin/upload.py");
>     applet.getAppletContext().showDocument(url,"_blank");


Try something like this: url="http:/foo/upload.py?id=" + id + "&nodeID="
+ nodeID ....

 thomas


From boud at valdyas.org  Thu Sep 26 05:10:06 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Thu, 26 Sep 2002 11:10:06 +0200
Subject: unicode woes
References: <amuhrk$4cn$05$1@news.t-online.com>
Message-ID: <3d92d120$0$76313$e4fe514c@dreader4.news.xs4all.nl>

Ulli Stein wrote:

> 
> Some weeks ago the management decided to deploy unicode so that we can
> handle every charset uniformly. The problem which showed at first: we
> realized that we cannot change the encoding in site.py afterwards, i.e. we
> have to specify the encoding all the way. 

Use a sitecustomize.py file that contains:

import sys

sys.setappdefaultencoding=sys.setdefaultencoding

Then, in the application starter script use:

if hasattr(sys, 'setappdefaultencoding'):
    sys.setappdefaultencoding('utf-8')
elif sys.getdefaultencoding() != 'utf-8':
    print 'Warning: encoding not set for unicode - see ReadMe file'

That should solve a lot of your problems. Then you need to carefully 
determine how the data that enters and leaves your application should
be encoded. For instance, you might be getting data from files, databases,
network or paste operations. If that data is many different encodings, you
need to check the encoding at every entry point. And whenever you write 
data, you need to be aware of the encoding.

But at least _inside_ your application all data will be utf-8, which makes
it easy to handle.

Don't be too sanguine about the unicode handling of Java, by the way. Lots 
of room for hard-to-find unicode-related bugs there.

> It got worse and worse because
> the unicode encoding spread like a virus through all the source code.
> Simple Exceptions which we wanted to print to the console throw another
> Exception:
> UnicodeError: ASCII decoding error: ordinal not in range(128)

Other people will be able to tell you why it was a good design decision to
make it difficult to set a default application encoding different from 
ASCII. I'm still not convinced, but that might be just my denseness.

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


From solonas73 at hotmail.com  Tue Sep 17 03:05:52 2002
From: solonas73 at hotmail.com (Solon Kazantzidis)
Date: 17 Sep 2002 00:05:52 -0700
Subject: upload/download to a palm os device with python
References: <24165ed4.0209160304.7a3c08af@posting.google.com> <mailman.1032177101.12986.python-list@python.org>
Message-ID: <24165ed4.0209162305.3925451e@posting.google.com>

Hello,

thank you for the information. 

Grzegorz Makarewicz <mak at trisoft.com.pl> wrote in message news:<mailman.1032177101.12986.python-list at python.org>...
> sf.net/projects/pilot-link - for Palm palms with http://www.pyrite.org/ 
> for offline database manipulations.
>
> wincerapi for Windows CE devices

I tried to locate the pilot-link package. It is nomore at
sourcefourge. Now you can find it at: www.pilot-link.org

I downloaded the pilot-link-0.11.4.tar.gz file. 
Configured/Build/Installed then build the python module pisock but
somehow I miss a documentation or examples for its usage.

Any ideas where to find programming examples which show the usage of
the pisock lib ?

Thanks for your postings,

Solon.


From wlfraed at ix.netcom.com  Mon Sep 30 13:55:59 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Mon, 30 Sep 2002 10:55:59 -0700
Subject: Catching user defined exceptions from foreign modules
References: <mailman.1033401642.9268.python-list@python.org>
Message-ID: <f73ana.vu4.ln@ix.netcom.com>

Derek Basch fed this fish to the penguins on Monday 30 September 2002 
08:59 am:

> to have to import user defined exceptions? I dont
> rememeber having to import any of the built in
> exceptions. So, here is the output of the fixed test1:
>
        "built-in" -> already there... <G>

 
> test2.TestError
> fudge
> 
> Am I doing this all wrong?

        You want to use items defined in an imported module. I see three 
alternatives...

import module

        all uses then become module.item

from module import item1, item2, ... itemN

        only the named items are useable, but can be used as itemM

from module import *

        practically everything within the module is available as item

The last is not recommended as you can encounter unexpected 
redefinition of names... In the last week we had one such...

from os import *

replaces the Python open() (which I understand is being phased out for 
file()) with the low-level (C-style) open -- totally different 
arguments needed.

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From jb at cascade-sys.com  Thu Sep 26 02:19:06 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Wed, 25 Sep 2002 23:19:06 -0700
Subject: Getting started
References: <00a801c26511$4e084750$df7e4e18@brianspiv1700>
Message-ID: <3D92A6DA.3020501@cascade-sys.com>

Brian Quinlan wrote:

>>Some use "weak typing" to mean any type system where the typing can be
>>foiled by casts;
>>    
>>
>
>What does "foiled by" mean in this context?
>
Bypassed or otherwise defeated.  A "cast" in C or C++ is a way to 
override the type system.  Sometimes this is meaningful and useful but 
it also can result in catestrophic type errors.

Types are otherwise pretty secure in C++ except the "cast" mechanism 
"foils" the type system and thus allows type errors to occur.

An earlier example was:

    CEdit* edit = (CEdit*)GetDlgItem( ID_EDIT );

This fetches a pointer to user interface "dialog item" from the existing 
GUI.  Such objects may be of different types.  GetDlgItem() returns a 
pointer to a generic CWnd type, which is the super class to all the 
possible window objects..  The (CEdit*) is a cast says to convert the 
pointer from whatever it may have been (one for a generic windows 
object) to specifically be an Edit Box object.  If all goes well then 
this works just fine.  However, if the dialog item happens NOT to be an 
edit box, say if it was a list box instead, then the resulting object 
pointer would point to the wrong object type and likely result in a 
fatal error when list box operations are applied to the edit box object.

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From misterxx at uni-koblenz.de  Mon Sep  2 04:39:23 2002
From: misterxx at uni-koblenz.de (Oleg Seifert)
Date: Mon, 02 Sep 2002 10:39:23 +0200
Subject: wie greife ich mit Python auf die c-bibliotheken zu ?
Message-ID: <3D7323BB.4070200@uni-koblenz.de>

hallo Leute,

k?nnte vielleicht jemand es mir an einem Beispiel zeigen. Ich habe als 
Attachments zwei kleine Anwendungen dazu geschickt.

danke im Vorraus,

Oleg
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: punkt.cpp
URL: <http://mail.python.org/pipermail/python-list/attachments/20020902/da40d35f/attachment.ksh>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: punkt.py
URL: <http://mail.python.org/pipermail/python-list/attachments/20020902/da40d35f/attachment-0001.ksh>

From sschwarzer at sschwarzer.net  Wed Sep 11 17:03:15 2002
From: sschwarzer at sschwarzer.net (Stefan Schwarzer)
Date: Wed, 11 Sep 2002 23:03:15 +0200
Subject: "str.contains(part)" or alternatives?
References: <alo8jr$l04$1@news.online.de> <eWNf9.165073$l_4.88540@atlpnn01.usenetserver.com>
Message-ID: <aloav7$mdc$1@news.online.de>

Hi Steve :)

Steve Holden wrote:
>> >>> s = 'Hello'
>> >>> if s.contains('ll'):
>>...     do_something_if_found()
>>... else:
>>...     do_something_if_not_found()
>>
>>Do you have other suggestions how to do the test easily without using special
>>values? Should a .contains method be part of string objects in future Python
>>versions? What do you think?
> 
> I think the current proposal is to move to
> 
>     if ll in s:
> 
> for this construct. Happy?

I would like .contains more, but at least the proposal is _much_ better than
"if s.find(part) != -1". :-)

"if part in s" disturbs the pattern "if item in sequence" somewhat but that has
already been weakened by "if key in dict" (that I like, nonetheless).

Stefan



From skpeternospam at ucdavis.edu  Mon Sep 30 00:19:39 2002
From: skpeternospam at ucdavis.edu (Sam Peterson)
Date: 30 Sep 2002 00:19:39 -0400
Subject: Subclass SGMLParser or HTMLParser?
Message-ID: <slrnapfllu.eo2.skpeternospam@hip.ucdavis.edu>

Hello, I've just started doing some python programming with
htmllib.HTMLParser to spider a website of mine and grab all of the
images and download them to disk, as well as collecting reference
counts for my hyperlinks.  It works pretty well, except on a few web
pages that were generated with Word and most of these pages don't
contain images or anchor tags and I imagine the HTMLParser module
meant for XHTML documents will handle those just find once I get
around to playing with it.

My questions is, after having looked around on the web for examples,
I've noticed that most people seem to use sgmllib.SGMLParser instead.
I know that htmllib.HTMLParser is just a subclass of SGMLParser,
therefore I was wondering what the pros and cons are to using one or
the other.  Any recommendations?  Thanks in advance.

-- 
Sam
"Da Man"
s/nospam/son/ -- to email me.


From DennisR at dair.com  Tue Sep 17 10:59:32 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Tue, 17 Sep 2002 14:59:32 GMT
Subject: Windows AutoRunning Python Hello -- feedback requested
References: <Hmuh9.70$Ha6.5066815@newssvr14.news.prodigy.com> <3D86DE3A.2040903@mxm.dk>
Message-ID: <orHh9.168$UJ1.17889587@newssvr14.news.prodigy.com>

> To high level to be a short learning example,
In time, I can fix this.

> It will probably be interresting to only a small subset of developers.

That seems to be the story of my life   :-)

Thanks for your comments.
--

Dennis Reinhardt

http://www.dair.com




From warkid at storm.ru  Fri Sep  6 08:17:15 2002
From: warkid at storm.ru (Kerim Borchaev)
Date: Fri, 6 Sep 2002 16:17:15 +0400
Subject: .py to .pyc
In-Reply-To: <OF7755EF7E.E950A855-ONC1256C2C.002DD99D@redturtle.it>
References: <OF7755EF7E.E950A855-ONC1256C2C.002DD99D@redturtle.it>
Message-ID: <2678.020906@storm.ru>

Hello andrea,

Friday, September 06, 2002, 12:25:14 PM, you wrote:

abrn> Hi everybody,

abrn> I'm a python newbie and i've a problem.

abrn> I want to compile a python file called File.py and i want to produce a file
abrn> called File.pyc, so i do

abrn> python File.py

abrn> I always find the following error:

abrn> Traceback (most recent call last):
abrn>   File "File.py", line 18, in ?
abrn>     import Globals, StructuredText, string, utils, re
abrn> ImportError: No module named Globals

abrn> Can anybody help me to find the problem? I think it's a problem of
abrn> $PYTHONPATH and $PYTHONHOME but i'm not sure.

abrn> Please...help!

abrn> Thanx in advance

abrn> Andrew

You could use py_compile module like this:

##test.py############
fname = 'test_module.py'
src = 'import this_module_does_not_exist\n'
open(fname, 'w').write(src)
import py_compile 
py_compile.compile(fname)
print `open(fname+'c').read()`
#############

so just do:
import py_compile
py_compile.compile('File.py')

Best regards,
 Kerim                            mailto:warkid at storm.ru





From tjreedy at udel.edu  Sat Sep  7 18:20:04 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 07 Sep 2002 22:20:04 GMT
Subject: Converting python object reference to int and back
References: <3D79AC95.1000000@NOSPAMbitbuilder.com>
Message-ID: <oYue9.370424$Aw4.15595936@bin2.nnrp.aus1.giganews.com>

"Ahmad Baitalmal" <ahmad at NOSPAMbitbuilder.com> wrote in message
news:3D79AC95.1000000 at NOSPAMbitbuilder.com...
> Hi,
> I need to keep references to python objects in a tree control's
item.

This is what Python does for you behind the scenes, automatically,
without error, every time you assign abject to a name or slot.

> The items can hold an integer as data.
> How can I get an id (pointer/reference) from a python object and
then
> get that object back using that integer?

You almost certainly do not want to do this since it amounts to
duplicating, in a clumsy way, what the interpreter already does for
you.

Terry J. Reedy





From alanmk at hotmail.com  Thu Sep 26 10:06:56 2002
From: alanmk at hotmail.com (Alan Kennedy)
Date: 26 Sep 2002 07:06:56 -0700
Subject: Page describing http compression with gzip.
Message-ID: <f72dac0d.0209260606.4dab583e@posting.google.com>

Hi All,

A few months ago, after spending a couple of days troubleshooting gzip
compression over HTTP, I put together a page describing how to do HTTP
compression in python, which included some sample code for CGI and
mod_python.

For completeness, I have now added some sample code for jython
servlets to that page.

Also, I have moved the page to a new location, namely

http://xhaus.com/alan/python/httpcomp.html

Lastly, I have created a facility for checking HTTP headers, at the
following address

http://xhaus.com/headers

Sincerely,

alan kennedy
----------------------------------------------
check http headers here: xhaus.com/headers
email alan:              xhaus.com/mailto/alan


From 2002 at weholt.org  Thu Sep 19 18:48:15 2002
From: 2002 at weholt.org (Thomas Weholt)
Date: Thu, 19 Sep 2002 22:48:15 GMT
Subject: Suggesting a filename on a HTTP-request by sending someting in the header of the reply
Message-ID: <Pusi9.20169$0p1.363761@news2.ulv.nextra.no>

I got a simple webserver serving files, but the files are found using
lookup based on ids ( like http://localhost/dl?id=10, where 10 is a key in a
dictionary pointing to a file ), not their filename so when I send
content-type = 'application/octet-stream' in the headers to the client I get
the proper "Save as ..."-dialog, but the filename is all screwed up. It seem
to be the "dl"-part of the url above. I've tried to read the RFC
2616-document and sent "Content-Disposition"-headers etc. but with no luck.

What headers do I need to send for it to get the proper filename in the
"Save as ..."-dialog? Say for instance that the filename I wanted the client
to save the file as was "test.zip".

Any hints?

Best regards,
Thomas




From mstenner at phy.duke.edu  Wed Sep  4 20:31:47 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Wed, 4 Sep 2002 20:31:47 -0400
Subject: simple xml pickle
In-Reply-To: <20020904132951.A7928@phy.duke.edu>; from mstenner@phy.duke.edu on Wed, Sep 04, 2002 at 01:29:51PM -0400
References: <mailman.1031104505.31870.python-list@python.org> <QEjd9kKkXEbH092yn@gnosis.cx> <20020904132951.A7928@phy.duke.edu>
Message-ID: <20020904203147.A9032@phy.duke.edu>

On Wed, Sep 04, 2002 at 01:29:51PM -0400, Michael Stenner wrote:
> The xml marshalling does almost exactly what I want, and the
> modifications to make it "standalone" for Python 2.2 were very easy
> (patch forthcoming) so I may just stick with that.  Nonetheless, I'll
> look at xml_pickle again and see what I can see.

Results:

  With David's hint, xml_pickle.py was also fairly easy to make work
  with python 2.2 in a standalone fashion.  Here are the patches to
  make both the PyXML marshaller and xml_pickle.py work as standalone
  files with python 2.2 (probably 2.x, but I haven't tested it).

  I haven't really tested either of these.  I just tweeked them so
  that they passed their own built-in tests.

  generic.py.diff should be applied to xml.marshal.generic.py from
  PyXML 0.8.

  xml_pickle.py.diff should be applied to xml_pickle.py 0.51.

Now I just need to decide which I want to use :)

					-Michael
-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305
-------------- next part --------------
--- generic.py.orig	Wed Sep  4 20:19:58 2002
+++ generic.py	Wed Sep  4 10:34:45 2002
@@ -6,7 +6,12 @@
 
 from types import *
 import string
-from xml.sax import saxlib, saxexts
+import xml.sax
+from xml.sax.handler import EntityResolver, DTDHandler, ContentHandler,\
+     ErrorHandler
+
+class HandlerBase(EntityResolver, DTDHandler, ContentHandler,\
+                  ErrorHandler): pass
 
 # Basic marshaller class, customizable by overriding it and
 # changing various attributes and methods.
@@ -20,7 +25,7 @@
         return cmp(ta.__name__, tb.__name__)
     return cmp(a,b)
 
-class Marshaller(saxlib.HandlerBase):
+class Marshaller(HandlerBase):
     # XML version and DOCTYPE declaration
     PROLOGUE = '<?xml version="1.0"?>'
     DTD = ""
@@ -251,7 +256,7 @@
 
 TUPLE = {} ; LIST = {} ; DICT = {}
 
-class Unmarshaller(saxlib.HandlerBase):
+class Unmarshaller(HandlerBase):
     # This dictionary maps element names to the names of starting and ending
     # functions to call when unmarshalling them.  My convention is to
     # name them um_start_foo and um_end_foo, but do whatever you like.
@@ -314,9 +319,9 @@
 
     def _load(self, file):
         "Read one value from the open file"
-        p=saxexts.make_parser()
-        p.setDocumentHandler(self)
-        p.parseFile(file)
+        p=xml.sax.make_parser()
+        p.setContentHandler(self)
+        p.parse(file)
         assert len(self.data_stack) == 1
         # leave the instance in a steady state
         result = self.data_stack[0]
@@ -358,9 +363,9 @@
         sm, em = self.unmarshal_meth[ name ]
         if sm is not None: return sm(name,attrs)
 
-    def characters(self, ch, start, length):
+    def characters(self, ch):
         if self.accumulating_chars:
-            self.data_stack[-1].append(ch[start:start+length])
+            self.data_stack[-1].append(ch)
 
     def endElement(self, name):
         # Call the ending method
-------------- next part --------------
--- xml_pickle.py.orig	Tue Sep  3 18:38:48 2002
+++ xml_pickle.py	Wed Sep  4 20:18:34 2002
@@ -206,7 +206,7 @@
                      '<!DOCTYPE PyObject SYSTEM "PyObjects.dtd">\n',
                      '<PyObject class="%s" id="%s">\n'
                          % (_klass(self.py_obj),id_)]
-        for name in dir(self.py_obj):
+        for name in self.py_obj.__dict__.keys():
             xml_lines.append(_attr_tag(name, getattr(self, name)))
         xml_lines.append('</PyObject>\n')
         return string.join(xml_lines,'')
@@ -226,36 +226,36 @@
             # Add all the subnodes to PyObject container
             container = thing_from_dom(node, obj_from_node(node))
             try:
-                id = node.attributes[('', 'id')].value
+                id = node.attributes['id'].value
                 visited[id] = container
             except KeyError:
                 pass
 
         elif node.nodeName == 'attr':
             try:
-                node_type = node.attributes[('','type')].value
+                node_type = node.attributes['type'].value
             except:
                 print "node", node.attributes, repr(node.attributes)
                 print node.attributes.keys()
                 raise ValueError   # WHAT?!
-            node_name = node.attributes[('', 'name')].value
+            node_name = node.attributes['name'].value
             if node_name == '__parent__' and EXCLUDE_PARENT_ATTR:
                 # Do not pickle [xml_objectify] bookkeeping attribute
                 pass
             elif node_type == 'None':
                 setattr(container, node_name, None)
             elif node_type == 'numeric':
-                node_val = safe_eval(node.attributes[('','value')].value)
+                node_val = safe_eval(node.attributes['value'].value)
                 setattr(container, node_name, node_val)
             elif node_type == 'string':
-                node_val = node.attributes[('','value')].value
+                node_val = node.attributes['value'].value
                 node_val = unsafe_string(node_val)
                 setattr(container, node_name, node_val)
             elif node_type == 'list':
                 subcontainer = thing_from_dom(node, [])
                 setattr(container, node_name, subcontainer)
                 try:
-                    id = node.attributes[('', 'id')].value
+                    id = node.attributes['id'].value
                     visited[id] = subcontainer
                 except KeyError:
                     pass
@@ -263,7 +263,7 @@
                 subcontainer = thing_from_dom(node, []) # use list then convert
                 setattr(container, node_name, tuple(subcontainer))
                 try:
-                    id = node.attributes[('', 'id')].value
+                    id = node.attributes['id'].value
                     visited[id] = subcontainer
                 except KeyError:
                     pass
@@ -284,7 +284,7 @@
                 subcontainer = thing_from_dom(node, {})
                 setattr(container, node_name, subcontainer)
                 try:
-                    id = node.attributes[('', 'id')].value
+                    id = node.attributes['id'].value
                     visited[id] = subcontainer
                 except KeyError:
                     pass
@@ -292,16 +292,16 @@
                 subcontainer = thing_from_dom(node, obj_from_node(node))
                 setattr(container, node_name, subcontainer)
                 try:
-                    id = node.attributes[('', 'id')].value
+                    id = node.attributes['id'].value
                     visited[id] = subcontainer
                 except KeyError:
                     pass
             elif node_type == 'mxDateTime':
-                parts = node.attributes[('','value')].value.split()
+                parts = node.attributes['value'].value.split()
                 node_val = apply(mx.DateTime.DateTime,map(float,parts))
                 setattr(container, node_name, node_val)
             elif node_type == 'ref':
-                ref_id = node.attributes[('', 'refid')].value
+                ref_id = node.attributes['refid'].value
                 setattr(container, node_name, visited[ref_id])
 
         elif node.nodeName in ['item', 'key', 'val']:
@@ -310,12 +310,12 @@
             # all cases, elif's are defined for mutable types. Furthermore,
             # there should only ever be *one* item in any key/val list,
             # but we again rely on other validation of the XML happening.
-            node_type = str(node.attributes[('','type')].value)
+            node_type = str(node.attributes['type'].value)
             if  node_type == 'numeric':
-                node_val = safe_eval(node.attributes[('','value')].value)
+                node_val = safe_eval(node.attributes['value'].value)
                 container.append(node_val)
             elif node_type == 'string':
-                node_val = node.attributes[('','value')].value
+                node_val = node.attributes['value'].value
                 node_val = unsafe_string(node_val)
                 container.append(node_val)
             elif node_type in ('list',array_type):
@@ -331,7 +331,7 @@
                 subcontainer = thing_from_dom(node, obj_from_node(node))
                 container.append(subcontainer)
             elif node_type == 'mxDateTime':
-                parts = node.attributes[('','value')].value.split()
+                parts = node.attributes['value'].value.split()
                 node_val = apply(mx.DateTime.DateTime,map(float,parts))
                 container.append(node_val)
 
@@ -350,7 +350,7 @@
     # Get classname of object (with fallback to 'PyObject')
     try:
         if node.attributes:
-            klass = node.attributes[('','class')].value
+            klass = node.attributes['class'].value
         else:
             klass = 'PyObject'
     except KeyError: klass = 'PyObject'
@@ -452,7 +452,7 @@
             visited[id(thing)] = 1
             start_tag = start_tag + 'type="PyObject" class="%s" id="%s">\n' \
                                               % (_klass(thing), id(thing))
-            for name in dir(thing):
+            for name in thing.__dict__.keys():
                 tag_body.append(_attr_tag(name, getattr(thing, name), level+1))
     elif '%s' % type(thing) == "<type 'SRE_Pattern'>":
         # SRE_Pattern objects are extension objects, so not

From mertz at gnosis.cx  Wed Sep 25 12:25:25 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Wed, 25 Sep 2002 12:25:25 -0400
Subject: Python Compiling
References: <amfveg$5g64k$1@ID-53055.news.dfncis.de>
 <80a8af7d.0209250113.28ae4c72@posting.google.com>
 <pdgk9.140947$ub2.3077676@news1.tin.it>
Message-ID: <1Nek9kKkX0+P092yn@gnosis.cx>         

Alex Martelli <aleax at aleax.it> wrote previously:
|Or Haskell, www.haskell.org -=- HUGS inteprets it, the Edinburgh
|compiler compiles it.  Scheme has _many_ interpreters and compilers
|too.  Hey, so does Visual Basic... it's quite the thing these days.

Wrong Scottish city, I'm afraid.  Alex probably means the GLASGOW
Haskell Compiler (GHC).  There's also NHC and HBC available, but GHC
seems like where the real focus is.

Yours, Lulu...

--
Keeping medicines from the bloodstreams of the sick; food from the bellies of
the hungry; books from the hands of the uneducated; technology from the
underdeveloped; and putting advocates of freedom in prisons.  Intellectual
property is to the 21st century what the slave trade was to the 16th.



From sholden at holdenweb.com  Wed Sep 25 09:12:54 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 25 Sep 2002 09:12:54 -0400
Subject: Getting started
References: <uojinmcfapjf45@corp.supernews.com> <mailman.1032442884.18747.python-list@python.org> <H3mi9.133635$pX1.4751676@news2.tin.it> <mailman.1032567390.29374.python-list@python.org>
Message-ID: <AFik9.51323$b%3.25248@fe03>

"James J. Besemer" <jb at cascade-sys.com> wrote in message
news:mailman.1032567390.29374.python-list at python.org...
>
> Alex Martelli wrote:
>
> >$a = "2.3";
> >$b = 45;
> >print $a+$b;
> >
> >emitting 47.3 is pretty close to what I consider "weak typing".
> >
>
> I don't see why you call this "weak typing".  In Perl, the "+" operator
> is defined to work with strings or numbers.

Well, it seems pretty weak to me that the *operator* has to examine the
types of its arguments in order to know which operation to perform. Can we
say polymorphism?

>                                                          It may seem more
confusing
> at first if you expect "+" also to be the string concatenation operator,
> which it is not.  In the above example,
>
>     print $a . $b;
>
> will print the string "2.345".
>
How very intuitive. I am happy to say I have now forgotten enough Perl for
this to seem totally unreasonable. I usually don't say these things about
Perl because it, too, is a pragmatist's language, and Perl is fine for the
things Perl is fine for. If it's working for someone I can't in all
conscience disturb their busy working day to suggest they rewrite it all in
Python.

> This all is no different in principle from Python's:
>
>     a = 2.3
>     b = 45L
>     print a + b
>
> and getting 47.3.  The "+" operator is defined for floats and also for
> longs.  Since they're all "numbers" it would seem really strange if they
> did not interact naturally like this.  But fundamentally they're
> distinct types as "type( 2.3 )" and "type( 45L )" illustrate.
>
It seems very different to me. For example, how does the Perl "+" operator
find out about a new addable type? Is Perl even able to integrate its
intrinsic data objects with user-defined types? Can we say inheritance
hierarchy? In Python it's easy to define methods to integrate your own
classes into the language's architecture. I could never discern enough
architecture in Perl to see how to do that, though I willingly accept that
this deficiency might be in me rather than in Perl.

> Perhaps an even better example is the "*" operator:
>
>         print  2  * 5     # integer x integer -> integer
>         print [2] * 5     # list x integer -> list
>         print "2" * 5     # string x integer -> string
>
> This all makes sense even though radically different results and result
> types are produced each time.  It's not "weakly typed" but rather
> strong, dynamic typing.  The outcome is completely determined by the
> operator and the types of the operands and there is no way to break the
> rules.  Since we know Python is strongly typed then it follows that your
> Perl example must be also.

Not really. Python has strongly-typed data with late binding of methods and
operators. This is an organised way to obtain the results you have so far
quoted.

>
> Anyway, why should "2" + 5 == "7" be all that much stranger than "2" * 5
> == "22222"?  If the operation is unambiguous and useful what rule would
> it break?  Personally, I think Python would be improved slightly by
> adding this conversion.  It's convenient and there's ample precedent in
> other languages (Awk, Snobol, Icon, Perl, etc.).  But I am sure that
> notion will provoke howls of protest from the True Believers.
>
Not at all. If I thought such a suggestion would be taken seriously I might
go so far as to advise against that, but I don't see this being a religious
debate.

> There are a lot of reasons why Python is better than Perl but strongly
> vs. weakly typed is not one of them.  I don't see a material difference
> in the "strength," per se, of typing between Perl and Python.  They're
> both "strong" and "dynamic", though Python's type model is better
> thought out, particularly since object/type unification.
>
[C/Fortran irrelevancies]
>
> These types of errors are not possible in Python or Perl, which
> illustrates that they're both "strong".
>
For some meaning of strong, I suppose, but please don't bother to define
what this is.

> Curiously, in Python
>
>     print "%s" % 2.5
>
> prints the string "2.5" instead of, say, raising an exception.  So
> there's some small precedent for treating numbers as strings in a string
> "context" even in Python.
>
Well, considering that the %r and %s conversion types are currently
documented as

"""
r String (converts any python object using repr()).
s String (converts any python object using str()).
"""

it should hardly be surprising that conversions take place. Such
conversions, however, are not intrinsic to the *language*, but to the *data*
(where the type is significant) - Python goes so far as to expose explicitly
how the conversions are performed, and they will clearly work for any object
with a suitable (returns a string) __repr__() and/or __str__() method. Just
to clarify the nature of this, try

    "%6.2f" % 2.6

and then try

    "%6.2f" % "2.6"

and we'll see if you still feel that string formatting 'treats numbers as
strings in a string "context"'.

By contrast Perl seems a little muddled, and I personally think it's no
accident that Wall's own writing tends to stress the significance of
*context*. The rules are quite explicit in Python, clouded in Perl. Of
course you must take this as a dim memory frmo one who never claimed to
understand Perl that well in the first place (I used to program Perl, but
I'm all right now ;-).

> I'll go out on a limb and say that most "dynamic typed" languages are
> also "strong" by necessity.  If the user isn't keeping track of the
> types then the system HAS to.  An exception I can think of would be an
> ancient dialect of Lisp where integers also were machine addresses and
> thus callable without restriction.  E.g., if a function name evaluated
> to a number the interpreter performed a subroutine call to that address
> instead of, say, applying a lambda.  It was the mechanism through which
> system builtins were accessed.  However,  nothing prevented a user from
> "calling" an arbitrary integer and leaping off into never never land.
>
I seem to have lost track of what "dynamic typed" and "strong" mean. Sorry.

> "Strong typing" (the "static" kind) is also the norm in Pascal and C++.
>  I like to think of strong typing as the norm in ANSI C but there is
> still optional and thus may be considered "weak".  Furthermore, C
> programmers are on the "honor" system across module boundaries, as
> there's no link time checking.  There even are ways to cheat in C++, so
> some might still class it as a "weakly typed" language.  Of course, with
> discipline, it's possible to avoid the pitfalls and always use the
> strong typing facilities of C and with effort enjoy most of the benefit.
>
It's also possible to enjoy some of the benefits of prison by simply
refusing to go outside. Your very confession of C's and C++'s weaknesses
highlights the fact that even "strongly typed" languages have problems
enforcing the rules (whose value I anyway challenge except for the specific
promotion of execution efficiency). Python is just a bit less anal :-)

> >I
> >find it very hard to produce substantial, large programs using
> >weakly-typed languages.
> >
> Given my definition of "weakly typed," I agree it's more difficult than
> in strongly typed languages.
>
> On the other hand, isn't it the case that much of Python is implemented
> in C, which is weakly typed?  So "weakly typed" makes it harder but not
> impossible, at least not for some.
>
You can write structured programs in assembler. This does not make it the
ideal vehicle for a GUI-based commercial accounting system. Python is an
object-oriented environment implemented in C. Ultimately C is compiled into
machine code. What's your point? Once the Turing completeness of von Neumann
architectures is accepted the rest is merely linguistic anthropology: what a
system can do is, ultimately, more important to its users than how it does
it (which is a significant concern of the implementors).

> > Python is strongly, albeit dynamically, typed,
> >
> FWIW, so is Perl.
>
But not, IMHO, in the same way.

> >Unfortunately, some seminal material of a few
> >decades ago confused terminology (I vaguely remember something
> >about it in some early article about Scheme, for example).
> >
> I think you're right.  I originally learned a different taxonomy for
> many of these concepts.
>
Well, we all have to watch out for Humpty Dumpty in our own output.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From amk at amk.ca  Thu Sep 19 19:51:55 2002
From: amk at amk.ca (A.M. Kuchling)
Date: Thu, 19 Sep 2002 23:51:55 -0000
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org>
Message-ID: <uokoorcumdrmc6@corp.supernews.com>

In article <mailman.1032293422.22920.python-list at python.org>, 
	Michael Chermside wrote:
>   (5) Good reasons exist why Stackless CANNOT be implemented in Jython,
>       in Python for .NET, or other Python implementations -- it's a
>       C-Python-only thing.

Note that the Parrot VM is already stackless, so a hypothetical Python
running on Parrot would be stackless from the start.

--amk


From hellprout at yahoo.com  Sat Sep 21 05:55:38 2002
From: hellprout at yahoo.com (hellprout)
Date: 21 Sep 2002 02:55:38 -0700
Subject: siebel and python
Message-ID: <5b8834c2.0209210155.5ae21270@posting.google.com>

hi , i'm working on siebel , i use vb to charge data in siebel 
but i want to use python with win32com
i can't load the cfg file and login on siebel 
what can i do ?
i just want to have example
thanks a lot


From rsrchstr at msn.com  Sun Sep 29 21:23:07 2002
From: rsrchstr at msn.com (mike henley)
Date: Mon, 30 Sep 2002 02:23:07 +0100
Subject: idea for a much needed application
References: <576c1752.0209291326.56dcdf31@posting.google.com> <an87db$bjtvs$1@ID-157863.news.dfncis.de>
Message-ID: <an8924$bujmm$1@ID-157863.news.dfncis.de>

Just brainstorming on one possible way to do it; please advise if you have:

what would such a script/application do?

it would start up with my computer, simply insert itself in my start-up
folder or whatever
it would just wait and detect when i go online
then it when i do get online it would detect what my IP is
and when if i'm disconnected then it will detect that too
and when if i'm reconnected it would detect that too and detect my IP
or it can just simply do this at regular intervals

then it would use jabber/xml-rp or whatever to make a presence for
me/my-webapplication on IM networks
using a username or account name that i can give to others

so that this way the webapplication will have "presence"
so others will know when the webapplication is online

and then if they IM it it'll reply with my IP so it appears in the IM as a
hyperlink that they can click to launch their default browser to a url
consisting of http://myIP

and then perhaps if they wish to be automatically notified of the
webapplication's presence they can insert their username on a form on the
webapplication when they get their which will add it to its buddylist

or even within the first IM they get from it there can be a link that if
they wish to click on will detect and add their username automatically (as
an option)

or they reply with a message such as "add me" which will do that

and either way they can be added, if they wish, to the webapplications
buddylist so next time it goes online it can send them an IM automatically

of course it should enable them to remove themselves from the buddylist
later on if they wish

sounds very much like a jabber bot
perhaps i can get an open source one and extend or customize it


got any better ideas or there's something wrong in what i have written?
please tell me, i have read about jabber yet.




"mike henley" <rsrchstr at msn.com> wrote in message
news:an87db$bjtvs$1 at ID-157863.news.dfncis.de...
> i guess i'll rephrase; what i wish for is an "application-to-person" (P2A)
> IM-based "presence", so that a web-application that i host on a personal
> webserver, on my own machine, can let others become aware of its
"presence"
> easily and be easy to connect to with a simple
> click-to-launch-browser-and-go-to-my-ip. This should be easy to do. I just
> don't have experience with Jabber or xml-rpc, or other potential useful
> technologies.




From eppstein at ics.uci.edu  Mon Sep 16 11:58:17 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Mon, 16 Sep 2002 08:58:17 -0700
Subject: Are generator functions thread-safe?
References: <T%dh9.6427$Lg2.1180859@news2.news.adelphia.net> <am4t4l$b33$1@panix1.panix.com>
Message-ID: <eppstein-C9F3CA.08581716092002@news.service.uci.edu>

In article <am4t4l$b33$1 at panix1.panix.com>, aahz at pythoncraft.com (Aahz) 
wrote:

> >I know that a call to a generator function actually returns an independent
> >iterator object (mea culpa if I'm wrong on this), but are there any hidden
> >gotchas that would cause the use of the same generator function from
> >multiple threads, to result in a sharing (especially local storage) or
> >memory collision issue?
> 
> The same gotchas for any other function: you're screwed if you share
> mutable objects without locks.

Also, if a call to the generator object is running from one thread, you 
will get an exception if you attempt to simultaneously call it (e.g. 
from another thread).  I don't know how thread-safe is the code that 
prevents the generator from being called twice simultaneously, though.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From bdesth at nospam.free.fr  Fri Sep  6 21:30:46 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Fri, 06 Sep 2002 21:30:46 -0400
Subject: Grumble about too strict an attitude about backward compatibility...
References: <mailman.1031108345.32199.python-list@python.org>        <3D77F98B.3050607@nospam.free.fr> <mailman.1031255836.28274.python-list@python.org>
Message-ID: <3D7956C6.6040800@nospam.free.fr>

Skip Montanaro wrote:
>     >> What's the Python connection?  Other than a reminder not to get to
>     >> slavish about backward compatibility, I note that these same function
>     >> names will then go on to pollute the Python namespace because all
>     >> this Fortran code is automatically wrapped using f2py.
> 
>     laotseu> Well, that may be true, but the chances of a name collision are
>     laotseu> pretty low IMHO...
> 
> Name collisions weren't what I was getting at.  Python allows essentially
> unlimited identifier lengths, but people manipulating these functions from
> Python code still have to use the cryptic 4-6 character names.  Perhaps f2py
> allows mapping such names to something more meaningful.  (One can hope. ;-)
> 
> Skip
> 
> 
> 

mapping such names to something more meaningful ?

aFunctionWithAnySensibleName = lpni
aFunctionWithAnotherSensibleName = klvna

if aFunctionWithAnySensibleName(andFarTooMuchLettersInIt):
     resutl = aFunctionWithAnotherSensibleName(andAlsoManyUselessLetters)


Like this ?-)

laotseu
--
The Old Master said :
"function names are always too long... unless they are too short"
--



From aleax at aleax.it  Fri Sep 20 11:09:16 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 15:09:16 GMT
Subject: bit operations ?
References: <amfbuq$56f28$1@ID-146704.news.dfncis.de>
Message-ID: <wSGi9.119671$ub2.2539193@news1.tin.it>

Shagshag13 wrote:

> hello,
> 
> i'm looking for any libs which could handle bit operations on integer.
> (i must be able to get 0 to n bits from any int., only bit j, etc.)

gmpy does the two ops that you require -- functions lowbits and
getbit respectively.  No 'etc', though.


Alex



From jlkeating at clear.net.nz  Fri Sep  6 03:21:45 2002
From: jlkeating at clear.net.nz (jlkeating)
Date: Fri, 6 Sep 2002 19:21:45 +1200
Subject: A special  excite game
Message-ID: <20020906072045.RPBB9205.mta3-rme.xtra.co.nz@Hatmsluq>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020906/8c7ac539/attachment.html>
-------------- next part --------------

------------------  Virus Warning Message  ------------------

install.exe is removed from here because it contains a virus.

-------------------------------------------------------------
-------------- next part --------------
A non-text attachment was scrubbed...
Name: SQLSRDME.TXT
Type: application/octet-stream
Size: 13228 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020906/8c7ac539/attachment.obj>
-------------- next part --------------
------------------  Virus Warning Message  ------------------

Found virus WORM_KLEZ.H in file install.exe
The uncleanable file is deleted.

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

From david.g.morgenthaler at lmco.com  Tue Sep  3 22:03:12 2002
From: david.g.morgenthaler at lmco.com (Dave M.)
Date: 3 Sep 2002 19:03:12 -0700
Subject: Jython and zxJDBC package
References: <d76a3954.0208301123.331bdfbf@posting.google.com> <23891c90.0209020013.b05e81c@posting.google.com>
Message-ID: <71ec32f3.0209031803.2d0d245@posting.google.com>

> > Is anyone here using the zxJDBC package as described in "Jython
> > Essentials" to access a MySQL database?
> > 
> > When I try to run the sample program (Example 10-1, p. 164), I receive
> > a NoClassDefFoundError: org/gjt/mm/mysql/Statement in
> > MySQLDataHandler.  Decompiling MySQLDataHandler reveals an explicit
> > reference to org.gjt.mm.mysql.Statement; but this class isn't in the
> > Connector/J library.  Am I missing something here?
> > 
> > I'm using Jython 2.1 and Connector/J 2.0.14.
> 
> It looks like MySQLDataHandler wants to use the MM.MySQL package,
> which can be found here:
> 
>   http://sourceforge.net/projects/mmmysql
> 
The example in the book does indeed expect the use of MySQL. I don't
know about Connnetor/J, but, for example, the following works with
Sun's JDBC-ODBC bridge driver talking to an Access database:

----------code-------
import com.ziclix.python.sql as sql
dburl, user, pw, drv = ("jdbc:odbc:testTrack","",
"","sun.jdbc.odbc.JdbcOdbcDriver")
db = sql.zxJDBC.connect(dburl, user, pw, drv)
cursor = db.cursor()
#cursor.datahandler = sql.handler.MySQLDataHandler(cursor.datahandler)
cursor.execute("SELECT * FROM trackTable")

for item in cursor.fetchall():
    print item
-----------end code----------
Be sure that, e.g., sun.jdbc.odbc.JdbcOdbcDriver is on your Jython
path.

Note also the setting of a specific datahandler is commented out.
zxJDBC ships with custom drivers for
Oracle (sql.handler.OracleDataHandler),
Informix (sql.handler.InformixDataHandler),
Postgresql (sql.handler.PostgresqlDataHandler), and
SQLServer (sql.handler.SQLServerDataHandler) 
in addition to the MySQLDataHandler. If you're not using one of these,
the default data handler will work.


From andy at agmweb.ca  Tue Sep 24 23:51:36 2002
From: andy at agmweb.ca (Andy McKay)
Date: Tue, 24 Sep 2002 20:51:36 -0700
Subject: Zope Documentation and example sites
References: <503511c2.0209230758.74a64a3d@posting.google.com> <9bf19f33.0209241810.1fbd28f5@posting.google.com>
Message-ID: <00ea01c26446$d9a7db20$89505018@cr582427a>

> I would like to find some working examples that use CMF.  Does anyone
> know of any?

http://zopezen.org is in the CMF and has links to a bunch of other sites,
not necessarily using CMF.
--
  Andy McKay
  Agmweb Consulting
  www.agmweb.ca





From martin at v.loewis.de  Mon Sep  2 01:54:41 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 02 Sep 2002 07:54:41 +0200
Subject: va_list warnings on Solaris?
References: <mailman.1030555067.1038.python-list@python.org> <pan.2002.08.28.22.48.52.513247.2752@metaslash.com>
Message-ID: <m34rd97x7i.fsf@mira.informatik.hu-berlin.de>

Neal Norwitz <neal at metaslash.com> writes:

> I get this as well.  I believe the warning started after I upgraded Solaris,
> but I'm not sure.  I didn't build gcc locally.  If I remember correctly 
> the collision was between a gcc header file (stdarg.h shown above) and 
> a Solaris header file in /usr/include.

If you upgrade Solaris, you really need to rebuild gcc, for that
Solaris version. gcc installation will run fixincludes, copying the
system headers. It frequently happened that using fixincluded headers
from an older Solaris version caused very obscure bugs.

Skip's report seems to indicate that he indeed built gcc for his
system. I'd be curious to see one more line of error messages though
(hoping to get a line like "previous definition here").

Regards,
Martin


From michael at stroeder.com  Wed Sep 18 10:19:35 2002
From: michael at stroeder.com (=?ISO-8859-1?Q?Michael_Str=F6der?=)
Date: Wed, 18 Sep 2002 16:19:35 +0200
Subject: Earing money with Python software (was: More on Protecting Source
 Code)
References: <mailman.1032290902.29298.python-list@python.org>
Message-ID: <nkem5-fei.ln1@junker.stroeder.com>

brueckd at tbye.com wrote:
> free-for-personal-use-but-companies-must-pay approach when possible.

The problem with getting some bucks from a company is the amount of time 
spent during decision taking and ordering process. For you to earn 300 
bucks for your software the company spends another 500+ bucks for 
working time until the money gets out. IMHO that's why it's so hard to 
sell small packages of rather cheap software to big companies.

Ciao, Michael.



From Tom_Good1 at excite.com  Wed Sep 11 16:25:18 2002
From: Tom_Good1 at excite.com (Tom Good)
Date: 11 Sep 2002 13:25:18 -0700
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com>
Message-ID: <ac677656.0209111225.4a141bc5@posting.google.com>

steve at ferg.org (Stephen Ferg) wrote in message news:<b16e4ef7.0209062009.26e218e2 at posting.google.com>...
> A small page devoted to "Thinking in Tkinter"
> at http://www.ferg.org/thinking_in_tkinter/index.html

Cool stuff.  A couple of comments:

1. You might want to point out that we can enter and experiment with
simple Tkinter scripts interactively in PythonWin or IDLE.  This may
not be obvious to users who are new to Python.  There is something
satisfying about typing "root.mainloop()" into the interactive window
and seeing the Tkinter window immediately pop up :-)

2. In section (1) of tt020.py, it says 'The constructor method
recognizes the root under the name "parent".'  I think you meant to
say, 'under the name "myParent".'


Tom


From kwdotbenninkNOSPAM at 12move.nl  Fri Sep 20 11:07:32 2002
From: kwdotbenninkNOSPAM at 12move.nl (Highwaykind)
Date: Fri, 20 Sep 2002 17:07:32 +0200
Subject: Newbia alert .. working my way through Easy Python Montana Edu page, need help..
References: <MPG.17f540bc75626c66989683@news.worldonline.nl> <mTFi9.119296$ub2.2532256@news1.tin.it>
Message-ID: <MPG.17f5581f3a5ca93989684@news.worldonline.nl>

Thank you very much !
It's working now :)
After I finished the tutorial I'll try to translate it into Dutch ( 
which makes it even easier for me and probably other Dutch people .. ) 
Didn't know what an indentation was ( desperately looking for a 
dictionary LOL ).
Will go on with the tutorial .. and if I get stuck again and can;t 
figure out what I've done wrong this time I'll be back .. on my way to 
becoming a progammer .. in 2010 or so :)

Corien


> Highwaykind wrote:
>         ...
> > def print_options() :
> 
> You define this function, but never call it anywhere.
> 
> > choice = 'p'
> > while choice != 'q':
> >     if choice == 'r':
> >         a=input ("width")
> >         b=input ("height")
> >     print "result",rect (width,height)
> 
> Note that the print is NOT aligned with the input calls, but
> rather with the if statement.  Thus, the print (and its call
> to rect) is not conditional -- it happens even when choice
> is NOT 'r'.  This is unlikely to be what you want, so, indent
> this print statement 4 more spaces rightwards.
> 
> However, the crux of your problem is another:
> 
> > But the thing keeps telling me width is not defined, also I can;t use
> 
> Inddeed, width is not a defined variable: what you obtained
> from the call to input("width") you did not assign to the
> name width, but rather to the name a.  Therefore, the variable
> that is defined once this statement has executed is named a.
> 
> Therefore, change your call to rect  to rect(a,b) -- and
> similarly for the following passages in your code.
> 
> > the ELIF command .. THink I could do with some help :)
> 
> The reason you can't use elif (not a 'command' -- a clause
> in an if statement) is probably the indentation problem I
> mentioned earlier -- once you fix that you can probably
> also change the following if's to elif's (not the first one
> of course).  When indentation returns to the same level as
> the word 'if', it means the if statement is over: therefore
> you cannot "reprise" it later, continuing it, so to speak,
> with an elif clause.  Once you correct the indentation of
> the print statements, this will change.
> 
> 
> Alex
> 
> 

-- 
For the correct e-mail address : remove NOSPAM and replace dot with .


From mcfletch at rogers.com  Thu Sep 12 22:24:42 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Thu, 12 Sep 2002 22:24:42 -0400
Subject: Piper Pickling?
References: <dg9g9.2373$Lg2.427869@news2.news.adelphia.net>
Message-ID: <3D814C6A.6050902@rogers.com>

Strangely enough, I just spent the afternoon working to plan something 
like this.  I believe this is called an object-relational adapter in 
case you want to look for other implementations (there's lots for Java). 
  Basically, this is most easily done with objects that have 
static-typed properties (such as I have in the basicproperty package). 
You can construct a schema from the class+properties.

I'd estimate this will take a few weeks of solid programming to make 
work properly (i.e. make the data-type mappings all reasonable and 
efficient with a decent interface).

Only relevant pre-cursor project I've seen is Python Data Object (PyDO) 
which is part of the SkunkWeb system.  That system doesn't solve the 
"create schema from object" problem AFAICS, and it's that problem that 
I'm interested in solving.

For instance, with a data-model like this:

Event:
	mxDateTime startDate
	mxDateTime stopDate
	str name
	User user
	Collection{type Resource} resources
Resource:
	str name
User:
	str name

(simplified example from one of my projects), you'd want the data-store 
to create 3 class-specific tables with the Event table having a resource 
sub-table for the one-to-many relationship.  You'd want it to manage all 
the persistentIDs, and to properly re-constitute property values on 
access to be references to objects (rather than persistentIDs).

The model I'm currently laying out seems quite capable of supporting 
this, though there will need to be some additions to the basicproperty 
class to support selecting among the various many-to-one / one-to-many / 
one-plus-sub-key-to-many / etceteras data models for certain properties.

The thing is, to be really useful, you need to make a lot of simplifying 
assumptions (such as that all objects of a type/class have the same 
data-storage requirements).  I can make those assumptions in my apps, 
but the result won't allow for anything like the full range of Python's 
dynamic nature.

Anyway, I'm going back to my SQL lessons now :),
Mike

Robert Oschler wrote:
> Forgive me :)
> 
> If Peter Piper pickles a peck of Python packages, can he pack the pickled
> Python peckers in a proper packing place?
> 
> Translated: Anybody seen a module that can override the standard pickling
> process and store and retrieve Python objects to a MySQL database?
> 
> thx
> 
> P.S. - Whell hell the language is named after the world's best comedy
> troupe, what did you expect, green eggs and spam? :)
> 
> 

-- 
_______________________________________
   Mike C. Fletcher
   Designer, VR Plumber, Coder
   http://members.rogers.com/mcfletch/





From m2 at plusseven.com  Wed Sep 18 15:46:47 2002
From: m2 at plusseven.com (Alex Polite)
Date: Wed, 18 Sep 2002 21:46:47 +0200
Subject: debugging "Too many open files"
Message-ID: <20020918194647.GB2682@matijek.plusseven.com>

I'm getting the "Too many open files" error.
I've reset /proc/sys/fs/file-max to 100000 and the count in
/proc/sys/fs/file-nr is nowhere close.

I'm not sure that python is the root of the problem. Is there anyway
to see which processes own the active file descriptors?


-- 
Alex Polite
http://plusseven.com/gpg/



From duncan at NOSPAMrcp.co.uk  Mon Sep  9 10:36:27 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Mon, 9 Sep 2002 14:36:27 +0000 (UTC)
Subject: xml.minidom is stripping out my CRLF's in attrib values!!
References: <3D7C80C4.9070204@NOSPAMbitbuilder.com> <j4n0qr9yeo.fsf@informatik.hu-berlin.de> <3D7C8C76.8060700@NOSPAMbitbuilder.com>
Message-ID: <Xns92849D1FAE9D8duncanrcpcouk@127.0.0.1>

Ahmad Baitalmal <ahmad at NOSPAMbitbuilder.com> wrote in
news:3D7C8C76.8060700 at NOSPAMbitbuilder.com: 

> That's not what my problem is, that I know about ( crlf's between
> nodes),, 
> 
> Here is the deal:
><cow>
>     <hide value="spotted
> with black and white"></hide>
></cow>
> 
> After the word "spotted" there is a crlf, -inside- the attribute
> value. Sholdn't it be treated as part of the value?
> 
> The value now comes stripped of that crlf.

The XML specification, para 3.3.3 specifies that attribute values must
be normalised. The normalisation will convert your newline to a space and 
may also remove duplicate spaces:

> 3.3.3 Attribute-Value Normalization
> Before the value of an attribute is passed to the application or
> checked for validity, the XML processor must normalize the attribute
> value by applying the algorithm below, or by using some other method
> such that the value passed to the application is the same as that
> produced by the algorithm. 
> 
> All line breaks must have been normalized on input to #xA as described
> in 2.11 End-of-Line Handling, so the rest of this algorithm operates
> on text normalized in this way. 
> 
> Begin with a normalized value consisting of the empty string.
> 
> For each character, entity reference, or character reference in the
> unnormalized attribute value, beginning with the first and continuing
> to the last, do the following: 
> 
> For a character reference, append the referenced character to the
> normalized value. 
> 
> For an entity reference, recursively apply step 3 of this algorithm to
> the replacement text of the entity. 
> 
> For a white space character (#x20, #xD, #xA, #x9), append a space
> character (#x20) to the normalized value. 
> 
> For another character, append the character to the normalized value.
> 
> If the attribute type is not CDATA, then the XML processor must
> further process the normalized attribute value by discarding any
> leading and trailing space (#x20) characters, and by replacing
> sequences of space (#x20) characters by a single space (#x20)
> character. 
> 
> Note that if the unnormalized attribute value contains a character
> reference to a white space character other than space (#x20), the
> normalized value contains the referenced character itself (#xD, #xA or
> #x9). This contrasts with the case where the unnormalized value
> contains a white space character (not a reference), which is replaced
> with a space character (#x20) in the normalized value and also
> contrasts with the case where the unnormalized value contains an
> entity reference whose replacement text contains a white space
> character; being recursively processed, the white space character is
> replaced with a space character (#x20) in the normalized value. 
> 
> All attributes for which no declaration has been read should be
> treated by a non-validating processor as if declared CDATA. 
> 

So the only way to get a newline into an attribute is to escape it in using 
an entity reference.

-- 
Duncan Booth                                            
duncan at rcp.co.uk int month(char
*p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3" 
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure? 


From benji_75 at hotmail.com  Thu Sep 19 00:23:39 2002
From: benji_75 at hotmail.com (Ben Gerblich)
Date: Thu, 19 Sep 2002 13:53:39 +0930
Subject: try import except want lineno from Exception
Message-ID: <F98k7LiHdlQgdYXSVvR000003f9@hotmail.com>

I am importing module files, and want to catch an error (if any) and
print the filename and line number of the error.
The "Exception ,e" does not return, or does not have the attirbutes listed 
above. It only returns e="line() takes exactly 2 arguments (3 given)"

I think I want to use SyntaxError as it has the attributes: filename, 
lineno, offset and text.
I have found no info on traceback.
With the code below, how do I find the **line_number**?
Thanks.
--<snip>--
[13:45:40]~/py/dev/import2:cat runme.py
from progfile import *

[13:45:44]~/py/dev/import2:cat progfile.py
class Foo:
    def __init__(self):
        self.lines = []

    def line(self,x):
        self.lines.append( x )

print 'start'
try:
    klass = __import__('module')
    code = "object = klass.%s()" % 'Bar'
    exec code
except TypeError, e:
    print 'TypeError\n%s' % e
except Exception, e:
    print 'Exception\n%s' % e
print 'end'

[13:45:51]~/py/dev/import2:cat module.py
from progfile import Foo

class Bar(Foo):
    def __init__(self):
        Foo.__init__(self)
        print 'line-1'; self.line(20)
        print 'line-2'; self.line(40,60) # <--- an error!!!
        print 'line-3'; self.line(80)

[13:45:57]~/py/dev/import2:python2 runme.py
start
line-1
line-2
TypeError
line() takes exactly 2 arguments (3 given)
end

[13:46:07]~/py/dev/import2:


_________________________________________________________________
Join the world?s largest e-mail service with MSN Hotmail. 
http://www.hotmail.com




From bokr at oz.net  Tue Sep  3 12:28:15 2002
From: bokr at oz.net (Bengt Richter)
Date: 3 Sep 2002 16:28:15 GMT
Subject: getch in python
References: <75ca4e69.0209022326.4cde72e8@posting.google.com>
Message-ID: <al2nuv$fp1$0@216.39.172.122>

On 3 Sep 2002 00:26:13 -0700, vvenka1 at hotmail.com (Venkat Venkataraju) wrote:

>Hi All
>
>Is there any function that behaves like getch() function in C? 
>
If you are on windows,

 >>> import msvcrt
 >>> while 1:
 ...     c=msvcrt.getch()
 ...     print '%02X: "%s"' % (ord(c),c)
 ...     if c=='q': break
 ...
 31: "1"
 32: "2"
 33: "3"
 61: "a"
 41: "A"
 40: "@"
 01: "?"
 71: "q"

(The 01 was a ctrl-A, and on a console window you see a happy face
instead of the question mark ;-)

For more info, see

    http://www.python.org/doc/current/lib/msvcrt-console.html

On *nix look into curses.

Regards,
Bengt Richter


From loewis at informatik.hu-berlin.de  Fri Sep 20 16:20:11 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 20 Sep 2002 22:20:11 +0200
Subject: Python 2.2.1 compatible with Tcl/Tk 8.4?
References: <yu9965x05vkr.fsf@europa.research.att.com>
Message-ID: <j4ptv89zwk.fsf@informatik.hu-berlin.de>

Andrew Koenig <ark at research.att.com> writes:

> Now that Tcl/Tk 8.4 is out, can anyone tell me whether Python 2.2.1
> will work with it?  

It appears to work fine here, and I can't see a reason why it would be
incompatible. Notice that the setup.py of 2.2.1 is already prepared to
search for Tcl/Tk 8.4 on its own.

Regards,
Martin


From donn at u.washington.edu  Thu Sep  5 15:48:50 2002
From: donn at u.washington.edu (Donn Cave)
Date: 5 Sep 2002 19:48:50 GMT
Subject: str() of a tuple
References: <H1xnHJ.1zE1@beaver.cs.washington.edu> <UtMd9.9506$HY3.2176138@newsc.telia.net>
Message-ID: <al8cf2$ibi$1@nntp6.u.washington.edu>

Quoth "Fredrik Lundh" <fredrik at pythonware.com>:
| Douglas Zongker wrote:
|> Why, oh why does the str() of a container use the repr()s of the
|> objects inside?  This seems to violate what the docs claim str() does
|> -- "return a nicely printable representation of an object."
|
| "an object" refers to the container itself, not the objects
| in the container.

And the expression "nicely printable" is so ambiguous that it can
mean whatever you like.  I like to define it as what str() does,
which automatically resolves any question about whether str could
be violating this provision.

	Donn Cave, donn at u.washington.edu


From mennosimons at gmx.net  Thu Sep 26 09:57:28 2002
From: mennosimons at gmx.net (Ulli Stein)
Date: Thu, 26 Sep 2002 15:57:28 +0200
Subject: unicode woes
References: <amuhrk$4cn$05$1@news.t-online.com> <3d92d120$0$76313$e4fe514c@dreader4.news.xs4all.nl> <amuldd$bur$01$1@news.t-online.com> <s2eb6-433.ln1@atlas15.dnp.fmph.uniba.sk>
Message-ID: <amv3o4$p1s$02$1@news.t-online.com>

Radovan Garabik wrote:

> Ulli Stein <mennosimons at gmx.net> wrote:
>  : But nevertheless you would have to use everywhere the Python unicode()
>  : function.
> 
>  : Or how would you do this (nonsense string):
>  : str = str + "bla?" + str[:5] + "??
> 
>  : Would you write then:
>  : str = str + unicode("bla?") + str[:5] + unicode("??)?
> 
> I ran into the same problem and find out that python2.3
> handles this.
> Just add the line
> # -*- coding: UTF-8 -*-
> as the second line of your python script and then
> you can use all your strings in UTF-8 encoding
> (or any other, whatever fits you)
> 

>From where do you have this information? Is there some official changelog 
for python2.3?

Thanks.


From skip at pobox.com  Tue Sep 10 15:14:32 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 10 Sep 2002 14:14:32 -0500
Subject: read() file??
In-Reply-To: <-1076673467.1031684821080.JavaMail.nobody@webmail2.brturbo.com>
References: <-1076673467.1031684821080.JavaMail.nobody@webmail2.brturbo.com>
Message-ID: <15742.17560.693117.961029@12-248-11-90.client.attbi.com>

    Juliano> I have a file with assembler instruction and i want read the
    Juliano> lines of the file and put in a list of string, this is easy to
    Juliano> do, but i don?t want to consider the caracteres "/" and the
    Juliano> othes after him.

    Juliano> file.asm:
    Juliano> ADD D1  /coment1
    Juliano> LDA D2  /coment2

    Juliano> list of strings:
    Juliano> X=[ 'ADD D1' , 'LDA D2' ]
	
How about:

import re
x = []
for line in file("file.asm"):
    line = re.sub(r"\s*/.*", "", line.strip())
    x.append(line)

?

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From rjones at ekit-inc.com  Thu Sep 19 08:59:26 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Thu, 19 Sep 2002 22:59:26 +1000
Subject: StructuredText -> HTML module??
In-Reply-To: <imji9.23500$sR2.405010@news4.ulv.nextra.no>
References: <wbii9.19869$0p1.359879@news2.ulv.nextra.no> <3d89b69d$1@news.sentex.net> <imji9.23500$sR2.405010@news4.ulv.nextra.no>
Message-ID: <200209192259.26556.rjones@ekit-inc.com>

On Thu, 19 Sep 2002 10:24 pm, Thomas Weholt wrote:
> All I'm actually after is to be able to write stuff like :
>
> **Header**
>
> List:
> 1. Foo
> 2. Bar
> 3. Python
>
>     Heres an indented section of text.
>
> And turn this into
>
> <h1>Header</h1>
>
> List:
> <ol>
>    <li>foo</li>
>   <li>bar</li>
>   <li>Python</li>
> </ol>
>
> <P>
> <blockquote>
>     Heres an indented section of text.
> <blockquote>
> </P>
>
> Or something similar. Since there's allready something that I believed was
> standardized it would be easier to use modules allready available than
> doing it all from scratch.

If you're really able to go with "something similar" then look at the docutils 
project (http://docutils.sf.net/ ... start with the "primer" document). It's 
like StructuredText, but much better in a lot of ways (see that website for 
detail if you need it). The above text would be written with a minor 
change... the heading would be:

Header
------

instead of **Header** (which denotes generic bolding in ReStructuredText).
   


    Richard




From vinodvas at pacific.net.sg  Mon Sep  2 04:24:49 2002
From: vinodvas at pacific.net.sg (Vinod Vasnani)
Date: 2 Sep 2002 01:24:49 -0700
Subject: PyQt: Win Systray issue
References: <5695be92.0208291731.6f3d696@posting.google.com> <mailman.1030722464.24485.python-list@python.org>
Message-ID: <5695be92.0209020024.60258514@posting.google.com>

Phil Thompson <phil at river-bank.demon.co.uk> wrote in message news:<mailman.1030722464.24485.python-list at python.org>...
> 
> 
> Hmm - the second request for this in a week.
> 
> I'll implement winEventFilter() for the next version - sometime in 
> September.
> 
> Phil

Thanks in advance!

Vinod


From timh at zute.net  Sun Sep  8 09:53:20 2002
From: timh at zute.net (Tim Hoffman)
Date: Sun, 08 Sep 2002 21:53:20 +0800
Subject: python program to watch a file - example.
References: <mailman.1031486051.16547.python-list@python.org>
Message-ID: <3D7B5650.4050304@zute.net>

Wayne R wrote:
> Sorry for the html,  I forgot to turn it off (as I usually turn it off
> as soon as I install a new system, forgot this time).
> 
> The platform is Gentoo Linux,  I am writing a piece of code for the
> Twiggi project to patch in adduser support.
> 
> 
> |-> > >I am trying to write a python program that will run as a deamon
> and
> |-> > >watch a file for data to arrive and then wake up and perform
> duties
> |-> on
> |-> > >that data.   I found ways to make a python program run as a
> deamon
> |-> but I
> |-> > >am at a loss to find anything related to watching a file for data
> so
> |-> > >that it can then process the data.  I didn't really want to do a
> |-> sleep
> |-> > >loop or something like, I am sure there is a better way.
> |-> > >
> |-> > >Any ideas or links to more info?
> |-> > >
> |-> > Don't annoy people you want help from, for starters.
> |-> > What system are you targeting? (Outlook in the headers says you're
> on
> |-> windows
> |-> > generating this abominable mix of actual message and useless
> cruft).
> |-> >
> [Wayne R] 
> 
> 
> 

Are you talking about the same file always (ie you don't have a new 
file, overwriting the old one ?).  Anyway you really need to look at
how the underlyng OS, can handle such an event.

If it is always the same file and seeing you are looking at *nix 
varient, one of the unix like ways would be to tail -f the file, and 
hook it up to your process which just does blocking reads on stdin.

ie tail -f somfile | python my_watch.py

Rgds

Tim



From eddie at holyrood.ed.ac.uk  Fri Sep 13 07:45:47 2002
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Fri, 13 Sep 2002 11:45:47 +0000 (UTC)
Subject: Expect() from telnetlib giving me problems.
References: <4378fa6f.0209121325.130766fc@posting.google.com>
Message-ID: <alsj5b$8de$1@scotsman.ed.ac.uk>

mnations at airmail.net (Marc) writes:

>Hi,

>This is the error I'm getting:

>Exception in thread Thread-3:


>Traceback (most recent call last):
>  File "C:\Python22\lib\threading.py", line 408, in __bootstrap
>    self.run()
>  File "C:\Python22\lib\threading.py", line 396, in run
>    apply(self.__target, self.__args, self.__kwargs)
>  File "main.py", line 85, in run
>    command[0](*command[1:3])
>  File "key.py", line 94, in command
>    read_result(conn)
>  File "key.py", line 104, in read_result
>    buffer = conn.expect(alist, 5)
>  File "C:\python22\lib\telnetlib.py", line 541, in expect
>    list[i] = re.compile(list[i])
>  File "C:\Python22\lib\sre.py", line 178, in compile
>    return _compile(pattern, flags)
>  File "C:\Python22\lib\sre.py", line 228, in _compile
>    raise error, v # invalid expression
>error: nothing to repeat


>I've tried several different ways to get the expect command to work. I
>can't find any examples and only limited documentation. Here is how
>I'm trying to implement it:
> 

>    alist = [ 'COMPLD', '*/' ]
>    buffer = conn.expect(alist, 5)

> 
>Can someone please tell me what is wrong or provide me with some
>working examples of using the command.

As previously mentioned '*/' is not a regular expression.  If you want to
search for fixed strings without bothering to learn abot regular expressions
or if you're not sure exactly what the input will be and whether it may
contain special characters you could do:

  import re
  alist = map (re.escape, [ 'COMPLD', '*/' ])

Eddie


From ehagemann at comcast.net  Wed Sep 25 05:38:45 2002
From: ehagemann at comcast.net (Eric Hagemann)
Date: Wed, 25 Sep 2002 09:38:45 GMT
Subject: help with print function in extension
References: <tA8k9.270440$AR1.11448440@bin2.nnrp.aus1.giganews.com> <3D912DC3.7010302@something.invalid>
Message-ID: <Fufk9.211854$216.7261698@bin4.nnrp.aus1.giganews.com>

Its my understanding that str() should return a PyObject * to a string


"Greg Ewing" <see_reply_address at something.invalid> wrote in message
news:3D912DC3.7010302 at something.invalid...
> Eric Hagemann wrote:
>
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in ?
> > SystemError: error return without exception set
>
>
> Check what your C function implementing str() is
> returning. My guess is that you've forgotten a
> "return 0" statement, and it's returning some junk
> value that in some cases the Python interpreter
> thinks is signalling an error.
>
> --
> Greg Ewing, Computer Science Dept,
> University of Canterbury,
> Christchurch, New Zealand
> http://www.cosc.canterbury.ac.nz/~greg
>




From eric.brunel at pragmadev.com  Thu Sep  5 06:11:52 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Thu, 5 Sep 2002 12:11:52 +0200
Subject: Privacy and Inheritance
References: <mailman.1031180226.13060.python-list@python.org>
Message-ID: <al7a71$hnh$1@wanadoo.fr>

Delaney, Timothy wrote:

>> From: Eric Brunel [mailto:eric.brunel at pragmadev.com]
>> 
>> In languages that have actual "private", "protected" and "public"
>> attributes, private attributes are not inherited. They are
>> visible only in
>> the class that define them, not outside it and not in its
>> sub-classes. What
>> you want is a protected attribute, that is visible in the
>> defining class
>> and in its sub-classes.
>> 
>> Unfortunately, Python doesn't know anything about protected
>> attributes.
>> AFAIK, there's no simple way to do what you want: in Python,
>> attributes are
>> either private or public; there's nothing in between...
> 
> Actually, there is.
> 
> Anything beginning with a leading underscore is not imported via a 'from
> module import *' (not that anyone should in general use this ;) Thus
> anything with a leading underscore is something other than completely
> "public".
> 
> There is a (strong) convention that anything beginning with a *single*
> leading underscore is "protected". No name mangling is performed, but the
> name is flagged as "special".

I just meant that Python has no way of enforcing "protectedness", as 
leading double underscores (kind of) enforce privateness. As you say, using 
a single leading underscore in protected attribute names is merely a 
convention, even if it's a strong one: Python won't prevent any use of the 
attribute, neither from sub-classes, nor from the "outside world". Anybody 
not knowing the convention can use these attributes in any way without any 
problem.
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From tjreedy at udel.edu  Tue Sep 10 21:08:59 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 11 Sep 2002 01:08:59 GMT
Subject: Python primer - comments appreciated!
References: <aljdhe$1q63hj$1@ID-77524.news.dfncis.de> <w3cf9.535714$m91.21695020@bin5.nnrp.aus1.giganews.com> <allpms$1r005a$1@ID-77524.news.dfncis.de>
Message-ID: <LIwf9.447074$Aw4.18721391@bin2.nnrp.aus1.giganews.com>

"Thorsten Kampe" <thorsten at thorstenkampe.de> wrote in message
news:allpms$1r005a$1 at ID-77524.news.dfncis.de...
> * Terry Reedy
...
> >      result.setdefault(func(item),[]).append(item)
>
> To make every item hashable, I made this:
> result.setdefault(repr(func(item)),[]).append(item)
>
> >      # {}.setdefault(key,default) was added for just this sort of
usage
>
> Sorry, I just don't get it: your mixing (chaining?) a dict method
and
> a list method as if it was a f(g(x)) and /it works/!?! Where is that
> documented? How is is that evaluated (left to right)?

Yes, left to right.  Dict.setdefault returns the value corresponding
to the key after setting d[key]=default if key not already in d.  In
this case, *all* values are initialized to [] by such a call.  Then
append to the list.  This working depends on there being *two*
references to the list.  The one in the dict (possibly new, possible
not) and the one returned by the method and used by append().  See
Python Library Reference 2.2.7 for dicts.

> > >>> f = lambda x: x % 3
> > >>> equiv_class([0,1,2,3,4,5,6,7,8,9],f,part='bool')
>
> Now this is:
> quotient_set([0, 1, 2, 3, 4, 5, 6, 7, 8, 9], lambda x: bool(x % 3))
>
> (Mathematically spoken, the quotient set is the set of all
equivalence
> classes induced by func, which is what we return)

Since the equivalence lists may have duplicates, they are not,
strictly speaking, sets.  If the set of return values were made to be
dicts (or, in 2.3, sets) instead of lists, then they would be.  Such
details can, of course, be adjusted for a particular application.

Terry J. Reedy





From tjreedy at udel.edu  Wed Sep  4 11:39:11 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 04 Sep 2002 15:39:11 GMT
Subject: Please help in  calling python fucntion from 'c'
References: <mailman.1031142606.11252.python-list@python.org>
Message-ID: <zOpd9.291180$Aw4.12264386@bin2.nnrp.aus1.giganews.com>

> [ The information contained in this e-mail is confidential and is
intended for the named recipient only. If you are not the named
recipient, please notify us by telephone on +44 (0)1249 442 430
immediately, destroy the message and delete it from your computer.

I will gladly delete it, but expensive call? no.  Pretty whacko.




From SBrunning at trisystems.co.uk  Wed Sep 25 09:41:52 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 25 Sep 2002 14:41:52 +0100
Subject: missing functions in os.path
Message-ID: <31575A892FF6D1118F5800600846864DCBD876@intrepid>

> From:	Jens Bloch Helmers [SMTP:helm at dnv.com]
> I hope that a future version of Python will include the following
> two functions:
> 
>      os.path.localpath( file )       # Returns the non-UNC name 
>                                      # of file.
> 
> >>>file = r'\\computer\shared\myfile.txt'
> >>>os.path.localpath( file )
> 'd:\somedir\shared\myfile.txt'
 
A word of advice - don't use 'file' as a label - it clobbers the built-in.
 
>      os.path.relpath( file, dir )    # returns the relative path of
>                                      # file relative to directory dir
> 
> >>>file = r'C:\test\mydir\hum.txt'
> >>>dir  = r'C:\otherdir'
> >>>os.path.relpath( file, dir )
> '..\test\mydir\hum.txt'
 
All patched gratefully received, I'm sure!

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




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



From erikprice at mac.com  Tue Sep 17 10:04:47 2002
From: erikprice at mac.com (Erik Price)
Date: Tue, 17 Sep 2002 10:04:47 -0400
Subject: Good book choice?
In-Reply-To: <2393234a.0209160513.469a4357@posting.google.com>
Message-ID: <6CF1C3E4-CA46-11D6-888F-00039351FE6A@mac.com>

On Monday, September 16, 2002, at 09:13  AM, davisman wrote:

> This book includes a CD with
> all the scripts used in the book

<vent>
This isn't directed at the book mentioned (which is why I didn't quote 
its name in this rant), rather at all books which come with CDs.  I 
hate it when books come with CDs.  I'm perfectly capable of downloading 
the source code used in the book, and in fact I prefer to hand-type it 
while reading from the book anyway, to make sure I see every part of 
the code.  When a book comes with a CD, it is cumbersome to read 
portably, which is the whole advantage of books, unless you take the CD 
out.  And if I take the CD out, then there is no doubt that it will get 
lost, or I will be unable to return the book if dissatisfied, etc.  
Plus, you can't tell me that the inclusion of a CD and the 
envelope/publishing used to include the CD doesn't add a few units of 
currency to the book's cost.

Poke this opinion full of holes, but when I buy a book, I want a book, 
not a software package.  The book should be good enough to stand on its 
own without the "bonus" material on CD.  I think twice about buying a 
book that comes with a CD, to tell you the truth.
</vent>



Erik






--
Erik Price                                                  (zombies 
roam)

email: erikprice at mac.com
jabber: erikprice at jabber.org




From ehagemann at comcast.net  Sun Sep 29 14:11:26 2002
From: ehagemann at comcast.net (Eric Hagemann)
Date: Sun, 29 Sep 2002 18:11:26 GMT
Subject: help with writing extension
Message-ID: <inHl9.24658$Ii4.986202@bin2.nnrp.aus1.giganews.com>

I have been handcrafting an extension module creating a new type

I am making use of the numeric methods (+,-,* etc)

I solved the problem of the object interating with other types by working
through the __coerce__() function and have no issue with operations like
<obj>+2 but I am getting an error on 2-<obj>.

In Beazley's book there are descriptions of functions like __radd__() when
defining a new type in python, but I cannot find reference to these
functions in C ?

I have checked the doc's (not to say that I did not miss something ;) )

What is interesting is that both <obj>+2 and 2+<obj> both end up in the
__coerce__ function.  Whats even more interesting is that the arguments to
__coerce__ are in the same order for both function calls, the first argument
is <obj> the second is a PyInt.  The Error I get (for 2+<obj>) is

Traceback (most recent call last):
  File "C:\py_lib\test.py", line 30, in ?
    print "2+d = ",2+d
TypeError: unsupported operand types for +

that occurs after the __coerce__ function finishes.

Has anybody worked through this before?









From aleax at aleax.it  Wed Sep 25 06:28:37 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 10:28:37 GMT
Subject: Python Compiling
References: <amfveg$5g64k$1@ID-53055.news.dfncis.de> <80a8af7d.0209250113.28ae4c72@posting.google.com>
Message-ID: <pdgk9.140947$ub2.3077676@news1.tin.it>

Klaus Momberger wrote:

> "Terje Johan Abrahamsen" <spoermeg at voldelig.com> wrote in message
> news:<amfveg$5g64k$1 at ID-53055.news.dfncis.de>... .....
>> 
>> Is it not possible to write a complete compiler like the C compiler for
>> Python? Or is it just that nobody has done it? Or must the language be
>> constructed differently?
>> 
> Slightly off-topic, but if you are looking for a language which can
> be run by an interpreter and can be compiled as well, maybe you should
> take a look at Ocaml:
> 
> http://www.ocaml.org

Or Haskell, www.haskell.org -=- HUGS inteprets it, the Edinburgh
compiler compiles it.  Scheme has _many_ interpreters and compilers
too.  Hey, so does Visual Basic... it's quite the thing these days.


Alex



From jb at cascade-sys.com  Tue Sep 17 11:18:42 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 17 Sep 2002 08:18:42 -0700
Subject: threads or queue for this task
References: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com> <mailman.1032241763.6933.python-list@python.org> <d7Ah9.103593$ub2.2156613@news1.tin.it> <mailman.1032260122.29578.python-list@python.org> <6dFh9.120102$pX1.4321019@news2.tin.it>
Message-ID: <3D8747D1.6050304@cascade-sys.com>

Alex Martelli wrote:

>Taking it for granted that one wouldn't pontificate without vast
>relevant experience,
>

A rather rash assumption for internet communications.  ;o)

But I appreciate your giving me benefit of the doubt.

>Perhaps the hint is in Mr Pettersen's mention of
>"processing too many 16-GB files" vs your insistence on "capacity
>to handle the expected traffic", "time-critical requests", "the
>work requests may be created on a different machine", etc.
>
>I.e., you're focusing (perhaps because of how life shaped your
>experience of threading) on a "server" process or machine that
>is just responding to traffic originating elsewhere, while I'm
>giving at least equal weight to a process that's "just" doing a
>lot of batch-like work, e.g. generated by splitting up the task
>of processing a "16-GB file" or the like, and using threading
>to ensure the best throughput (implies that substantial parts of
>processing each work request happen in C code that release the
>GIL, be it within the Python core or in extensions).  
>

It does sound like a big part of our disagreement is past experience 
with different types of appellations. 

The bulk of my past experience almost always involved multiple processes 
and often multiple processors and usually highly complex, dedicated 
applications.  Generally, lost data was never an option, though as I 
enumerated some of the systems in my previous note, a few did include 
bounded queues.

More specifically to Python, IIRC, whenever I've used threads, the 
application always involved input via sockets, ultimately from client 
apps, over which the server app had no direct control.  In those cases, 
I've always used unbounded queues to connect threads and they worked 
just fine.

I never wrote a stand-alone Python app like you describe.  I agree if 
you're writing a self-contained application then it makes sense to used 
bounded queues to help throttle the generator thread.

>If this is the root of our disagreement, I suggest you reconsider
>whether "the general case" need ONLY include "networked"
>situations, which seem to be what you istinctively think of
>in this context, 
>

In a cursory review of the record, I don't see where I made that 
particular assertion.

>or "threading to slice up a big batch of work" too.  
>

Seems at every opportunity I agree that your scenarios are valid and 
never once said or implied they were not.  I think part of my problem 
here is I don't have the experience to pick words that exactly match 
your expectations.  

I think the number of times the following sentence has been referred to 
in our discussion may point to the crux of our problem.

>>>>>>Generally, Queues never get "full" and you generally don't care if they're empty.
>>>>>>            
>>>>>>

I intended this statement to be a gross simplification in the interest 
of directing Robin away from what I regarded to be unnecessary 
complexities.  

Recall that it was an answer to the specific question of  "I am put off 
by [...] empty() [...] how can you code if you don't know if the queue 
is empty or full?"  It was intended as a way for the beginner to begin 
to use queues and not worry about unnecessary details.  

Out of context the statement is a much easier nit to pick.  In context, 
it's as much a part of an elaboration on  "multithreading semantics" as 
anything else.  The last sentence is the most important part.

    Generally, Queues never get "full" and you generally don't care if
    they're empty.  The import thing about Queues is that the reader
    thread will automatically be suspended when reading from an empty
    queue and furthermore it will automatically wake up when the queue
    becomes non empty.  All you need from Queues is to create them and
    to get and put data.

I submit that the oversimplification applies, in a way, to your bounded 
queue scenario -- the programmer generally shouldn't "CARE" if the 
queues are empty or full -- the important thing is that they block and 
unblock when appropriate.

I hardly intended this statement to stand as an accurate 
characterization of all queuing systems, which appears to be a big part 
of your complaint.  

In any case, I hereby withdraw the entire statement altogether and 
invite you to propose instead whatever alternative summary you think 
best serves the the OP and the Python community generally.  In this I 
invite you to have the last words on the subject.

>...but setting a Queue to bounded, and implicitly sleeping when
>trying to add one more item to a full Queue, is NOT significantly
>more complicated.  It's strongly parallel to sleeping just as
>implicitly when trying to peel an item from an empty Queue.  The
>symmetry, when applicable, may be considered more elegant than
>an *asymmetric* solution, after all.
>

Looking at queues in isolation, I agree an upper bound is not a deep 
concept.

Whether bounded or unbounded is more "elegant" is a value judgment where 
reasonable people may disagree.  E.g., your discourse about Rosetti's 
Ferrara seemed to imply that asymmetry was more elegant than symmetry.

In your scenario of a self-contained process, a bounded queue is simpler 
solution.  

In my scenario where incoming traffic may have to be discarded or dealt 
with in some out of band fashion a bounded queue is more complicated 
than unbounded queues.

>But at least we can agree to
>disagree civilly, whence my apologies for appearing unwilling to
>do so in my earlier post.
>

This is progress, I suppose.

With my latest concessions and stipulations I hope we're no longer all 
that far apart.

>>[...] Extreme Programming [..]
>>    
>>
>But from my POV this generally goes for *threading in general*!  
>

True and I believe I was the only person to mention previously that 
multi threading may be overkill.

However, the original request was specifically about threads and without 
more detail, "don't use threads" seemed an inappropriate response.

>I always consider a first-cut solution WITHOUT multi-processing, which
>is a huge complication in itself, more often than not.  
>

While generally I would agree, I find that many TCP/IP applications are 
somewhat  intrinsically concurrent.  Within that, threads are much 
simpler than, say, using Select().  For starters, it's sometimes easier 
to have separate reader and writer threads than to rely on incoming in 
outgoing data to be perfectly interleaved.   A common paradigm is a 
listener thread accepts connects from a socket and launches new reader 
threads to handle each connection.  Then maybe the reader threads queue 
some incoming requests to server threads so that the socket never fills 
or blocks.  In one application, I had two listeners listening on two 
separate sockets, each spawning per-client reader threads, each of which 
funneled some subset of requests to a common processing thread guarding 
a critical resource.  On the client side, sometimes it's easier having a 
separate thread monitoring return traffic.  In another app, I have a 
listener thread which spawns multiple connection threads, one for each 
external client.  Connection threads process commands, some of which are 
queued to a "command processor" thread and others of which queued to a 
serial port writer thread.  A serial port reader thread handles the back 
traffic, some of which gets forwarded via selected connection threads 
back to interested client processes.  Although theoretically possible, I 
shudder to think of how difficult that would be to implement using 
Select() instead of threads.  

In all of these cases, I have encountered no need for limiting the size 
of the queues used.  But a key ingredient is there is no 'generator' 
thread to overflow the queues.  Then too, it's also always been a given 
that overall service utilization was very low.

>I do disagree, and this may be a philosophical point in part.  
>

That would explain the amount of energy being expended by both parties.

The argument is so intense because the stakes are so low.  ;o)

>But the ability to make SOME Queue's bounded,
>and still handle them without fuss, IS important in a vast
>enough number of cases that ruling this possibility out is
>a serious didactic mistake, IMHO.
>

I stipulated this much from the beginning and never ruled it out.

>In particular, when preparing work-requests is part of the
>design's task (e.g. by analyzing pieces of a huge existing
>file), a bounded Queue can simply and without problems do
>some appropriate balancing within the system between the
>amount of resources devoted to preparing work requests,
>and the amount devoted to processing them.  It's as simple
>as that!
>

You've made this point more than once in this exchange.  I've never 
disputed it.

>Perhaps, but this is a straight quote from Albert Einstein, of
>course, therefore you should take the matter up with him...
>

The important thing is we both agree it should be as simple as possible.  

We simply don't agree what that means, exactly.

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From karczma at info.unicaen.fr  Wed Sep  4 07:09:59 2002
From: karczma at info.unicaen.fr (Jerzy Karczmarczuk)
Date: Wed, 04 Sep 2002 13:09:59 +0200
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com> <unbp765rbsn809@corp.supernews.com>
Message-ID: <3D75EA07.D5EB1F5A@info.unicaen.fr>

RPM1 wrote:

> If you really want to get a feel for a language, read a fairly large
> program that somebody else has written in that language, (not
> just little snippets).
> 
> Patrick

I'll tell you an anecdote. 100% authentic and meaningful.
Long time ago a friend of mine, a physicist, had to read, understand,
correct and extend a really huge program in Fortran, something within
the realm of High Energy Physics, with Monte-Carlo, modelling of
resonances, and other succulent stuff.

The suicide was not far, but finally he managed.
And then he told me:
  "What a disgusting, unreadable programming style!
   It was a nightmare to read all that!
   But now I became a competent, experienced Fortran programmer!"

And, yes..., he continued to program in Fortran for many years.

So, if somebody suggests that a newbie should learn "C" or assembler
just in order to appreciate the elegance of Python or other, even nicer
(yes, there are...) languages, beware. The result might be unexpected.

You know, I am a peaceful, animal-friendly guy who thinks that corridas
are a pure barbarism. But I refused myself the temptation to go and see
one, because I am afraid that, all philosophy apart, I may discover
that - after all - perhaps I adore the public slaughter of bulls?

Jerzy Karczmarczuk
Caen, France


From aleax at aleax.it  Fri Sep 20 03:23:46 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 07:23:46 GMT
Subject: implementing reduce protocol in C
References: <3D8A7DA1.7090506@hotmail.com>
Message-ID: <62Ai9.136323$pX1.4852975@news2.tin.it>

Tom Widrick wrote:
        ...
> static PyObject *
> myobj_reduce(self)
> MyObject *self;
> {
> PyObject *args;
> PyObject *result;
> 
> args = PyTuple_New(5);
> if(args == NULL)
> return NULL;
> 
> PyTuple_SET_ITEM(args, 0, self->po_name);
> PyTuple_SET_ITEM(args, 1, self->po_parents);
> PyTuple_SET_ITEM(args, 2, self->po_children);
> PyTuple_SET_ITEM(args, 3, self->po_doc);
> PyTuple_SET_ITEM(args, 4, self->po_dict);

This is stealing references to the five objects you pass as items.
I'm not sure what effects this might have, but doesn't look good.
Perhaps it's a contributing factor to strange behavior (though it
would cause anomalies at pickling time, not when unpickling...).

> result = Py_BuildValue("(OOO)", self->ob_type, Py_None, args);
> return result;
> };

Looks like a leak (no decref of args) but that's not the specific
problem's that's biting you, I think.

> The type constructor wouldn't have worked as a call to unpickle the
> object so I've used basicnew instead.
> 
> static PyObject *
> myobj_basicnew()
> {
> MyObject *self;
> 
> self = PyObject_New(MyObject, &MyObject_Type);
> return (PyObject *)self;
> }

Do you expose this as the __basicnew__ static method of your type?
I'm not sure how you do so, as the definition of staticmethod
lives in funcobject.c, not exposed in the .h, I think.  Or, ditto
for classmethod.  Haven't tried exposing a __basicnew__ method
from a C-coded type, yet.  If I had to, I think I'd try to run
from C the Python-level exposed staticmethod built-in -- but since
as I said I didn't try that yet, I may be missing something.


Alex



From pan-newsreader at thomas-guettler.de  Tue Sep 24 15:15:30 2002
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Tue, 24 Sep 2002 21:15:30 +0200
Subject: isFloat: Without Exception-Handling
References: <mailman.1032531448.12021.python-list@python.org>
Message-ID: <pan.2002.09.24.21.15.30.363977.730@thomas-guettler.de>

On Fri, 20 Sep 2002 16:15:55 +0200, Mark McEahern wrote:

> [Magnus Lycka]
>> Something like this perhaps:
>>
>> import re
>>
>> def isFloat(S):
>>      floatRE = r"^[-+]?(\d+\.?\d*|\.\d+)([eE][-+]?\d+)?$" return
>>      re.match(floatRE, str(S)) is not None
>>
>> I'm not sure it's faster though.
> 
> If you were going to run this lots of times, you might want to compile
> the pattern.  Anyway, the exception based code is faster, by my testing
> (harness at bottom):
> 
>        isFloatExcept 0.220
>            isFloatRE 0.521
>    isFloatRECompiled 0.400

I thought exceptions are slow. But as 
someone mentioned, exception in python
seem to be different from exceptions in 
java or c++.

Thank you for the performance test!

 thomas


From jon+usenet at unequivocal.co.uk  Mon Sep 16 10:14:11 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Mon, 16 Sep 2002 14:14:11 -0000
Subject: html template system?
References: <btug9.688$Ee4.1090@news-server.bigpond.net.au>
Message-ID: <slrnaobppi.jgm.jon+usenet@snowy.squish.net>

In article <btug9.688$Ee4.1090 at news-server.bigpond.net.au>, :B nerdy wrote:
> what are good templating systems avaliable for python?
> 
> ive been at http://zebra.sourceforge.org and the documentation is kinda sus.
> zope DocumentTemplate is included with zope. but i want a smaller footprint
> system.

Try http://jonpy.sf.net/ , it does what you describe.


From max at alcyone.com  Wed Sep 18 17:42:13 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 18 Sep 2002 14:42:13 -0700
Subject: ANN: empy 1.5 -- Embed Python in template text as markup
Message-ID: <3D88F335.FC5B4E49@alcyone.com>

Summary

    empy is a system for embedding Python expressions and statements
    in template text; it takes empy source files, processes them, and
    produces output.  This is accomplished via expansions, which are
    special signals to the empy system and are set off by a special
    prefix (by default the at sign, '@').  empy can expand arbitrary
    Python expressions and statements in this way, as well as a
    variety of special forms.  Textual data not explicitly delimited
    in this way is sent unaffected to the output, allowing Python to
    be used in effect as a markup language.  Also supported are
    recording and playback via diversions and dynamic, chainable
    filters.  The system is highly configurable via command line
    options.

    Expressions are embedded in text with the '@(...)' notation;
    variations include conditional expressions ('@(...?...:...)') and
    the ability to handle thrown exceptions ('@(...$...)').  As a
    shortcut, simple variables and expressions can be abbreviated as
    '@variable', '@object.attribute', '@function(arguments)',
    '@sequence' [index], and combinations.  Full-fledged statements
    are embedded with '@{...}'.  Forms of conditional and repeated
    expansion are available via '@[...]'.  A '@' followed by a
    whitespace character (including a newline) expands to nothing,
    allowing string concatenations and line continuations.  Comments
    are indicated with '@#' and consume the rest of the line, up to
    and including the trailing newline.  '@%' indicate optional
    "significators," which are special forms of variable assignment
    intended to specify per-file identification information in a
    format which is easy to parse externally.  Finally, a '@@'
    sequence expands to a single literal at sign.


Getting the software

    The current version of empy is 1.5.

    The latest version of the software is available in a tarball here:
    "http://www.alcyone.com/pyos/empy/empy-latest.tar.gz",
    http://www.alcyone.com/pyos/empy/empy-latest.tar.gz.

    The official URL for this Web site is
    "http://www.alcyone.com/pyos/empy/",
    http://www.alcyone.com/pyos/empy/.


Requirements

    empy should work with any version of Python from 1.5.x onward.


License

    This code is released under the "GPL",
    http://www.gnu.org/copyleft/gpl.html.

	...

Release history [since 1.4]

    - 1.5; 2002 Sep 18.  Escape codes ('@\...'); conditional and
      repeated expansion substitutions via '@[if E: ...]', '@[for X in
      E: ...]', and '@[while E: ....]' notations; fix a few bugs
      involving files which do not end in newlines.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ I am the king of sorrow
\__/ Sade
    EmPy / http://www.alcyone.com/pyos/empy/
 A system for embedding arbitrary Python in template text as markup.


From cnetzer at mail.arc.nasa.gov  Sat Sep 28 22:35:34 2002
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Sat, 28 Sep 2002 19:35:34 -0700
Subject: Best way to create a class dynamically on the fly?
In-Reply-To: <Rgtl9.13598$Lg2.3950931@news2.news.adelphia.net>
References: <tFml9.13043$Lg2.3865279@news2.news.adelphia.net> <3D96511D.2040407@4-am.com> <Rgtl9.13598$Lg2.3950931@news2.news.adelphia.net>
Message-ID: <200209290235.TAA10390@mail.arc.nasa.gov>

On Saturday 28 September 2002 19:08, Robert Oschler wrote:
>
> I understand but what I'm after is building the class itself
> dynamically, akin to the way you can manufacture a prolog predicate
> on the fly.  For example,  if wanted to create a class whose name was
> provided in a string variable called ClassName, at runtime.  Or did I
> misunderstand your reply?

Well, you could create a class at runtime, given a class name, using 
several techniques; But what would be the code?  Do you mean selecting 
from various classes (pre-made) given a string?  Or would the class 
code be generated on the fly?

Perhaps you mean you want to choose, at runtime, which class to 
instantiate given a string, such as in a Factory pattern.

ie:  (extremely simple example)

class Foo:
    def __init__( self ):
        return

def factory( name ):
    if name == 'dict':
        return {}       # An instance of a class
    elif name == 'list':
        return []       # Another instance
    elif name == 'foo':
        return Foo()    # A user defined class
    else:
        return None

This is creating an instance on the fly, given a name at runtime, 
rather than a class, but unless you are doing meta programming, I think 
perhaps it is what you mean (and I apologize if otherwise)

Here is a Python Cookbook entry on Factory patterns, BTW.:
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/86900

BTW, rather than returning the instance, you could return the class 
object itself, to be instantiated later (ie. "return Foo", instead of 
"return Foo()")

-- 

Chad Netzer
cnetzer at mail.arc.nasa.gov



From kahanpaa at gstar.astro.helsinki.fi  Mon Sep  2 07:05:33 2002
From: kahanpaa at gstar.astro.helsinki.fi (Jere Kahanpaa)
Date: 2 Sep 2002 11:05:33 GMT
Subject: file processing question
References: <akvg3g$5bn$04$1@news.t-online.com>
Message-ID: <akvglt$aln$1@oravannahka.helsinki.fi>

Hi.


Axel Grune <NineOfSix at gmx.de> wrote:
> Hi, Im trying to read the data of a file with following method

> file = open( inputfile, "r")
> data = file.read()
> file.close()

> but 'data' is always smaller then the file size, e.g. I have a file 
> which is 25677824 bytes but file.read() reads only 4295 (file.tell() = 
> 25677824).
> Im using python 2.2 under win2k.

Is your data binary or plaintext? Have you tried reading in binary mode? 

file = open(inputfilename,"rb")
data = file.read()
file.close()


Jere
-- 
It's hard to think outside the box when you ARE the box.
                            - unknown, alt.religion.kibology 


From albgrig at tiscalinet.it  Mon Sep 30 07:54:33 2002
From: albgrig at tiscalinet.it (Alberto Griggio)
Date: 30 Sep 2002 04:54:33 -0700
Subject: Python and GUI question
References: <3d96b808_3@news.bluewin.ch> <3d97459e_2@news.arcor-ip.de>
Message-ID: <6a8dcd75.0209300354.696c43b0@posting.google.com>

> [...] WxPython looks good on Win32 and Linux,
> and it pretty powerful. The downside is it doesn't have a (free) GUI
> designer like Gtk or Qt.

Well, it has (at least) three: boa constructor
(boa-constructor.sourceforge.net - also an IDE),
PythonCard (www.pythoncard.org - this is rather a complete framework
than just a GUI designer), and wxGlade (wxglade.sourceforge.net)

Ciao,
Alberto


From marklists at mceahern.com  Wed Sep 11 18:29:37 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 11 Sep 2002 17:29:37 -0500
Subject: Restricting methods in derived classes
In-Reply-To: <slrnanve8n.1to.huaiyu@gauss.almadan.ibm.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOAEADDOAA.marklists@mceahern.com>

[Huaiyu Zhu]
> What I'm looking for is a way to say which methods from dict are inherited
> in MyClass at the time the class is defined, without incuring a cost when
> the methods are called.  Is this possible?

Would something like this work?

$ python
Python 2.2.1 (#1, Jun 25 2002, 10:55:46)
[GCC 2.95.3-5 (cygwin special)] on cygwin
Type "help", "copyright", "credits" or "license" for more information.
>>> class foo(dict):
...     def not_defined(self, *args, **kwargs):
...             raise NotImplementedError
...     update = not_defined
...
>>> f = foo()
>>> f['a'] = 1
>>> f
{'a': 1}
>>> d = {'a': 2}
>>> f.update(d)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 3, in not_defined
NotImplementedError
>>>

// m

-




From peter at engcorp.com  Tue Sep  3 09:01:20 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 03 Sep 2002 09:01:20 -0400
Subject: Design question.... parent/child class
References: <3d74b021$0$234$4d4ebb8e@news.nl.uu.net>
Message-ID: <3d74b283$1@news.sentex.net>

Gumuz wrote:
> hi all,
> 
> i have a SessionManager class, which contains a list of Session
> objects(instantiated from a Session class). I have a design problem.
> 
> Each session object has a small 'queue' of messages for the specific
> session. How can a session send a message to another session without having
> some kind of parent-reference to the SessionManager object?
> 
> i am a bit puzzled, or is this parent-reference not such a bad idea after
> all?

Increasing coupling is always bad, but you could do this in a way that
doesn't really increase coupling (at least in Python).  The Session
objects don't need to know about SessionManagers specifically, just
about some thing that has a particular interface.  In this case, maybe
it's a "post" method.  Pass in a reference to the SessionManager,
or even the post method itself (a "bound" method, bound to a specific
SessionManager instance), but don't do an "import sessionmanager"
or anything like that in the session.py module itself.

Note that passing messages between Sessions could be considered a
function not appropriate for something called merely "SessionManager".
Maybe you want to do this with another mechanism in parallel to this,
or rename that class.  Personally, I wouldn't do that unless the
whole design was starting to get complex.

-Peter



From jb at cascade-sys.com  Sat Sep 28 04:24:54 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sat, 28 Sep 2002 01:24:54 -0700
Subject: Getting started
References: <3D911306.7080609@ucsd.edu> <mailman.1032982114.4627.python-list@python.org> <e15b6fa3.0209261128.4edcdfe8@posting.google.com> <mailman.1033119160.31301.python-list@python.org> <e15b6fa3.0209271155.53ac7fcf@posting.google.com>
Message-ID: <3D956756.7010502@cascade-sys.com>

Adam Taylor wrote:

>Python is in the same square as Lisp (which I mentioned).  So is Perl.
>At least using this terminology... ;)
>
Agreed.

>Okay, that's true, he does define weakly-checked only in terms of
>statically checked languages, which I hadn't appreciated:
>
>But the clear generalization of this (it seems to me) is to define
>weakly checked languages as languages in which some unsafe operations
>are detected and some are not detected.  In other words, just delete the
>word "statically" from the last sentence in the quote above.
>
Exactly.  I think I said as much earlier, though not as succinctly.

>Of course, it doesn't sound like any existing languages are both
>dynamically typed and weakly checked (and I can't think of any myself):
>
I recall a Lisp implementation from long ago that did not distinguish 
between integers and machine addresses.  When applying a function to an 
arg list, if the function name evaluated to a number the interpreter 
blithely JSRed to that address.  This was a Feature, in that it was how 
system built-ins were linked with the higher level stuff.  User code did 
not normally do this and it was not a big issue in practice.  However it 
WAS a loophole that allowed type errors to occur, as it allowed code to 
jump to any random address in the system.  

But I think the generalization holds for most dynamic languages.

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From ianholmes01 at lycos.com  Fri Sep 13 11:58:22 2002
From: ianholmes01 at lycos.com (Ian Holmes)
Date: 13 Sep 2002 08:58:22 -0700
Subject: [newbie] Buying an introductory/reference text
References: <Mu5g9.1702$U_.1021@news04.bloor.is.net.cable.rogers.com> <a38d43b0.0209121413.1bf274c9@posting.google.com>
Message-ID: <b31eb63a.0209130758.652049c3@posting.google.com>

jetman516 at hotmail.com (The Jetman) wrote in message news:<a38d43b0.0209121413.1bf274c9 at posting.google.com>...
> Can't go wrong w/ Programming Python 2ndEd by Lutz from O'Reilly.  Here's 
> a sample chapter.  If you can only afford a single book, this mite be it....Jet
> 
> http://www.oreilly.com/catalog/python2/chapter/ch15.html

Agree with that totally - there is not a bad set of tutorials at 
http://www.devshed.com/Server_Side/Python
that gets you to about the level the books starts at. Python 101 parts
1-8 and OO programming 1 and 2


From martin at v.loewis.de  Wed Sep 18 17:23:32 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 18 Sep 2002 23:23:32 +0200
Subject: debugging "Too many open files"
References: <mailman.1032378623.20448.python-list@python.org>
Message-ID: <m3y99zj8kr.fsf@mira.informatik.hu-berlin.de>

Alex Polite <m2 at plusseven.com> writes:

> I'm getting the "Too many open files" error.

What "the" error? An exception? A message in the system log? An SMS on
your mobile phone?

> I've reset /proc/sys/fs/file-max to 100000 and the count in
> /proc/sys/fs/file-nr is nowhere close.
> 
> I'm not sure that python is the root of the problem. Is there anyway
> to see which processes own the active file descriptors?

I recommend you list /proc/<pid of python program>/fd/

Regards,
Martin


From shalehperry at attbi.com  Tue Sep 10 16:08:53 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 10 Sep 2002 13:08:53 -0700
Subject: "str"(expr)"str" interpolation syntax
In-Reply-To: <Pine.GSO.4.33.0209102221070.10002-100000@techst02.technion.ac.il>
References: <Pine.GSO.4.33.0209102221070.10002-100000@techst02.technion.ac.il>
Message-ID: <200209101308.53220.shalehperry@attbi.com>

On Tuesday 10 September 2002 12:35, Beni Cherniavksy wrote:
> Hi all.
>
> I was searching for a new interpolation syntax that would use quotes
> for starting/ending the interpolated part (I don't feel comfortable
> putting code inside unclosed quotes...)
>
> After contemplating some horrors like $"""str"expr"str""" or
> $("str" expr "str"), I noticed that currently both  "(  and  )"  are
> illegal (first because strings are not callable, second is a syntax
> error).  So I propose that a parenthesized expresion near a string would
> imply concatenation.  In other words,
> 	"foo"(x * 3)"bar"
> would be equivallent to
> 	"foo" + str(x * 3) + "bar".
> This works of course with only one side, e.g. "foo"(x) or (y)"bar" but in
> longer cases it would be good style to add empty strings on either end,
> e.g. ""(y)"bar".  And of course it all works with single quotes too.  One
> of the benefits over current interpolation PEPs is that you can change the
> quote type in the middle...
>
> What do you think?

having not read the PEPs, why is this needed?

s = "foo%dbar" % (x*3,) # this works

and as you showed above so does str(x*3).



From alet at unice.fr  Wed Sep 18 06:56:38 2002
From: alet at unice.fr (Jerome Alet)
Date: Wed, 18 Sep 2002 12:56:38 +0200
Subject: [ANNOUNCE] DebSync v0.1
Message-ID: <3D885BE6.2ADF0055@unice.fr>

Hi,

I'm pleased to announce DebSync v0.1

DebSync is a Python command line tool which helps to synchronize a bunch
of Debian GNU/Linux machines wrt the list of packages installed.

DebSync retrieves the list of installed packages from a master host, and
then install or remove packages on any number of hosts to make them have
the very same installed packages list as the master host.

Of course DebSync doesn't deal with configuring these packages, this
must
be done as usual.

DebSync can work over ssh (default) or rsh.

DebSync is different from fai (Full Automated Installation) in that, to
my knowledge, fai only deals with machines you plan to install, while
DebSync
deals with existing and certainely differently installed Debian
GNU/Linux machines.

Debsync is distributed under the GNU GPL of the Free Software
Foundation,
and is available from :

	http://www.librelogiciel.com/software/

Remark : it's only a 0.1 so don't expect miracles, yet, but it seems to
work reasonably well on the three hosts I've tried. Also be sure to have
the same version of Debian installed and do an apt-get update on each
host before launching DebSync.

Any comment, contribution, or bug report is more than welcome !

Thanks for reading.

Jerome Alet


From lists at gregfortune.com  Wed Sep 18 18:07:17 2002
From: lists at gregfortune.com (Greg Fortune)
Date: Wed, 18 Sep 2002 15:07:17 -0700
Subject: Catching results of an exec
Message-ID: <3d88f701$0$304$39cecf19@nnrp1.twtelecom.net>

Is there any way to "catch" the value produced by an exec?  I'm considering 
things like print statements that don't really assign a value, but do send 
a result to stdout.

For example, if I execute

exec('print dir()')

from an interpreter, there will be some stuff spit out on the console.  
What if I really wanted to display the list from dir() in a different way?  
I'm working on a really dinky server that will sit on a machine and execute 
arbitrary python code.  I want the results from the code execution to be 
returned to client.

Even just eval('dir()') exhibits the desired behavior, but only from an 
interactive interpreter.

Right now I've got a nasty hack that redirects stdout, uses an 
InteractiveInterpereter from the code module to execute the code, writes to 
an output file, reads the result from the file, restores stdout and then 
sends the result to the client.  Hopefully, that's a terrible way to do it. 
 It just occured to me that I can probably use a pipe in place of the 
output file, but still....

It would be great if exec and eval set something somewhere that held the 
results of the code execution.  I looked in the InteractiveInterpreter 
(which, at least, kicks the results of the code execution out to stdout), 
but all it does is

exec code in self.locals

Did I miss something?

Greg Fortune


From op73418 at mail.telepac.pt  Tue Sep 10 09:14:50 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Tue, 10 Sep 2002 14:14:50 +0100
Subject: Graph algorithms - minimisation
References: <cf0ad9fb.0209092301.44288bc7@posting.google.com> <tYif9.3454$cP3.6958@news.iol.ie>
Message-ID: <8vrrnuckib7gbuodhrmbs2f88nq8r5r5vu@4ax.com>

On Tue, 10 Sep 2002 09:30:01 GMT, Padraig Brady <Padraig at Linux.ie>
wrote:

>Jesper Olsen wrote:
>> Do you know of python extentions that implement efficient graph
>> algorithms?
>> 
>> In particular I am interested in minimising the number of nodes & arcs
>> used for representing directed graphs.
>> 
>> If you know of implementations in other languages than python - e.g.
>> C/C++, I am interested as well.
>> 
>> Cheers
>> Jesper
>
>Guido wrote an essay on graphs @
>http://www.python.org/doc/essays/graphs.html
>In there is a method to return the shortest path
>which may be useful?

I remember seeing a recipe in ActiveState cookbook with Dijkstra's
algorithm for finding shortest paths. You can then adapt it to your own
representation of a graph. 

>
>P?draig.

Best regards,
Gon?alo Rodrigues


From claird at starbase.neosoft.com  Tue Sep 17 12:52:09 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 17 Sep 2002 11:52:09 -0500
Subject: can wxpython do this?
References: <737aefee.0209151702.3f735252@posting.google.com> <spcaou804m5oo6gah31oqaeuqsuj60h0ip@4ax.com> <am3d0b$jj$1@news01bc.so-net.ne.jp> <am7hrt$9b3$1@news01bb.so-net.ne.jp>
Message-ID: <20929D5514805227.48A29064FC554044.5053D887939EB774@lp.airnews.net>

In article <am7hrt$9b3$1 at news01bb.so-net.ne.jp>,
Erick <erick_papadakis at REMOVEFORSPAM.yahoo.com> wrote:
>sorry for restating, but is there a way for me to connect to the Internet,
>make a call to a webserver as a browser would do, retrieve data and display
>it inside my application?
			.
			.
			.
There are too many ways to do this.

You might start with
  from urllib import urlopen
  
  URL = "http://www.python.org"
  page = urlopen(URL).read()    
  print page
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From codeapocalypse at msn.com  Wed Sep 18 13:23:43 2002
From: codeapocalypse at msn.com (Brian)
Date: 18 Sep 2002 10:23:43 -0700
Subject: Win32COM & Detaching from automation servers
References: <ef3269df.0209171441.2d6f8e59@posting.google.com>
Message-ID: <ef3269df.0209180923.2a8f3eba@posting.google.com>

codeapocalypse at msn.com (Brian) wrote in message news:<ef3269df.0209171441.2d6f8e59 at posting.google.com>...
> Perhaps the solution is obvious and I'm just
> not seeing the forest for the trees...)

This appears to be the case - maybe too much recycled office air
yesterday for me?  I ran my code later in the evening yesterday and it
is detached without me needing to do anything.  Maybe too much stress
was causing me to hallucinate in the morning yesterday.  :)


From phr-n2002b at NOSPAMnightsong.com  Tue Sep  3 18:21:38 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 03 Sep 2002 15:21:38 -0700
Subject: https in Python?
References: <ac259413.0209031134.3778c6ac@posting.google.com>
Message-ID: <7x8z2ipvd9.fsf@ruckus.brouhaha.com>

s_gherman at yahoo.com (Sorin Gherman) writes:
> Is it possible to handle https using Python's libraries? If so, what's
> involved with using Python to read my hotmail messages, for instance?

The urllib module can handle https but you're going to need a bunch of
other hair to read hotmail messages too.  Hotmail uses a lot of
javascript to control the browser and you're going to have to figure
out what it's doing to get at the relevant info in it.



From DennisR at dair.com  Tue Sep 17 14:04:24 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Tue, 17 Sep 2002 18:04:24 GMT
Subject: using python as cgi on apache unix
References: <am6fm301477@enews2.newsguy.com> <orHh9.167$UJ1.17889587@newssvr14.news.prodigy.com> <am7nv0075d@enews2.newsguy.com>
Message-ID: <I8Kh9.215$gp5.14020635@newssvr13.news.prodigy.com>

> I do get the correct results [from shell]
> I must
> have Apache configured in such a way on the one machine that it is
confused.
Sounds right to me.  The command line test was to see if the problem was
unix related.  It does not sound like it.  Note that on a virtual host, "/"
can be a different directory at command line vs. in a cgi file.

--

Dennis Reinhardt

http://www.dair.com




From aleax at aleax.it  Tue Sep 17 17:50:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 17 Sep 2002 23:50:24 +0200
Subject: Will "Python in a Nutshell" be too short?
In-Reply-To: <5.1.1.6.0.20020917181205.035d0a58@spey.st-andrews.ac.uk>
References: <5.1.1.6.0.20020917181205.035d0a58@spey.st-andrews.ac.uk>
Message-ID: <02091723502402.21243@arthur>

On Tuesday 17 September 2002 20:46, Hamish Lawson wrote:
> O'Reilly doesn't yet list the forthcoming "Python in a Nutshell" by Alex

It's being final-edited right in these days.

> Martelli, but according to Amazon.com's page for the book, it will have
> 400 pages. Is that indeed the plan? I know that O'Reilly aims to keep fat
> out of its books, and that Alex is striving to be comprehensive yet
> concise, but I worry that the book's proposed length will simply be too
> short to cover the material I'm hoping will be included.

That was my original dream.  I think that, to some extent, the usefulness
of a book can be inversely proportional to its bulk.  However, the need to
cover a wide variety of subjects, AND to provide tutorial info and examples
as well as reference info, bids fair to put paid to my dreams.  I fear the
book will be closer to 600 pages than to 400:-(.

> I regularly consult David Beazley's "Python Essential Reference", but the
> usefulness to me of this otherwise excellent book is compromised by the
> number of (platform-independent and not obsolescent) standard modules and
> packages that had to be left out of the book's 380 pages (htmllib,
> telnetlib, urllib2, CongigParser, unittest, tkinter, xml, etc.).

Out of those you list, I don't cover CongigParser (non-existing) nor even
ConfigParser (existing, but I had to weigh its importance vs a zillion
others).  The number of pages I had to devote, in particular, to tkinter and
xml, still boggles my mind -- yet, I do see that they add to the book's
usefulness more than they take away from it.


> Many of us were hoping that "Python in a Nutshell" would not only be a
> comprehensive reference of pretty much all of the modules in the standard
> library that aren't obsolescent or platform-specific, but would also
> cover some of the most popular third-party libraries - e.g. DB-API, PIL,
> mxDateTime, Numeric, win32all, wxPython, mod_python, ReportLab.

I cover some of these, but not nearly all -- DBAPI yes, PIL no, mxDateTime
yes, Numeric yes (and then some!), win32all no (_no_ platform-specific 
stuff!), and so on.  And after mature deliberation I decided I just *had* to
cover many you don't mention, such as twisted.internet, Cheetah, Jython's
interfacing to Java libraries, etc.


> My congratulations to Alex and O'Reilly if they reckon they can somehow
> manage to fit this all into just 400 pages; but if this won't be
> possible, as I worry must be the case, then may I make a plea for the
> book's length to be reconsidered. It would be good to have a one-stop
> Python reference book that would be a suitable companion to the
> task-oriented "Python Cookbook" - I'm hoping "Python in a Nutshell" can
> still be that book.

I think you'll be happier with the book size than I'll be -- even though
not fully satisfied of course (1000 pages would apparently not suffice
for your desires, definitely not when including examples and discursive/
tutorial coverage of at least some parts).  _My_ personal dream is more
of "SQL in a Nutshell"... the most value-packed 300 pages I've ever seen!
But I won't reach it in "Python in a (cocoa)Nut-shell" - that's for sure!-)


> By comparison, O'Reilly's "Perl in a Nutshell" is 800 pages and has the

I used it throughout as my bete noire:-).  Py iaN won't be THAT large --
not QUITE that large -- though it will maybe end up closer to it than
to my dream 400 pages.


Alex



From brian at sweetapp.com  Tue Sep 24 14:38:08 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Tue, 24 Sep 2002 11:38:08 -0700
Subject: ANN: Pyana 0.6.0
Message-ID: <001c01c263f9$879487e0$df7e4e18@brianspiv1700>

Pyana 0.6.0 has been released. Source and binary distributions are
available at:
http://sourceforge.net/project/showfiles.php?group_id=28142

Pyana is an extension module that allows Python scripts to access the
Apache Group's Xalan XSLT transformation engine.

For usage examples and other information, see:
http://pyana.sourceforge.net/

What's new in this release?

- Updated to use Xalan 1.4/Xerces 2.1
- A default ErrorHandler and ProblemListener are automatically installed
  when a Transformer instance is created. The global variables
  defaultErrorHandlerFactory and defaultProblemListenerFactory determine
  which classes are used.
- Messages generated with xsl:message now appear correctly
- Exception decoding is now done using a preprocessor macro instead of
an
  include file. This should fix a build problem on some platforms e.g.
  HP-UX.
- Better exception decoding of XSLExceptions

Cheers,
Brian  




From nhaggin at attbi.com  Tue Sep 17 22:42:41 2002
From: nhaggin at attbi.com (Nicholas Haggin)
Date: 17 Sep 2002 19:42:41 -0700
Subject: Yet another Tkinter and threads question
Message-ID: <4936aa67.0209171842.12cacf2a@posting.google.com>

After a short perusal of the group archives I haven't found anything
answering the following, so here goes:

I am polishing up the GUI for a client that uses FTP to submit MVS
jobs to an IBM S/390, and am adding message boxes for exceptional
circumstances i.e., user mistyped password, connection timeout, etc.
To prevent locking up the GUI, all submissions spawn a thread, which
calls a wrapper function around ftplib that does the right
site-specific hanky-panky to make everything go smoothly. In the
thread function, the code goes something like this:

try:
    submit_job()
except socket.error:
    showerror("Socket has barfed")
except ftplib.error_perm:
    showerror("You mistyped your password")

The call to showerror locks the program. Any thoughts? Older posts on
other threading+GUI questions had mentioned that Tk was not
thread-safe; has it become so, or is this still a tar-pit?

Nick

A.M.D.G.


From henrik.motakef at web.de  Mon Sep 23 14:20:22 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 23 Sep 2002 20:20:22 +0200
Subject: Session handling
References: <c5b3549a.0209230511.5bbe6849@posting.google.com> <amn7dp$6cn$1@mamenchi.zrz.TU-Berlin.DE> <IWHj9.264$841.144712@newssrv26.news.prodigy.com>
Message-ID: <871y7k1sbd.fsf@pokey.henrik-motakef.de>

"dsavitsk" <dsavitsk at e-coli.net> writes:

> If the only state that needs to be maintained is a language variable, this
> seems like a lot of work when a cookie will work fine.

Not to mention that HTTP clients are supposed to indicate the user's
preferred language in the Accept-Language header.


From aleax at aleax.it  Wed Sep 25 05:14:14 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 09:14:14 GMT
Subject: Removing all occurences of a character from a string.
References: <amr8nh$rd8$1@driftwood.ccs.carleton.ca> <DL9k9.201596$216.7001230@bin4.nnrp.aus1.giganews.com> <amrav1$4rk$1@driftwood.ccs.carleton.ca> <mailman.1032928353.13514.python-list@python.org>
Message-ID: <G7fk9.163231$pX1.5840844@news2.tin.it>

Mike C. Fletcher wrote:

> Under later Python versions (such as 2.2.1), you can do the following:
> 
> attr = [[item.replace("'","") for item in element] for element in attr]
> 
> Which is rather elegant IMO.  It also has the side-effect of _not_
> rewriting the data in-place in your original list, instead just
> replacing that list when we are finished creating the new list, which
> may or may not be desirable for you.

If it's NOT desirable, just assign to attr[:] tather than to attr.


Alex



From aleax at aleax.it  Fri Sep 27 03:09:37 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 27 Sep 2002 07:09:37 GMT
Subject: comparing all values of a list to regex
References: <mailman.1033083698.8166.python-list@python.org>
Message-ID: <RuTk9.149029$ub2.3288430@news1.tin.it>

Delaney, Timothy wrote:
        ...
>> > elif matches != len(alist):
>> >     match_list = match_some
>> > 
>> > match_list.append(alist)
>> 
>> Yes, choosing a list first and then appending to the chosen
>> list is nicer, but you cannot do that within the original
>> specs (for all I know the match_some list might grow
>> unbearably large over time, in this case).
> 
> Hmm - why not? Ignoring that the original specs didn't actually *have*
> that case, 

Not ignoring it, one would think the specs must be to do nothing
special in that case.

If your program was specified to "append to list1 if the time is
exactly 11:22:33, to list2 if the time is exactly 17:26:35", would
you consider it reasonable to make your program appent to list3
if it's any other time of day?

The problem could be that the program is called a billion times,
of which, say, about seven fall in the first case, about three in
the second one, and all the rest in the "neither" category.  If you
extend the program's spec to append to another list when "neither"
applies, your program will overflow memory and crash, while any
proper implementation of the specs would work correctly.

> choosing the list to append to, or doing it in-place is merely
> an implementation detail. 

Except that, to meet the specs, you must NOT append to match_some,
ever.  It's quite possible of course that the specs where incorrect
or incomplete, but just assuming that seems strange to me.

> Of course, you would need instead to do something like:
> 
> match_list = None
> 
> if not matches:
>     match_list = match_none
> elif matches == 1:
>     match_list = match_one
> elif matches == len(alist):
>     match_list = match_all
> 
> if match_list is not None:
>     match_list.append(alist)

Right (I did much the same in my dictionary-version of this
snippet) although of course the need for the final if is not
most elegant.

A good design pattern for such cases is Null Object, which
Dinu Gherman showed in recipe:
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/68205

To quote him,
"""
Null objects are intended to provide first-class citizens as
a replacement for the primitive value None. Using them you
can avoid conditional statements in your code and express
algorithms with less checking for special values.
"""

With this DP, your snippet would become:

import null
match_list = null.Null

if not matches:
    match_list = match_none
elif matches == 1:
    match_list = match_one
elif matches == len(alist):
    match_list = match_all

match_list.append(alist)


The need for the guard goes away because null.Null is
basically designed so you can call any method on it
innocuously.  Of course, you get more mileage out of
it in more complicated situations.

{0:match_none, 1:match_one, len(alist):match_all}.get(
    matches, null.Null).append(alist)

gets greater benefit wrt the None-plus-guard.  Of course,
this more-concise snippet has a problem when len(alist)
is 0 -- your snippet, like my original one, is not
ambiguous in this case (appends to match_none), while a
dictionary-display with two identical keys has behavior
that is less obvious (I don't recall whether the Python
language specifies which of the two repetitions of the
same key "takes", but in any case most readers and
maintainers of this code would be puzzled).  Fortunately,
the OP clarified (if I read his followup post correctly)
than len(alist)==0 is a "can't happen" in his case.


Alex



From walto at vaag.cx  Thu Sep 26 08:09:34 2002
From: walto at vaag.cx (walter hoolwerf)
Date: 26 Sep 2002 05:09:34 -0700
Subject: html in xml problem
Message-ID: <9a1d8b0f.0209260409.3b759cba@posting.google.com>

Hello,

I need to place some HTML code in an XML file en process it with
python.

The output is a bit weird. I've also used this way to place html code
in XML with flash, worked fine.

as you can see, I already tried replacing the newlines with nothing
(or space), didn't work. The output should be <a> instead of:
<
a
>
My browser wont accept the last.

Who knows? Thanx in advance, 
Walto

--------

xml file:
<?xml version="1.0" encoding="iso-8859-1"?>
<HAN>
   <info>&lt;B&gt;Welcome&lt;/B&gt;</info>
</HAN>

code:
#!/usr/bin/python2
from xml.dom import minidom, Node
import string

doc = minidom.parseString(open('han.xml').read())
for c in doc.documentElement.childNodes:
        if c.nodeType == Node.ELEMENT_NODE:
                for d in c.childNodes:
                        print d.nodeValue.replace("\n", "
").replace("\r", " ")

output:
<
B
>
Welcome
<
/B
>


From pedro_rodriguez at club-internet.fr  Thu Sep 26 14:36:40 2002
From: pedro_rodriguez at club-internet.fr (Pedro RODRIGUEZ)
Date: Thu, 26 Sep 2002 20:36:40 +0200
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <amltd3$3k$1@glue.ucr.edu> <3D911FE1.1050801@something.invalid> <amvcev$cf6$1@s1.read.news.oleane.net> <7nHk9.172208$pX1.6117207@news2.tin.it>
Message-ID: <amvjr0$f1v$1@s1.read.news.oleane.net>

On Thu, 26 Sep 2002 19:22:11 +0200, Alex Martelli wrote:

> Pedro RODRIGUEZ wrote:
>         ...
>>>> But why use grotesque hacks in the first place?  "a ? b : c" is a
>         ...
>> Ada has two short-circuits operators : "and then" and "or else" that do
>> the trick :
>>     if A and then B then
>>         ...
>>     end if;
>> 
>> Looks cleaner to me.
> 
> Ada's "and then" is like Python's "and"; Ada;s "or else" is like
> Python's "or".  Neither is like C's "a?b:c" (ternary operator), though
> with some tricks boolean shortcircuiting operators can be used to
> simulate ternary operators (a bit cumbersome in the general case).
> 
> The reason
>         zz = a and b or c
> is not the same thing as
>         zz = a ? b : c
> is that, when a is true and b is false, the former returns c, while the
> latter returns b.
> 

I stand corrected the value of b should not be part of the condition but
only of the result.


> The tricks deal with ways to work around this problem, by ensuring that
> and's RHO can never be false, e.g.
>         zz = (a and [b] or [c])[0]
> 

If python had short-circuiting operators ('and then', 'or else'), this
will be the ternary operator :
    zz = (a and then [b] or else [c])[0]

The problem when talking about conditionnal expression should be to know
what really matters in the context :
- the short-circuit evaluation (only (a,b) or (a,c) are evaluated)
- or the ability to pick a value in a list

The first one isn't possible in a Python expression AFAIK, 
and the second one can be expressed as (b,c)[not a].


Pedro


From Andreas.Leitgeb at siemens.at  Fri Sep 20 17:18:19 2002
From: Andreas.Leitgeb at siemens.at (Andreas Leitgeb)
Date: Fri, 20 Sep 2002 21:18:19 +0000 (UTC)
Subject: lists in dictonaries ?
References: <mailman.1032426629.1933.python-list@python.org> <slrnaojc79.c05.Andreas.Leitgeb@pc7499.gud.siemens.at> <d111eafa.0209191646.6429dcb9@posting.google.com>
Message-ID: <slrnaon52k.c05.Andreas.Leitgeb@pc7499.gud.siemens.at>

xtian <xtian at toysinabag.com> wrote:
>> > You can put any python objects into a python dictionary, 
>> > even the dictionary itself...
>> Oh, can I create a dictionary in Python that contains all those
>> dictionaries that don't contain themselves ... ?-)
>> *scnr*'ly yours.
> Would it contain itself?

if dict in dict: print "yeah"

*scnr*'ly yours.


From martin at v.loewis.de  Wed Sep  4 03:17:58 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 04 Sep 2002 09:17:58 +0200
Subject: MD4 and DES for unicode strings (samba passwd file manipulation)
References: <uoit1mxubi.fsf@bashful.cdf.toronto.edu>
Message-ID: <m37ki2te8p.fsf@mira.informatik.hu-berlin.de>

Arcady Genkin <agenkin at cdf.toronto.edu> writes:

> Perhaps the problem lies with Unicode encoding?  Samba's man page (see
> the link below) says that the string is obtained by taking MD4 hash on
> the password, represented as a string of 16-bit little-endian unicode
> characters.

There is a contradiction in this statement: You can't really MD4-hash
"Unicode", as Unicode characters are values in range(0,2**21) - to
MD4-hash, you need a byte string. However, the statement means that
they use the UTF-16 encoding in its little-endian variant, which, in
Python, is implemented in the utf-16le codec:

>>> m = MD4.new()
>>> m.update( u'111222333'.encode("utf-16le") )
>>> print m.hexdigest().upper()
B66B458E27BD714F99A05EE3C479FBF1

m.update expects a byte string. If you pass a Unicode string, that
gets converted with the system default encoding, which is "ascii", and
in this case not appropriate.

HTH,
Martin


From member at dbforums.com  Mon Sep 16 22:09:17 2002
From: member at dbforums.com (hpyhpy)
Date: Tue, 17 Sep 2002 02:09:17 +0000
Subject: string type in SWIG.
Message-ID: <1822835.1032228557@dbforums.com>

I define a class,it has string type member,and include ,how SWIG wrap
it?Such as:
in example.cpp
////////////////////////////
#include
using namespace std;
class a
{
private:
string aaa;
};
////////////////////////////
How to do example.i?

--
Posted via http://dbforums.com



From kennethg at pd.jaring.my  Fri Sep 27 23:44:29 2002
From: kennethg at pd.jaring.my (Kenneth Gomez)
Date: Sat, 28 Sep 2002 03:44:29 GMT
Subject: Local Python Server Installation
References: <3d8b47c6.277669@news.jaring.my> <slrnaomn6m.39s.gerhard.haering@haering.opus-gmbh.net> <3d8c993c.2190426@news.jaring.my> <mailman.1032625889.13783.python-list@python.org> <3d8d4009.375344@news.jaring.my> <bQYk9.1630$6g7.1090@fe03> <3d947c26.2656654@news.jaring.my> <JR3l9.2341$79.998@fe03>
Message-ID: <3d952575.1651269@news.jaring.my>

Thanks Steve. At least now I know that my code is OK. I'll go look
into the APache config Bible.

Thanks again.
Kindest regards,
Kenneth.
Cheers,
Kenneth Gomez.


From gerhard.haering at gmx.de  Sat Sep 14 16:25:11 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sat, 14 Sep 2002 22:25:11 +0200
Subject: newbie-question: abstract classes
In-Reply-To: <3D839852.2060405@snafu.de>
References: <3D839852.2060405@snafu.de>
Message-ID: <20020914202510.GA28307@lilith.ghaering.test>

* T. Kaufmann <merman at snafu.de> [2002-09-14 22:13 +0200]:
> Hi everybody,
> 
> I have a problem:
> 
> I want to create an abstract class but it doesn't work.
> What' s wrong here?
> 
> class Abstract:
> 
>     def __init__(self): return

Better style is to use "pass" instead of "return" here.

>     def method(self):   raise NotImplementedError
> 
> class Concrete: [...]

Use "class Concrete(Abstract):" instead.

HTH,

-- Gerhard



From rob at diespammerdieuselesspython.com  Wed Sep 18 16:38:39 2002
From: rob at diespammerdieuselesspython.com (Rob Andrews)
Date: Wed, 18 Sep 2002 20:38:39 GMT
Subject: linux python ideas
Message-ID: <3D88E578.6010001@diespammerdieuselesspython.com>

I'll be giving a presentation on Python to my local LUG 
(http://lugoj.org) in a few weeks, and would like to point out at least 
a few things that would be of particular interest to linux users.

The idea is not "how to program in Python", which would take more than 
one quick presentation, but a one-off demonstration of ways in which 
Python is particularly useful for such a group.

Any suggestions would be appreciated, since my notes are pretty generic 
so far.

regards,
Rob Andrews



From jepler at unpythonic.net  Wed Sep  4 13:57:47 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Wed, 4 Sep 2002 12:57:47 -0500
Subject: Program for ... HELP!!!
In-Reply-To: <knqd9.25$%z1.943@twister1.libero.it>
References: <knqd9.25$%z1.943@twister1.libero.it>
Message-ID: <20020904175739.GE2111@unpythonic.net>

kyro,
One method is to perform LU decomposition of the matrix, then take
det(U) which is simply the product of the elements on the diagonals.
(det(A) = det(L)det(U) since A=LU, but det(L)=1 and U is triangular).
You can find a description of LU decomposition in any decent numerical
analysis text, or online.  You could also try the search term "gaussian
elimination".

LU decomposition is probably expressible most clearly as iteration, not
recursion.  There is a method whose name escapes me which involves
taking the sum of factors like
    (-1)^(j+k) * A_jk * det(A with the row and column passing through jk removed)
which is naturally recursive, but this is not a "good" algorithm: it'll
take much longer than the O(n^3) operations of LU decomposition.

I don't know if there are better methods for finding determinants,
the LU method is the only one explored at length in the text I have on
hand.

If you're having trouble representing matricies in Python, well, that's
another matter.  As another poster suggested, you should probably get
the "Numeric" package.

Jeff



From jb at cascade-sys.com  Wed Sep 25 15:25:52 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Wed, 25 Sep 2002 12:25:52 -0700
Subject: Getting started
References: <3D911306.7080609@ucsd.edu>
Message-ID: <3D920DC0.8040808@cascade-sys.com>

Adam Taylor wrote:

> This last is available at:
> http://research.microsoft.com/Users/luca/Papers/TypeSystems.pdf 

I believe your interpretation is accurate for the most part.

> Cardelli doesn't use the term "weakly typed" himself, but on page 5 he
> seems to be say that "weakly typed" is the commonly used term for what
> he calls "weakly checked".  C is weakly checked, in his terminology, and
> that's why it's unsafe. 


My read of page 5 is he equates his internal use of "weakly checked" 
with "weakly typed" and further suggest that the latter term is commonly 
used "in the literature," which it does seem to be.

> But people on this thread seem to be using "weakly typed" to mean
> roughly the same thing as Cardelli's "untyped".  Which is fine, and
> pretty close to Cardelli's terminology.  But clearly the terms "untyped"
> (=="weakly typed" on this thread) and "weakly checked" (=="weakly typed"
> in common usage) are very different things. 


Well, some people here use the term casually and incorrectly, as a 
measure of quality of a language's object model, which I am sure you'll 
agree is completely inconsistent with most uses of the term, including 
Cardelli's.

Interestingly, a google search on the term revealed that about 25% of 
the meaningful hits equated "weakly typed" and "dynamically typed" (the 
latter of which is what Cardelli calls "untyped").  The other 75% 
generally characterized strong vs. weak type systems in terms of "type 
safety", i.e., a mechanism for preventing  "type errors," regardless of 
whether the checking is done statically or dynamically.  [The list of 
references appears in a prior post to this thread.]

I think your read of Cardelli overall is correct -- that Python, like 
Lisp, would be classified as "untyped" because it has no static type 
system.  A literal reading furthermore defines "weakly typed" only in 
terms of static checking.

However, I submit that his use is inconsistent with the more common use 
of the term.  Too, I think this is actually a minor oversight in 
Cardelli's paper.  His focus overall was static systems and I think he 
simply didn't give enough consideration to dynamically typed languages. 
 E.g., he says "untyped" languages such as Lisp can be "safe" but he 
defines "weakly typed" only in terms of languages that do static type 
checking.  I would argue that in the first place he misclasifies Lisp as 
"untyped".  I suspect he took the naive view of Lisp, in which 
everything is a list, when in fact a rather wide variety of "objects" 
actually exist in a typical, mature Lisp system.  Of course, Python and 
Smalltalk are better examples, in that they both have robust object type 
systems and are in principle very secure from a type checking sense. 
 (Of course, there are probably things you can do with the compiler and 
debug class objects that violate type safety, making them both unsafe in 
general.)

The primary distinction between safe and unsafe, strong and weak type 
systems revolves around the notion of whether or not "type errors" can 
be prevented by the type system.  The distinction between typed and 
untyped is less interesting, except to use "untyped" as a way to 
eliminate a class of languages from further discussion.  Within typed, 
there should be two sub-classes, based on whether type checking is 
primarily static or dynamic.  This is more interesting and useful, and 
is consistent with "the literature" as far as I can determine.  

I furthermore have a conjecture that much of Cardelli's work on 
"judgments" would apply equally well to static and to dynamic systems. 
 The important criterion is that the type system be secure and that the 
total collection of function x arguments definitions be unambiguous (at 
least during the sequence of computation of interest).  Whether the 
checks take place at compile time or at runtime I don't think materially 
affects his theories.  But I could be overlooking something in this regard.

In any case, my own use of the term "weakly typed" and associated terms 
(strongly typed, type safety, etc.) in this thread have been consistent 
with "the literature" (though slightly inconsistent with Cardelli), in 
that I class Python to be "strongly typed" using "dynamic checking".   
That seems to be the consensus here.  Only confusion has been about 
"weakly typed," and hopefully we're close to exhausting that topic..

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From edream at tds.net  Thu Sep 12 16:24:26 2002
From: edream at tds.net (Edward K. Ream)
Date: Thu, 12 Sep 2002 20:24:26 GMT
Subject: Cursed newlines and readline()
References: <6u5g9.11312$Da3.1133980@kent.svc.tds.net> <mailman.1031860578.17383.python-list@python.org>
Message-ID: <_J6g9.11$rJ3.11804@kent.svc.tds.net>

Oh joy.  Thanks very much for this reply.

Edward




From shalehperry at attbi.com  Fri Sep  6 16:58:36 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 6 Sep 2002 13:58:36 -0700
Subject: Beginners Question - use of variables
In-Reply-To: <1031345108.547.165.camel@martin.kleinerdrache.org>
References: <1031345108.547.165.camel@martin.kleinerdrache.org>
Message-ID: <200209061358.36506.shalehperry@attbi.com>

On Friday 06 September 2002 13:45, Martin Klaffenboeck wrote:
> Hello,
>
> Are newbies questions ok here?
>
> I've got a struktur like this:
>
> string string int timestamp
> string string int timestamp
> string string int timestamp
> ...
>
> in Perl I used to save such a thing in a multidimensional array like
>
> array(
>       array(string, string, int, timestamp),
>       array(string, string, int, timestamp),
>       array(string, string, int, timestamp))
>
> So how do I save that in python and how do I get fast to one of this
> strings or int or timestamp... ?
>
> Martin

my_list = [
           ['sean', 'perry', 25, '20020906'],
           ['Martin', 'Klaffenboeck', 20, '20020826'],
          ]

for entry in my_list:
    print entry[0], entry[2] # prints first string and the int

tutor at python.org is a list especially for those new to the language.



From pinard at iro.umontreal.ca  Sun Sep  1 17:16:46 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Sun, 01 Sep 2002 17:16:46 -0400
Subject: Making regex suck less
In-Reply-To: <874rd9phtd.fsf@pokey.henrik-motakef.de> (Henrik Motakef's
 message of "01 Sep 2002 22:37:34 +0200")
References: <3d725881.345921@news.t-online.de>
	<874rd9phtd.fsf@pokey.henrik-motakef.de>
Message-ID: <oqadn19zr5.fsf@titan.progiciels-bpi.ca>

[Henrik Motakef]

> There is an Emacs Lisp package called "symbolic regexps" (or sregexp.el)
> that lets you write regular expressions in standard lisp syntax, like
> (sregexq bol (or "abc" "def")) instead of "^(abc|def)" (bol meaning
> "beginning of line"). I don't see how that maps elegantly to Python syntax,
> however.

PLEX has a Python way to describe regular expressions.  It is likely available
stand-alone.  I found the copy I use within the Pyrex distribution, see:

   http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



From glingl at aon.at  Sat Sep 21 14:52:01 2002
From: glingl at aon.at (Gregor Lingl)
Date: Sat, 21 Sep 2002 20:52:01 +0200
Subject: Python 2.2.1 compatible with Tcl/Tk 8.4?
References: <yu9965x05vkr.fsf@europa.research.att.com> <j4ptv89zwk.fsf@informatik.hu-berlin.de>
Message-ID: <3D8CBFD1.4010808@aon.at>

Martin v. L?wis schrieb:
> Andrew Koenig <ark at research.att.com> writes:
> 
> 
>>Now that Tcl/Tk 8.4 is out, can anyone tell me whether Python 2.2.1
>>will work with it?  
> 
> 
> It appears to work fine here, and I can't see a reason why it would be
> incompatible. Notice that the setup.py of 2.2.1 is already prepared to
> search for Tcl/Tk 8.4 on its own.
> 
> Regards,
> Martin

1. When working with Python 2.2.1, are there any advantages of
using Tcl/Tk 8.4  over the version which comes with Python2.2.1 for Window?
2. Which steps would I have to perform, to change to the new version?

Gregor Lingl



From whisper at oz.net  Sat Sep 28 21:34:09 2002
From: whisper at oz.net (David LeBlanc)
Date: Sat, 28 Sep 2002 18:34:09 -0700
Subject: Best way to create a class dynamically on the fly?
In-Reply-To: <3D96511D.2040407@4-am.com>
Message-ID: <GCEDKONBLEFPPADDJCOEKEAJFEAA.whisper@oz.net>

I understood the question as 'what do you stuff into a variable to make it
behave like a class at runtime', not 'when is a class declaration is used to
create a class instance.'

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Evan
> Sent: Saturday, September 28, 2002 18:02
> To: python-list at python.org
> Subject: Re: Best way to create a class dynamically on the fly?
>
>
> Robert Oschler wrote:
> > What is the best way to create a Python class dynamically at runtime?
>
> ALL Python classes are created dynamically at runtime.  The following is
> not a declaration in the sense that you're accustomed to from C/C++, it
> is normal code executed when the containing file is run (or imported):
>
<snip>

> Cheers,
>
> Evan @ 4-am




From bobx at linuxmail.org  Fri Sep 27 18:05:54 2002
From: bobx at linuxmail.org (Bob X)
Date: Fri, 27 Sep 2002 22:05:54 GMT
Subject: Is Boost Python version 2 ready?
References: <MvZk9.12124$Lg2.3584877@news2.news.adelphia.net> <CG2l9.263192$216.10040127@bin4.nnrp.aus1.giganews.com>
Message-ID: <6D4l9.12184$Lg2.3654433@news2.news.adelphia.net>

"Terry Reedy" <tjreedy at udel.edu> wrote in message
news:CG2l9.263192$216.10040127 at bin4.nnrp.aus1.giganews.com...
>
<snip>
> Not to exactly answer... from post on PyDev list, I believe they want
> to test it with Python 2.2.2 (due in about 10 days?) before calling it
> finished.
>
> TJR
Is there a page up saying what 2.2.2 will give us?

Bob




From jepler at unpythonic.net  Mon Sep  2 20:37:40 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Mon, 2 Sep 2002 19:37:40 -0500
Subject: Making regex suck less
In-Reply-To: <1031008966.343893@ampungk.ozonline.com.au>
References: <3d725881.345921@news.t-online.de> <mailman.1030907703.26266.python-list@python.org> <akuifb$h0e$2@solaris.cc.vt.edu> <1030962064.628232@ampungk.ozonline.com.au> <al0djh$a33$1@solaris.cc.vt.edu> <1031008966.343893@ampungk.ozonline.com.au>
Message-ID: <20020902193731.A948@unpythonic.net>

On Mon, Sep 02, 2002 at 09:23:18PM +1000, John La Rooy wrote:
> Carl Banks wrote:
> 
> >>It would be more likely to look like this (I haven't put too much 
> >>thought into this)
> >
> >
> >No kidding.
> >
> >
> >
> >>"anything,anything,anything,same_as_3rd,same_as_2nd,same_as_1st"
> >>or would you like to suggest something else?
> >
> >
> >How about:
> >
> >pattern = Group(Any()) + Group(Any()) + Group(Any()) \
> >          + GroupRef(3) + GroupRef(2) + GroupRef(1)
> >
> Err symantically that's exactly the same as the re and my suggestion
> only the syntax is different. It's still nothing like saying
> 
> pattern = "6 character palindrome"

Do you mean something like this?

    def palindrome_re(n):
	pat = ["(.)" * ((n+1)/2)]
	for i in range(n/2, 0, -1):
	    pat.append("\\%d" % i)
	return "".join(pat)

With a little work, you can extend this to use named groups and named
backrefs as well, so that you can use it as a building block for larger
patterns:

    def Any(): return "."
    def Group(s, g): return "(?P<%s>%s)" % (g, s)
    def Backref(g): return "(?P=%s)" % g
    def Or(*args): return "|".join(args)

    def palindrome_re(n, p):
	pat = [Group(Any(), "%s%d") % (p, i+1) for i in range((n+1)/2)]
	for i in range(n/2, 0, -1):
	    pat.append(Backref("%s%d" % (p, i)))
	return "".join(pat)

I think that building REs in functions is a great approach for more
complex REs.

>>> q = re.compile(palindrome_re(7, "a") + palindrome_re(6, "b"))
>>> q.match("abcdcbaxyzzyx")
<_sre.SRE_Match object at 0x401c4f00>
>>> _.groupdict()
{'l4': 'd', 'l2': 'b', 'l3': 'c', 'l1': 'a', 'i1': 'x', 'i3': 'z', 'i2': 'y'}
>>> q = re.compile(Or(palindrome_re(7, "a"), palindrome_re(6, "b")))
>>> q.match('abccbb')
>>> q.match("abcdcba")
<_sre.SRE_Match object at 0x401c4f00>
>>> q.match("abccba")
<_sre.SRE_Match object at 0x402e5020>

Jeff



From aahz at pythoncraft.com  Thu Sep 19 10:29:02 2002
From: aahz at pythoncraft.com (Aahz)
Date: 19 Sep 2002 10:29:02 -0400
Subject: where is the __dict__ ?
References: <mailman.1032430104.28760.python-list@python.org> <3d89b79b$1@news.sentex.net>
Message-ID: <amcmve$aat$1@panix1.panix.com>

In article <3d89b79b$1 at news.sentex.net>,
Peter Hansen  <peter at engcorp.com> wrote:
>black wrote:
>>
>>   my book said all dictionaries have __dict__ method but couldnt find
>>   it by using dir(), where is it ?
>
>Your book is probably wrong, since dictionaries don't have a __dict__ 
>method.  In 2.2 their *classes* do, although dir() still doesn't show
>it.

__dict__ is an *attribute*, not a method:

class C:
    a = 1

print C.__dict__
print C().__dict__

x = C()
x.foo = 5

print x.__dict__
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From tchur at optushome.com.au  Fri Sep 20 17:57:57 2002
From: tchur at optushome.com.au (Tim Churches)
Date: Sat, 21 Sep 2002 07:57:57 +1000
Subject: saving state in Python/CGI
References: <5.0.0.25.2.20020920134912.00ad3230@hms.harvard.edu>
Message-ID: <3D8B99E5.B80E5287@optushome.com.au>

Frank Gibbons wrote:
> 
> Hi,
> 
> I've been using Python (and Jython) for about 5 months now. So far, so
> great. I'm building a CGI app that has about 5 different stages, and I need
> to maintain state between them. Specifically, it's a scientific tool that
> requires the user to upload a (possibly quite large) datafile, processes
> it, does some computation, then hands it back (formatted) with results.
> 
> I need access to the data on each page, yet don't want to pass it as a
> hidden field, because of its potential size. In Perl, I could just use the
> save() method from CGI.pm to save the form on the server side, pass a
> session key in a hidden field, then retrieve the data using the key. For
> the life of me, I can't find an equivalent way in Python. cgiFormStorage
> looks like a dictionary, but it's immutable, so you can't retrieve data
> from disk and add it to the form. cgi.py appears to have no analog to
> CGI.pm's save() method, at least not that I've been able to find. I've
> tried pickling the form, but that doesn't work either (it saves something
> to disk alright, but not the large data retrieved from a multi-part form
> file upload).
> 
> Perhaps I've missed something. Maybe that's just not how you do things in
> Python. Can anyone help?

You need to use one of the (many) Python Web app frameworks which help
(to greater
or lesser degrees) with storing state and other typical Web app tasks.
There is a 
list of such frameworks available at
http://www.python.org/cgi-bin/moinmoin/WebProgramming

Of these, I have only used Albatross, but I can heartily recommend it -
it was specifically
designed with highly stateful applications in mind, is elegantly
designed and well written and
has excellent documentation and the authors are very helpful. We use it
for a range of
quite complex epidemiological applications.

However, many of the other frameworks list on the above site are also
reputed to be very good and
you should consider all of them before making a choice.

Tim C

> 
> Thanks,
> 
> -Frank Gibbons
> 
> PhD, Computational Biologist,
> Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA.
> Tel: 617-432-3555       Fax:
> 617-432-3557       http://llama.med.harvard.edu/~fgibbons
> 
> --
> http://mail.python.org/mailman/listinfo/python-list



From rsrchstr at hotmail.com  Sat Sep 28 11:18:47 2002
From: rsrchstr at hotmail.com (mike henley)
Date: 28 Sep 2002 08:18:47 -0700
Subject: 'File name' is not recognized as an internal or external command, operable program
Message-ID: <576c1752.0209280718.331fe456@posting.google.com>

rsrchstr at hotmail.com (mike henley) wrote in message news:<576c1752.0209271443.1ea61518 at posting.google.com>...
> i am running windows xp which was working fine until the following
> problem occured...
> 
> in the command prompt window (as well as run...), whatever program,
> exe, batch file... etc i enter returns the following message
> 
> 'File name' is not recognized as an internal or external command,
> operable program or batch file
> 
> these are my environment variables
> 
> ComSpec 
> %SystemRoot%\system32\cmd.exe
> Path 
> C:\ruby\bin;C:\Python22;C:\Perl\bin;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem
> PATHEXT
> .COM;.EXE;.BAT;.CMD;.VBS;.VBE;.JS;.JSE;.WSF;.WSH;.py;.pyc;.pyo;.pyw;.pys;.RB;.RBW
> 
> previously the command prompt worked just fine... this happened in the
> past few days; my recent installations include activestate's perl,
> python, tcl, pragmatic programmer's ruby, gnu cygwin, apache 2, zope,
> realone player
> 
> does anyone know what i should do to fix this and restore command
> prompt functionality?

problem solved; activestate tcl binary binstallation is the offending
app. windows XP has user specific environmental variables and
universal environmental variables that affect all users, and while all
other applications mentioned above make changes to universal
variables, - as you can see the path above is mostly correct and looks
right when u look at it, but doesn't work, which was puzzling -
activestate's tcl installation makes changes to user specific
environmental variables, which override and make universal variables,
it seems, unusuable, and apparently, not read by the machine, as not
only the command prompt and "start|run..." were not working, but every
other .com, .exe, .bat that is called through them, (.exe would still
work if clicked on from explorer, but not through command prompt or
"start|run...")

the solution that worked was to delete the path and pathtext entries
from the user specific variables that were entered by activestate's
tcl binary installation and move them into the universal variables,
and with this the command prmopt and "start|run..." both worked, as
well as the beloved cygwin bash

i searched the internet, and usenet, for a solution to this problem, i
found none, nothing on microsoft knowledgebase or newsgroups, i even
went on irc to many technical newsgroups and no one knew what to do,
the replies i had on usenet groups, though i thank all who tried to
help, didn't work. (p.s. advising the poster to start a new command
windows for the environmental effects to take place is not a good
reply when the poster said he already restarted the machine, but
thanks anyway to those who suggested it for their wish to help :):)  )

i had to think up the solution myself, which took some time, and i
post it here if anyone encounters this problem in the future, as
windows XP no longer leads you to directly edit autoexec.bat, (that,
and also even if it did, regedit or sysedit no longer work when the
command prompt and start|run no longer work), go to start | control
panel | system | advanced | environmental variables  and look at not
only system variables, which are universal, but notice too the user
specific variables and move any entries not preceded by "documents and
and settings" (well, this is a generalization, it was true in my case,
but may not be so in urs, so try to work out which should belong in
the system variables instaed of the user variables) to the system
variables, restart the machine and see if it works

this seems a rather simple thing, but it is a rather unexpected one,
as system variables look right when you look at them, and if you're
not experienced with windows xp, you may not suspect that user
specific variables override system variables and render them
unreadable


p.s. i'll email this to activestate, but as i heard that most of the
code in activestate distribution is community contributed i post this
to comp.lang.tcl too


From phr-n2002b at NOSPAMnightsong.com  Fri Sep 20 20:29:34 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 20 Sep 2002 17:29:34 -0700
Subject: e vs exp()?
References: <mailman.1032550470.8080.python-list@python.org> <iiOi9.173334$5r1.7100346@bin5.nnrp.aus1.giganews.com>
Message-ID: <7xr8fofamp.fsf@ruckus.brouhaha.com>

"Terry Reedy" <tjreedy at udel.edu> writes:
> What was your basic approach?  Power series?  Rational function?
> Continued fraction? ???
> http://functions.wolfram.com/ElementaryFunctions/Power/
> gives 343 formulas (many only relevant to complex domain) but doesn't
> seem to indicate which Mathematica actually uses (trade secret I
> presume).

Mathematica has to work to arbitrary precision and would use totally
different methods than what a fixed-precision routine would use.

See "Numerical Recipes" by Press, Teukolsky, et al, for a reasonably
digestible explanation of how to do this stuff.  Note that true
numerics gurus have some problems with this book, but it will at least
give you an idea of what's going on.


From jslowery at hotmail.com  Sat Sep 14 16:35:08 2002
From: jslowery at hotmail.com (Jeremy Lowery)
Date: Sat, 14 Sep 2002 15:35:08 -0500
Subject: CPython detecting read-only attributes.
References: <qktg9.1085$AY5.284165@e3500-atl1.usenetserver.com>
Message-ID: <z%Mg9.122810$2L.6135100@e3500-atl2.usenetserver.com>

I believe I found a hacky way to do this. I am sure that this isn't the
fastest way.  I wonder if it can easily be shot down.

static int
proto_setattro(HyProtoObject *self, PyObject *name, PyObject *value)
{
 int res;
 PyObject *tmp;
 PyObject *etype, *evalue, *etb;

 res = PyObject_GenericSetAttr((PyObject *)self, name, value);
 if(res != -1)
  return res;
 /* if we can get the attribute with GetAttr, then it must be readonly.
  */

 PyErr_Fetch(&etype, &evalue, &etb);
 if(etype == NULL) {
  PyErr_SetString(PyExc_SystemError, "No error object found after a failed "
   "attribute lookup.");
  return -1;
 }

 tmp = PyObject_GenericGetAttr((PyObject*)self, name);
 if(tmp != NULL) {
  /* reference found, readonly */
  Py_DECREF(tmp);
  PyErr_Restore(etype, evalue, etb);
  return -1;
 }

 PyErr_Clear();

/* continue with custom processing */

J Lowery






From marklists at mceahern.com  Mon Sep 30 17:41:18 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 30 Sep 2002 16:41:18 -0500
Subject: a question about mailman
In-Reply-To: <m3it0n2n48.fsf@mira.informatik.hu-berlin.de>
Message-ID: <JHEOKEOOLIGLDHCMAHMOCEKNECAA.marklists@mceahern.com>

[Martin v. Loewis]
> I'm not sure I understand this question: Do you want to add the URL of
> the list archive to the footer, or the URL for the specific message?

The URL for the specific message.

> Adding the URL for the entire archive is easy (it only varies with the
> list, not with the message), so I assume you want to add the URL for
> the message itself.

Correct.

> Yes, yes, and no. I believe the trick is that, at the point of sending
> the message, you cannot know the URL that the message will get in the
> archive. Finding out that URL is particularly tricky in presence of
> external archivers.

Hmm, when I glanced at the code it looked like the order of processing in
the pipeline was for the message to pass through the Archiver before the
Dispatcher, so I just simplistically assumed: 1) that the Archiver added
enough information to the message so that; 2) by the time the message got to
the Dispatcher, you'd have enough information to format a pseudo string like
this in the footer:

  %(base_archive_url)s/%(message_archive_relative_url)s

Sadly, I haven't installed Mailman yet, so I can't test my hypothesis.

Thanks for the feedback.

Cheers,

// mark




From fetus at AUTO  Thu Sep 26 16:32:06 2002
From: fetus at AUTO (A. Dimsdale and A. Moser)
Date: Thu, 26 Sep 2002 20:32:06 GMT
Subject: Pythonwin (win32all build 148) on Python 2.2.1 on Win2K, bogus syntax e
References: <u47k9.46401$V7.11091779@twister.socal.rr.com> <CJ8k9.16479$6g7.52569@news-server.bigpond.net.au> <sHJk9.66556$U7.23747920@twister.socal.rr.com>
Message-ID: <a9Kk9.53648$V7.13380285@twister.socal.rr.com>

A. Dimsdale Fetus A. Moser <fetus at AUTO> writes:

>"Neil Hodgson" <nhodgson at bigpond.net.au> writes:

>>Aaron Dimsdale and Alex Moser:

>>> Pythonwin barfed on the quote marks marked by carets, returning: "Failed
>>to run script -
>>> syntax error - invalid syntax".

>>   Try turning on View | End-of-line markers to see if you have non-standard
>>line endings such as pure carriage returns, which cause just this sort of
>>error. On Windows, you want line endings to be carriage return + line feed.

>>   Neil

[snip]

Well, I went in and manually edited the EOLs on every code line in the file.
(Fun fun fun.) As it turns out, the cause of the problem is likely the
difference between DOS-style and Unix-style EOLs. Alex and I jointly own an
account on my Linux box at home, where we keep the latest version of the script.
Alex does most of the coding on our script, and I just do other stuff like
looking over the code when it slaps him, adding fixes and/or minor features,
etc. Well, since he does most of the coding, he had the script on his work comp
when we were at work on Tuesday, and I had him upload it to my Linux box so I
could grab it from my work box by FTP, and through the Win-to-Linux-to-Win
transfer the EOLs apparently got funkified.
Thanks again for the EOL tip.

--Aaron L. Dimsdale


From BPettersen at NAREX.com  Wed Sep 18 22:20:39 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Wed, 18 Sep 2002 20:20:39 -0600
Subject: How to build Python 2.2.1 on Windows?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201253451@admin56.narex.com>

> From: David Lees [mailto:debl2nonospammmmyy at bellatlantic.net] 
> 
> I have been playing with Python for awhile and would like to 
> know how to try building Python on a Windows machine that has 
> MS C++ version 6.  Is there a standard download that is 
> configured to do this?  If not could someone point me in the 
> direction of information on builds?  I downloaded 
> Python-2.2.1.tgz.tar on my Win98 box and WinZip 8 does not 
> even seem to know how to unzip it.  I just transfered the tar 
> file over to a linux box and it unzips without and problem 
> using tar.  The README seems to only talk about linux/unix 
> type systems.

The PCBuild directory contains pcbuild.dsw.

-- bjorn



From josh_py at sautez.com  Tue Sep 17 03:35:23 2002
From: josh_py at sautez.com (josh_py at sautez.com)
Date: Tue, 17 Sep 2002 00:35:23 -0700
Subject: using python as cgi on apache unix
References: <am6fm301477@enews2.newsguy.com> <mailman.1032245299.5563.python-list@python.org>
Message-ID: <am6lvq01e4k@enews2.newsguy.com>

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote in message
news:mailman.1032245299.5563.python-list at python.org...
> * josh_py at sautez.com <josh_py at sautez.com> [2002-09-16 22:48 -0700]:
> > All,
> >
> > I have developed a python solution to run and connect to a mysql
database.
> > I did all my development under windows apache 1.3.26 and python 2.2.1.
I
> > know want to move this to my unix server.  Unfortunately, no matter what
I
> > do I can not get apache on my linux box to execute any python scripts.
I
> > get the following errors in apache:
> >
> > Premature end of script headers: /var/www/html/cgi-bin/test.py
> >
> > I have all the settings in apache I believe to be correct.  It complains
> > that the file is not executible if that is not set.  Other CGI works
fine.
> >
> > test.py looks like:
> >
> > #!/usr/bin/python2
> >
> > print "\n\n"
> > print "Content-Type: text/html\n\n"
> > print "hello"
>
> The first print "\n\n" is an error. It was simply luck that this was
> running anywhere else as expected. Just delete this line and your hello
> world cgi script should run fine.
>
> Also, to test wether Python CGI scripts work, you can always simply use:
>
> #! [path to python ...]
> import cgi
> cgi.test()
>
> -- Gerhard
>

Gerhard,

Thanks for the test scipt.  Unfortunely that runs fine at a command prompt
but same error results under apache.

Josh




From sismex01 at hebmex.com  Mon Sep  9 16:26:50 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Mon, 9 Sep 2002 15:26:50 -0500 
Subject: a re question
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062330237CF33@mail.hebmex.com>

> 
> Hi,
>   I have a file with lines of the format:
> 
> 001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23
> 001 Xya FGh   143557789   7 100 09/05/2002 20:23:23
> 


Are the fields are always fixed length?


> I am trying to extract the 9 digit field and the single digit field
> immediatley after that.
> 
> When I use Visual Regexp to try out the regexp 
> 
> (\d{9,} {3,}\d)
> 
> it highlights the 2 fields exactly. 
> 
> But when I use the following Python code I get None:
> 
> >> s='001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23'
> >> p = re.compile(r'(\d{9,} {3,}\d)')
> >> print p.match(s)
> >> None
> 
> Could anybody point out where I'm going wrong?
> 
> Thanks,
> 

You could be a bit more explicit with your regexp.
Using the input you gave:

> 001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23
> 001 Xya FGh   143557789   7 100 09/05/2002 20:23:23

I'd think something like this is better:

rx = re.compile(r"\d+\s\D{9}\s(\d{9})\s+(\d+)")

And that gives us:

>>> import re
>>> rx = re.compile(r"(\d{9})\s+(\d+)")
>>> s = """
... 001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23
... 001 Xya FGh   143557789   7 100 09/05/2002 20:23:23
... """
>>> rx.findall(s)
[('123456789', '7'), ('143557789', '7')]
>>> 

Something like that.

-gus
--

Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From lee100 at sport.rr.com  Wed Sep 18 19:10:36 2002
From: lee100 at sport.rr.com (Lee Gray)
Date: Wed, 18 Sep 2002 23:10:36 GMT
Subject: Python GUI app to impress the boss?
References: <mailman.1032381445.13579.python-list@python.org> <Xns928DF32683D0cliechtigmxnet@62.2.16.82>
Message-ID: <MJ7i9.293795$Yd.11298232@twister.austin.rr.com>

Chris Liechti wrote:

> "Gray, Lee P." <lpgray at UOP.com> wrote in
>> Yeah, I tried to download from CVS, but being a complete noob
>> couldn't figure out how, short of tracking down and copying and pasting
>> every updated file.  Any tips would be appreciated.
> 
> you're on windows right? then grab http://tortoisecvs.sf.net and install
> the latest version. after that right click on an empty folder in the
> windows explorer and select "cvs checkout", copy and paste the "CVSROOT"
> in the dialog. you can find the cvsroot, or at least how it looks like on
> the "browse cvs" page that you have apparently found.
> 
> the home of cvs is http://www.cvshome.org and these are interesting sites:
> http://www.loria.fr/~molli/cvs-index.html, http://cvsbook.red-bean.com/

Yep, Windows at work, mostly Linux at home.  I'll check this out at work 
tomorrow.

Thanks for the info!
Lee


From gerhard.haering at opus-gmbh.net  Thu Sep 19 04:36:02 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 19 Sep 2002 08:36:02 GMT
Subject: monetary applications [was: Python GUI app to impress the bos s?]
References: <31575A892FF6D1118F5800600846864DCBD817@intrepid> <mailman.1032362843.10509.python-list@python.org> <23891c90.0209190014.26c5bb9@posting.google.com>
Message-ID: <slrnaoj369.28c.gerhard.haering@haering.opus-gmbh.net>

Paul Boddie wrote:
> Taking a look at pyPgSQL, as the only database module I have installed
> that actually readily documents such things (and that should tell you
> something), it would seem that monetary amounts are converted to floating
> point numbers - surely the first thing one would want to avoid in
> manipulating financial data.

Yeah, we should probably change PgMoney to use PgNumeric, which is a
fixed-point type that uses Python's long internally. I've added it to the
TODO list [1].

Btw. the MONEY type is deprecated in favour of NUMERIC and DECIMAL under
PostgreSQL [2]

-- Gerhard

[1] http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/pypgsql/pypgsql/TODO?rev=1.2&content-type=text/vnd.viewcvs-markup

[2] http://www.postgresql.org/idocs/index.php?datatype-money.html


From eric.brunel at pragmadev.com  Fri Sep 13 09:06:24 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Fri, 13 Sep 2002 15:06:24 +0200
Subject: Pmw.ScrolledText programatically scrolling
References: <3D81D44E.C5FC4A8B@mill.co.uk>
Message-ID: <alsndq$25r$1@wanadoo.fr>

Joe Connellan wrote:

> I've got a Pmw.ScrolledText() that I'm using as a log and I've written
> an addToLog() method like so:
> 
>  def addToLog(self, logString):
>   self.logScr.configure(text_state = 'normal')
>   self.logScr.insert(END, logString)
>   self.logScr.configure(text_state = 'disabled')
> 
> I now want it to scroll when I add a line so the user doesn't have to
> constantly scroll down to view the most recent lines.
> 
> I can't find any info on doing this with ScrolledText or its text
> component - do I need to reposition the scrollbars for it to scroll?

Adding:
self.logScr.see(END)
should do the trick.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From i.linkweiler at web.de  Tue Sep 17 14:18:35 2002
From: i.linkweiler at web.de (Ingo Linkweiler)
Date: Tue, 17 Sep 2002 20:18:35 +0200
Subject: Logical Programming with Python?
Message-ID: <3D8771FB.4000600@web.de>

Hello,

i am looking for examples for logical programming (Prolog like) with 
Python. I have found some projects, but no examples how to use them.
Most of them seem "unuseable" or only can solve equations:

* PyLog (provides a simple way to write logic terms, variables and 
atoms)- http://christophe.delord.free.fr/en/pylog/logic.html

* Logilab constraint package: a constraint satisfaction problem solver 
written in 100% pure Python - 
http://www.logilab.org/python-logic/constraint.html

* Holmes (dead)

* Formula (not available)

* CLIPS Wrapper (unuseable)

Do you know more sources for logical programming, or do you have some 
examples ?

Ingo



From contact at realspace.org  Sat Sep 21 19:13:23 2002
From: contact at realspace.org (Bernhard Glueck)
Date: Sun, 22 Sep 2002 01:13:23 +0200
Subject: Very complex Python/C++ embedding/extension question(s)
Message-ID: <3d8cfd3c$1@news.swissonline.ch>

Hi there
I am developing a new multimedia/game/vr architecture, and at the moment i
am trying to embed python as a scripting language. I have a slew of
questions about the Python/C API..

At first let me elaborate on my unique architecture. In my
"engine" as i call it an application is built using "Actors" which are
Classes that interact over a "reflection system".  This has been implemented
using a custom template based code on the C++ side, but was written so that
any language could interact with it.

Every Actor derives from teh C++ class "Actor" which defines some basic
operations and the call protocl..
e.g actor.Call("MyMethod",arg1,arg2,arg3); will call the method "MyMethod"
on the actor using the arguments passed.  This is achieved by having a type
descriptor for every Actor that is stored in a global list. The type
descriptor dervies from the base class "Type" and knows which methods a
class has.

e.g
class MyActor: public Actor
{
}

has a type descriptopr

class MyActorType : public Type
{
};

that is registered into the global type list.
Then when one calls a method on an actor the appropriate type descriptor is
retrieved from the global list and the method is caleld using the
descriptor.

So far so good .. This works like a charm, and since the type descriptors
are defined using some easy to use macros, writing new classes is very easy,
and allows actors to be defined in extension modules (DLL files on windows )
for my engine. In fact the main engine only contains one actor for loading
DLL files .. all the actors are in other DLLs... this way i can seperate my
modules very nice, and extend the engine at runtime..

Actors are then instantiated using XML and put together to achieve
functionality.

Again this works nice.. Actors are by the way internally stored using a
smartPointer class of mine for automatic reference counting/garbage
collection..

Now i want to extend my engine using Python so that Actors can also be
written in Python , and existing actors can be derived from in Python, as
well as Python Actors can be called from C++ in the same way as others..

I wanted to do this that way derive a "PythonType" class from "Type" and
store any classes that derive directly or indirectly from C++Actor in it ..
so that PythonType works the same way as MyActorType but for all Python
classes in the script files.This would allow me to treat python actors like
C++ actors ( for my tools especially like the editor for game levels )

How would one go about this ? I would need to find out all classes that are
currently in the (embedded ) python interpreter, check if it derives from
the C++Actor class,and if yes generate an instance of PythonType for it
which would know the methods of the python class..

How to do this ?

Second, i would need to export the C++ Actor base class and all other C++
Actor derived classes to python, so that one can write an actori n python
that derives from
one of the already written C++ actors..
I know of all the Wrapper generates etc, and read the tutortials on
embedding and extending python, but note that i do not know at COMPILE Time
of the C++ part which actors are there ( since new modules can be loaded at
runtime ) so this export process needs to be dynamic..... Any ideas ?

And third.. Since the engine deals with actors only, and does not know the
exact type, the python actors would have to have methods like

def MyMethod( actor )

which is just an Actor Smart Pointer wrapped....
Is it possible to have a "method hook" in (with the PythonC API)  that would
be caleld every time i try to invoke a method on this actor ?  This way i
could check the actor concrete type, and determine which methods it has ...
e.g so
that
actor.DoSomething() Would internally use the actor.Call("DoSomething" )
protocol in C++ ..... just translating native Python code to my own method
of calling methods on classes(actors) Remeber i do not know at compile time
which methods an actor supports..... so this would need to be done using
some sort of MethodCallHook.....




From tjreedy at udel.edu  Sat Sep 21 13:49:05 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 21 Sep 2002 17:49:05 GMT
Subject: if...else on the same line?
References: <mailman.1032622182.9002.python-list@python.org>
Message-ID: <li2j9.183619$z91.7670812@bin3.nnrp.aus1.giganews.com>

"David Garamond" <davegaramond at icqmail.com> wrote in message
news:mailman.1032622182.9002.python-list at python.org...
> i'm sorry for this terrorizingly basic question, but is it possible
at
> all to write an if ... else statement on a single line in python?

If you mean something like

if x == 1: y=3; else z=4

then no.  Keywords that introduce blocks (suites) must be first word
on line:  I believe this includes try, except, finally, for, while,
if, elif, else, def, and class.

A statement using a conditional expression is something else:
y = (x==1) and 3 or 4 #for example

Terry J. Reedy




From michaels at one.net  Thu Sep 12 04:08:25 2002
From: michaels at one.net (Michael Schneider)
Date: Thu, 12 Sep 2002 04:08:25 -0400
Subject: iterators and generators, is this the Python Way???
References: <3D7FF921.2040305@one.net> <just-DB3861.09023412092002@news1.xs4all.nl>
Message-ID: <3D804B79.9070003@one.net>

Just,

Thank you very much, I like this approach.

I was under the mistaken impression what the
iterator needed to do.

Mike

Just wrote:

>In article <3D7FF921.2040305 at one.net>,
> Michael Schneider <michaels at one.net> wrote:
>
>>class FastFile:
>>    def __init__(self, filename):
>>        """ Read in entire file into memory"""
>>        f = open(filename, 'r')
>>        fileContents = f.read();
>>        self.lines = fileContents.split('\n')
>>        f.close()
>>   
>>    def  __iter__(self):
>>        return self.fastFileGen()
>>
>>    def fastFileGen(self):
>>        for line in self.lines:
>>                yield line
>>
>
>I would write the __iter__ like so:
>
>    def  __iter__(self):
>        for line in self.lines:
>                yield line
>
>Same effect, less code...
>
>Just
>



From sholden at holdenweb.com  Mon Sep 30 13:41:45 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 30 Sep 2002 13:41:45 -0400
Subject: McMillian installer + win32com.client.constants - how to use them together?
References: <9ff6b66.0209300856.5fcdf961@posting.google.com>
Message-ID: <i60m9.6298$_32.1254@fe04>

"vtail" <vtail at yandex.ru> wrote ...
> Greetings,
>
> can anybody please explain me the Right Way to use McMillian installer
> with application that uses constants from MSWord library?
>
Gordon MacMillan reads this list when he has time, but you should be aware
that there is a mailing list just for the users of the installer package --
see

    http://trixie.triqs.com/mailman/listinfo/installer

for details.

> I've generate them with win32com/client/makepy.py and use them in my
> code via
> ====>
> import win32com.client
> ...
> a = win32com.client.constants.wdBlack  #  application crashed here
> ...
> <===
>
> When I build .exe from my applications it is crashed on marked line
> with AttributeError exception.
>
> I've tried to study this subject myself, but now I'm totally confused
> with all this gencache.EnsureModule(), prog_id, CLSID etc. I can't
> find any clear documentation on the subject.
>
> I'll greatly appreciated any help on this, even RTFM links ;).
>
I'm *guessing* that the installer hasn't been able to define that the
win32com.client package's submodules need to be included. To verify this is
the error you might try including an explicit

    import win32com.client.constants

in the appropriate module and see if the error goes away. There are things
you can do in installer if that's the case, but I'm not the best person to
advise.

regards
------------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From joe.sixpack at quartercase.com  Tue Sep 24 17:39:19 2002
From: joe.sixpack at quartercase.com (joe sixpack)
Date: Tue, 24 Sep 2002 21:39:19 GMT
Subject: Running with arguments in IDLE?
Message-ID: <3D90DB87.4797EA7@quartercase.com>

Probable dumb question of the day...

I do most of my work in Python with IDLE and love it.  If I need to 
change a parameter in a script I've written, I usually just change a 
variable by editing the script, then run it.

Now I've gotten to where I need to write a script that runs from a 
shell (Windows2000) command line.  I know to use sys.argv[x] to get
the arguments.  What I don't know is how to run the script from 
within IDLE and specify the argument to use when it is run.  In
Pythonwin, 'RUN'ning a script brings up the dialog box in which
arguments may be specified.  How is this done in IDLE?  

Thanks!


From peter at engcorp.com  Sun Sep 22 14:23:51 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 22 Sep 2002 14:23:51 -0400
Subject: sys.ps2 assignment has no effect
References: <uos1np8t53k396@corp.supernews.com>
Message-ID: <3d8e0aae@news.sentex.net>

Lance wrote:
> I'm running the latest Python on Windows XP. The default installed secondary
> string appears to be blank spaces rather than an ellipse. The following has
> no effect:
> 
> import sys
> sys.ps2 = '... '
What does "print '...'" show you?  Maybe this is just a phenomenon of 
your current font?





From a.clarke11 at pop.ntlworld.com  Mon Sep 16 18:23:58 2002
From: a.clarke11 at pop.ntlworld.com (a.clarke11)
Date: Mon, 16 Sep 2002 23:23:58 +0100
Subject: doh
References: <3d83fa2c$0$24041@echo-01.iinet.net.au> <mailman.1032061218.4602.python-list@python.org> <3d84a3ed$0$24051@echo-01.iinet.net.au>
Message-ID: <3D8659FE.D406C4F8@pop.ntlworld.com>


> A deer...



From r.kalbermatter at hccnet.nl  Wed Sep 18 17:33:22 2002
From: r.kalbermatter at hccnet.nl (Rolf Kalbermatter)
Date: Wed, 18 Sep 2002 23:33:22 +0200
Subject: Python as shared library on Unix
In-Reply-To: <EKEJLGEFHHFGJOBEGJECGEIICFAA.r.kalbermatter@hccnet.nl>
Message-ID: <EKEJLGEFHHFGJOBEGJECEEIMCFAA.r.kalbermatter@hccnet.nl>

Please disregard this post. I just found out that the current 2.3 CVS
version supports shared library builds. I was working with a 2.2.1c2
source code package.

There is probably little chance that this will be officially backported
to the 2.2.1 branch of Python? 

> I'm currently in the process of porting a Python embedding from
> Windows to Linux and have come across an appearent problem.
> 
> Under Windows I have linked my app dynamically to the pythonXX.dll using
> LoadLibrary/GetProcAddress and that works very nicely. Under Linux I see
> no option to build a python.so.2.2.1 or similar to link my application
> to with the dl* functions.

Rolf Kalbermatter



From eugene1977 at hotmail.com  Sat Sep 14 22:52:20 2002
From: eugene1977 at hotmail.com (eugene kim)
Date: Sun, 15 Sep 2002 02:52:20 +0000
Subject: question about dictionary type..
References: <alt22f$pf8$1@newsreader.mailgate.org> <alv4op$m67$1@newsreader2.netcologne.de> <61Jg9.16334$5r1.665293@bin5.nnrp.aus1.giganews.com>
Message-ID: <am0390$cvf$1@newsreader.mailgate.org>

thank you.. i just checked replies..
i haven't tried them..
this is what i came up with..
(making needed tables first.. as carl and padraig said)
i'm having one problem(described below)


d4 = {}
for line in stdin.readlines():
    words = split(line)
    if len(words) > 0 :
        if words[0] == 'call' :
            games = games + 1
        elif words[0][0] == ')':
            table1 = {}
            table2= {}
            table3 = {}
            table1[dataList[3]] = 0
            table2[dataList[2]] = table1
            table3[dataList[0]] = table2
            table3[dataList[1]] = table2

            for i in [dataList[0], dataList[1]]:
                if not d4.has_key(i) :
                    d4[i] = table2
                else:
                    if not d4[i].has_key(dataList[2]):
                        d4[i][dataList[2]] = table1
                    else:
                        if not d4[i][dataList[2]].has_key(dataList[3]):
                            d4[i][dataList[2]][dataList[3]] = 0
            print d4
            dataList = []

        else:
            aList = []
            for i in words:
                i = (upper(replace(replace(i, '\'', ''), ',' ,'')))
                aList.append(i)
            
            dataList.append(join(aList))

input looks like this..
-------------------------------------
call games_view_insert (
'California',
'Baylor',
'football',
'm',
'31-Aug-02',
'Berkeley, Calif.',
'-1',
'-1'
);

call games_view_insert (
'Baylor',
'Samford',
'football',
'f',
'07-Sep-02',
'Waco, Texas',
'-1',
'-1'
);

call games_view_insert (
'New Mexico',
'Baylor',
'football',
'm',
'14-Sep-02',
'Albuquerque, N.M.',
'-1',
'-1'
);
--------------------------------------------
this is output..in each loop ( print d4)
problem is california should have had 'M' only, not 'F'emale team..

{'BAYLOR': {'FOOTBALL': {'M': 0}}, 'CALIFORNIA': {'FOOTBALL': {'M': 0}}}
{'BAYLOR': {'FOOTBALL': {'M': 0, 'F': 0}}, 'CALIFORNIA': {'FOOTBALL': {'M': 
0, 'F': 0}}, 'SAMFORD': {'FOOTBALL': {'F': 0}}}
{'BAYLOR': {'FOOTBALL': {'M': 0, 'F': 0}}, 'CALIFORNIA': {'FOOTBALL': {'M': 
0, 'F': 0}}, 'NEW MEXICO': {'FOOTBALL': {'M': 0}}, 'SAMFORD': {'FOOTBALL': 
{'F': 0}}}





From brueckd at tbye.com  Fri Sep 27 10:40:02 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Fri, 27 Sep 2002 07:40:02 -0700 (PDT)
Subject: _winreg.DeleteValue - Permission denied ...
In-Reply-To: <an108i$ddv$03$1@news.t-online.com>
Message-ID: <Pine.LNX.4.44.0209270736540.25858-100000@ns0.tbye.com>

On Fri, 27 Sep 2002, Harald Schneider wrote:

> When I try to delete a registry key, I get a WindowsError (Errno5) /
> Permission denied, even thought I am logged in as admin.
> Deleting the same key with regedit works.
> 
> The function looks like this:
> 
>     def del_cfg_path(self):
>         key =
> _winreg.OpenKey(_winreg.HKEY_LOCAL_MACHINE,'SOFTWARE\TEST\SUBTEST)
>         _winreg.DeleteValue(key,"TESTVAL")
>         _winreg.CloseKey(key)
> 
> Any ideas ?

Hi Harald,
Perhaps you need to open the registry key with write permissions, e.g.:

key = _winreg.OpenKey(_winreg.HKEY_LOCAL_MACHINE, \
                      'SOFTWARE\\TEST\\SUBTEST', \
                      0, _winreg.KEY_ALL_ACCESS)

(Also note that you should use double backslashes in your strings or use 
raw strings)

HTH,
Dave




From phr-n2002b at NOSPAMnightsong.com  Wed Sep  4 06:32:49 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 04 Sep 2002 03:32:49 -0700
Subject: encryption (passwords)
References: <3d75d115$0$221$4d4ebb8e@news.nl.uu.net>
Message-ID: <7xd6ru2gfi.fsf@ruckus.brouhaha.com>

"Gumuz" <gumuz at looze.net> writes:
> My limited understanding of encryption tells me that I need a 'key' to
> decrypt stuff. So I figure that this key will bey written(hard-coded?)
> somewhere in my Python-script and it needs to be known at both sides in
> order to encrypt and decrypt. Somehow, I can't believe this is true. I can't
> get over the idea that if my application becomes open-source anyone could
> see the key and decrypt my messages.
> 
> I am fairly certain that it probably doesn't work this way, but I have
> really no idea. Cann someone clarify this for me?

OK, first of all, that type of question is probably better asked on
sci.crypt rather than the Python group.  

Second, a word of advice: don't try to implement encryption protocols
with your present state of knowledge, and especially don't try to
design protocols yourself.  It's extremely easy to make mistakes that
mess up security.  Even experts make those mistakes, all the time.

I'll sketch out answers to your questions but basically I think your
best approach is to use an unencrypted protocol between your Python
scripts, and tunnel it through an encrypting socket layer that someone
else has already debugged.  The two most popular choices of encryption
layers are SSL and SSH.  SSH has a "port forwarding" feature that lets
you set up local ports on both machines, that communicate through an
encrypted tunnel.  Your Python program would just open the local port
and send unencrypted data to it, and the same data would appear at the
other end transparently decrypted.  SSL is a fancier protocol that
doesn't incorporate forwarding, but there's a program called stunnel
that implements forwarding over SSL.  

Between the two, I'd say SSH is a easier to set up if you're just
trying to secure a single socket pair between two machines.  If you're
trying to have a bunch of machines that talk to each other, and/or
they're not all controlled by the same people (e.g. you're a vendor
that needs to exchange secure traffic with various different
customers), you probably want to use SSL.  That type of network is
called an extranet and various books have been written about how to
build and manage them.  If you say more specifically what you're
trying to do, I or someone else here might be able to advise you more
usefully.

Finally, if you're building any kind of secure system that's intended
to protect valuable data, I recommend you read "Security Engineering"
by Ross Anderson all the way through before deploying your system.  It
will give you an idea of what you're up against and what kinds of
measures you can take.

Re your basic questions:

Yes you're correct that you need a secret decryption key.  The key
should be unique for your specific installation, NOT part of the
program that gets distributed.  It's like the password that you log
into your computer with.  The login program is distributed freely as
source code, but your password is not part of it.  You install the
program, THEN set your password for your specific account.

Then there's the problem of how to store the key, so your program can
use it.  There are several approaches, each with good and bad points:

    1) Include the key, in cleartext form, as part of the program
    configuration that gets loaded at program startup.  Good points:
    convenient.  Bad point: anyone with access to the config file
    (possibly gotten by breaking into your system) gets the key.
    This method is not recommended if the data being protected is valuable.

    2) Use a passphrase as a key that you enter when you start the
    program, so it stays in memory and doesn't get stored on disk.
    Good point: better security (key not stored).  Bad points: you
    have to enter the key every time you start the program.  If you
    want someone else to start the program for you, you have to tell
    them the passphrase.  If the program has to start without human
    attention at reboot if the system crashes, you're hosed.

    3) Encapsulate the key in a special hardware device that does the
    cryptography.  This can range from a $5 smart card or USB dongle, to a
    multi-kilobuck sealed module on your PCI or SCSI bus.  This is the
    preferred method for high security applications (banking, etc.).  Good
    points: best security of all, key can't be stolen without physical
    access to the machine, can implement its own access controls, etc.
    Bad points: special hardware required; cheaper devices are slow.


From gerhard.haering at opus-gmbh.net  Thu Sep  5 03:10:12 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 5 Sep 2002 07:10:12 GMT
Subject: Database BSD on Mac?
References: <3144be3c280c062e5e71a98ab191f664.98737@mygate.mailgate.org> <mailman.1031095985.2997.python-list@python.org> <_7vd9.57894$ub2.928923@news1.tin.it> <slrnancvnh.107.gerhard.haering@lilith.my-fqdn.de> <tDvd9.57994$ub2.932923@news1.tin.it>
Message-ID: <slrnane0nf.1v8.gerhard.haering@haering.opus-gmbh.net>

Max wrote:
>> > I read shelve can be used for small amount of data... will it work with
> 5mb?
^^^

Your newsreader is broken.

> [...] I was reading on my book it's possible to create encrypted modules,
> but don't explain much how.  My idea is to create a exe file in C joust
> to execute the python encrypted modules stored with shelve.  Is it
> possible? Can you point me to a tutorial/example code?

Hooray. Closed source at its best.

It's possible using import hooks. You can search groups.google.com for
previous discussions of this. 

-- Gerhard


From jknapka at earthlink.net  Thu Sep 12 00:18:34 2002
From: jknapka at earthlink.net (Joseph A. Knapka)
Date: Wed, 11 Sep 2002 22:18:34 -0600
Subject: "str.contains(part)" or alternatives?
References: <B43D149A9AB2D411971300B0D03D7E8BF0A592@natasha.auslabs.avaya.com>
Message-ID: <3D80159A.C21B7002@earthlink.net>

"Delaney, Timothy" wrote:
> 
> > From: Joseph A. Knapka [mailto:jknapka at earthlink.net]
> >
> > So in 2.3, can I do this and get the expected result?
> >
> > >>> "" in "abc"
> > 1
> 
> Yes - this has been hotly debated and defined as true.
> 
> > >>> for subst in "abc":
> > ...   print subst
> > ...
> > <empty string>
> > a
> > ab
> > abc
> > b
> > bc
> > c
> 
> No. Iterating over a string is defined to return each single character in
> turn. Remember:
> 
>     for subst in "abc":
> 
> is equivalent to
> 
>     for subst in iter("abc"):

I didn't know that (never saw "iter()" before).

> and iter("abc") returns "a", "b", "c"
> 
> whilst
> 
>     if subst in "abc":
> 
> is equivalent to
> 
>     if "abc".__contains__(subst):
> 
> (falling back to if subst in iter("abc") if "abc" does not define
> __contains__)
> 
> and str.__contains__ performs a substring match.

Ah. So we have:

not(X in Y <==> X in [Z for Z in Y])

But only if Y is a string. That strikes me as a trifle odd,
but nevertheless I agree that the "substring in string"
behavior seems more useful.

-- Joe
  "I'd rather chew my leg off than maintain Java code, which
   sucks, 'cause I have a lot of Java code to maintain and
   the leg surgery is starting to get expensive." - Me



From sholden at holdenweb.com  Mon Sep 30 14:00:27 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 30 Sep 2002 14:00:27 -0400
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05> <3D988FCE.6040502@cascade-sys.com>
Message-ID: <042e01c268ab$427dc430$6300000a@holdenweb.com>

>
> Steve Holden wrote:
>
> ><sismex01 at hebmex.com> wrote...
> >
> >
> >[ ... ]
> >OK, so don't use whole pennies (or smallest currency) as your
> >base, use 10,000ths of a penny, but still use integer arithmetic
> >for your calculations.
> >
> Normal Integers aren't big enough.  2^31 / 10K == 215K.
>

Please note that the remark you are responding to was made by the poster you
quote me as quoting.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------






From cnetzer at mail.arc.nasa.gov  Mon Sep 30 18:04:03 2002
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Mon, 30 Sep 2002 15:04:03 -0700
Subject: Help please with menubars
In-Reply-To: <cb035744.0209290318.2b1fe7ef@posting.google.com>
References: <000f01c266f3$2903dbc0$2502a8c0@emily.ewndsr01.nj.comcast.net> <mailman.1033266279.6222.python-list@python.org> <cb035744.0209290318.2b1fe7ef@posting.google.com>
Message-ID: <200209302204.PAA27060@mail.arc.nasa.gov>

On Sunday 29 September 2002 04:18, Will Stuyvesant wrote:
>
> There is no visible difference between such an ``Exit'' entry
> (calling a function directly) and a normal dropdown menu entry.  Do
> you know how to display ``Exit'' like a Button?

You could use the image option to make a button image, or change the 
background color to differentiate it.

But otherwise, not really.  You would have to manage your own menubar 
(by making a frame w/ buttons and menubuttons, etc.  But that won't 
have the advantages of using the Tk Menu (ie. it'll stick out like a 
sore thumb on Macs, among other things).

But, I'll defer to Jeff Epler <jepler at unpythonic.net>, in his post, and 
say that probably none of this really makes sense.  You could make a 
button that isn't in the menu bar to quit (as I do in some 
applications), or capture the window manager close window event to 
clean up when the system provided window close is clicked on, etc.  And 
just make "Quit" a real menu item (under File, say).

-- 

Chad Netzer
cnetzer at mail.arc.nasa.gov



From kseehof at neuralintegrator.com  Thu Sep 12 01:34:40 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Wed, 11 Sep 2002 22:34:40 -0700
Subject: Larry Wall's comment on python...
In-Reply-To: <ald3g5$1p8lpp$1@ID-77524.news.dfncis.de>
References: <wkr8g7rn4c.fsf@yahoo.com>
 <3D7958DB.50903@nospam.free.fr>
Message-ID: <4.3.1.0.20020911222726.00c00e90@pop.hurrah.com>

At 07:46 AM 9/7/2002 Saturday, Thorsten Kampe wrote:
>* laotseu
> >Rod Stephenson wrote:
> >> Slashdot has a list of questions posed to Larry Wall (perl). When
> >> asked his thoughts on other scripting languages, he makes the
> >> following observation about python
> >>
> >> "Python is cool to look at small bits of, but I think the "outline"
> >> syntax breaks down with larger chunks of code. I'm with Aristotle on
> >> the structure of discourse--a story should have a beginning, and
> >> middle, and an end. So should blocks"
> >> [...]
> >
> > def fun(args):
> > #BEGIN
> >      if ceci or cela:
> >      #BEGIN
> >          do_this(with_that)
> >          [snip a lot of code]
> >      #END
> > #END
> >
> > So what the problem with blocks ?
>
>Nope. Bad, bad readability, sorry.
>
>#||<--- LOOK HERE, LARRY, THIS IS WHERE THE FUNCTION _*BEGINS*_!
>def fun(args):
>     #||<--- LOOK HERE, LARRY, THIS IS WHERE THE IF-CLAUSE _*BEGINS*_!
>     if ceci or cela:
>         do_this(with_that)
>         [snip a lot of code]
>     #||<--- LOOK HERE, LARRY, THIS IS WHERE THE IF-CLAUSE _*ENDS*_!
>#||<--- LOOK HERE, LARRY, THIS IS WHERE THE FUNCTION _*ENDS*_!
>
>
>Thorsten

Definitely not.  Too wide by far.  You need to reduce the indent to one space.

#||<--- LOOK HERE, LARRY, THIS IS WHERE THE FUNCTION _*BEGINS*_!
def fun(args):
  #||<--- LOOK HERE, LARRY, THIS IS WHERE THE IF-CLAUSE _*BEGINS*_!
  if ceci or cela:
   do_this(with_that)
   [snip a lot of code]
  #||<--- LOOK HERE, LARRY, THIS IS WHERE THE IF-CLAUSE _*ENDS*_!
#||<--- LOOK HERE, LARRY, THIS IS WHERE THE FUNCTION _*ENDS*_!

- Ken





From bac at OCF.Berkeley.EDU  Tue Sep  3 20:41:35 2002
From: bac at OCF.Berkeley.EDU (Brett Cannon)
Date: Tue, 3 Sep 2002 17:41:35 -0700
Subject: Python-dev summary for 2002-08-16 - 2002-09-01
Message-ID: <mailman.1031253706.7156.clpa-moderators@python.org>

This is a summary of traffic on the python-dev mailing list between August
16, 2002 and September 1, 2002 (exclusive).  It is intended to inform the
wider Python community of ongoing developments.  To comment, just post to
python-list at python.org or comp.lang.python in the usual way. Give your
posting a meaningful subject line, and if it's about a PEP, include the
PEP number (e.g. Subject: PEP 201 - Lockstep iteration) All python-dev
members are interested in seeing ideas discussed by the community, so
don't hesitate to take a stance on a PEP if you have an opinion.

This is the first summary written by Brett Cannon.
Summaries are archived no where at the moment.  =)   They will be, though,
so stay tuned for the URL in future summaries.



================
Type Categories
================
This VERY long thread was sparked by Andrew Koenig asking if a discussion
of making type categories more explicit had ever occured (Andrew meant for
category to mean "the set of all types that implement a particular marker
interface").  As Andrew later pointed out, he was asking about  "a way of
making notions such as 'file-like object' more formal and/or automatic".
The discussion quickly started using the term interface to mean defining a
way to specify that an object implemented certain methods (think of it in
terms of Java's 'implements' mechanism).

Once definition was out of the way, the discussion took off.  Zope's
implementation was pointed out (
http://cvs.zope.org/Zope3/lib/python/Interface/ ) very quickly.  PEP 245
(Python Interface Syntax) was also brought to the attention of the list.
The idea of using inheritance to handle interfaces was brought up.  Guido
said that he hasn't "given up the hope that inheritance and interfaces
could use the same mechanisms.  But Jim Fulton, based on years of
experience in Zope, claims they really should be different" in terms of
how interfaces should be handled in objects.  Jeremy Hylton tried to
channel Jim's opinion by pointing out that "We'd like to use interfaces to
make fairly strong claims.  If a class A implements an interface I, then
we should be able to use an instance of A anywhere that an I is needed."
But "the inheritance mechanism is too general" because if a class A
implements interface I and then a class B, which does not implement I,
subclasses class A we end up with a class B that claims it has a certain
interface which it doesn't actually have.  Guido understood the point, but
still thought inheritence could be used "if there was a way to "shut off"
inheritance as far as isinstance() (or issubclass())" is concerned.  Guido
asked the simple question, "Why do keep arguing for inheritance?  (a) the
need to deny inheritance from an interface, while essential, is relatively
rare IMO, and in *most* cases the inheritance rules work just fine; (b)
having two separate but similar mechanisms makes the language larger."

Samuele Pedroni asked that any implementation "allow also for refering to
anonymous super-interfaces of an interface in terms of the interface plus
a subset of its signatures, also e.g. FileLike and just 'write'.  [that
means an interface can be thought to correspond to a set of
(tag,signature) tuples, where tag identifies the interface, and one can
also just consider subsets of it]".

The thread has finally seemed to have stopped (for now) with Guido saying
he is mulling the whole thing in the back of his head.  This is a very
sticky topic because of the number of design decisions required and how it
might change the way people program in Python.

There was also a partial sub-thread in this whole discussion about
multimethods; basically a way to do overloading of methods based on
parameter signature.  Most of the discussion was over syntax and such and
how to handle resolution order.  It then seemed to go to the wayside when
the main part of the thread took over again.

==============================
type categories -- an example
==============================
This thread was starteed when Andrew Koenig said that the reason he
brought up his type category question was because he wanted a way so as to
be able to identify members of a type easily.  He now had an example in a
program he was writing where what the type of the argument was varied and
thus what needed to be done to the data changed accordingly.  Jermey
Hylton suggested the isinstance(obj, type(re.compile(''))) idiom.  Andrew
asked if this was guaranteed to work, which Jeremy said no.  I asked why
this was not guaranteed, and Frederick Lundh said because re.compile() is
a factory fxn and it is possible that a future version could return a
different object based on the pattern.

===============================================
Python build trouble with the new gcc/binutils
===============================================
Andrew Koenig said that he couldn't compile Python using the newest gcc
(this was the day after the latest release hit servers).  With help from
Zack Weinberg of Code Sourcery (who also recently rewrote the tempfile
module), the problem was tracked down to binutils 2.13. being the culprit
and was not Python's fault.

===================================
Last call: mortal interned strings
===================================
The patch http://python.org/sf/576101 removes the default immortality of
interned strings.  I believe it was in early August (possibly spilled over
from late July) when Oren Tirosh proposed the idea and wrote the above
mentioned patch.  There had been some discussion over whether any 3rd
party code was reliant upon interned strings being immortal; none was
found (MacPython was reliant upon it, but since it is under Python core
control it was considered a moot point since it could be changed).  It has
been checked in.  With the patch the way to make a string immortal is to
call PyString_InternImmortal(); no code in the core uses this function.

=====================================
PEP 218 (sets); moving set.py to Lib
=====================================
Thanks to Greg Wilson (for writing the PEP), Alex Martelli (for writing
the module initially), Guido (for refactoring Alex's code), Raymond
Hettinger (for writing the docs and playing with the code himself), and
Tim Peters (for some speedup code) the stdlib has now gained a sets
module.  It has both the notion of mutable and immutable sets (the latter
used when you have a set of sets).  There was discussion about how sets
should print (sorted or not; unsorted was chosen) and what operators
should be overloaded for working on sets (| and & were chosen).  The
module is a beautiful chunk of code and I highly recommend reading its
source.

===========================================
A few lessons from the tempfile.py rewrite
===========================================
Zack Weinberg, after rewriting the tempfile module, brought up three
points:
1) Lack of dummy threads, 2) lack of a pthreads_once equivalent, and 3)
lack of a way to skip tests from unittest.py via some built-in method.
Guido responded accordingly: 1) since some code uses the idiom of trying
to import thread and catching the exception if it fails, Guido said he
would be willing to accept a dummy_thread.py that would allow:

try:
    import thread as _thread
except ImportError:
    import dummy_thread as _thread

to work.  No word on whether this is being written at the moment.  2)
Guido said the method was, in his opinion, overkill.  He said to "be
Pythonic, live dangerously, accept the risk that a ^C can screw you.  It
can anyway. :-)".  And as for 3) Guido deferred Zack to the PyUnit list
and Steve Purcell since Python just tracks Steve's code (pyunit.sf.net).
Guido's suggestion was to stick code that was reliant on some other code
in a separate testing suite that is only run when the reliant code is
available.

===========================
Standard datetime objects?
===========================
Kevin Jacobs asked what stage the new datetime object was at.  Guido said
it is in python/nondist/sandbox/datetime/ in CVS which also has comments
pointing to a wiki containing the current work on it.  Fred L. Drake, Jr.
is working on the C re-implementation and has now checked it into the
sandbox in CVS.

===================
PEP 269 versus 283
===================
Jonathan Riehl noticed that PEP 283 said PEP 269 was dead; not good
considering he was close to having a patch for PEP 269 (pgen module to
interface with the C version).  Guido said he will revive the PEP.  The
patch has since been put on SF at http://python.org/sf/599331 .

==============================
What is a backport candidate?
==============================
Since Python 2.2 is going to be around for a long time, the question was
brought up of what constitutes code that should be backported.  Guido made
the following three points:

1) code trivial to backport should always be backported

2) code patching 2.3 code should obviously not be backported

3) 2.2 code requires changes to use patch, but applies; gradients of this
exist.

So please, when submitting patches, mention whether you think the patch
should be backported to the 2.2 tree and any possible dependencies it
might have in a backport.

=================================
python/nondist/sandbox/spambayes
=================================
In response to Paul Graham's spam filter written using Baye's Rule
(Slashdot post on it is at
http://developers.slashdot.org/article.pl?sid=02/08/16/1428238&tid=156 ),
a
thread spawned around this checkin of code that followed that paper's
suggestions.  This thread quickly jumped into discussions on data
structures, Baye's Rule, and a whole lot of talk about spam.  Very
interesting if spam filtering interests you.  Tim Peters has been leading
the drive on this chunk of code (and thanks to his illness that befelled
him in late August which he has subsequently gotten over he had a few days
of major hacking on it; Tim showed he is a performance stats whore
<wink>).

A very cool quote came out of this thread from Eric S. Raymond when
discussing the spam filter he has been working on: "This is actually the
first new program I've coded in C (rather than Python) in a good four
years or so".

====================
Parsing vs. lexing.
====================
In response to a question by Aahz about what the differences were between
a lexer, parser, and tokenizer, Eric Raymond posted a good overview of the
differences.  Guido later commented in an email mentioning SPARK and about
how Python's lexer (pgen) works and why he wrote it.  He also made some
other comments on lexers.  Jeremy Hylton pointed out a "neat new paper
about an old algorithm for recursive descent parsers with backtracking and
unlimited lookahead" by Bryan Ford at http://www.brynosaurus.com/pub.html
.  Alex Martelli pointed out that this discussion reminded him of "a
long-ago interview with Borland's techies"  in which they said they were
able to make Borland PASCAL fit on a floppy while MS PASCAL took multiple
floppies.  Their trick was "we just did everything by the Dragon Book --
except that the parser is a hand-written recursive descent parser [Aho &c
being adamant defenders of Yacc & the like], which buys us a lot".
Someone named Noah also emailed a discussion on lexers and parsers pulling
in Finite State Machines, Push Down Autonoma, and Turing Machines in his
discussion.

Martin Sj?n says that Haskell's pattern matching and lazy evaluation makes
lexers easy (even a Recursive-Descent parser), but unfortunately Haskell
does not play with other languages nicely.  Haskell is where Python got
it's list comprehension idea.

=========================================
[Python-Dev] Fw: Security hole in rexec?
=========================================
It was brought to the attention of the list that deleting __builtins__
allowed a compromise in rexec.  Guido pointed out that
http://python.org/sf/577530 reports this.  He also said don't trust rexec.

A patch was submitted and checked in to document the view that rexec is
really not that safe.

=================
A `cogen' module
=================
Francois Pinard asked about Cartesian products using the new sets module.
Guido didn't think people would in general need it.  Francois quickly
started this thread of discussing a cogen module to generate Cartesian
products and other ways of operating on sets.  The thread quickly died
when Tim Peters posted "his elaborate state-of-the-art code", as Guido
called it.  But Francois said he would be back for more discussion on
this.

=================
Mersenne Twister
=================
Raymond Hettinger volunteered to implement the Merseene Twister algorithm
(one in Python exists at http://www.math.keio.ac.jp/~matumoto/emt.html).
While discussing to implement in C or Python, Guido noticed that
random.Random re-implements whrandom.  Guido then came up with the idea of
writing a base random class that is subclassed where .random() can be
implemented;  Tim Peters agreed and suggested more methods to subclass.

=================================
New PEP Format: reStructuredText
=================================
David Goodger and Barry Warsaw have now gotten reST as a usable syntax for
PEPs.  Read the PEPs on the subject to learn more:

- PEP 12 -- Sample reStructuredText PEP Template
  (http://www.python.org/peps/pep-0012.html)

- PEP 258 -- Docutils Design Specification
  (http://www.python.org/peps/pep-0258.html)

- PEP 287 -- reStructuredText Docstring Format
  (http://www.python.org/peps/pep-0287.html)

It has been suggested that the summaries try using reST; I am considering
it.

====================================
tiny optimization in ceval mainloop
====================================
Jeremy Hylton noticed that in ceval that their is a test of whether the
ticker was 0 or if things_to_do was set to true (explanation of the
ticker, checkinterval, and the GIL follow this paragraph).  Jeremy
wondered if we could just drop the ticker to 0 when things_to_do is true.
Jack Jansen, though, pointed out that clearing it is not guaranteed since
there may be an interrupt routine when "we fiddle things_to_do".  Skip
Montanaro then pointed out that since neither ticker nor things_to_do is
fiddled with unless the GIL is held that instead of causing each thread to
execute this test that they could be made globals instead; he did a patch
that implements this ( http://python.org/sf/602191 ).  Guido then said
that if there wasn't a decent speed improvement, then no patch would be
checked in.  He then changed his mind when it was pointed out that it
actually simplified the code.  Skip tested anyway, though, and there is a
speed improvement.  This also brought up whether the default value of 10
for checkinterval was reasonable.  It was then agreed to be bumped up to
100.  Jack ran some code and said he noticed a definite improvement.

Python's version of threading is not like in C.  There is something called
the GIL (Global Interpreter Lock) which any thread wishing to execute
Python code or play with Python objects must hold.  This means that when
you have Python threads running (using the thread or threading module)
they are usually all waiting in line to get the GIL.  Now for Python to
decide when to release the GIL for another thread to grab it, it uses the
ticker.  This variable counts down to zero by being decremented every time
a Python opcode is executed (originally defaulted to 10, now defaulted to
100).  The ticker's starting value after each release of the GIL is what
sys.checkinterval() sets.

To get a better understanding of threading under Python I recommend
reading Aahz's tutorials on threading.





From mcfletch at rogers.com  Wed Sep 25 23:18:12 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Wed, 25 Sep 2002 23:18:12 -0400
Subject: Is there a way to find the name of the method currently being
 executed?
References: <3D9278D0.9060307@arielpartners.com>
Message-ID: <3D927C74.1040107@rogers.com>

This is a massive hack, and I'm not sure if it covers the corner cases, 
but oh well :),

PythonWin 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32.
Portions Copyright 1994-2001 Mark Hammond (mhammond at skippinet.com.au) - 
see 'Help/About PythonWin' for further copyright information.
 >>> import traceback
 >>> class a:
... 	def func1( self ):
... 		return traceback.extract_stack(limit=1)[0][2]
... 	def func2( self ):
... 		return traceback.extract_stack(limit=1)[0][2]
... 	
 >>> a().func1()
'func1'
 >>> a().func2()
'func2'
 >>>

HTH,
Mike

Craeg K Strong wrote:
> Hello:
> 
> class a:
> 
>    def func1(self):
>        print "func1 being called"
> 
>    def func2(self):
>        print "func2 being called"
> 
> In the code sample above, I would like to replace the "func1"
> and "func2" strings with constants so the same code could
> be copied verbatim into multiple functions and do the right
> thing.
> 
> Is there a way to do this in Python?  For example, maybe you could
> get your hands on the call stack and print out stack[0].__name__
> or something....
> 
> Thanks in advance,
> 
> --Craeg
> 
> BTW, I could not find this in the python.org FAQ....
> 
> 
> 
_______________________________________
   Mike C. Fletcher
   Designer, VR Plumber, Coder
   http://members.rogers.com/mcfletch/





From Hobbes2176 at yahoo.com  Thu Sep 19 15:43:25 2002
From: Hobbes2176 at yahoo.com (Mark)
Date: Thu, 19 Sep 2002 19:43:25 GMT
Subject: Minimal debug/rep functionality
References: <mailman.1032406044.16132.python-list@python.org> <3d89b2e9$1@news.sentex.net>
Message-ID: <xNpi9.16332$t%6.1471@nwrddc02.gnilink.net>

Peter Hansen wrote:
> True... presumably either the OP is not concerned about standard
> library modules anyway, or he would check the archives and add
> a simple way of doing this.  YAGNI...

Correct, I'm currently trying to figure out how to ignore them (ie have it 
step "over" instead of "into" standard library calls (in particular, the 
linecache.getline(foo, x) call).

Also, I had quite a time figuring out the damn name for what I was trying to 
do.  My first thought was implementing a REP loop of my own.  Then I 
started looking at the debugging code.  Finally, I was put on the correct 
trail by looking at "tracing".

> Hmmm... I didn't see any implication that speed was even remotely
> a concern for the OP.  After all, he did ask to *print* each line
> as it executed.  What could possibly be slower than that?
> (I doubt he would care to deal with the volume of information
> generated when using this with a large program anyway... needs
> a way to turn it on and off, at least.)

Yup, for now I simply need to look at traces.  Speed is not important (1. 
this is a prototype and 2. speed may never be important).  Further, I'm 
only looking at single function (aside from library function calls) 
algorithms for the moment.

Thanks for the input.

Regards,
Mark


From thisisabogusemail at bogus.thisisabogusdomain.com  Tue Sep 10 12:26:50 2002
From: thisisabogusemail at bogus.thisisabogusdomain.com (Jack D. Hammer)
Date: Tue, 10 Sep 2002 16:26:50 GMT
Subject: Larry Wall's comment on python...
References: <mailman.1031579787.7378.python-list@python.org>
Message-ID: <3D7E1E28.5040709@bogus.thisisabogusdomain.com>

One good argument for having block delimiters is for embedded situations like 
ASP.  One can conditionally put code into a block as such (ASP is rusty):

<%
		if (some_condition) {
%>
	<tr>
		<td>some data</td>
		<td>some more data</td>
	</tr>
<%
	} // closing curly
%>

I would think that for embedding python, having block delimiters would be 
beneficial.

brueckd at tbye.com wrote:
> On 8 Sep 2002, Brian Slesinsky wrote:
> 
> 
>>He's repeating the standard newbie complaint.  However, I think the
>>fact that so many people have repeated this for so long is
>>significant.  Long-term Python programmers tend to forget how weird it
>>is to some newcomers to end a function or class just by *stopping*. 
>>It feels abrupt and rather fragile, particularly when you don't have a
>>firm grasp of the language.  If newbies have complained for this long,
>>the complaint's not going away, and if we're serious about making
>>Python easy to pick up it might be good to listen.
> 
> 
> Yes, the fact that so many newbies complain about it is very significant.  
> Perhaps we need to do a better job of presenting the concept. For example,
> instead of tossing it out there and then defending against complaints
> perhaps it would be better to take the offensive: first make a case for
> why the "normal" way to delimit blocks has its drawbacks, and then present
> the Python way as a solution to a problem.
> 
> But just because people consistently complain about something does not
> mean the answer is to bow to their wishes. From an arrogant point of view,
> they're newbies, so what do they know? ;-) Seriously though, I enjoy
> Python because it _is_ different - and I usually derive benefit from the
> ways in which it is different (i.e. - it's not different for arbitrary or
> random reasons). It's significant that lots of newbies complain about "no
> curly braces", but it's as least as significant that once people use it
> for awhile they tend to end up liking it.
> 
> 
>>Of course it's unnecessary and should be entirely optional.  Putting
>>an "end" after every block would certainly be overly verbose and bad
>>style.  But ending major blocks in a way that's syntax-checkable might
>>catch the occasional indentation error and/or make error messages
> 
> 
> Well, as you know you can always do an "# end" comment which really gives 
> much of the same benefit. At the bottom of my Emacs window it tells me 
> which block I'm closing (e.g. "Closes block: ...try:"), other editors 
> probably do the same. Also, it seems like whenever it's not readily 
> apparent which block I'm in, then the _real_ problem is sloppy code on my 
> part.
> 
> -Dave
> 
> 




From tdelaney at avaya.com  Thu Sep 19 20:47:54 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 20 Sep 2002 10:47:54 +1000
Subject: Minimal debug/rep functionality
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5B9@natasha.auslabs.avaya.com>

> From: Peter Hansen [mailto:peter at engcorp.com]
> 
> I'm a commercial Python user as well, so I can imagine the kinds of
> things your employer think about whether to release or not.  We have

As I said, I *expect* to be able to release (there's no commercial reason
not to, and we have done so with such things in the past).

Tim Delaney



From gurubert-dated-1032624215.ifbmbckh at epigenomics.com  Wed Sep 25 08:19:10 2002
From: gurubert-dated-1032624215.ifbmbckh at epigenomics.com (Robert Sander)
Date: Wed, 25 Sep 2002 12:19:10 +0000 (UTC)
Subject: ANN: Python FAM
Message-ID: <mailman.1032957124.29556.clpa-moderators@python.org>

Hi!

I would like to announce release 1.0.1 of Python FAM.

http://python-fam.sourceforge.net/

Python FAM is a port (or wrapper module) of libfam from the File
Alteration Monitor project at http://oss.sgi.com/projects/fam/.

The included test.py does everything test.c++ (the original example)
does, so I am quite confident that it works. ;-)

This is my first adventure in the large world of extending Python with
C modules. Therefore feedback is very appreciated.

Greetings
-- 
Robert Sander
Manager
Information Systems        www.epigenomics.com        Kastanienallee 24
+493024345330                                              10435 Berlin



From SBrunning at trisystems.co.uk  Thu Sep 12 10:26:27 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 12 Sep 2002 15:26:27 +0100
Subject: Best tutorial
Message-ID: <31575A892FF6D1118F5800600846864DCBD79B@intrepid>

> From:	Jordan Faris [SMTP:jfaris at sbcglobal.net]
> 
> Of the several fine tutorials on the Web, which one would be the most
> direct
> in getting a novice up to snuff on Python?
 
A novice programmer, or novice with Python?

For the former, I'd recommend How to Think Like a Computer Scientist -
<http://www.ibiblio.org/obp/thinkCSpy/>, and for the latter, Dive into
Python - <http://diveintopython.org/>.

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




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



From tjreedy at udel.edu  Mon Sep 23 09:12:32 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 23 Sep 2002 13:12:32 GMT
Subject: How to call a procedure in exe file
References: <mailman.1032767369.29784.python-list@python.org>
Message-ID: <4rEj9.225917$z91.9596261@bin3.nnrp.aus1.giganews.com>

"A" <printers at sendme.cz> wrote in message
news:mailman.1032767369.29784.python-list at python.org...
> from Python but is it possible to call a procedure in an exe
program?
> I have a program that downloads some files( that was written in
Delphi,
> we bought that  program compiled already in exe).
> I must start the program and then push a button that runs a
procedure for downloading.
> And I must do it al the time when I want to download a file. So I am
thinking about
> downloading automatically from Python. I can start the program
> with os.startfile but I do know know how to start the procedure
> that is started when I push
> the button for downloading.(I may possiblly find address of the
procedure

Lots of IFs -- in my order of preference:

IF the program has a command-line option to download without a 'button
push' (ie, --download <list of files>), use that.  IF is doesn't, ask
that such be added if you can.  (Utility programs ought to have such
options just so they can be used programmatically without operator
present, just like you want to do.)

IF you can start download with keys instead of just mouse, and if
program takes keypresses through generic standard input functions
(like C getchar()) rather than specific getkey function, start it at
the other end of a pipe.

IF program has COM interface, get PythonWin stuff and use that.

IF you can find (or write or modify open-source code to make) a
mouse-click simulator program thats responds to dynamic input via
stdin (rather than only responding to prewritten script read as
startup), use that.

Terry J. Reedy




From martin at v.loewis.de  Sat Sep 21 06:51:05 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 21 Sep 2002 12:51:05 +0200
Subject: Solaris 64 bit compilation
References: <c2595393.0209201844.2ed62265@posting.google.com>
Message-ID: <m3sn03k44m.fsf@mira.informatik.hu-berlin.de>

phoebe_1 at att.net (Holden Caulfield) writes:

>  This is my first attempt at compiling Python v2.2.1 in *64bit* mode 
> on a Solaris 5.9 system. I am using the Sun Compiler (the latest I do not
> remember the version). I am migrating a few of my apps (all developed on
> 64bit platforms) to Solaris, so I *need* the 64bit mode.


I can't follow your conclusion - if you migrate applications which
have been developed on 64-bit platforms, there is no inherent need to
compile them as 64-bit binaries on Solaris 9.

> But, it is failing miserably. It compiles *fine* in 32bit mode (the default),
> but it does me no good. As some of the extensions library that I have are
> 64bit. Before, I start hacking I thought I would give a query here and see
> if anyone had any success?

It compiles fine for me, with SunPRO 6U1.

> 
> Here are my preliminary results:
> - Flags used "-xarch=v9 -xcode=pic32".
> The first thing I get is the bail-out in "pyport.h":
>  Where the #LONG_BITS != (8 * sizeof(LONG)) fails!!!

Please try to report failures carefully. It does not say sizeof(LONG),
instead, it says SIZEOF_LONG. Which gives rise to the question: What
value does SIZEOF_LONG have, in pyconfig.h?

> So, I commented it out, first to proceed and see what happens. Most of them
> compiled fine except, I got some serious errors in the:
> fileobject.c file probably related to the LP64 model.
> 
> I will try to add more info, when I get around to it.

More info would be required if you want help. Start with reporting the
error messages as the compiler produces them, instead of interpreting
them, and reporting your interpretation.

Regards,
Martin


From rolf.kalbermatter at citeng.com  Tue Sep 17 06:14:55 2002
From: rolf.kalbermatter at citeng.com (Rolf Kalbermatter)
Date: Tue, 17 Sep 2002 12:14:55 +0200
Subject: More on Protecting Source Code
In-Reply-To: <GCEDKONBLEFPPADDJCOEMEEBFAAA.whisper@oz.net>
Message-ID: <EKEJLGEFHHFGJOBEGJECEEFECFAA.rolf.kalbermatter@citeng.com>

> Well, I think of it this way: machine coded binaries are more
> like a 128 bit
> key and Python is more like a 40 bit key. I agree that nothing is
> safe from
> reverse engineering; it's a matter of how much pain and money it takes to
> unscrut the inscrutable. Python could do better - how I'm not
> totally sure.

Personally I feel that getting at the idea of (parts) of a 32 bit PE
executable code in the environments I usually work in is easier for me
than finding out what the Python byte code in a compiled module would
do ;-)

That left beside I really wonder if Python should cater to people feeling
their development is so valuable that everybody wants to steal it in huge
masses and ignore the existing copyright laws. The incidential hacker who
looks at the code and may get a good idea is probably much less dangerous
in terms of profitloss than the average user posting its license key or
whatever in some public forum so that everybody may be installing and
using the software illegal. Obfuscating your code will maybe cause the
hacker a few hours extra work but neverhtless not stop him to find out
eventually what you have done. On the other hand it will not stop the
average user to post his license keys in any way or just copy the pirated
software, but most probably even promote it.

I also believe that the very nature of Python being an Open Source project
doesn't lend itself naturally to usage in a product from someone being kind
of paranoic and overly concerned with obscure protection of his ideas. And
I'm sure that the time spent to try to make Python more obfuscating its byte
code can be spent much more productive in improving Python itself and adding
new features and extensions instead.

The decision in the Python developer community to not spend a lot of efforts
in Python to obfuscate its code may lead to less commercial usage of Python,
that is true. However I wonder if that is a problem at all as it seems to me
that Python has a very active developer and user community already. Enough
anyhow to guarantee that it will be around for many years to come.
If this community decides that the addition of code obfuscation does not
justify the possible gain in usage by commercial entities who are most happy
to profit from the Python efforts but being the paranoids they are, also
quite
likely to not contribute back to Python in some ways, it is their full
right!

>> Anyway, good luck!  If your software is any better than your protection
>> scheme I'll download it from http://www.warez.com when you release it ;)
>
>I'd like it: it would means my software it's worth to crack.

Which again comes down to the point: If it's worthful it will be cracked
anyhow eventually, and otherwise why bother?

Rolf Kalbermatter




From duduca007 at yahoo.com  Tue Sep  3 16:31:02 2002
From: duduca007 at yahoo.com (Carlos Moreira)
Date: Tue, 3 Sep 2002 13:31:02 -0700 (PDT)
Subject: Distributed Shared Memory
Message-ID: <20020903203102.88032.qmail@web11401.mail.yahoo.com>

I need a Python bind for some lib that supports
"Distributed Shared Memory". Is there?

        []'s

Cadu Moreira
KaduSoft President
--------------------------
[...yes I'm a Sith...]

__________________________________________________
Do You Yahoo!?
Yahoo! Finance - Get real-time stock quotes
http://finance.yahoo.com



From tmohr at s.netic.de  Tue Sep 17 17:40:34 2002
From: tmohr at s.netic.de (Torsten Mohr)
Date: Tue, 17 Sep 2002 23:40:34 +0200
Subject: OpenGL Bindings for Tcl/Tk 8.4
Message-ID: <am87gi$o9$1@schleim.qwe.de>

Hi,

i just downloaded and installed Tcl/Tk 8.4 and PyOpenGL-2.0.44.

Sadly it doesn't compile, though i edited "togl.c", as mentioned
in the #error directive.

Also the latest CVS snapshot doesn't compile.


Has anybody got some patches/hints that will help me compiling
and installing OpenGL bindings for python?


Best regards,
Torsten.



From duncan at NOSPAMrcp.co.uk  Fri Sep 13 09:01:15 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Fri, 13 Sep 2002 13:01:15 +0000 (UTC)
Subject: wrapping yield ?
References: <alqehl$9qv$1@nntp0.reith.bbc.co.uk> <Xns92885F082A21Cduncanrcpcouk@127.0.0.1> <alsdpt$2nq$1@nntp0.reith.bbc.co.uk> <Xns9288759449799duncanrcpcouk@127.0.0.1> <alsg06$3e2$1@nntp0.reith.bbc.co.uk>
Message-ID: <Xns92888DEA3BB2duncanrcpcouk@127.0.0.1>

"Michael Sparks" <Michael.Sparks at rd.bbc.co.uk> wrote in
news:alsg06$3e2$1 at nntp0.reith.bbc.co.uk: 

> "Duncan Booth" <duncan at NOSPAMrcp.co.uk> wrote in message
> news:Xns9288759449799duncanrcpcouk at 127.0.0.1...
>> The syntax wouldn't look quite so bad if your function wasn't
>> returning a value.
> 
> Problem is, it would in most situations, also I'm trying to keep the
> number of generators to a minimum whilst a lower overhead than things
> like threads and processes if I had several 10s of thousands going,
> keeping throughput up may become an issue. (Hence the reason for
> wanting the functions called part of the same generator rather than a
> separate, extra one) 

Don't forget that iterating with a generator is *much* more efficient than 
calling a Python functions thousands of times.

Python has a big overhead on each function call, but you only pay that 
overhead when you first call a generator and the subsequent iterations have 
virtually no overhead at all. 

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From gerhard.haering at gmx.de  Tue Sep 17 02:46:48 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Tue, 17 Sep 2002 08:46:48 +0200
Subject: using python as cgi on apache unix
In-Reply-To: <am6fm301477@enews2.newsguy.com>
References: <am6fm301477@enews2.newsguy.com>
Message-ID: <20020917064648.GB10169@lilith.ghaering.test>

* josh_py at sautez.com <josh_py at sautez.com> [2002-09-16 22:48 -0700]:
> All,
> 
> I have developed a python solution to run and connect to a mysql database.
> I did all my development under windows apache 1.3.26 and python 2.2.1.  I
> know want to move this to my unix server.  Unfortunately, no matter what I
> do I can not get apache on my linux box to execute any python scripts.  I
> get the following errors in apache:
> 
> Premature end of script headers: /var/www/html/cgi-bin/test.py
> 
> I have all the settings in apache I believe to be correct.  It complains
> that the file is not executible if that is not set.  Other CGI works fine.
> 
> test.py looks like:
> 
> #!/usr/bin/python2
> 
> print "\n\n"
> print "Content-Type: text/html\n\n"
> print "hello"

The first print "\n\n" is an error. It was simply luck that this was
running anywhere else as expected. Just delete this line and your hello
world cgi script should run fine.

Also, to test wether Python CGI scripts work, you can always simply use:

#! [path to python ...]
import cgi
cgi.test()

-- Gerhard



From a at b.c  Fri Sep 27 06:54:36 2002
From: a at b.c (Tkinter stumped)
Date: 27 Sep 2002 10:54:36 GMT
Subject: TK canvas text bounding box question
Message-ID: <an1ddc$us5$0@216.39.145.102>

Perhaps I am just stupid, but I have spend significant amounts of time
banging my head against this wall.

Using the canvas widget text item (so I can have per pixel control of text
placement), I can get the bounding box for the entire text item, but would
like (and cannot seem to get) the bounding box for a charater in the text
object.

This seems like something that would have come up before (vis a vis Tkinter)
but a search of archives gave me nothing. I did find this page:
http://mini.net/tcl/1172, but if I am reading that code correctly, it seems
to be doing something like canvas.index(tagOrId, x, y) => characterIndex,
which won't work.

What am I missing (or just plain unaware of)? Am I missing something? Or is
there no way to do this short of the painful, ugly and slow brute-force
workaround that presents itself:

# x, y somewhere safely out of view
text1 = canvas.create_text(x,y, text="foo", anchor="nw", font=f)
bb_1 = canvas.bbox(text1)
text2 = canvas.create_text(x,y, text="foob", anchor="nw", font=f)
bb_2 = canvas.bbox(text2)

now subtracting bb_1 from bb_2 gives me a bounding box for 'b' in "foobar"
relative to x,y , and modulo the extra space on the end of a text item.

please tell me there is a better way to do this....




From mstenner at phy.duke.edu  Tue Sep 10 08:57:37 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Tue, 10 Sep 2002 08:57:37 -0400
Subject: Pickle problem
In-Reply-To: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220D10@hendrix.empolisuk.com>; from hst@empolis.co.uk on Tue, Sep 10, 2002 at 01:44:57PM +0100
References: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220D10@hendrix.empolisuk.com>
Message-ID: <20020910085737.A4329@phy.duke.edu>

On Tue, Sep 10, 2002 at 01:44:57PM +0100, Harvey Thomas wrote:
> Erik Price wrote:
> > 
> > On Tuesday, September 10, 2002, at 06:13  AM, Harvey Thomas wrote:
> > 
> > > I'm trying to pickle a complicated data structure which includes a 
> > > dictionary whose entries are created by (simplified):
> > >
> > > a[x] = re.compile(y).match
> > >
> > > The pickle fails, however, with the error:
> > >
> > > PicklingError: Can't pickle <built-in method match of 
> > _sre.SRE_Pattern 
> > > object at
> > >  0x00838700>: it's not found as __main__.match
> > 
> > Are you trying to store the MatchObject returned by a call to 
> > a regular 
> > expression object's match() method?  If so, it looks like you're 
> > forgetting the parens at the end of "re.compile(y).match".
> > 
> No, I haven't forgotten the parentheses. I'd like to pickle the match method of the compiled regular expression.

Are you interested in storing the "match results" or the "match
method"?  I think Erik was suggesting that you may want

a[x] = re.compile(y).match()

That will pickle the match object.  What you're doing is trying to
pickle the method (function) that gets called to return the result.

Example:

class foo:
    def bar(): return "some data"

f = foo()

pickling "f.bar" is very different from pickling "f.bar()" because
the former is a method, and the latter is a string (specifically,
"some data").

					-Michael
-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From ben at remove.Stanford.EDU  Tue Sep 17 18:28:15 2002
From: ben at remove.Stanford.EDU (Ben Escoto)
Date: Tue, 17 Sep 2002 15:28:15 -0700
Subject: Large file support and stat vs stat64 in extension modules
References: <pan.2002.09.10.16.03.39.756929.4320@remove.Stanford.EDU> <pan.2002.09.16.22.58.21.620527.4504@remove.Stanford.EDU> <m3znuhb174.fsf@mira.informatik.hu-berlin.de>
Message-ID: <pan.2002.09.17.15.28.15.278428.10231@remove.Stanford.EDU>

On Mon, 16 Sep 2002 23:05:35 -0700, Martin v. Loewis wrote:

> If Python was configured for LFS, then those defines appear in
> pyconfig.h. So including Python.h before any system headers is the
> recommend, and better, way.

Ahh, this makes the most sense.  I had just mixed up the order.  Thanks.

-- 
Ben Escoto


From achim.domma at syynx.de  Tue Sep 24 09:16:49 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Tue, 24 Sep 2002 15:16:49 +0200
Subject: Python.Net ?
Message-ID: <ampnua$nti$00$1@news.t-online.com>

Hi,

a german IT magazin (iX) mentioned a .Net version of python, developed by
ActiveState. My last information was, that there will be no .Net version
with produces IL code, but only an extension for the new Visual Studio.
Am I missing something or is the article simply wrong?

regards
Achim




From marklists at mceahern.com  Thu Sep 19 12:57:54 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 19 Sep 2002 11:57:54 -0500
Subject: strong/weak - dynamic/static [Was: Getting started]
In-Reply-To: <15753.57546.804290.415645@12-248-11-90.client.attbi.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOOEPGDPAA.marklists@mceahern.com>

[Skip Montanaro]
> That doesn't make Python weakly typed.  You have to be explicit about the
> change.

All true.  I should have been more explicit about the part of what I was
replying to that I thought wasn't true.  It was the tail end of this:

[sismex01 at hebmex.com]
> A variable, being only a "reference to *any* object"
> may refer to any object, but that object (see the difference?)
> may not change it's functionality or type during runtime.
  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

The fact that I can change the instance's __class__ demonstrates pretty
clearly that that's not the case.

Cheers,

// mark

-




From b.maryniuk at forbis.lt  Thu Sep  5 08:56:30 2002
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Thu, 5 Sep 2002 14:56:30 +0200
Subject: How do i print to a printer using python?
In-Reply-To: <al63r8$3bo$1@news7.svr.pol.co.uk>
References: <6afb9d2c.0209041044.43c5f5f1@posting.google.com> <al63r8$3bo$1@news7.svr.pol.co.uk>
Message-ID: <200209051456.30340.b.maryniuk@forbis.lt>

On Thursday 05 September 2002 01:09, G. Willoughby wrote:
> I would really like to know how to do this too, basically i want to print
> to a standard bubblejet print (non postscript) from a Tkinter Text widget,
> any ideas?

Write canvas to PS, send to the printer via LP (UNIX).

-- 
Regards, Bogdan

The advertisement is the most truthful part of a newspaper.
                -- Thomas Jefferson




From cbbrowne at acm.org  Sun Sep 29 13:06:53 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: 29 Sep 2002 17:06:53 GMT
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05>
Message-ID: <an7bvd$burft$1@ID-125932.news.dfncis.de>

Quoth "Steve Holden" <sholden at holdenweb.com>:
> <sismex01 at hebmex.com> wrote...
>> >
> [ ... ]
> OK, so don't use whole pennies (or smallest currency) as your
> base, use 10,000ths of a penny, but still use integer arithmetic
> for your calculations.
>
> Floating point has too many rounding errors to be trustworthy;
> or, looking at it from another angle, floating-point calculations
> cannot be 100% accurate because of the way it's implemented.

The problem with FP is that FP involves the use of binary
approximations, which commonly cannot exactly represent decimal
values.

Notably, the decimal fraction 0.1, which is /exactly/ 1/10, does not
have any exact encoding in binary.  Its representation in binary is a
repeating binary fraction.

When the (not unreasonable!) expectation is that computers ought to be
able to deal /exactly/ with plain simple decimal values turns out to
be wrong, that ought to lead people to think.  

But all too often it doesn't, as we get the typical reaction "Well,
IEEE FP values can be sufficiently exact for this range of numbers;
just use them right!"

There's thirty years of experience now of programmers being able to
use COBOL and PL/1 to manipulate decimal values without having to go
into the numerical analysis effort to "just use FP right!"

The "FP way" is that once in a while, you've got to insert operations
to round things appropriately.  The "old school" folks can
legitimately gripe that they had decimal types that never required any
of that nonsense, and that they had those types before many of the
modern developers were even born.
-- 
(concatenate 'string "aa454" "@freenet.carleton.ca")
http://cbbrowne.com/info/linuxxian.html
Rules of the Evil Overlord #195. "I will not use hostages as bait in a
trap.  Unless  you're going  to use them  for negotiation or  as human
shields, there's no point in taking them."
<http://www.eviloverlord.com/>


From andrew.thompson at ashecastle.com  Tue Sep 24 05:11:23 2002
From: andrew.thompson at ashecastle.com (Andrew Thompson)
Date: Tue, 24 Sep 2002 10:11:23 +0100
Subject: instance creation
In-Reply-To: <265e380f.0209240052.4c2988f3@posting.google.com>
Message-ID: <014801c263aa$5bf5f0f0$4802a8c0@blair>

Dear Paul,

One simple way to do this is to define your single __init__ method to be
polymorphic, switch on the type of the first argument, and call the
appropriate method to complete the initialization.

'
class foo:
	def __init__(self, arg1 , arg2=None):
		if type(arg1)==type(string):
			self.init2(arg1,arg2)
		else:
			self.init3(arg1)

	def init2(name,details):
		blah blah

	def init3(id):
		blah blah

'



-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]
On Behalf Of Paul MG
Sent: 24 September 2002 09:52
To: python-list at python.org
Subject: instance creation

Hi

I am dabbling with Python and have run into a bit of an oddness. I
have a class Appointment, with fields Title and Details. I want to
allow two means of construction:

1) Providing Title and Details strings explicitly, to create a new
instance.
2) Providing an ID, which causes that Appointment to be loaded from a
backing store.

Now in C++ or smalltalk, I would do one of

- Provide two constructors:

    def __init__(self, name, details):
      self.name = name
      self.details = details

    def __init__(self, id):
      load(id)

    def load(self, id):
      i = open("appointment"+str(id))
      self = pickle.load(i)

- Provide a class method (creation method) to instantiate:

    def create(id):
      i = open("appointment"+str(id))
      return pickle.load(i)

However, much of this seems not to work.

- Having two constructors seems to be illegal - or at least the second
overwrites the first.
- Assigning to 'self' in the load method seems to really not work!
- Putting the create() method outside of the class works fine - but I
want it to be a class method on Appointment, goddammit!:) Is there no
way to do this?

Any help would be appreciated, basically i am looking for the python
idiom to do this, as my C++ and Smalltalk idiom's clearly aren't in
the grain of the language!

Thanks very much
Paul MG
-- 
http://mail.python.org/mailman/listinfo/python-list




From bogusdrop at myself.com  Tue Sep 24 02:42:56 2002
From: bogusdrop at myself.com (TuxTrax)
Date: 23 Sep 2002 23:42:56 -0700
Subject: cursor positioning and color
References: <slrnaoqv2q.hn4.TuxTrax@fortress.tuxnet> <1b6dca50.0209231022.6fae7f6d@posting.google.com>
Message-ID: <30770aa4.0209232242.1e91d9d5@posting.google.com>

bfarwick at juno.com (brrrent) wrote in message news:<1b6dca50.0209231022.6fae7f6d at posting.google.com>...
> TuxTrax at fortress.tuxnet.net (TuxTrax) wrote in message news:<slrnaoqv2q.hn4.TuxTrax at fortress.tuxnet>...
> > Being new to python, I would like to know how one goes about such 
> > tasks as precise cursor positioning and output to screen in color.
> > 
> 
> Hi Mathew,
> 
> Curses is the way to go when using Linux/Unix.
> 
> Wconio is the way to go when using Microsoft.
> You can find in the Vaults of Parnassus.  If you have any trouble
> locating it, let me know, I can email it to you.  Small library, works
> well.
> 
> All the best,
> 
> brrrent

Thanks Brent. Appreciate it. Am I to understand that there are no
python libraries that do this as a standard part of the python
distribution?

That would be preferable, as it would eliminate the requirement of
making
the user have curses on their system before they can use the code.

Cheers,

Mathew


From see_reply_address at something.invalid  Wed Sep 18 20:02:59 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Thu, 19 Sep 2002 12:02:59 +1200
Subject: READ THE ONLY SITE THAT HAS THE GUTS TO CALL FOR THE MURDER OF JEWS AND BLACKS
References: <am9grc$bhs$12304@news1.kornet.net>
Message-ID: <3D891433.8050909@something.invalid>

Hal.Turner at HalTurnerShow.com wrote:

> She wants to taste 
> easy tyrants behind Pearl's corner.


Assuming this is a typo for "Perl's corner"... if
Guido is a benevolent dictator, could Larry be
considered an easy tyrant?

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From mnations at airmail.net  Thu Sep 12 17:25:31 2002
From: mnations at airmail.net (Marc)
Date: 12 Sep 2002 14:25:31 -0700
Subject: Expect() from telnetlib giving me problems.
Message-ID: <4378fa6f.0209121325.130766fc@posting.google.com>

Hi,

This is the error I'm getting:

Exception in thread Thread-3:


Traceback (most recent call last):
  File "C:\Python22\lib\threading.py", line 408, in __bootstrap
    self.run()
  File "C:\Python22\lib\threading.py", line 396, in run
    apply(self.__target, self.__args, self.__kwargs)
  File "main.py", line 85, in run
    command[0](*command[1:3])
  File "key.py", line 94, in command
    read_result(conn)
  File "key.py", line 104, in read_result
    buffer = conn.expect(alist, 5)
  File "C:\python22\lib\telnetlib.py", line 541, in expect
    list[i] = re.compile(list[i])
  File "C:\Python22\lib\sre.py", line 178, in compile
    return _compile(pattern, flags)
  File "C:\Python22\lib\sre.py", line 228, in _compile
    raise error, v # invalid expression
error: nothing to repeat


I've tried several different ways to get the expect command to work. I
can't find any examples and only limited documentation. Here is how
I'm trying to implement it:
 

    alist = [ 'COMPLD', '*/' ]
    buffer = conn.expect(alist, 5)

 
Can someone please tell me what is wrong or provide me with some
working examples of using the command.

Thanks,

Marc


From thoa0025 at mail.usyd.edu.au  Sun Sep 15 23:40:47 2002
From: thoa0025 at mail.usyd.edu.au (:B nerdy)
Date: Mon, 16 Sep 2002 03:40:47 GMT
Subject: possibly OT - delete char?
Message-ID: <3pch9.10253$Ee4.26056@news-server.bigpond.net.au>

im making a console app. i wish to clear the line. ive got 2 ideas

- does python have a clear line method to clear the current text of line?
- is there a character which deletes the previous printed character when
printed?

cheers




From phr-n2002b at NOSPAMnightsong.com  Fri Sep 20 20:33:46 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 20 Sep 2002 17:33:46 -0700
Subject: Decimal arithmetic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amabpb$5rm$1@news.nems.noaa.gov> <3D8B1537.4010803@thinkware.se> <ad052e5c.0209201128.1a53108@posting.google.com> <amgcgf$5e1fd$1@ID-125932.news.dfncis.de>
Message-ID: <7xn0qcfafp.fsf@ruckus.brouhaha.com>

Christopher Browne <cbbrowne at acm.org> writes:
> The COBOL (or PL/I, or anyone else accustomed to BCD data types) guy
> will justifiably look at the bizarro FP rounding as being, well,
> bizarre.
> 
> There's _no_ good reason for subtracting two perfectly good decimal
> values from one another to result in having weird artifacts appear in
> some 17th decimal place, aside from having a willingness to admit
> fairly deranged arithmetic.

Um, put $1000.00 in an imagined bank account and have it collect 5%
annual interest, compounded daily for 25 years.  How much money is in
the account at the end?

The "true" answer is (1+(.05/365))**(365*25) dollars, assuming 365 day
years.

The IEEE double precision answer will be extremely close to that
amount.

What happens if you do that calculation in BCD?  How would you even
specify how to do it?  I don't think you will get nearly as much
accuracy.


From aleax at aleax.it  Fri Sep 20 15:00:37 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 19:00:37 GMT
Subject: Destructor never called ???
References: <mailman.1032506132.22396.python-list@python.org> <amfqoq$5iect$1@ID-59885.news.dfncis.de>
Message-ID: <pfKi9.140227$pX1.4976402@news2.tin.it>

Thomas Heller wrote:
        ...
> So what are the recommended practices if my program creates datastructures
> containing cyclic references?
> 
> - avoid using __del__ in these objects by  cleaning them 'manually'?
> - only using weak references to break these cycles?
> 
> - and finally using the debugflags of the gc module to ensure everything
> is cleaned correctly?

These are not mutually exclusive, and they all look good to me.  If
you can avoid __del__, do; if you can avoid reference loops (e.g thanks
to weak references) do that, too.  And checking that some leak didn't
happen anyway is sound.


Alex



From sholden at holdenweb.com  Wed Sep 18 22:09:56 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 18 Sep 2002 22:09:56 -0400
Subject: How to build Python 2.2.1 on Windows?
References: <3D892BBD.5B38DD49@bellatlantic.net>
Message-ID: <unai9.6131$bh1.1173@fe03>

"David Lees" <debl2nonospammmmyy at bellatlantic.net> wrote ...
> I have been playing with Python for awhile and would like to know how to
> try building Python on a Windows machine that has MS C++ version 6.  Is
> there a standard download that is configured to do this?  If not could
> someone point me in the direction of information on builds?  I
> downloaded Python-2.2.1.tgz.tar on my Win98 box and WinZip 8 does not
> even seem to know how to unzip it.  I just transfered the tar file over
> to a linux box and it unzips without and problem using tar.  The README
> seems to only talk about linux/unix type systems.
>

There's a whole directory concerned with building for the PC architrecture.
See whether the tarball's src directory contained a PCBuild directory, and
if so start with the readme.txt file.

If no such directory exists then you might need to access the CVS repository
for it (though I can't imagine that all PC builders do this).

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From bobx at linuxmail.org  Thu Sep 12 21:19:01 2002
From: bobx at linuxmail.org (Bob X)
Date: Fri, 13 Sep 2002 01:19:01 GMT
Subject: [ANN] PySQLite 0.3.0 released
References: <slrnao0e6p.23c.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <92bg9.2397$Lg2.449907@news2.news.adelphia.net>

Gerhard H?ring wrote:
> NAME:
>     pysqlite - Extension module for SQLite databases.
> 
> DESCRIPTION:
>     A Python-DB API 2.0 compliant extension module that connects to
>     SQLite databases.
> 
>     SQLite is a powerful, embedded relational database in a compact C library.
>     It supports a large subset of SQL92, multiple tables and indices,
>     transactions, and triggers. It has a simple C/C++ interface requiring only
>     three functions to perform queries. It has TCL bindings and an ODBC
>     driver. Sources are uncopyrighted and can be used for any purpose. More
>     information can be found at <http://www.hwaci.com/sw/sqlite/index.html>.
> 
> USAGE:
>     import sqlite
> 
>     conn = sqlite.connect("db")
>     cursor = conn.cursor()
>     SQL = """ select category, family, genus, species
>               from calflora order by genus, species limit 10"""
> 
>     cursor.execute(SQL)
> 
>     for col in cursor.description:
>             print "\t %12s - %3s bytes" % (col[0], repr(col[3]))
> 
>     row = cursor.fetchone()
> 
>     while row != None:
>             print "%14s, %15s, %19s, %8s, %25s" % tuple(row)
>             row = cursor.fetchone()
> 
>     SQL = "insert into calflora (family,genus,species) values(%s,%s,%s)"
>     cursor.execute(SQL, ('greenus', 'weedus', 'maximus')) 
> 
>     conn.close()
> 
> HOMEPAGE:
>     <http://pysqlite.sourceforge.net>
> 
> DOWNLOAD:
>     Source and Windows binary downloads are available at 
>     <http://www.sourceforge.net/projects/pysqlite>
> 
> AUTHORS:
>     Gerhard H?ring <gerhard.haering at gmx.de>
>     Michael Owens <mike at mikesclutter.com>
>     William Trenker <wtrenker at shaw.ca>
> 
> LICENSE:
>     Python
> 
> NOTES:
> PySQLite has been tested on FreeBSD, Linux, and Windows with both Python 2.1 
> and 2.2. The latest versions of SQLite (v. 2.6 and greater) are recommended. 
> 
> Version 0.3.0 fixed a lot of bugs, while also adding new features. Users of
> PySQLite are recommended to upgrade.
> 
> The following are some of the changes and enhancements since the last 
> release:
> 
> - Squashed a few memory leaks.
> 
> - Exposed the sqlite_exec, sqlite_last_insert_rowid and
>   sqlite_changes functions of SQLite to Python as methods of the
>   connection object.
> 
> - Add support for Date types, if mxDateTime is
>   available.
> 
> - Support for optional DB-API extensions from PEP 0249
> 
> - Added files for creating a PySQLite Debian package.
> 
> - setup.py: - Added Cygwin as platform that's supported by default
> 
> - Added third example: program to dump a table in XML format.
> 
> - Use bool type and custom bool converter for boolean fields.
> 
> - Set cursor.rowcount appropriately after DML (insert/update/delete) 
>   statements.
> 
> - Fixed a bug with the SQL-quoting of longs.
> 
> - Inline documentation improvement.
> 
> - Change invocation of ReferenceError to work with Python 2.1
> 
> - (really) implemented weak references from cursors to connections
> 
> - Added new test suite.
> 
> - Adapted test for threadsafety attribute.
> 
> - Added checks with user-defined float and string functions.
> 
> - Removed Connection.begin(). Transactions are now started
>   implicitely.  - Use DB-API quoting instead of manual Python
>   quoting.  - Use string methods instead of the string module.
> 
> - Added checks for functions/aggreagates that raise exceptions or
>   return None (NULL).
> 
> - Added tests for proper commit/rollback behaviour. Added tests
>   for autocommit feature.
> 
> - Implemented autocommit feature. Renamed the parameter of
>   Connection.__init__ from "filename" to "db". Commit and rollback
>   now check if the connection is open at all. Started adapting the
>   weak reference code from pyPgSQL to loosely couple cursors and
>   connections. This is needed when you want to check if there are
>   any open cursors for a given connection, for example.
> 
> - Simplified version numbers, removed the check for equal version
>   numbers in _sqlite and sqlite. This looked like overhead to
>   me. The version number scheme is now only: major.minor.micro Set
>   threadsafety to level 1. I believe we can guarantee than sharing
>   the module among threads will do no harm. Sharing the connection
>   would.
Windows binary?



From simonb at webone.com.au  Sun Sep  1 20:18:01 2002
From: simonb at webone.com.au (Simon Burton)
Date: Mon, 02 Sep 2002 10:18:01 +1000
Subject: yield keyword reserved
References: <mailman.1030895343.10143.python-list@python.org>
Message-ID: <pan.2002.09.02.10.17.50.429883.21014@webone.com.au>

you forgot this:

>>> from __future__ import generators


On Mon, 02 Sep 2002 01:57:06 +1000, Erik Price wrote:

> I'm sorry if this is an easily answered-question.  I have Python2.2
> running on my system (it is installed on Darwin by default in Mac OS X
> 10.2), and am learning about iterators from
> <http://python.org/doc/current/whatsnew/node5.html>, but the keyword
> doesn't appear to be present on my system:
> 
> Python 2.2 (#1, 07/14/02, 23:25:09)
> [GCC Apple cpp-precomp 6.14] on darwin Type "help", "copyright",
> "credits" or "license" for more information.
>  >>> def generate_ints(N):
> ...  for i in range(N):
> ...   yield i
> <string>:3: Warning: 'yield' will become a reserved keyword in the
> future
>    File "<string>", line 3
>      yield i
>            ^
> SyntaxError: invalid syntax
>  >>>
>  >>>
>  >>>
> Is my Python not a true 2.2 release, or do I have something wrong in my
> syntax?
> For that matter a recompile wouldn't be so bad since this build doesn't
> have readline :(
> 
> 
> Erik
> 
> 
> 
> 
> 
> 
> --
> Erik Price
> 
> email: erikprice at mac.com
> jabber: erikprice at jabber.org


From sholden at holdenweb.com  Tue Sep 10 18:27:55 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 10 Sep 2002 18:27:55 -0400
Subject: How do i print to a printer using python?
References: <6afb9d2c.0209041044.43c5f5f1@posting.google.com> <al73gj$lbt$1@wanadoo.fr> <6afb9d2c.0209091641.1d9b652e@posting.google.com> <allofr$lgscc$2@hades.rz.uni-sb.de>
Message-ID: <dnuf9.122228$l_4.76576@atlpnn01.usenetserver.com>

"Uwe Schmitt" <uwe.schmitt at procoders.net> wrote in message
news:allofr$lgscc$2 at hades.rz.uni-sb.de...
> Naveed Iqbal <iqbal at colorado.edu> wrote:
> > Thank you all for your help. I am really a novice when it comes to
> > python. Will someone tell me how to install the win32ui module in
> > windows?
> > thanka
> > naveed
>
> I suceeded without win32all: I generated a postscript file
> and copied it to the 'file' "prn:" using shutil.copyfile().
>
Disk Operating System Rules OK!

As a pythonista of somewhat longer standing I found PythonWin a delight. I
don't overstress it, and it works.

Naveed: while you are likely to stay more up-to-date by installing an
updated base Python and installing updated Win32 extensions, there's no real
reason for a "novice" to need (or even want :-) to be so close to the
(b)leading edge. I haven't heard anything to suggest the ActiveState
distribution isn't a perfectly practical solution to everyday needs.

Uwe: you'll clearly do well with such lateral thinking abilities. What
software did you use to generate the PS?

nostalgical-ly y'rs  - ste ve
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------






From oliver.eichler at dspsolutions.de  Tue Sep 10 08:22:14 2002
From: oliver.eichler at dspsolutions.de (Oliver Eichler)
Date: Tue, 10 Sep 2002 14:22:14 +0200
Subject: Howto debug self?
References: <alhivr$fom$04$1@news.t-online.com> <j47khvbjq1.fsf@informatik.hu-berlin.de>
Message-ID: <alko1u$ne4$05$1@news.t-online.com>

> print self.__dict__

Yes, this would work. However if self has some classes you have to iterate 
over the dictionaries. I tried to make my own abstraction from bdb to play 
with the eval() command, to display the complete tree. But I ran into 
recursion problems (enless loops).  

Somehow it is hard to imagin this has not been implemented sofar, as I would 
consider it as a very basic debugging function. Are there better frontends 
to bdb than pdb? 

Oliver




From syver-en+usenet at online.no  Thu Sep 12 16:19:48 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Thu, 12 Sep 2002 20:19:48 GMT
Subject: web crawler help?
References: <JARd9.6927$yt3.3340577@newssrv26.news.prodigy.com> <ac259413.0209101122.54cc774c@posting.google.com>
Message-ID: <ud6rjx8oh.fsf@online.no>

s_gherman at yahoo.com (Sorin Gherman) writes:

> "koko" <kokohh at hotmail.com> wrote in message
> news:<JARd9.6927$yt3.3340577 at newssrv26.news.prodigy.com>...
> 
> > is there any sample for basic web crawler, that ask for a starting
> url and
> 
> > log the url and extract the hyperlinks?
> > thx
> 
>     It's a very simple one in Mark Pilgrim's "Dive into Python" book,
> whose text is freely available at: http://diveintopython.org/
>     Check the "HTML processing" chapter. It contains a urllister.py 9
> lines program, followed by a 7 lines usage example which does just
> that: given a URL for a HTML file, it lists the hyperlinks inside it.
> 
> /sorin gherman

You could also check out Aahz's slides about multithreading, they show
the development of different threading architectures for a spider.

http://starship.python.net/crew/aahz/OSCON2001/index.html

-- 

Vennlig hilsen 

Syver Enstad


From b.maryniuk at forbis.lt  Wed Sep 11 10:37:51 2002
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Wed, 11 Sep 2002 16:37:51 +0200
Subject: UNIX processes
In-Reply-To: <kuhegycb8r.fsf@lasipalatsi.fi>
References: <mailman.1031654172.15956.python-list@python.org> <kuhegycb8r.fsf@lasipalatsi.fi>
Message-ID: <200209111637.51785.b.maryniuk@forbis.lt>

On Tuesday 10 September 2002 13:52, Erno Kuusela wrote:
> use the ps program (ps -e works on linux). there is no other
> standard api for it.

I already made native Python module for this, because I don't like 'ps' in 
this case. Well, module works even faster. ;-)

-- 
Regards, Bogdan

I doubt, therefore I might be.




From tjreedy at udel.edu  Mon Sep  9 21:39:40 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 10 Sep 2002 01:39:40 GMT
Subject: Python primer - comments appreciated!
References: <aljdhe$1q63hj$1@ID-77524.news.dfncis.de>
Message-ID: <w3cf9.535714$m91.21695020@bin5.nnrp.aus1.giganews.com>

"Thorsten Kampe" <thorsten at thorstenkampe.de> wrote in message
news:aljdhe$1q63hj$1 at ID-77524.news.dfncis.de...
> "Python, anyone?"... Yep, this is my first one. Comments on
> style/readability and technique/flow appreciated!
>
> #v+
> def quotient_set(seq, func, partition='nonbool'):
>     """ partition <seq> into equivalence classes

To my mind, you should return a dictionary mapping each realized value
of func (assuming hashable)  to the list of items for which func has
that value.  In other words, result.values() would be the list of
equivalence classes.  Let caller toss away info of value defining
class.  In any case, using a dict to sort items into such classes
should be much faster that what you wrote.

def equiv_class(seq, func, part = 'nonbool'):
  if part == 'bool':
    result = {False:[], True:[]}
    for item in seq:
      result[bool(func(item))].append(item)
  else:
    result = {}
    for item in seq:
      result.setdefault(func(item),[]).append(item)
      # {}.setdefault(key,default) was added for just this sort of
usage
  return result

>>> f = lambda x: x % 3
>>> equiv_class([0,1,2,3,4,5,6,7,8,9],f)
{0: [0, 3, 6, 9], 1: [1, 4, 7], 2: [2, 5, 8]}
>>> equiv_class([0,1,2,3,4,5,6,7,8,9],f,part='bool')
{0: [0, 3, 6, 9], 1: [1, 2, 4, 5, 7, 8]}


> def set(seq):
>     """ make <seq> a true set by removing duplicates """
>     # Shortest way to do a 'null function'? 'None' doesn't work...
>     f = lambda x: x  # identity function
>
>     return [item[0] for item in quotient_set(seq, f)]

No, no.  If items are hashable, standard method is...

def set(seq):
  d = {}
  for item in seq:
    d[item] = 0
  return d.keys()

Terry J. Reedy





From phncontact.libero.it at   Thu Sep  5 11:16:48 2002
From: phncontact.libero.it at  (s1kboy)
Date: Thu, 05 Sep 2002 15:16:48 GMT
Subject: Servizi Internet
References: <mailman.1031056925.7383.python-list@python.org> <f9fbnu8dghga0tbpvf7nsv64jo3csrlsm1@4ax.com> <y9Ed9.2904$%z1.72415@twister1.libero.it>
Message-ID: <katenucl66j2kgtlnvie5g212tc7n84tns@4ax.com>

ma ciao ^_^

On Thu, 05 Sep 2002 07:59:26 GMT, Paolo Invernizzi
<paoloinvernizzi at dmsware.com> wrote:

>Benjamin wrote:
>> by the way, how many people understand italian in
>> this nesgroup? ^_^
>
>** a little hand is raised! **



From tadin at cirus.dhz.hr  Thu Sep 26 04:04:34 2002
From: tadin at cirus.dhz.hr (Marijan Tadin)
Date: Thu, 26 Sep 2002 10:04:34 +0200
Subject: Scientific Python for Win32
References: <3D92082C.3050707@rmc.ca>
Message-ID: <amui5e$6l79$1@as201.hinet.hr>

1. I am far away from expert but I've tried out netCDF module from
Scientific Python , you can download necessery dll-s from
http://basic.astrian.net/godzilla/  . Unfortunately, I am afraid that the
modules from Scientific Python that you need are not there, but you schould
check.

2. actualy you need *.dll and not *.exe files if you want to use compiled
binaries in Python on windows. (or maybe I just didn't understand your
message wright).

3. you can find more about extending Python on http://www.swig.org/, (check
swig manual). I think it is more understadable for easy start then
information on python.org, unfortunately only info about use of VC++ and not
Borland for Windows (even that info was difficult to find on net).

4. I feel that extending python is not so easy (if you are not experienced
C-programmer, like I am not), the main problem is not Python but knowlege of
C and details of particular compiler even more. If you know how to use dll-s
from Python (try some simple example from swig manual), then try to find
dll-s you need on Internet.


"bourassa" <bourassa-m at rmc.ca> wrote in message
news:3D92082C.3050707 at rmc.ca...
> Hi there.  I'm a Python and programming newbie.  I was looking for at
> Scientific Python esp. for its use of VRML for plotting output.  Unlike
> many of the other Python modules I have, there are no .exe binaries.
> So, I have Borland C++Builder 5 on a Win2K P4.  How would I compile(?)
> my own Scientific Python?  Would it work with my Python 2.2 which I just
> installed from a .exe file?  Any help would be appreciated.  Thanks.
>
> Mike
>
> PS - I spoke with the creator of Scientific Python.  He was very helpful
> but admits to not having expertise in this aspect, hence my question.
>




From doru-cat at ifi.uio.no  Sat Sep  7 08:57:21 2002
From: doru-cat at ifi.uio.no (Doru-Catalin Togea)
Date: Sat, 7 Sep 2002 14:57:21 +0200 (MET DST)
Subject: Python, CGI, MySQL -> 'Error 2004: Can't create TCP/IP socket (10106)'
Message-ID: <Pine.GSO.4.21.0209071452450.11315-100000@tva.ifi.uio.no>

Hi!

I am using Python 2.2 on a win 2000 Pro computer to interact with mysql,
by means of the MySQLdb module. My non-CGI scripts work as expected. 

I have now installed a webserver (Savant 3.1) am I try to connect to mysql
by means of CGI and still MySQLdb. My scripts fail now with the 'Error
2004: Can't create TCP/IP socket (10106)' error message.

Can anyone tell me what these errors are caused by, and who is it that
generates the error messages? They do not seem to originate from the MySQL
server as it's error checking utility (perror) does not recognize any of
the 2004 or 10106 error codes. Neither did I find any mention of them in
the documentation comming with PythonWin. This leaves me with the theory
that they originate with the web server.

Anyone who can recommand a free webserver which interacts well with MySQL
by means of pythonic CGI scripts on Win 2000 Pro?

What I am trying to do is first to establish a connection to mysql-server
and then to have it execute a sql statement using that connection. 

I have a module called MyDatabase.py. It defines the following functions.	
	
# ********************************************************************
def openDBConnection(h, u, p, d): # host, user, password, database
	try:
		conn = MySQLdb.connect(host = h, user = u, passwd = p, db
= d)
	except MySQLdb.Error, e:
		result = pname + "Error %d: %s" % (e.args[0], e.args[1])
		#sys.exit(1)
		return result
		
	return conn
	
# ********************************************************************
# executes an SQL statement using the given connection
def execSQL(connection, sqlCommand):
	cursor = connection.cursor()
	
	cursor.execute(sqlCommand)	
	rows = cursor.fetchall()
	
	cursor.close()
	return rows
	
These functions work when the script connects to the mysql-server
directly, but fail when a webserver is in between. I call my functions as
follows:
	
import MyDatabase

...
conn = MyDatabase.openDBConnection(host, user, password, database)
sqlCommand = "..."
rows = MyDatabase.execSQL(conn, sqlCommand)
...


Thanks if you can help.

Best regards,
Catalin



	<<<< ================================== >>>>
	<<     We are what we repeatedly do.      >>
	<<  Excellence, therefore, is not an act  >>
	<<             but a habit.               >>
	<<<< ================================== >>>>




From loewis at informatik.hu-berlin.de  Thu Sep 26 07:04:44 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 26 Sep 2002 13:04:44 +0200
Subject: unicode woes
References: <amuhrk$4cn$05$1@news.t-online.com> <3d92d120$0$76313$e4fe514c@dreader4.news.xs4all.nl> <amuldd$bur$01$1@news.t-online.com>
Message-ID: <j465wt810z.fsf@informatik.hu-berlin.de>

Ulli Stein <mennosimons at gmx.net> writes:

> Or how would you do this (nonsense string):
> str = str + "bla??" + str[:5] + "???"
> 
> Would you write then:
> str = str + unicode("bla??") + str[:5] + unicode("???")?

If you want to use Unicode in source code, and you want to use
non-English text: yes, you have to define a helper function, e.g.

def uni(x):
  return unicode(x, "iso-8859-1")

In this case, the system encoding might be incorrect; Python cannot
know what encoding you use in your source code: If you move your
source code from one machine to another, the system encoding may
change, but your source code won't.

If you want to write internationalized code, I recommend to avoid
using non-English string literals. Instead, use English text, and
translate this at run-time using message catalogs (e.g. gettext).

> And what about the "encode" variable in site.py: If another Python 
> application running in parallel to our application changes the encode 
> string, will it affect our app? Or has every application its own encode 
> string?

If you keep your sitecustomize.py in a location that is private to
your application, there is a good chance that it won't interfere with
another application.

Regards,
Martin


From david at fielden.com.au  Tue Sep 24 20:30:34 2002
From: david at fielden.com.au (DG)
Date: Wed, 25 Sep 2002 10:30:34 +1000
Subject: Strange python and cgi!
In-Reply-To: <3D9101AC.842F0513@alcyone.com>
Message-ID: <001a01c2642a$c3efec80$3101a8c0@borg.fielden.com.au>

And/or put this at the start of your script:

	import cgitb; cgitb.enable()

Rowdy

> -----Original Message-----
> On Behalf Of Erik Max Francis
> Sent: Wednesday, 25 September 2002 10:22 AM
> Subject: Re: Strange python and cgi!
> 
> 
> CheapSkate wrote:
> 
> > when I go to the webpage, it stops halfway and in "view source: it
> > only show
> > the code till the open "pre".
> 
> It's likely some other error occurred.  Check your Web server 
> error log
> for the final word.




From brian at sweetapp.com  Wed Sep  4 17:49:53 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Wed, 04 Sep 2002 14:49:53 -0700
Subject: str() of a tuple
In-Reply-To: <H1xnHJ.1zE1@beaver.cs.washington.edu>
Message-ID: <000901c2545d$0085b6d0$df7e4e18@brianspiv1700>

> Why, oh why does the str() of a container use the repr()s of the
> objects inside?  This seems to violate what the docs claim str() does
> -- "return a nicely printable representation of an object."  Would a
> sane implementation break something?

(Before I tell you, could you please tell me what this tuple contains? I
bet that you can't figure it on in less than, say, 5, days)

Cheers,
Brian




From martin.klaffenboeck at gmx.at  Fri Sep  6 16:45:08 2002
From: martin.klaffenboeck at gmx.at (Martin Klaffenboeck)
Date: 06 Sep 2002 22:45:08 +0200
Subject: Beginners Question - use of variables
Message-ID: <1031345108.547.165.camel@martin.kleinerdrache.org>

Hello,

Are newbies questions ok here?

I've got a struktur like this:

string string int timestamp
string string int timestamp
string string int timestamp
...

in Perl I used to save such a thing in a multidimensional array like

array(
      array(string, string, int, timestamp),
      array(string, string, int, timestamp),
      array(string, string, int, timestamp))

So how do I save that in python and how do I get fast to one of this
strings or int or timestamp... ?

Martin


-- 
     Virtuelle Jugendarbeit auf http://create.kleinerdrache.org
            ...  bald kommt ein neues Layout  ...




From thorsten at thorstenkampe.de  Tue Sep 10 07:55:47 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Tue, 10 Sep 2002 13:55:47 +0200
Subject: Python primer - comments appreciated!
References: <aljdhe$1q63hj$1@ID-77524.news.dfncis.de> <mailman.1031619853.30153.python-list@python.org>
Message-ID: <alkmjt$1prs25$1@ID-77524.news.dfncis.de>

*  James J. Besemer
> Thorsten Kampe wrote:
>
>> "Python, anyone?"... Yep, this is my first one. Comments on
>> style/readability and technique/flow appreciated!
> [...]
>>        # put this in one line?
>>        true_class  = []
>>        false_class = []
> [...]
>        #### if you are thinking "true_class = false_class = []"
>        #### then it would be a BIG NO-NO.
>        #### the latter form initializes both variables to the SAME 
> (empty) list

Well, uh, ah, eh, errm, good to know, thanks!

>>                quotient_set.append([item])
>
> #### teeny tiny nit:  I prefer "quotient_set.extend( item )" to the above.

That's not the same: I'm appending a new equivalence list to the 
quotient set by making item a list and your version would append a 
list as individual items.

>>                # This somehow "inelegant", because all I want to do is:
>>                # "seq[index] = seq[index].append(item)" Suggestions?
>>
> ####  Define your set or sequence types to be a class or classes.  
> ####  You can then define [] and .append() operations
> ####  to be whatever you want (e.g., silently exclude duplicates, etc.)
>
> ####  Even the most trivial data abstractions usually benefit by being 
> made an
> ####  explicit class/object.

Well I RTFM'd and 'tutorialed' a lot, but I cannot find any concrete 
use for this OOP thing. Could you elaborate?!

> ####  if you are 'indexing' you may want to use a dictionary base type 
> instead of a list.

I think I'm not "indexing" but "hashing" over the representants, so a 
dictionary (associative array) should be the natural thing, thanks.

> ####  if you only want to test membership, instead of the 'try' you can 
> test:
>
>                  if representative in representative_class:  ....  

Well, even "if item in seq: seq.index(item)" is faster than "try: 
seq.index(item)" 


Thorsten


From mark.charsley at REMOVE_THIS.radioscape.com  Mon Sep 23 11:22:00 2002
From: mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley)
Date: Mon, 23 Sep 2002 16:22 +0100 (BST)
Subject: What the heck? open() requires an integer?
References: <fec163ad.0209201606.6fff11c0@posting.google.com>
Message-ID: <memo.20020923162223.544A@a.a>

In article <fec163ad.0209201606.6fff11c0 at posting.google.com>, 
mgarcia at cole-switches.com (Manuel M. Garcia) wrote:

> O'Reilly publishes two excellent books for people who wish to learn
> Python and write applications quickly: Programming Python, 2nd
> Edition, and Python Cookbook.

Programming Python 2nd ed is _not_ much good for a complete beginner at 
Python - I speak from experience (and I'm not new to programming). 

The gentle intro from the 1st edition (which was good for beginners from 
the reviews I've read) has been taken out, and expanded into Learning 
Python also by O'Reilly.

Once one has read Learning Python, then Programming Python 2nd ed makes a 
lot more sense.

-- 

Mark - personal opinion only, could well be wrong, not representing 
company, don't sue us etc. etc.


From pyth at devel.trillke.net  Sun Sep  1 19:00:50 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 2 Sep 2002 01:00:50 +0200
Subject: Add methods to a class at runtime?
In-Reply-To: <Xns927CF1E86FFE8cliechtigmxnet@62.2.16.82>; from cliechti@gmx.net on Sun, Sep 01, 2002 at 11:47:01PM +0200
References: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com> <slrnan4dl3.18h.gerhard.haering@lilith.my-fqdn.de> <wmvc9.21757$FJ1.3181934@e3500-atl1.usenetserver.com> <Xns927CF1E86FFE8cliechtigmxnet@62.2.16.82>
Message-ID: <20020902010050.G3893@prim.han.de>

Chris Liechti wrote:
> "Robert Oschler" <Oschler at earthlink.net> wrote in
> news:wmvc9.21757$FJ1.3181934 at e3500-atl1.usenetserver.com: 
> > Some interesting tidbits I
> > found while trying out your suggestions,  as you will see in the two
> > ".py" files that follow, I discovered that to create a method you can
> > dynamically add to a class _instance_, you need to omit the 'self'
> > parameter in the method declaration.  For a method you add dynamically
> > to a class definition itself, then you must include the 'self' 
> > parameter in the method declaration.  If not, you will get an argument
> > count mismatch complaint from the interpreter.  Fascinating stuff.
> 
> that's just the differece between a "function" and a "method". a method 
> works on an object and that is passed explicit as first argument. "explicit 
> is better than implicit" (you can read this and some other design ideas by 
> typing ">>> import this" ;-)
> 
> > I'm running 2.1.1, anybody know if 2.2.2 is going "disable" either of
> > these abilities? (class or class instance dynamic method addition).  I
> > saw a note about such activities requiring a "dynamic" keyword, which
> > is fine.  I'd really miss either of these abilities.
> 
> nah. "classic classes" stay the way they are and are available until 3.0
> new style classes, those that inherit from "object" or "type", will need 
> the marking as "dynamic".

Can you give a pointer for further explanation?  Can't believe it :-)

thanks,

    holger




From maxm at mxm.dk  Tue Sep  3 03:39:25 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 03 Sep 2002 09:39:25 +0200
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <un50ccotaak86b@corp.supernews.com> <al0e96$oms$0@216.39.172.122> <1031008206.347446@ampungk.ozonline.com.au>
Message-ID: <3D74672D.8050600@mxm.dk>

John La Rooy wrote:
> Bengt Richter wrote:
>>
> It is unfortunate that often a re that starts off as a simple idea
> turns out to be a big mess. I think some sort of graphical tool
> would be even better so you could see what the re does. Especially
> if it could convert both ways and allow you to edit the re.


There used to be a tkinter tool like this in Python under 1.52 written 
by Guido. I don't know what happened to it though.

regards Max m



From tim.one at comcast.net  Thu Sep 19 14:20:22 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 19 Sep 2002 14:20:22 -0400
Subject: monetary applications [was: Python GUI app to impress the bos s?]
In-Reply-To: <23891c90.0209190014.26c5bb9@posting.google.com>
Message-ID: <BIEJKCLHCIOIHAGOKOLHAENCDKAA.tim.one@comcast.net>

About
    http://starship.python.net/crew/aahz/FixedPoint.py

If someone is willing and able to do the work (sorry, I can't make time for
it), Guido will accept this into the standard library for 2.3.  It's a
small, self-contained, pure Python module; it's been in use for a few years
with no bugs reported; and is unencumbered by any license (a lawyer has told
me that an individual in the USA cannot meaningfully disclaim copyright, so
fine, the copyright I didn't want is hereby transferred to the PSF <wink>).

What work remains:

+ Write docs for the Library Reference manual.  I expect the existing
  module docstring will be a good start.

+ Create a test driver for Python's regression suite.

+ Have fun modernizing it, if you like (for example, "/" should be
  changed to "//" where integer division is intended; playing with
   __slots__ may yield some nice savings; whatever).

This should be a fun and comparatively easy task -- if you ever wanted to
get your name into the ACKS file, here's one way to do it without tithing
you income directly to Guido <wink>.

although-tithing-saves-me-time-ly y'rs  - tim




From sholden at holdenweb.com  Wed Sep 11 16:42:17 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 11 Sep 2002 16:42:17 -0400
Subject: "str.contains(part)" or alternatives?
References: <alo8jr$l04$1@news.online.de>
Message-ID: <eWNf9.165073$l_4.88540@atlpnn01.usenetserver.com>

"Stefan Schwarzer" <sschwarzer at sschwarzer.net> wrote in message
news:alo8jr$l04$1 at news.online.de...
> Hello all
>
> I would like to get your thoughts on testing if a string contains another.
>
> To do something dependent on whether a part is contained in another
string, the
> following code is familiar to me:
>
>  >>> s = 'Hello'
>  >>> if s.find('ll') != -1:
> ...     do_something_if_found()
> ... else:
> ...     do_something_if_not_found()
>
> However, I dislike the test for the special value -1; that reminds me of
error
> code checking. On the other hand, using exceptions could look like
>
>  >>> try:
> ...     s.index('ll')
> ...     do_something_if_found()
> ... except ValueError:
> ...     do_something_if_not_found()
>
> This has the problem that a ValueError raised in do_something_if_found()
may give
> the false impression that the substring isn't contained in the string s.
Moreover,
> that code doesn't represent my thoughts very well. (Maybe I could say that
better
> if I were a native English speaker :) ).
>
> I would like something analogous to .startswith and .endswith:
>
>  >>> s = 'Hello'
>  >>> if s.contains('ll'):
> ...     do_something_if_found()
> ... else:
> ...     do_something_if_not_found()
>
> Do you have other suggestions how to do the test easily without using
special
> values? Should a .contains method be part of string objects in future
Python
> versions? What do you think?
>

I think the current proposal is to move to

    if ll in s:

for this construct. Happy?

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From theller at python.net  Fri Sep 27 16:06:41 2002
From: theller at python.net (Thomas Heller)
Date: Fri, 27 Sep 2002 22:06:41 +0200
Subject: watching mutables?
References: <mailman.1033137396.14482.python-list@python.org> <an211j$oi$1@news.hccnet.nl> <Xns9296C58C9ADFCcliechtigmxnet@62.2.16.82> <k7l7qpvf.fsf@ion-tof.com> <Xns9296D3AB53777cliechtigmxnet@62.2.16.82> <mailman.1033154856.20758.python-list@python.org>
Message-ID: <an2doh$ajo3p$1@ID-59885.news.dfncis.de>

"Michael Stenner" <mstenner at phy.duke.edu> wrote in message news:mailman.1033154856.20758.python-list at python.org...
> On Fri, Sep 27, 2002 at 08:48:10PM +0200, Chris Liechti wrote:
> > Thomas Heller <thomas.heller at ion-tof.com> wrote in
> > news:k7l7qpvf.fsf at ion-tof.com:
> > > Chris Liechti <cliechti at gmx.net> writes:
> > [triggered by *any* change of object m]
> > >> see EventObj:
> > >> http://oomadness.tuxfamily.org/p-pyobjtools.php
> > >>
> > > This is what I get when clicking on the download link:
> > > <title>404 Not Found</title>
> >
> > works for me.
>
> I find that rather surprising since that file isn't there.  Are you
> sure you clicked on the download link?  For your convenience, I
> provide that link here:
>
> http://oomadness.tuxfamily.org/downloads/EventObj-0.2.tar.gz
>
> I am also interested in this module (curiosity, mostly) and would
> appreciate a working link if you have one.
>
Here's what seems to be available:
http://oomadness.tuxfamily.org/downloads/
and this is obviously where the EventObject is contained in:
http://oomadness.tuxfamily.org/downloads/EditObj-0.3.tar.gz

Thomas




From dsavitsk at e-coli.net  Wed Sep 11 21:18:21 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Thu, 12 Sep 2002 01:18:21 GMT
Subject: ntpath bug or expected?
Message-ID: <xXRf9.7957$yt3.4120228@newssrv26.news.prodigy.com>

python 2.1.3.  this is pretty minor ... os.path.join() on win32 fails if the
first path component is C: (or any other letter).  that is, if the first
component is a drive letter, it must contain either '\\' or '/' or else the
path is wrong.  this makes it the only component which must have slashes.

example
>>> os.path.join('C:', 'path', 'to', 'file.txt')
C:path\\to\\file.txt

the offending code is below and can be easily fixed by removing the ':' from
'/\\:'

def join(a, *p):
    """Join two or more pathname components, inserting "\\" as needed"""
    path = a
    for b in p:
        if isabs(b):
            path = b
        elif path == '' or path[-1:] in '/\\:':
            path = path + b
        else:
            path = path + "\\" + b
    return path




From doyle at nebcorp.com  Mon Sep 30 19:13:05 2002
From: doyle at nebcorp.com (Rev. Joe Doyle)
Date: Mon, 30 Sep 2002 23:13:05 -0000
Subject: Stow and Python
Message-ID: <slrnaphmk1.ks5.doyle@pianosa.catch22.org>

	Hello, I'm trying to build a Python package for a large 
installation, where packages are of the following form:

1) /usr/local is a symlink to /tools/foo

2) packages live in /tools/bar/package-name

3) within each package, there are "bin", "lib", etc. that get linked into 
/usr/local/ (using GNU stow).

	My problem is that the Python interpreter looks in the absolute 
real path for modules, rather than the symlinked "proper" path 
("/usr/local/lib" etc.).  How do I get it to look in the right place?  
Thanks in advance for any hints.


-Joe Doyle


From mnations at airmail.net  Mon Sep  2 12:24:13 2002
From: mnations at airmail.net (Marc)
Date: 2 Sep 2002 09:24:13 -0700
Subject: Problems with file.write()
References: <4378fa6f.0208291457.ba738cb@posting.google.com> <mailman.1030669067.7832.python-list@python.org> <4378fa6f.0208300731.18105bc7@posting.google.com> <akqf9u$1kjdm1$1@ID-114079.news.dfncis.de>
Message-ID: <4378fa6f.0209020824.3e658da8@posting.google.com>

> Note that all buffered data is flushed automatically when you close()
> an output file (you *are* closing the file, aren't you?).

Actually I wasn't closing the file originally but it had never been a
problem before. But closing the file now causes it to spill it's guts
and gives me everything I was looking for.

However, that doesn't explain why it wasn't working in the first
place. I originally had it in unbuffered mode and wasn't getting all
my text. As I increased the buffering from 1 to 2 I kept getting more
text. But never all of it until I closed the file. It seems like it
shouldn't be necessary to flush the buffer after every write if you're
in unbuffered mode. This seems like a problem.


From wlfraed at ix.netcom.com  Mon Sep 30 01:54:00 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Sun, 29 Sep 2002 22:54:00 -0700
Subject: Catching user defined exceptions from foreign modules
References: <mailman.1033352618.4920.python-list@python.org>
Message-ID: <oto8na.bnn.ln@ix.netcom.com>

Derek Basch fed this fish to the penguins on Sunday 29 September 2002 
07:22 pm:

> I know that TestError has been imported because I can
> put a "print TestError" before the except clause in
> test1 and I get "test2.TestError". Can anyone tell me
> why I am having to put the test2.TestError instead of
> just TestError? It confuses me even further that my
> small test modules work like I would expect and it
> doesn't work on the larger modules.
>
 
        I don't even see why "test1" is working. You explicitly import the 
exception by name, but you didn't bring in "TryItOut"...  That alone 
should have raised an exception.


[wulfraed at beastie wulfraed]$ python test1.py
Traceback (most recent call last):
  File "test1.py", line 14, in ?
    letitrip.run()
  File "test1.py", line 7, in run
    tryitout = TryItOut()
NameError: global name 'TryItOut' is not defined

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From loewis at informatik.hu-berlin.de  Mon Sep  2 09:37:49 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 02 Sep 2002 15:37:49 +0200
Subject: version independant object usage counting
References: <mailman.1030956725.28058.python-list@python.org>
Message-ID: <j465xoldg2.fsf@informatik.hu-berlin.de>

"Rolf Kalbermatter" <rolf.kalbermatter at citeng.com> writes:

> However although this loader works quite fine now, I found out that
> I gained little with it because the current Python architecture is
> not really designed to link dynamically to it.

Yes, this is the case - mainly because the API changes in slight ways,
and because therefore, extension modules explicitly link to
pythonxy.dll.

> The problem arises in fact with the refcounting of Python
> objects. The macros Py_INCREF and Py_DECREF access directly
> structure members in the object layout and this layout is different
> for debugging and non debugging versions and potentially different
> between Python versions as well.

That is only a minor problem: those macros are correct across all
released Python versions, back to 1.4 or earlier.

> Is there a specific reason that these two API functions are not
> available and one is forced to use the according macros to do proper
> refcounting, or did I miss those functions completely?

Nobody has proposed to introduce them, because there is no need to do
so, and because other ABI changes break the ABI more often - INCREF
and DECREF have never done so, and it is unlikely that they ever will.

> Any thought about this?

Feel free to submit a patch to SourceForge. However, you might want to
consider the following issues:

- various places assume presence of the ob_type field:
  * the Py*_Check functions (e.g. PyList_Check) directly access the
    ob_type field.
  * PyObject_INIT relies on initializing the object members.
- PyType_HasFeature relies on the position of the tp_flags field
- various accessor macros rely on the internal structure, e.g.
  * PyTuple_GET_ITEM
  * PyString_AS_STRING
  * ...

Regards,
Martin



From stepanov at oitunix.oit.umass.edu  Tue Sep  3 17:51:47 2002
From: stepanov at oitunix.oit.umass.edu (Alexander Stepanov)
Date: Tue, 3 Sep 2002 17:51:47 -0400
Subject: Drawing graphs/newtorks with Python
Message-ID: <Pine.OSF.4.10.10209031748180.23796-100000@wilde.oit.umass.edu>

Hi Pythonians,

I work on program which performs some calculations over graphs. 
(shortest paths). I would like to draw an image/ps_document with 
Python. Does anybody use some kind of modules to draw graphs?
I will appreciate if you can point me out to the source. 

Thank you, 

Alexander Stepanov

astepano at ecs.umass.edu



From anthony at computronix.com  Thu Sep 19 10:25:03 2002
From: anthony at computronix.com (Anthony Tuininga)
Date: 19 Sep 2002 08:25:03 -0600
Subject: surface of pygame
In-Reply-To: <20020919101158.58549.qmail@web21305.mail.yahoo.com>
References: <20020919101158.58549.qmail@web21305.mail.yahoo.com>
Message-ID: <1032445504.17506.58.camel@chl0151.edmonton.computronix.com>

Look at the documentation for pygame at www.pygame.org, specifically the
introduction regarding programming in pygame. Specifically, you have to
blit the image on to the screen before you will see it.

On Thu, 2002-09-19 at 04:11, black wrote:
> Hi,all~
> 
> I touched something about pygame today and confused with surface type. I
> didnt remember surface data type when learning python, and it displays
> strange to me, for example, I create a new surface by loading a picture
> as follows:
> 
> pygame.image.load("myPic.bmp") 
> 
> it works fine but I coulnt see any surface appears on the screen.
> any ideas ?
> 
> 
> 
> 
> 
> 
>   _____  
> 
> Do you Yahoo!?
> New DSL Internet Access
> <http://rd.yahoo.com/evt=1207/*http://sbc.yahoo.com/>  from SBC & Yahoo!
-- 
Anthony Tuininga
anthony at computronix.com
 
Computronix
Distinctive Software. Real People.
Suite 200, 10216 - 124 Street NW
Edmonton, AB, Canada  T5N 4A3
Phone:	(780) 454-3700
Fax:	(780) 454-3838
http://www.computronix.com




From export at hope.cz  Fri Sep 20 16:08:15 2002
From: export at hope.cz (A)
Date: Fri, 20 Sep 2002 22:08:15 +0200
Subject: How to upload a file using http protocol
Message-ID: <3D8B9C4F.14553.1E62DBD@localhost>

Hi,
Does anyone know how to upload a file through HTTP protocol?Is that a possible?
Thank you for help.
ladislav
 




From fabien.henon at caramail.com  Tue Sep 10 17:58:38 2002
From: fabien.henon at caramail.com (Fabien HENON)
Date: Tue, 10 Sep 2002 23:58:38 +0200
Subject: Maybe a dumb question
Message-ID: <3D7E6B0E.4060709@caramail.com>

I have written an editor for Povray using Python, Tkinter and Pmw

It is available at :

http://pyvon.sourceforge.net

The software includes one main file and 6 others which are for the 
language customization. People can choose between French, English, 
German,...
All the translations for these languages are stored in files called 
menu_francais.py, menu_english.py, menu_deutsch.py ....when the editor 
is started.

I would like to clean the directory a bit and put all these 'languages 
files' into a sub-directory called Lang.
How do I tell python to look into a particular sub-directory when 
importing a file.

I tried the os.path.join syntax, but to no avail.

Below is the bit of code that imports the language file from the current 
  working directory :


try :
   homedir = os.path.expanduser('~')
   fd = open(homedir+"/.pyvonrc", 'r')
   for line in fd.readlines():
      if line.startswith("Language :"):
	line=line[11:]		
	if   line.startswith("francais"):from menu_francais import *
         elif line.startswith("english"):from menu_english import *
         elif line.startswith("deutsch"):from menu_deutsch import *
         ...............


Thanks,


Fabien



From mertz at gnosis.cx  Wed Sep 25 20:57:06 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Wed, 25 Sep 2002 20:57:06 -0400
Subject: a python puzzle
Message-ID: <itlk9kKkXwvD092yn@gnosis.cx>         

I think the various responders did this "by hand."  Probably having the
Python source code made it even easier to just eyeball.  In a more
general sense (just in case anyone doesn't know), the way to solve a
Caesar cipher like this is to start counting letter frequencies.
Comparing the frequencies to well-known English letter frequencies like:

    http://gnosis.cx/download/letterfrequency.gif

lets you start matching letters pretty quickly.  Of course, if you don't
know it's English, the frequency table isn't as accurate.

I was thinking of writing a little solver script to actually assign
likely permutations based on the frequencies.  But a dozen people had
already solved it before I saw the problem... so I gave up on that :-).

Yours, Lulu...

--
Keeping medicines from the bloodstreams of the sick; food from the bellies of
the hungry; books from the hands of the uneducated; technology from the
underdeveloped; and putting advocates of freedom in prisons.  Intellectual
property is to the 21st century what the slave trade was to the 16th.



From a-steinhoff at web.de  Thu Sep 12 05:24:47 2002
From: a-steinhoff at web.de (Armin Steinhoff)
Date: 12 Sep 2002 02:24:47 -0700
Subject: Multi-threading python
References: <almvvt$34p$1@news.netpower.no>
Message-ID: <ddc19db7.0209120124.6c72c8c4@posting.google.com>

Chris de Souza wrote:

>We require the ability to run multiple Python sessions in parallel
>under VxWorks.

Python 'sessions' ... what does it mean ??
Well .. it"s a pity that VxWorks isn't a multi user OS. With other
OS'es you can run Python in different terminal sessions ...

>Has anyone developed a version of Python which supports
multi-threading
>different Python sessions?

Python does multi threading at OS level ..

We have re-compiled the stackless version of Python for QNX6 ... it
looks very interesting. Is Stackless Python an option ??? (w/o OS
threads, w/o the GIL)

Armin


> Chris de Souza    email: chrisd at hpato.aus.hp.com
>Australian Telecom Operation
>Hewlett-Packard Australia Ltd
>31 Joseph St, Blackburn 3130, Australia


From chris.gonnerman at newcenturycomputers.net  Mon Sep 23 22:10:39 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Mon, 23 Sep 2002 21:10:39 -0500
Subject: cursor positioning and color
References: <slrnaoqv2q.hn4.TuxTrax@fortress.tuxnet> <1b6dca50.0209231022.6fae7f6d@posting.google.com>
Message-ID: <00a701c2636f$9daa5260$0101010a@local>

----- Original Message -----
From: "brrrent" <bfarwick at juno.com>


> TuxTrax at fortress.tuxnet.net (TuxTrax) wrote in message
news:<slrnaoqv2q.hn4.TuxTrax at fortress.tuxnet>...
> > Being new to python, I would like to know how one goes
> > about such tasks as precise cursor positioning and output
> > to screen in color.
>
> Hi Mathew,
>
> Curses is the way to go when using Linux/Unix.
>
> Wconio is the way to go when using Microsoft.
> You can find in the Vaults of Parnassus.  If you have any
> trouble locating it, let me know, I can email it to you.
> Small library, works well.

Hey, thanks for the nice plug!  Actually, WConio is my
Pythonic adaptation of TCCONIO.C by Daniel Guerrero Miralles.
It's at:

    http://newcenturycomputers.net/projects/wconio.html

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From peter at engcorp.com  Sat Sep 14 16:44:30 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 14 Sep 2002 16:44:30 -0400
Subject: Converting a string to time - simple task but don't know how
 to do it!
References: <7A6D2E53461F2BF2.AC83A3D5E1A01468.48CC780BD6530537@lp.airnews.net>
Message-ID: <3d839f9d$1@news.sentex.net>

Richard Kessler wrote:
> I have a string returned from a database expressing the time i.e. 2002-09-14
> 10:30:34.  I need to compare this value to the current time to determine the
> amount of time elapsed between the two. I get the current time using
> time.time() but I cannot find a way to convert the string from the database
> into a time value (seconds from epoch) to do the simple subraction to find
> seconds elapsed.

If the format is exactly that, you could use something primitive like this:

 >>> s = '2002-09-14 10:30:34'
 >>> import re, time
 >>> tt = tuple(map(int, re.findall('(....)-(..)-(..) (..):(..):(..)', 
s)[0]) + [0,0,0])
 >>> time.mktime(tt)
1032017434.0

-Peter



From syver-en+usenet at online.no  Tue Sep 17 08:48:38 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Tue, 17 Sep 2002 12:48:38 GMT
Subject: faster way for adding many strings
References: <am6tgt$lqm$01$1@news.t-online.com> <3d8720b8$1@news.sentex.net>
Message-ID: <uu1ko4wa7.fsf@online.no>

Peter Hansen <peter at engcorp.com> writes:

> Ulli Stein wrote:
> > Hi,
> > What is the faster way: appending many strings to a list and then
> > joining them, or writing to StringIO and then getvalue()?
> 
> Alex' detailed response notwithstanding, there's another usual
> answer to your question:
> 
> "Premature optimization is the root of all evil in programming"
> (D. Knuth)
> 
> It means you usually shouldn't care which is faster until (a) your
> software is working fully, and (b) you have profiled the code
> and determined that in fact you have an unacceptable bottleneck
> which must be optimized to meet your requirements.
> 
> Don't get me wrong, there _are_ times to optimize.  You just
> shouldn't see them very often... and life is simpler that way. :)

Summarized: make it work, make it right, make it fast. (In that order).
 

-- 

Vennlig hilsen 

Syver Enstad


From andrew-pythonlist at puzzling.org  Fri Sep 27 21:59:26 2002
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sat, 28 Sep 2002 11:59:26 +1000
Subject: Python threading?
In-Reply-To: <AGpk9.11786$Lg2.3232428@news2.news.adelphia.net>
Message-ID: <20020928015926.GB450@frobozz.local>

Robert Oschler wrote:
> Having written _many_ state machines in my life, which async programming
> boils down to in one form or another, I feel it depends on the task.  When
> you have an async event driven model, the code that services a
> "transaction-package" ends up spreading functionality that services the
> "transaction-package" application's code landscape.  This can frequently
> be a headache.  As an example of a "transaction-package", take all the
> steps involved in servicing an incoming TCP connection, with custom
> authentication of a user, and perhaps a header and data object
> transacttion or two, finishing with a proper ending and cleanup of the
> transaction, and with subsequent notification of all necessary modules and
> data structures of the results of the transaction.  For dessert add a
> centralized clean error handler that can log errors propertly and
> gracefully recover the app regardless of where in the distributed event
> flow the error took place.

You want Twisted.  http://twistedmatrix.com/

It's an asynchronous networking framework which has an elegant abstraction
of callbacks and error handlers that I find make event-based programming a
breeze.  See http://twistedmatrix.com/documents/howto/defer for a detailed
description of how this works.  It makes it trivial to sequence a series of
operations on some event, regardless of when it happens.

In fact, here's a breakdown of your steps and Twisted:

1. "servicing an incoming connection": 
  twisted.internet does that for you.  It's dead easy.

2. "custom authentication of a user":
  Use the twisted.cred package.  This is exactly what it is designed for.
  It has a standard asynchronous interface, with ready-made implementations
  involving a simple python dict, and also a database-based implementation.

3. "perhaps a header and a data object transacttion":
  There isn't general transaction support in Twisted, but I imagine any
  object database you'd be using would support them, e.g. the ZODB.  And
  again, Deferreds make it easy to catch exceptions and so on (see my next
  points).

4. "centralized clean error handler than can log errors properly and
   gracefully recover":
  Twisted's Deferreds make that trivial.  In fact, if you forget to install
  an error handler, Twisted will log the error for you automatically.
  Deferreds callbacks and errbacks in many ways mirror your usual
  try/except constructs, so structured error handling is straightforward.
  See the doc I linked to above for more details.  Also, Twisted's main loop
  will never be crashed by an exception in one of the connections it
  services... in general, if something causes one connection to crash, the
  rest of Twisted merrily keeps on going regardless.

> Contrast this with a simple procedure call where a linear series of
> function calls are made, each with appropriate error handling and status
> adjustment, in a pleasant centrally located manner.

For instance, talking to a database asynchronously, Twisted-style:

    # See also http://twistedmatrix.com/documents/howto/enterprise
    deferred = db.runQuery('SELECT * FROM table')

Then hooking up a bunch of callbacks:
    deferred.addCallbacks(gotResult, somethingBroke)
    deferred.addCallbacks(doStep2, step2Rollback)
    deferred.addCallback(doStep3)   # step 3 doesn't care about rollbacks
    # ...etc...
    
    # Catch any uncaught errors
    deferred.addErrback(logError)

The results of each callback get fed into the next, so you can perform a
series of data transformations or whatever, e.g. perhaps you want something
like:
    # Note that addCallback returns the Deferred for your convenience
    deferred.addCallback(recordSetToXML).addCallback(applyXSLT)  # etc

Which looks like a linear series of function calls to me, and it certainly
has error handling (.addErrback) and status adjustment (you could easily be
updating an instance variable from within the callbacks).

And there's even more to Deferreds than I've described here!  Go read the
documentation :)

> There is an upcoming massively multiplayer game called EVE which is using
> Stackless Python which makes this point quite well.  Game AI suffers from
> the same design considerations as mentioned above.  Here's the link if
> you're interested (I'm not associated with them in any way, just
> considering taking a look at Stackless).  Look for question "8.5 How is
> the game logic implemented in EVE?":

Interesting.  Twisted is actually designed with exactly this goal in mind
(MMORPG).  Stackless is nice, but I think for what you're talking about
Twisted already solves your problems in "pure" Python :)

Twisted is an excellent framework.  I highly recommend you download it
(version 0.99.2 is still fresh off the press -- grab it while it's hot!),
play with it, bug us on the mailing list and irc channel (#twisted on
irc.freenode.net), and let us know what you think.  We think you might just
like it.

-Andrew.




From gerhard.haering at opus-gmbh.net  Wed Sep  4 05:45:13 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 4 Sep 2002 09:45:13 GMT
Subject: encryption (passwords)
References: <3d75d115$0$221$4d4ebb8e@news.nl.uu.net>
Message-ID: <slrnanblej.1rs.gerhard.haering@haering.opus-gmbh.net>

Gumuz wrote:
> I am quite a newbie regarding encryption stuff. There are some issues
> boggling my head and I suppose these come from the fact that I have no
> clue how to apply encryption in a real-life situation.
> 
> I am using network-sockets to communicate between two(or more) python
> applications. I want to encrypt these messages, or at least a part of it
> e.g. password and such.

Ok. I'd recommend to use SSL instead of inventing your own crypo layer. I'd
probably go with PyOpenSSL.

> My limited understanding of encryption tells me that I need a 'key' to
> decrypt stuff.

There are two kinds of encryption:
- symmetric encryption
- assymemtric encryption

Symmetric works by both ends having the same key. The sender encrypts the
message with the key, the receiver decrypts it using the same key. The
problem is that you have to distribute the keys first, and have a secure
way of distributing the keys.

Asymmetric encryption works by splitting the key up into a private and a
public key. The private key can only be used for decrypting, the public key
only for encrypting messages.

-- Gerhard


From user at domain.invalid  Mon Sep  9 14:44:02 2002
From: user at domain.invalid (ryan)
Date: Mon, 09 Sep 2002 18:44:02 GMT
Subject: old ODBC module
References: <f5ae9.2146$Qx1.35346@news1.telusplanet.net> <albcmn$2ia2$1@norfair.nerim.net>
Message-ID: <SZ5f9.12921$Qx1.427342@news1.telusplanet.net>

Perfect! Thanks.

ryan.

Gillou wrote:
> Google for "Mark Hammond win32 python extensions".
> It's in the package.
> 
> --Gilles
> 
> mailto:gilles at pilotsystems.net
> "ryan" <user at domain.invalid> a ?crit dans le message de news:
> f5ae9.2146$Qx1.35346 at news1.telusplanet.net...
> 
>>Does anyone know where I can find the old free odbc module? Not the
>>mxODBC module, but the other (apprently not very well supported) one.
>>I've been looking for a while without success. Thanks!
>>
>>ryan.
>>
> 
> 
> 
> 



From Padraig at Linux.ie  Mon Sep 30 12:22:52 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Mon, 30 Sep 2002 16:22:52 GMT
Subject: redhat8
References: <873crr4fy6.fsf@flibuste.net>
Message-ID: <3D987A16.2070401@Linux.ie>

William wrote:
> hi,
> 
> anybody know wich version of python is /usr/bin/python on the new redhat
> release ?
> 

You only need the first 10M of the iso to get this info:
(mount -o loop psyche-i386-disc1.iso somedir)

pyOpenSSL-0.5.0.91-1.i386.rpm             482027 Sep  3 23:36 
pygtk2-1.99.12-7.i386.rpm                1249637 Sep  5 05:42 
pygtk2-libglade-1.99.12-7.i386.rpm        142016 Sep  5 05:42 
python-2.2.1-17.i386.rpm                 4371618 Sep  3 22:35 
python-devel-2.2.1-17.i386.rpm            571688 Sep  3 22:35 
python-optik-1.3-2.noarch.rpm              44828 Sep  3 22:35

P?draig.



From see_reply_address at something.invalid  Tue Sep 17 20:51:22 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Wed, 18 Sep 2002 12:51:22 +1200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com>	<mailman.1032107006.6411.python-list@python.org>	<am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net>	<j4znuio0kr.fsf@informatik.hu-berlin.de>	<mailman.1032204706.14737.python-list@python.org>	<m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de>	<1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org>
Message-ID: <3D87CE0A.9090609@something.invalid>

Michael Chermside wrote:

>  (4) Stackless works on some architechures (including at least X86,
>      Sparc, PPC and Cygwin). But it does NOT work on some other
>      architectures. So (rather like the thread package) it is not
>      as portable as Python itself.


Also, I believe that, even on the supported
architectures, it plays dangerous tricks with
the C stack. I'd prefer to see that fixed
before making it a standard part of the
distribution.


-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From findler_lambda at yahoo.com  Sun Sep  8 15:39:04 2002
From: findler_lambda at yahoo.com (Robert Hanlin)
Date: 8 Sep 2002 12:39:04 -0700
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com>
Message-ID: <c427d639.0209081139.df3b0e1@posting.google.com>

Rod Stephenson <viking_kiwi at yahoo.com> wrote:
> "Python is cool to look at small bits of, but I think the "outline"
> syntax breaks down with larger chunks of code. 

Looks like Larry is simplifying or making a novice mistake.  Knuth
wrote that whitespace indentation is good when languages can let you
write small blocks of code.  Python does this for you.  In fact,
Python is set up to gently prod you to write clear chunks of code.

Perl is not.  It's a different philosophy.  If he believes there's
more than one way to do things, he should realize that Python has its
own way.

Rob


From ken at hotmail.com  Fri Sep 20 20:34:26 2002
From: ken at hotmail.com (Ken)
Date: Sat, 21 Sep 2002 10:34:26 +1000
Subject: Python question on printout to HTML
Message-ID: <amgeok$5kg5c$1@ID-49758.news.dfncis.de>

Hi, I wonder how you make the program to print out onto HTML such that it
prints at a time delayed setting of about 5-10 seconds between each line of
text so that the steps can be more readable to the user of the interface?

Thanks




From cotabas at earthlink.net  Tue Sep  3 21:47:48 2002
From: cotabas at earthlink.net (CotaBas)
Date: Wed, 04 Sep 2002 01:47:48 GMT
Subject: Executable Python Programs
References: <1rtc9.108$LI2.7238@newsread2.prod.itd.earthlink.net> <mailman.1030908003.28854.python-list@python.org> <Xns927D61C17CF84duncanrcpcouk@127.0.0.1>
Message-ID: <8Ddd9.6311$6i4.491988@newsread1.prod.itd.earthlink.net>

> > My personal opinion is that py2exe/Installer/etc. are for
> > intermediate/advanced Python programmers. The reason is that it's
> > sometimes difficult to fix the dependencies manually, if these tools are
> > unable to hunt them down automatically for you.
>
> I think you may be overstating the difficulty here. The vast majority of
> programs run through py2exe etc. with no problems at all. I can only think
> ot a couple of times that I have even had to think about the dependencies:
>
> I had a case recently with py2exe was when it found some extra
dependencies
> that I didn't want: I was using jabber.py which for not very good reasons
> imports site.py which indirectly imports a whole load of stuff such as
> tkinter. Explicitly excluding site.py fixed this. Admittedly it took me a
> little while to track down which import was the problem. An option to
> py2exe to dump out why it decided to pull in each module would have been
> helpful.
>
> The only time I had trouble with an import that wasn't being found was
when
> I used py2exe on a program using PIL. PIL demand loads a lot of converters
> for different graphics formats which was beyond py2exe's capabilities.
> Explicitly importing all the formats I wanted into the main program fixed
> that one.

I tried useing Installer and could not get it to work, but when I used
py2exe I got it to work just fine Thank you for all your help and I'm sorry
I didn't read the FAQ.....









From park997 at aol.comnotospam  Wed Sep 18 15:26:45 2002
From: park997 at aol.comnotospam (Park997)
Date: 18 Sep 2002 19:26:45 GMT
Subject: Tkinter:select directory dialog?
References: <0x3i9.13118$HY3.2916659@newsc.telia.net>
Message-ID: <20020918152645.03154.00006132@mb-dh.aol.com>

Thanks this did the trick. I am still using 
Python 2.0. The addition of askdirectory
to Python 2.2 was a good idea :-).

Wendell Cropper


>> from tkCommonDialog import Dialog
>> class Chooser(Dialog):
>>     command = "tk_chooseDirectory"
>>     def _fixresult(self, widget, result):
>>         if result:
>>             self.options["initialdir"] = result
>>         self.directory = result # compatibility
>>         return result
>>
>> def askdirectory(**options):
>>     return apply(Chooser, (), options).show()
>
>(hmm.  I think I've seen that code before...)
>
>the askdirectory function and a Directory class (same as
>the Chooser above) was added to the tkFileDialog module
>in Python 2.2:
>
>    from tkFileDialog import askdirectory
>
>    directory = askdirectory(
>        title="select directory",
>        mustexist=1
>    )
>
>    if not directory:
>        print "cancelled"
>    else:
>        print "selected", directory
>
></F>
>
><!-- (the eff-bot guide to) the python standard library:
>http://www.pythonware.com/people/fredrik/librarybook.htm
>-->
>
>




From mhuening at zedat.fu-berlin.de  Wed Sep 18 15:00:42 2002
From: mhuening at zedat.fu-berlin.de (Matthias Huening)
Date: 18 Sep 2002 19:00:42 GMT
Subject: BibTeX module?
References: <mailman.1032359243.24708.python-list@python.org>
Message-ID: <Xns928DD5BFDCA47mhueningzedatfuberli@130.133.1.4>

Pearu Peterson <pearu at cens.ioc.ee> wrote in 
news:mailman.1032359243.24708.python-list at python.org:

>   http://cens.ioc.ee/~pearu/misc/readbib.py
> 
> that I use to maintain my bib file database.
> 

Thanks. 
- Matthias


From henrik.motakef at web.de  Fri Sep 27 16:19:20 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 27 Sep 2002 22:19:20 +0200
Subject: [XML-SIG] well-formed xml
References: <JHEOKEOOLIGLDHCMAHMOAEOPEBAA.mark@mceahern.com> <E17uppg-0007LF-00@smtp.web.de> <mailman.1033154622.13438.python-list@python.org>
Message-ID: <878z1nuqwn.fsf@pokey.henrik-motakef.de>

"Mike Rovner" <mike at bindkey.com> writes:

> IIRC, double quotes are the must:
> s = '<a href="http://google.com/search?hl=en&amp;q=foobar">search</a>'

No, single quotes are perfectly OK.


From florian.proff.schulze at gmx.net  Mon Sep 30 15:46:24 2002
From: florian.proff.schulze at gmx.net (Florian Schulze)
Date: Mon, 30 Sep 2002 21:46:24 +0200 (Westeuropäische Sommerzeit)
Subject: Printing with python
Message-ID: <Mahogany-0.64.2-1396-20020930-214624.00@>

Hi!

I would like to know what solutions exist for printing (on paper) with
python. Preferable would be a crossplatform approach, but only for Windows
would be good enough. I also thought about using the reportlab pdf lib and
using the free Acrobat Reader for preview and printing. It has a nice API
and I don't think it will be too hard to send it to Acrobat Reader.

Any suggestions or thoughts about this?

Regards,
Florian Schulze





From jepler at unpythonic.net  Tue Sep 17 08:08:44 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 17 Sep 2002 07:08:44 -0500
Subject: Difflib question
In-Reply-To: <886c5e4b.0209161208.25012958@posting.google.com>
References: <886c5e4b.0209161208.25012958@posting.google.com>
Message-ID: <20020917120842.GB18327@unpythonic.net>

On Mon, Sep 16, 2002 at 01:08:44PM -0700, Robin Siebler wrote:
> Is there a way to get difflib to output *only* the lines that are
> different (instead of all of the lines)?

you can almost certainly do this by going "lower-level" yourself, and
using difflib.SequenceMatcher directly.  roughly:
    # create a sequence matcher object
    m = difflib.SequenceMatcher(None, s1, s2)

    # get the opcodes
    ops = m.get_opcodes()

    for op, i1, i2, j1, j2 in ops:
	# Do something with each opcode
	pass

see my pyunidiff program at http://unpythonic.net/~jepler/pyunidiff.py
for a more complex example.

Jeff



From 53ab2750 at meowing.net  Sat Sep 28 06:24:29 2002
From: 53ab2750 at meowing.net (itsy bitsy meowbot)
Date: Sat, 28 Sep 2002 10:24:29 GMT
Subject: print a %
References: <ff8044a.0209280211.345dc898@posting.google.com>
Message-ID: <xrfl9.4767$Bj5.2508@nwrddc02.gnilink.net>

Claude wrote:
> When I do 
> print "image size=%d%" %(50)
> I've got an error
> 
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> ValueError: incomplete format
> 
> I do not know how to escape the second %

print "image size=%d%%" %(50)

<http://www.python.org/doc/current/lib/typesseq-strings.html>


From christina at iaeste.at  Wed Sep 11 06:15:25 2002
From: christina at iaeste.at (Christina)
Date: 11 Sep 2002 03:15:25 -0700
Subject: function pointer
Message-ID: <ffcedbc9.0209110215.644c1788@posting.google.com>

Hi!

Does Python have function pointers? If not, is there a possibility to emulate them?

Christina


From sismex01 at hebmex.com  Thu Sep 26 19:22:12 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Thu, 26 Sep 2002 18:22:12 -0500
Subject: Who knows somefunction?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233026399A9@mail.hebmex.com>

> 
> Marco Mariani
> > Now you can harm yourself. :-)
> > 
> > Tested with python 2.2.1
> > 
> 
> ;-)
> 
> FWIW, against current cvs it crashes when you do:
> 
> from pointer import pointer
> x = range(3)
> y = id(x)
> del x
> x1 = pointer(y)
> print x1
> 
> Where's Houdini when you need him...
> 
>  
> --
> 
> Emile van Sebille
> emile at fenx.com
> 

Well, that's more or less expected; when you
do the "del x", the object is deallocated,
and y contains the address of rubbish.

I haven't compiled it yet; what happens
when you don't "del x" before you declare x1?

-gus











Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From codeapocalypse at msn.com  Tue Sep 17 18:41:53 2002
From: codeapocalypse at msn.com (Brian)
Date: 17 Sep 2002 15:41:53 -0700
Subject: Win32COM & Detaching from automation servers
Message-ID: <ef3269df.0209171441.2d6f8e59@posting.google.com>

I've searched this newsgroup for an answer to my problem, but haven't
come up with anything.  (Perhaps the solution is obvious and I'm just
not seeing the forest for the trees...)

I want to write a Python program on Windows which will use the
automation interface for Excel.  What I want it to do is launch Excel,
open a template, stuff data into it (based on command line parameters
to the Python app), detach from Excel and exit, leaving Excel open.

I know how to do everything except for the detach part.  Any help is
much appreciated.


From hartmut.goebel at noris.de  Wed Sep 11 12:08:09 2002
From: hartmut.goebel at noris.de (Hartmut Goebel)
Date: Wed, 11 Sep 2002 18:08:09 +0200
Subject: ANN: shellwords.py 0.1
Message-ID: <3D7F6A69.C5CD2912@crazy-compilers.com>

Announcing:

                         shellwords.py
                         =============
                          Version 0.1


Parse a string into words like a POSIX shell does.


Why this module?
----------------

Out in the wild there are quite a few modules for executing commands
in a sub-process. Most of them take a string (the command-line) as
input and use os.exec* for executing the program. This requires
splitting the string into words _exacly_like_the_shell_ does.

If this parsing/splitting is incorrect you can have quite some funny
time debugging. ;-)

Since I didn't find found any other module that-like, I decided to
develop this module.


Benefits using this module
--------------------------

Using this module has the following benefits:

- Enables your application or module to mimic word splitting like a
  POSIX shell without effort.

- Saves yourself debugginf-time then doing word-splitting.

- Avoids confusin the users of you application/module when splitting
  shell command lines into words, since this module behaves _exactly_
  like a POSIX shell does.

- The Unittest-Suite proves the correct word-splitting. Currently 75
  command lines are used, each testing a special pattern. The input
  data for this test-suite consists of command-lines which are split
  ba the shell on-fly fly. You can add your own test-patterns without
  any hassle.


Semantics
---------

This module parses a string into words according to the parings-rules
of a POSIX shell. These parsing rules are (quoted after 'man bash'):

1) Words are split at whitespace charakters; these are Space, Tab,
   Newline, Carriage-Return, Vertival-Tab (0B) and Form-Feet (0C).

   NB: Quotes do _not_ separate words! Thus
        "My"Fancy"Computer"
   will be parsed into a single word:
        MyFancyComputer

2) A non-quoted backslash (\) is the escape character. It preserves
   the literal value of the next character that follows.

3) Enclosing characters in single quotes preserves the literal value
   of each character within the quotes. A single quote may not occur
   between single quotes, even when preceded by a backslash.

   This means: baskslash (\) has no special meaning within single
   quotes. All charakters within single quotes are taken as-is.

4) Enclosing characters in double quotes preserves the literal value
   of all characters within the quotes, with the exception of \. The
   backslash retains its special meaning only when followed " or \. A
   double quote may be quoted within double quotes by preceding it
   with a backslash.

Availablity
-----------

'shellwords' is available for download at
http://www.crazy-compilers.com/py-lib/#shellwords

Requirements
------------

Requires Python >= 2.0

Frequently Asked Questions
--------------------------

Q: Hey, there is 'shlex' coming with Python. Why there is a need for
   this module?

A: I know 'shlex' and I gave it a try. But 'shlex' takes quotes as
   word-delemiters which divers from the shell-semantic (see above).
   And even if 'shlex' would parse strings as needed, I would have
   written a (very, very) thin layer above, since 'shlex' is simple
   but seldomly used for this kind of job.


Copyright/License
-----------------

(C) Copyright 2002 by Hartmut Goebel <h.goebel at crazy-compilers.com>

License: Python Software Foundation License
         http://www.opensource.org/licenses/PythonSoftFoundation.html


From maxm at mxm.dk  Fri Sep  6 03:56:17 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 06 Sep 2002 09:56:17 +0200
Subject: What database should I use
References: <mailman.1031257921.29400.python-list@python.org>
Message-ID: <3D785FA1.6090201@mxm.dk>

A wrote:
> I am going to program a small application (in Python and wxPython) which consists about 
> 2000 - 3000 records . Can you please recommend which database is the best for that. I 
> use Windows.
> Maybe MS Access  or FoxPro or MySQL or better?
> I want to keep the program as small as possible.



In such a small program why don't you just use shelve from the standard 
library?

You can store your objects directly in it and you don't have to install 
any db.


regards Max M



From loewis at informatik.hu-berlin.de  Fri Sep  6 09:11:22 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 06 Sep 2002 15:11:22 +0200
Subject: Python and Jython inconsistencies when encoding strings
References: <ala00d$dij$1@wanadoo.fr>
Message-ID: <j4ofbbclfp.fsf@informatik.hu-berlin.de>

"Andre Michel Descombes" <amdescombes at qualicontrol.com> writes:

> I've noticed the following inconsistency between Python 2.1 and
> Jython 2.1 : when I do the following in Python :

> >>> s = u'?' # Alt-0163, pound sign
> >>> print s

Please try

>>> s
u'\x9c'

Notice that Unicode U+009C is a control character, STRING TERMINATOR,
and not U+00A3, POUND SIGN. You may wonder where it gets the 9C value
from: This is because, in a console window, Microsoft uses code page
850 (probably), which uses

<U00A3>     /x9c         POUND SIGN

So when you enter a Alt-0163, the terminal reports the byte 0x9c, not
the byte 0x00A3. In Python 2.x, you can only use non-ASCII in Unicode
literal if it is Latin-1. However, your pound sign is not entered in
Latin-1, but entered in cp850. To really get a pound sign into a
Unicode literal, you need to write

>>> s = u"\u00A3"
>>> print s.encode("cp850")

You should use cp850 here, since your terminal uses cp850, not
Latin-1.

> now when I do the same thing in Jython 2.1 I get this:
> 
> >>> s = u'?' # Alt-0163, pound sign
> >>> print s
> ?
> >>>
> 
> and if I do:
> >>> print s.encode('latin-1')

This is interesting, since you also get

>>> s
u"\u0153"

Now, U+0153 is LATIN SMALL LIGATURE OE. It so happens that \x9c (what
the terminal sends) is U+0153 in CP 1252 (which is the ANSI code page
on your Windows installation). This might be a bug in Java, which
assumes that bytes sent by the terminal are in the ANSI code page,
when they are really in the OEM code page.

> Does anybody know what is causing this inconsistency? Is there any way to
> avoid it?

Yes. Don't use the console.

Regards,
Martin


From oren-py-l at hishome.net  Tue Sep 17 08:30:28 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Tue, 17 Sep 2002 08:30:28 -0400
Subject: More on Protecting Source Code
In-Reply-To: <GCEDKONBLEFPPADDJCOEOEEKFAAA.whisper@oz.net>
References: <20020917062412.GA85457@hishome.net> <GCEDKONBLEFPPADDJCOEOEEKFAAA.whisper@oz.net>
Message-ID: <20020917123028.GA19214@hishome.net>

On Tue, Sep 17, 2002 at 03:22:20AM -0700, David LeBlanc wrote:
> Oh yeah, and my views of intellectual property seems to be a lot 
> closer to the law of the land then yours.

The Law is not a view, it is a set of rules and enforcement mechanisms.
Our views are personal interpretations of the Law in the context of the 
society in which we live and our moral values.

Here are two alternative interpretations of information restriction
laws:

1. Information and knowledge is a scarce resource and should be closely
held and treated as property just like physical goods. Information
belongs to its creator or any entity to which it has been sold.
Transmission, storage or use of information without the consent of its
owner is theft.  Expiration of copyrights and patents? Ummm.... must be
some bug in the system. It will be fixed ASAP.

2. Information and knowledge is abundant. Information belongs to all of
mankind. Sharing information promotes creativity and thus creates even
more information, knowledge and wealth. In order to encourage
individuals to create new valuable information and distribute it the
state offers them an arrangement whereby they can use the power of the
courts and police to restrict the rights of others to use this
information in certain ways.  These include restricting the right to
copy and distribute information for profit (copyright) and restricting
the right to build and sell goods based on the information (patents).
This extraordinary power must be limited: it is only granted for a
limited period and certain types of fair use are protected from it. In
the case of patents the the inventor must fully disclose his invention
in the patent application so when the restriction period expires it
enters the public domain for the benefit of society instead of being
lost as a trade secret. 

The first view may be a lot closer to the way the law of the land is
seen by corporations and the politicians they buy. The second view is
much closer to the intentions and beliefs of the noble people who
originally drafted these laws. It grieves me to see how these laws have
been perverted over time. I hope that my participation in the open
source community makes a tiny contribution to improving the situation.

We both respect the same laws, we just see them differently.  For me,
information restriction laws are a necessary compromise of my freedom
while you appear to treat them as an inalienable right.

	Oren




From pygta at engcorp.com  Mon Sep 16 21:46:25 2002
From: pygta at engcorp.com (Peter Hansen)
Date: Mon, 16 Sep 2002 21:46:25 -0400
Subject: Final reminder of PyGTA (Toronto Python/Zope group) first meeting
 Tuesday night
Message-ID: <3d868963$1@news.sentex.net>

This is the final reminder of the first meeting of the new
Toronto-area Python and Zope user group.  The meeting will
take place Tuesday, September 17 at 7:00 pm the Willowdale
United Church at 379 Kenneth Ave in North York (just to the
south-east of the Finch subway station).

Details including an agenda and a map are available here:

   http://web.engcorp.com/pygta/wiki/FirstMeeting

We hope to see you out there.  ActiveState has generously
donated some goodies which we'll distribute Tuesday and
at the second meeting.  (Currently we have roughly twenty
people registered to attend but you don't need to register
to come out!)

-Ian Garmaise, Peter Hansen, and the rest of the PyGTA
  organizing committee



From peter at engcorp.com  Wed Sep  4 08:35:04 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 04 Sep 2002 08:35:04 -0400
Subject: newsgroup for zope
References: <al4tru$o3j$03$1@news.t-online.com>
Message-ID: <3d75fddc@news.sentex.net>

stephan wrote:
> Hi perhaps a stupid question,
> 
> but does there exist a newsgrup for zope centric items?
                          ^^^^^^^^
You spelled "newsgroup" wrong.  It's spelled "mailing list" and
you can find out about them here:

    http://www.zope.org/Resources/MailingLists

:-)

-Peter

(There's no actual newsgroup though.)



From marco at sferacarta.com  Fri Sep  6 11:24:18 2002
From: marco at sferacarta.com (Marco Mariani)
Date: Fri, 6 Sep 2002 17:24:18 +0200
Subject: ttySx
In-Reply-To: <200209061708.25677.b.maryniuk@forbis.lt>
References: <200209061708.25677.b.maryniuk@forbis.lt>
Message-ID: <20020906152418.GA6937@sferacarta.com>

On Fri, Sep 06, 2002 at 05:08:25PM +0200, Bo M. Maryniuck wrote:

> I've tried do somethig like:
> 
> modem = open('/dev/ttyS0', 'r+')
> modem.write('ATZ\r')
> modem.read() # Hang here forewer
> 
> This is bad. Then I've tried use several modules for serial port, wheere I can 
> write, but if I read, always get ''.
> 
> Perhaps I should figure out with DTR status and so on.... Any ideas?


Here it is.
Of course, the code is old and sucks. If you have problems with lock
files, I don't want to know :-)

If you or anybody patch it, I would like to know..





#
# modem.py
#

import os, time, signal

from termios import *


def timeout(signum, frame):
    raise IOError, "Timeout!"

speeds = {
    2400:B2400,
    4800:B4800,
    9600:B9600,
    19200:B19200,
    38400:B38400,
    57600:B57600,
    115200:B115200,
    230400:B230400,
    460800:B460800
    }



class Modem:
    devname=None
    lockname=None
    locked=0
    fd=-1
    old_term=None
    new_term=[
    #       iflag       oflag       cflag
            IGNPAR,     0,          CRTSCTS | CS8 | CLOCAL | CREAD,
    #       lflag       ispeed      ospeed
            0,          0,          0,
    #       cc
            ['\000']*32
    ]
    answer=[]


    def __init__(self,name):
        self.devname = name
        self.lockname = '/var/lock/LCK..%s' % name

    def purge(self):
        if self.fd>=0:
            self.alarm_on()
            try:
                os.close(self.fd)
            except:
                pass
            self.alarm_off()

        if self.locked:
            os.unlink(self.lockname)


    #
    # 0:  OK
    # -1: Errore
    #
    def __lock(self,last_try=0):
        try:
            fd = os.open(self.lockname, os.O_RDWR | os.O_EXCL | os.O_CREAT, 0644)
            os.write(fd,'%10s\n' % os.getpid())
            os.close(fd)
            self.locked = 1
            return
        except:
            if not last_try:
                time.sleep(0.2)
                pid = int(open(self.lockname,'r').readline())
                try:
                    os.kill(pid,0)
                    return -1
                except OSError:
                    os.unlink(self.lockname)
                    return self.__lock(last_try=1)
            else:
                return -1

    def alarm_on(self):
        signal.signal(signal.SIGALRM, timeout)
        signal.alarm(1)

    def alarm_off(self):
        signal.alarm(0)


    #
    # 0:  OK
    # -1: Errore
    #
    def open(self):
        if self.__lock():
            return -1

        self.alarm_on()

        try:
            self.fd = os.open ('/dev/'+self.devname, os.O_RDWR | os.O_NOCTTY);
        except:
            self.fd=-1

        self.alarm_off()

        return self.fd<0

    def term_init(self):
        try:
            self.old_term = tcgetattr(self.fd)
            self.new_term[6][VMIN]=1
            self.new_term[6][VTIME]=0
            tcflush(self.fd,TCIFLUSH)
            tcsetattr(self.fd,TCSANOW,self.new_term)
            self.set_speed(B2400)
            return 0
        except:
            return -1

    def term_reset(self):
        self.alarm_on()

        try:
            tcsetattr(self.fd,TCSANOW,self.old_term)
            retval = 0
        except:
            retval = -1

        self.alarm_off()
        return retval

    def set_speed(self,rate):
        if rate in speeds.keys():
            rate = speeds[rate]
        self.new_term[4] = rate
        self.new_term[5] = rate
        tcsetattr(self.fd,TCSANOW,self.new_term)

    def sendline(self,line):
        if not os.write(self.fd,line+'\015'):
            return 1

        time.sleep(0.2)

        self.answer=[]
        s=''
        while 1:
            self.alarm_on()
            try:
                c = os.read(self.fd,1)
            except:
                retval = -1
                break

            if c=='\015' or c=='\012':
                if s:
                    self.answer.append(s)
                if s=='OK':
                    retval = 0
                    break
                if s=='ERROR':
                    retval = 1
                    break
                s=''
            else:
                s += c

        self.alarm_off()

        return retval




From kennethg at pd.jaring.my  Fri Sep 27 23:50:53 2002
From: kennethg at pd.jaring.my (Kenneth Gomez)
Date: Sat, 28 Sep 2002 03:50:53 GMT
Subject: Local Python Server Installation
References: <3d8b47c6.277669@news.jaring.my> <slrnaomn6m.39s.gerhard.haering@haering.opus-gmbh.net> <3d8c993c.2190426@news.jaring.my> <mailman.1032625889.13783.python-list@python.org> <3d8d4009.375344@news.jaring.my> <bQYk9.1630$6g7.1090@fe03> <3d947c26.2656654@news.jaring.my> <JR3l9.2341$79.998@fe03>
Message-ID: <3d952602.1791864@news.jaring.my>

Dear Steve,

I managed to get it to work!!

I didnt configure the ScriptAlias directory to where my scripts were.
How silly of me!

Thanks a lot Steve.

Regards,
Kenneth
Cheers,
Kenneth Gomez.


From aurin42 at yahoo.com  Tue Sep  3 04:06:05 2002
From: aurin42 at yahoo.com (Ante Bagaric)
Date: Tue, 3 Sep 2002 10:06:05 +0200
Subject: Add methods to a class at runtime?
References: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com> <slrnan4dl3.18h.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <al1qki$6aig$1@as201.hinet.hr>

> reduce(lambda x,y:x+y,map(lambda
x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))

I like your sig :)






From beall at psych.ucsb.edu  Thu Sep 26 14:16:03 2002
From: beall at psych.ucsb.edu (Andy Beall)
Date: Thu, 26 Sep 2002 11:16:03 -0700
Subject: distributing apps w/embedded python & missing modules
Message-ID: <amvihs$252l$1@yuggoth.ucsb.edu>

I working on an application in which I embed python to allow the end-user to
customize the apps behavior with scripts.  My question: Is there an easy way
to compile all of Python's modules into the app so that the user does not
need to separately download the core python package?  If there's standard
explanation for that which I've missed, just a reference to that would be
greatly appreciated.

Thanks,
Andy-





From tchur at optushome.com.au  Sat Sep  7 17:37:34 2002
From: tchur at optushome.com.au (Tim Churches)
Date: Sun, 08 Sep 2002 07:37:34 +1000
Subject: Smoothing a discrete set of data
References: <3csmdso5.fsf@morpheus.demon.co.uk> <B99F9C2F.1238E%pereira@cis.upenn.edu>
Message-ID: <3D7A719E.B8AE64E8@optushome.com.au>

Fernando Pereira wrote:
> 
> On 9/7/02 6:02 AM, in article 3csmdso5.fsf at morpheus.demon.co.uk, "Paul
> Moore" <gustav at morpheus.demon.co.uk> wrote:
> > I have a set of data, basically a histogram. The data is pretty
> > variable, and I'd like to "smooth" it to find trends. [...] When I've tried to
> formalise this, I've always stumbled over the fact
> > that it's a trade-off exercise (there most be some sort of "tolerance"
> > parameter in the algorithm) - on the one hand, the original data is
> > "best" (in the sense that it matches exactly) whereas on the other
> > hand, a single horizontal line at the average is "best" (fewest number
> > of steps).
> >
> > My instinct is that you're looking for something that balances number
> > of steps vs difference from the original data. But for the life of me
> > I can't see how to make this intuition precise.
> >
> > Can anyone help me? I can't believe that this problem has never come
> > up before, but I can't find any literature on it.
> You must have been using the wrong search engine <wink> This is the problem
> of *regression*, extensively studied in statistics and machine learning. The
> tradeoff you mention is discussed in the literature under headings like
> "bias-variance tradeoff", "generalization bounds", "structural risk
> minimization"... If you want a single good book that covers this and related
> topics with a minimum of prerequisites (elementary calculus and linear
> algebra, a teeny bit of probability), I recommend
> 
> <http://www-stat.stanford.edu/~tibs/ElemStatLearn/>

I agree that the problems with which Paul is wrestling are addressed by
the 
discipline of modern statistics. However, I suspect that the
abovementioned
book by Hastie, Tibshirani and Friedman, although truly excellent, might
be
a bit too specialised for Paul's needs. Some introductory texts which
cover
various regression techniques (linear, non-linear, local, polynomial
etc) might be
more appropriate.

BTW, Paul, you describe your data as a "histogram" but say you want to
find "trends". 
"Histogram" implies that the data represent a frequency distribution,
but looking for 
"trends" implies that they are a time-series - you need to be clear in
your own mind
about this because different smoothing and regression techniques are
used for each.

Of course, there is no need to write your own routines to implement
these 
statistical techniques. AFAIK, there is no native Python package which
implements a
comprehensive range of regression and smoothing techniques, but I can
highly
recommend the R package for statistics and statistical graphics - this
mature, free, open
source package has more statistical facilities than you are ever likely
to need. 
It runs on all major platforms, and some excellent introductory (free)
texts are available for it - see http://www.r-project.org

If you are using a Unix, Linux or Mac OS X system, you can even use R
from within Python, thanks to Walter Moreira's excellent RPy module  -
see http://rpy.sourceforge.net

Hope this helps,

Tim C



From aleax at aleax.it  Wed Sep 18 10:12:14 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 18 Sep 2002 14:12:14 GMT
Subject: monetary applications [was: Python GUI app to impress the boss?]
References: <mailman.1032348444.6783.python-list@python.org>
Message-ID: <2R%h9.109893$ub2.2324479@news1.tin.it>

Mark McEahern wrote:

> [Oren Tirosh]
>> I don't understand what exactly you are complaining about but is sounds
>> interesting - can you elaborate?
> 
> Perhaps the lack of a FixedPoint type?

http://starship.python.net/crew/aahz/FixedPoint.py


Alex



From dstuartkelly at online.ie  Thu Sep 12 05:09:51 2002
From: dstuartkelly at online.ie (Daniel Stuart Kelly)
Date: Thu, 12 Sep 2002 10:09:51 +0100
Subject: Anyone got a logfile script? - Help a n00b
Message-ID: <1031821791.3d8059df94ea8@mail.online.ie>

What I'm looking for is a little script when run will dig into a log file and 
look for a variable.

When it finds the variable I want it to print that line and the 20 or so right 
above it to either screen or a plaintext file. 

I know it sounds simple, but I've only started learning this and I've no idea 
where to begin.. Any tips or pointers would also be helpful...

Thanks
/Daniel





From eugene at ulka.com  Thu Sep 19 12:32:53 2002
From: eugene at ulka.com (Eugene Pervago)
Date: 19 Sep 2002 09:32:53 -0700
Subject: Loading modules from several directories
References: <a1247c5b.0209171610.243f3c02@posting.google.com> <33803989.0209172247.2e6afc1f@posting.google.com> <a1247c5b.0209180607.843526e@posting.google.com> <6Q1i9.1149$bh1.989@fe03> <a1247c5b.0209181438.3a1c0c9a@posting.google.com> <lS7i9.2127$ik2.2079@fe02>
Message-ID: <a1247c5b.0209190832.77989da@posting.google.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message news:<lS7i9.2127$ik2.2079 at fe02>...
> "Eugene Pervago" <eugene at ulka.com> wrote ...
> > They do not intersect. Actually the structure is something like this:
> > /onedir/foo
> >    dbutils.py
> >    utils.py
> >    auth.py
> >
> > /anotherdir/foo - Nothing
> > /anotherdir/foo/projectA
> >     User.py
> >     Product.py
> > [snip]
> 
> I'm at something of a loss to understand why both directories have to be
> called foo and what the relationship is between them. Why not foo1 and foo2?
> Am I missing something?

In this case 'foo' would be the company name. I just wanted to avoid
polluting top-level namespace and have all my modules within the foo.*
namespace.

Eugene Pervago


From jknapka at earthlink.net  Thu Sep  5 17:57:20 2002
From: jknapka at earthlink.net (Joseph A. Knapka)
Date: Thu, 05 Sep 2002 21:57:20 GMT
Subject: Python class ( Prolog ) ???
References: <mailman.1031234848.13265.python-list@python.org>
Message-ID: <3D77D340.A1D6641@earthlink.net>

> Luis wrote:
> 
> Does anyone have a class that provides the functionality of prolog ???

There's a SourceForge project that enables you to call
GNUProlog from Python:

<URL: http://bedevere.sourceforge.net/>

I've seen a similar product for SWI Prolog as well, but the
reference escapes me at present.

It would be possible to implement a Prolog-like inference
engine in Python, but it would be rather slow compared to
genuine Prolog. Of course, for simple things that might
not be a problem. I have a Prolog-like unifier written in
Python lying around somewhere, which I'll track down
and post if you're interested. It's not a complete
inference engine, just the unification piece, but
once you've got that building a simple resolver
is not too difficult. If you want to implement cut, or
things with side effects, it may get a bit hairy.

Cheers,

-- Joe
  "I'd rather chew my leg off than maintain Java code, which
   sucks, 'cause I have a lot of Java code to maintain and
   the leg surgery is starting to get expensive." - Me


From jepler at unpythonic.net  Thu Sep 19 18:19:53 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Thu, 19 Sep 2002 17:19:53 -0500
Subject: sorry for this stupid question about struct module
In-Reply-To: <3d8a4ae7$0$280$7a628cd7@news.club-internet.fr>
References: <3d8a4ae7$0$280$7a628cd7@news.club-internet.fr>
Message-ID: <20020919171953.A3473@unpythonic.net>

>>> '\x00\x00\x00\n' == '\x00\x00\x00\x0A'
True
'\x00\x00\x00\n'

Ascii code 10 is repr()'d as \n.

If you don't like this, you could use a different function to print the
quoted value of a string:

>>> quote = lambda s: "'" + "".join(["\\x%02x" % ord(c) for c in s]) + "'"
>>> print quote("\0\0\0\n")
'\x00\x00\x00\x0a'

Jeff



From edream at tds.net  Mon Sep  2 07:48:00 2002
From: edream at tds.net (Edward K. Ream)
Date: Mon, 02 Sep 2002 11:48:00 GMT
Subject: Does IDLE handle unicode?
Message-ID: <QdIc9.2412$Ak.1132837@kent.svc.tds.net>

IDLE took an exception writing a .py file containing the copyright
character.  This happened a while ago; IIRC, I lost the file.

Is this a known problem?  I see nothing about this in the FAQ.

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From tismer at tismer.com  Sun Sep  8 10:30:17 2002
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 08 Sep 2002 16:30:17 +0200
Subject: Python plugins
References: <alf7rc$1p06l1$1@ID-41653.news.dfncis.de>
Message-ID: <3D7B5EF9.60909@tismer.com>

Tobias Klausmann wrote:
> Hi there,
> 
> For an open source project that needs to be extensible by its
> users, I would like to create a plugin architecture. Mind you,
> I'm not trying to use Python as a plugin language only like XChat
> or Gnumeric do. Let me illustrate:
> 
> I have a directory somewhere in PYTHONPATH (let's just assume
> it's in the CWD) that contains lots and lots of *.py files with a
> defined set of functions (or, at least one: register()). What I'd
> like to do now is being able to load these plugins at program
> start and reload them if I'd like to later. Problem is that the
> program has to figure out by itself what *.py files are there -
> and I have to make sure somehow that I get a grip on the loaded
> files.

You can avoid exec at all.
The PIL package (Python Imaging Library) from Fredrik Lundh
has a similar problem:
There are lots of image drivers in a directory, and PIL
cannot know in advance what's there. Here an excerpt from
Image.py, that shows what it does:

"""
def init():
     "Load all file format drivers."

     global _initialized
     if _initialized >= 2:
         return

     visited = {}

     directories = sys.path + [os.path.dirname(__file__)]

     # only check directories (including current, if present in the path)
     for directory in filter(isDirectory, directories):
         fullpath = os.path.abspath(directory)
         if visited.has_key(fullpath):
             continue
         for file in os.listdir(directory):
             if file[-14:] == "ImagePlugin.py":
                 f, e = os.path.splitext(file)
                 try:
                     sys.path.insert(0, directory)
                     try:
                         __import__(f, globals(), locals(), [])
                     finally:
                         del sys.path[0]
                 except ImportError:
                     if DEBUG:
                         print "Image: failed to import",
                         print f, ":", sys.exc_value
         visited[fullpath] = None

     if OPEN or SAVE:
         _initialized = 2
"""

...

> Now I can just "import pi" and execute pi.rel() which in turn
> loads all the modules that are there ('cept the ones that only
> exist as pyc or pyo, but that's okay). But now I have two
> problems: first, I have no idea how to call, say
> pi.someplugin.dostuff(args) except constructing each and every
> call as a string and exec() it - which is both error-prone and
> awkward IMHO.

If you executed a proper __import__, all modules and funcs should
be accessible in the normal way.
There is no need to exec.

> Second, reloading just does not seem to work right as Python is
> too smart ;) for me: it knows all of them have been loaded. So I
> added the reload code above but to no avail.

Reloading should work. If you have imported some module into
your current one, reload(module) should work as expected.
You will get into trouble if you imported the contents of
a module like "from mod import *", since these identifiers will
not get refreshed when you reload. You would need to repeat
the from... import after reloading, since it copies the
identifiers again to your namespace.

ciao - chris
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From loewis at informatik.hu-berlin.de  Mon Sep  2 12:26:32 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 02 Sep 2002 18:26:32 +0200
Subject: version independant object usage counting
References: <mailman.1030977797.31291.python-list@python.org>
Message-ID: <j48z2kjr2f.fsf@informatik.hu-berlin.de>

"Rolf Kalbermatter" <rolf.kalbermatter at citeng.com> writes:

> > That is only a minor problem: those macros are correct across all
> > released Python versions, back to 1.4 or earlier.
> 
> But there is a serious difference between the DEBUG and NON-DEBUG
> PyObject_HEAD structure layout which I would like to not have to worry
> about if possible. 

I see.

> Does anyone see any problems with this code?

Looks fine.

> >   * PyObject_INIT relies on initializing the object members.
> > - PyType_HasFeature relies on the position of the tp_flags field
> 
> I'm not using these function at all until now. May have to do so when
> I start to add an extension module however?

You might be using HasFeature implicitly, e.g. in PyIter_Check, or
PyType_IS_GC. You might be using PyObject_INIT if you define types.

In short, I agree with you that it is desirable to have function
versions of everything that implies knowledge about the layout -
however, as soon as you start extension modules (e.g. sre, socket),
you will find that you require a specific version of the pythonxy.dll,
anyway.

Regards,
Martin


From mark at mceahern.com  Mon Sep  9 14:33:10 2002
From: mark at mceahern.com (Mark McEahern)
Date: Mon, 9 Sep 2002 13:33:10 -0500
Subject: shell expansion
Message-ID: <JHEOKEOOLIGLDHCMAHMOMEHFDNAA.mark@mceahern.com>

Is this supposed to work:

  filename = "$HOME/foo"
  f = file(filename)
  ...

It doesn't seem to.

// m
-




From lu_gio at hotmail.com  Sun Sep 22 19:55:44 2002
From: lu_gio at hotmail.com (Max)
Date: Sun, 22 Sep 2002 23:55:44 GMT
Subject: OpenGL alpha channel not working
Message-ID: <4Msj9.150251$pX1.5386352@news2.tin.it>

"""
Hi!
I wrote this little test... it's working without errors, only strange
behavor.
I have 2 texures RGBA, one loaded and blitted into another and one generated
with pygame.font.
Can someone tell me why the font over the picture is working and the picture
over the font is not??

If you want to test it, please sobstitute the GIFNAME and COLORKEY.
Thanks
"""
#-----------------------------------------------BEGIN
#-----------------------------------------------BEGIN
#-----------------------------------------------BEGIN

GIFNAME='fish1ne.gif'  #------------- PUT your image here!!! mine was 60x42
gif with colorkey (255,0,255)
COLORKEY=(255,0,255)

from pygame.locals import *
from OpenGL.GL import *
from OpenGL.GLU import *

import os.path
import sys
import pygame

def TestI():
    pygame.init()
    pygame.display.set_mode((640,480), OPENGL|DOUBLEBUF|HWSURFACE,32 )

    glMatrixMode(GL_PROJECTION)
    glLoadIdentity()
    glOrtho(0,640,480,0,0,-1)

    glMatrixMode(GL_MODELVIEW)
    glLoadIdentity()

    glEnable(GL_BLEND)
    glBlendFunc (GL_SRC_ALPHA, GL_ONE_MINUS_SRC_ALPHA)
    glEnable(GL_TEXTURE_2D)
    glEnable(GL_DEPTH_TEST)

    psurf=pygame.surface.Surface((16*8,16*8),SRCALPHA,32)
    psurf.set_colorkey((0,0,0,0))

    immge = pygame.image.load(GIFNAME)
    immge.set_colorkey(COLORKEY)

    psurf.blit(immge,immge.get_rect())
    spsurf = pygame.image.tostring ( psurf, "RGBA" )

    tx1 = AllocateTexture(psurf)

    mfont=pygame.font.Font(None,64)
    im = mfont.render('A',0,(255,255,255))
    im.set_colorkey((0,0,0,0))

    tx2 = AllocateTexture(im)

    while 1:
        event = pygame.event.poll()
        if event.type == QUIT or (event.type == KEYDOWN and event.key ==
K_ESCAPE):
            sys.exit()

        glClear(GL_COLOR_BUFFER_BIT|GL_DEPTH_BUFFER_BIT)

        glPushMatrix()
        glTranslate(100,100,0.5)
        mdraw(psurf, tx1)

        glPushMatrix()
        glTranslate(100,0,0)
        glScalef(5,5,0)
        mdraw(im, tx2)

        glPushMatrix()
        glTranslate(300,100,0)
        mdraw(psurf, tx1)

        glPushMatrix()
        glTranslate(300,0,0.5)
        glScalef(5,5,0)
        mdraw(im, tx2)

        pygame.display.flip()
        pygame.time.wait(10)

def mdraw(sf, txx):
    glBindTexture(GL_TEXTURE_2D, txx)
    w, h=sf.get_size()

    glBegin(GL_QUADS)

    glTexCoord2f(0, 0)
    glVertex2i(0, 0)

    glTexCoord2f(1, 0)
    glVertex2i( w, 0)

    glTexCoord2f(1,1)
    glVertex2i( w, h)

    glTexCoord2f(0,1)
    glVertex2i(0, h)

    glEnd()
    glPopMatrix()

def AllocateTexture(surface):
    image = pygame.image.tostring ( surface, "RGBA" )
    ix, iy = surface.get_size ()

    idImg = glGenTextures(1)
    glBindTexture(GL_TEXTURE_2D, idImg )
    glPixelStorei(GL_UNPACK_ALIGNMENT,1)

    gluBuild2DMipmaps(GL_TEXTURE_2D, 4, ix, iy, GL_RGBA, GL_UNSIGNED_BYTE,
image)
    return idImg

TestI()

#-----------------------------------------------END
#-----------------------------------------------END
#-----------------------------------------------END





From theller at python.net  Fri Sep 13 15:48:41 2002
From: theller at python.net (Thomas Heller)
Date: Fri, 13 Sep 2002 21:48:41 +0200
Subject: Problem with HTTPS request
References: <uo2fveepf88175@corp.supernews.com> <b0f083db.0209130558.2aa60398@posting.google.com> <uo45ara0l62cdc@corp.supernews.com> <slrnao4dkn.288.gerhard.haering@lilith.my-fqdn.de> <mYqg9.8349$WV.1529@fe03>
Message-ID: <altfeq$10gm9$1@ID-59885.news.dfncis.de>

"Steve Holden" <sholden at holdenweb.com> wrote in message news:mYqg9.8349$WV.1529 at fe03...
> "Gerhard H?ring" <gerhard.haering at gmx.de> wrote ...
> > Christian Reyes wrote in comp.lang.python:
> [...]
> > >> "Christian Reyes" <creyes at ea.com> wrote in message
> > > news:<uo2fveepf88175 at corp.supernews.com>...
> > ^^
> >
> > Broken quoting.
> >
> > >> > [...] Note: HTTPS support is only available if the socket module was
> compiled
> > > with
> > ^^
> >
> > Broken quoting again. Please throw away that OE crapware or fix it
> > using Morver or whatever.
> >
> Gerhard:
>
> You're showing distinct signs of having too much time on your hands :-)
> Can't you fix your brain instead so it will let you ignore other people's
> mailer problems?

At least the pointer to Morver was interesting...

Thomas




From 2002 at weholt.org  Thu Sep 19 08:24:46 2002
From: 2002 at weholt.org (Thomas Weholt)
Date: Thu, 19 Sep 2002 12:24:46 GMT
Subject: StructuredText -> HTML module??
References: <wbii9.19869$0p1.359879@news2.ulv.nextra.no> <3d89b69d$1@news.sentex.net>
Message-ID: <imji9.23500$sR2.405010@news4.ulv.nextra.no>

I need to take a piece of structured text as a string into a method and get
html out in the other end. As described on the page on Zope.org under
StructuredText. My hacking with PyWiki gave me a list where each element had
a '<pre>' in front of it etc. I might be using it the wrong way, but ....

I was under the impression that StructuredText as described on the site
mentioned earlier was sort of a standard, that the parser etc. would produce
the same output.

All I'm actually after is to be able to write stuff like :

**Header**

List:
1. Foo
2. Bar
3. Python

    Heres an indented section of text.

And turn this into

<h1>Header</h1>

List:
<ol>
   <li>foo</li>
  <li>bar</li>
  <li>Python</li>
</ol>

<P>
<blockquote>
    Heres an indented section of text.
<blockquote>
</P>

Or something similar. Since there's allready something that I believed was
standardized it would be easier to use modules allready available than doing
it all from scratch.

Hope that makes it clearer.

Best regards,
Thomas
"Peter Hansen" <peter at engcorp.com> wrote in message
news:3d89b69d$1 at news.sentex.net...
> Thomas Weholt wrote:
> > Need a module to parse StructuredText/Wiki into HTML. Looked at PyWiki,
but
> > the generated code was not what I expected. Does anybody have module or
lib
> > to do this?
>
> Maybe you could help us by telling us *what* you expected.
> Otherwise we're likely to waste our time making suggestions
> that are just as inadequate from your point of view.
>
> (For example, ZWiki... or the Zope structured text stuff.)
>
> -Peter
>




From ouhfos at dwdddaae.com  Wed Sep  4 21:17:41 2002
From: ouhfos at dwdddaae.com (ouhfos at dwdddaae.com)
Date: Thu, 5 Sep 2002 01:17:41 +0000 (UTC)
Subject: JA FOSTE HUMILHADO ???  3910
Message-ID: <1031187856.681346@spynews5>

>>>>   http://burlasdoleo.cjb.net    <<<<

ACTIVISM - HELP STOP A CRIMINAL ENTERPRISE OF DESTROYING PEOPLE
ACTIVISMO - AJUDE A PARAR UMA EMPRESA CRIMINOSA DE DESTROIR PESSOAS

>>>>   http://burlasdoleo.cjb.net    <<<<
mumk



From ark at research.att.com  Tue Sep 24 09:44:48 2002
From: ark at research.att.com (Andrew Koenig)
Date: Tue, 24 Sep 2002 13:44:48 GMT
Subject: confusion about opening files
References: <3d8fcda4$0$198$75868355@news.frii.net> <3D8FCFF3.FD52AFE5@alcyone.com> <3d8fd215$0$192$75868355@news.frii.net> <3D8FD2F3.81F3831C@alcyone.com>
Message-ID: <yu99n0q7h57z.fsf@europa.research.att.com>

Erik> Because there are two different open functions here.  One is in the
Erik> builtins, one is in the os module.  They are not the same:

In fact, the builtin "open" function is now called "file"; the name
"open" is there only for backward compatibility.

So the original example:

>>> import os
>>> file = os.open("e:\\d12", "r")

is bad news for two reasons:

        1) It uses os.open when it should use the builtin function;
        2) It uses a variable named "file", which will step on the
           name of the builtin function.

If you write it this way:

        # no need to import os here
        myfile = file("e:\\d12", "r")

you may find the resulting code easier to understand.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From uwe.schmitt at procoders.net  Wed Sep 11 05:42:28 2002
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 11 Sep 2002 09:42:28 GMT
Subject: How do i print to a printer using python?
References: <6afb9d2c.0209041044.43c5f5f1@posting.google.com> <al73gj$lbt$1@wanadoo.fr> <6afb9d2c.0209091641.1d9b652e@posting.google.com> <allofr$lgscc$2@hades.rz.uni-sb.de> <dnuf9.122228$l_4.76576@atlpnn01.usenetserver.com>
Message-ID: <aln364$liqog$1@hades.rz.uni-sb.de>

Steve Holden <sholden at holdenweb.com> wrote:
> "Uwe Schmitt" <uwe.schmitt at procoders.net> wrote in message
> news:allofr$lgscc$2 at hades.rz.uni-sb.de...
>> Naveed Iqbal <iqbal at colorado.edu> wrote:
>> > Thank you all for your help. I am really a novice when it comes to
>> > python. Will someone tell me how to install the win32ui module in
>> > windows?
>> > thanka
>> > naveed
>>
>> I suceeded without win32all: I generated a postscript file
>> and copied it to the 'file' "prn:" using shutil.copyfile().
>>
> Disk Operating System Rules OK!

shurely :-)

> Uwe: you'll clearly do well with such lateral thinking abilities. What
> software did you use to generate the PS?

I just found this solution because I was to lazy to read the
win32all ducomentation ....

At first I use pdflib to generate pdf-files (which can be stored,
send by email..) and then used ghostscript to convert pdf to ps.

The only "problem" I have, is that ghostscript opens a window, even
if I use it from the commandline...

Greetings, Uwe.


-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at procoders.net      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           


From bdelmee at advalvas._ZAPME_.be  Mon Sep  2 15:59:57 2002
From: bdelmee at advalvas._ZAPME_.be (Bernard Delmée)
Date: Mon, 2 Sep 2002 21:59:57 +0200
Subject: Why it is so dramatical?
References: <mailman.1030957743.4021.python-list@python.org> <Xns927D6DBDE418Cduncanrcpcouk@127.0.0.1> <200209021228.58453.b.maryniuk@forbis.lt> <mailman.1030963564.23463.python-list@python.org>
Message-ID: <3d73c295$0$30449$afc38c87@sisyphus.news.be.easynet.net>

> > $b = '';
> > $data =
'AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAA';
> > for $a (0..0xffffff) {
> >     $b += $data;
> > }

> That should result in a 1.3 GIGABYTE string.
> Watching the perl-process i see that perl doesn't use more than 1MB.
> So probably perl does some optimization for this in-fact meaningless code.

Nothing that magic: '+' does not concatenate strings. You need to use '.'
or '.=' for that. Otherwise $b keeps getting the numerical value 0 when
perl wants to use it with the arithmetic '+' addition.




From bobx at linuxmail.org  Sat Sep 14 12:32:23 2002
From: bobx at linuxmail.org (Bob X)
Date: Sat, 14 Sep 2002 16:32:23 GMT
Subject: [FEEDBACK] Is this script efficient...is there a better way?
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <L9Rf9.168314$l_4.127784@atlpnn01.usenetserver.com> <4ZRf9.1782$Lg2.248733@news2.news.adelphia.net> <3hSf9.168947$l_4.98525@atlpnn01.usenetserver.com> <8PSf9.1796$Lg2.257378@news2.news.adelphia.net> <_oUf9.170870$l_4.126599@atlpnn01.usenetserver.com> <Am7g9.2344$Lg2.402676@news2.news.adelphia.net> <3pkg9.3313$WV.2782@fe03>
Message-ID: <rwJg9.5283$Lg2.787978@news2.news.adelphia.net>

Just a quick test at home. On a 690K file the results were:

RE based script: 0.11
Original script: 0.12

I ran it a couple times and they were always with 1 or 2 ticks of each 
other but the RE one was always the faster.

I will try the same at work with the 30MB files to see if that makes a 
difference as well.



From anandpillai6 at yahoo.com  Wed Sep 25 11:02:49 2002
From: anandpillai6 at yahoo.com (Anand)
Date: 25 Sep 2002 08:02:49 -0700
Subject: Pygame event problem
Message-ID: <41c203e.0209250702.160102d@posting.google.com>

Hi !

  I wrote a simple image viewer in python using pygame image api. The
GUI
 is in wxPython. There are 2 classes, one for the window frame and one
for
 the pygame image processing.

   I added an effect in the pygame class which is called from a menu
in the
  wxPython GUI. Here is the code of the GUI class which makes the
call.

  def OnLiquidEffect(self, event):
          #imgview is of type MyPyGameImgViewer which is the pygame
          #class used to load the images.

          self.imgview.LiquidEffect()
          pass


  Here is the method which it calls on the MyPyGameImgViewer class.

 class MyPyGameImgViewer:

              ...
              ...
              ...

           def LiquidEffect(self):
          anim = 0.0

          if self.imgfile is not None:
               self.ShowImage(self.imgfile)
               self.flag = 0
          
               bitmap = pygame.Surface((self.h, self.w))
               bitmap.blit(self.screen, (0, 0))
               print self.h, self.w
          
               while 1:
                    anim = anim + 0.4
                    for x in range(0, self.h, 20):
                         for y in range(0, self.w, 20):
                              xpos = (x + (sin(anim + x * .01) * 15))
+ 20
                              ypos = (y + (sin(anim + y * .01) * 15))
+ 20
                              self.screen.blit(bitmap, (x, y), (xpos,
ypos, 20, 20))

                    #I think this code will remove all the
                    #events from the queue once it detects
                    #a mouse event.
                    if pygame.event.peek([MOUSEBUTTONDOWN]):
                         pygame.event.get([MOUSEBUTTONDOWN])
                         break
                    pygame.display.flip()
          else:
               print "No image loaded!"


   The effect is stopped by clicking with the mouse on the pygame
display
window. My problem is that if I try to invoke any other menu methods
in the
main GUI frame *without doing this*, the application ie both the
wxFrame and
the pygame display window, hangs.


  I have a gut-feeling that this is because I am calling a method on
the MyPyGameImgViewer class without exiting the while loop.

 Is there a way around this, ie to call any method in the
MyPyGameImgViewer class
from the GUI class, while one of its methods is executing a while loop
?

Thanks

Anand Pillai


From emile at fenx.com  Sun Sep 22 16:32:03 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 22 Sep 2002 13:32:03 -0700
Subject: sys.ps2 assignment has no effect
References: <uos1np8t53k396@corp.supernews.com> <3d8e0aae@news.sentex.net> <uos5dis0ogvc6f@corp.supernews.com>
Message-ID: <aml99h$8d4$1@main.gmane.org>

Lance:
> print '...' works correctly, showing ...

I can't duplicate this on winxp with 2.1.3, 2.2.1 or current cvs.  All
behaves as it should.

If you start an interpreter session, import sys, and check
ord(sys.ps2[0]) you should get 46.  Anything else might indicate
modifications to pythonrun.c, where ps1 & 2 are initialized, or that the
prompt is getting changed before you see it.  Try starting python
with -S (skips importing site) to see if that makes a difference.  Also,
what does set show?  Watch out in particular for PYTHONSTARTUP that may
also be a source of a change like this.

HTH,

--

Emile van Sebille
emile at fenx.com

---------






From cs26 at tecco.at  Mon Sep 16 03:21:50 2002
From: cs26 at tecco.at (Chris)
Date: 16 Sep 2002 00:21:50 -0700
Subject: concurrent programming
References: <fe31142f.0209130052.33174fcc@posting.google.com> <B7A63460808730BA.82D6BAEAFA5E4E78.4731C570A5FC8A2B@lp.airnews.net>
Message-ID: <fe31142f.0209152321.341fee8d@posting.google.com>

Thank you for the advice, but I have the instruction to write it in C++, so
this is a point I cannot change. Only for the scripts, Python should be
used.
Well, I know a little bit about how to do concurrent programming in Python,
and in C++ too, but my problem is that I don't know how to combine these two
technologies for the PARBEGIN, PAREND block.


From pyth at devel.trillke.net  Thu Sep 19 21:11:38 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 20 Sep 2002 03:11:38 +0200
Subject: Minimal debug/rep functionality
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8BF0A5B9@natasha.auslabs.avaya.com>; from tdelaney@avaya.com on Fri, Sep 20, 2002 at 10:47:54AM +1000
References: <B43D149A9AB2D411971300B0D03D7E8BF0A5B9@natasha.auslabs.avaya.com>
Message-ID: <20020920031138.C30315@prim.han.de>

Delaney, Timothy wrote:
> > From: Peter Hansen [mailto:peter at engcorp.com]
> > 
> > I'm a commercial Python user as well, so I can imagine the kinds of
> > things your employer think about whether to release or not.  We have
> 
> As I said, I *expect* to be able to release (there's no commercial reason
> not to, and we have done so with such things in the past).

Maybe we are turning circles [1] but why don't you do it, then? 

Maybe you need to invest some time in obfuscating your
source code and you don't want to do it? 

Why do you *expect* others to solve this problem for you?

    holger




[1] Sounds like a circular argument which might get GCed, soon :-)



From LogiplexSoftware at earthlink.net  Fri Sep 20 06:49:39 2002
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 20 Sep 2002 03:49:39 -0700
Subject: Larry Wall's comment on python...
In-Reply-To: <ambpgq$dre$1@nntp9.atl.mindspring.net>
References: <wkr8g7rn4c.fsf@yahoo.com>
	<bd751532.0209082142.54236307@posting.google.com>
	<allrrs$6hm$2@solaris.cc.vt.edu>  <ambpgq$dre$1@nntp9.atl.mindspring.net>
Message-ID: <1032518979.7312.8.camel@localhost.localdomain>

On Wed, 2002-09-18 at 23:06, Matt Gerrans wrote:
> One of the niftiest things about Python is you never have to argue about
> whether you should put your curlies in the wrong place, like this:
> 
> void foo() {
>    //...
> }
> 
> ...or whether you should place them correctly, like this:
> 
> void foo()
> {
>     //...
> }
> 
> It saves the world from a lot of unnecessary creation and release of hot air
> (so Python has probably done more to help reduce the greenhouse effect than
> any other software technology!).

Now all that's left to decide is whether to use tabs or spaces for
indentation...

;)

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From blarson at crary.com  Fri Sep  6 16:41:52 2002
From: blarson at crary.com (Bradley D. Larson)
Date: Fri, 06 Sep 2002 15:41:52 -0500
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <3D7958DB.50903@nospam.free.fr> <k08e9.25533$w51.7064239@news2.news.adelphia.net>
Message-ID: <3D79130F.73ED4E51@crary.com>

That is what is nice about Vi   paren,bracket matching!


> if (something == something_else)
> {
> } // if (something == something_else)
>
> As you can see, using comments to label my end blocks explicitly.  For me
> this issue was never an issue.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: blarson.vcf
Type: text/x-vcard
Size: 400 bytes
Desc: Card for Bradley D. Larson
URL: <http://mail.python.org/pipermail/python-list/attachments/20020906/51f50bf1/attachment.vcf>

From gua81 at XXXyahoo.com  Tue Sep 24 20:08:22 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Wed, 25 Sep 2002 10:08:22 +1000
Subject: Strange python and cgi!
Message-ID: <amquos$rcu$1@lust.ihug.co.nz>

Hi I made a cgi script in python and successfully compiled it.

the script is some thing like this

#! usr/lib.../python

import ...
def kickass(xxx)
...

...
print "<pre>"
kickass(xxx)
print "</pre>
...

when I go to the webpage, it stops halfway and in "view source: it only show
the code till the open "pre".
however when I output it to an html file:
python nameofscript.cgi >app.html
....
app.html works fine just as I want it to be.

What seems to be the problem here.

I think i''ve done evrything right

I've got that "Content-type" and also \n\n after that line.

Thanks in advance





From tjreedy at udel.edu  Mon Sep 23 21:33:03 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 24 Sep 2002 01:33:03 GMT
Subject: e vs exp()? / financial applications
References: <mailman.1032801271.6597.python-list@python.org>
Message-ID: <QgPj9.253053$AR1.10476629@bin2.nnrp.aus1.giganews.com>

"Andrew Thompson" <andrew.thompson at ashecastle.com> wrote in message
news:mailman.1032801271.6597.python-list at python.org...
> One must be aware of the representational issues and accuracy of
> arithmetic within a datatype when using numerical types.  This is
not so
> much a Python issue as one of representing the real number line with
> float types.
...
> W=0.89 / 4
> 1+w+w+w+w
> w+w+w+w+1
>
> which are both *different* according to the == operator, but for the
> purposes of numerical accuracy are identical.

Since 'W' != 'w' in Python, both give a NameError unless w defined
elsewise 8<).
However...
>>> w=.89/4
>>> w
0.2225
>>> 1+w+w+w+w
1.8899999999999997
>>> w+w+w+w+1
1.8900000000000001

Although w appears to be exactly represented, it must not be.  Cute
example.

Terry J. Reedy




From gerhard.haering at opus-gmbh.net  Fri Sep 20 13:36:21 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 20 Sep 2002 17:36:21 GMT
Subject: Local Python Server Installation
References: <3d8b47c6.277669@news.jaring.my>
Message-ID: <slrnaomn6m.39s.gerhard.haering@haering.opus-gmbh.net>

In article <3d8b47c6.277669 at news.jaring.my>, Kenneth Gomez wrote:
> Dear Group,
> 
> I would like to know wich software to install if I would like to learn
> to use Python as a web development tool? 
> 
> I have installed Python 2.2.1 and Apache but I belive I would need a
> Python Server Program to enable Apache to render .py files, am I
> right?
> 
> Could you kindly advise me? 

Add this line at the end of your .../Apache Group/Apache2/conf/httpd.conf:

    ScriptInterpreterSource Registry

Then you can write Python CGI scripts that have the .py extension. I
recommend to create a test.py script with the following two lines:

import cgi
cgi.test()

After you verify that this works (http://localhost/cgi-bin/test.py) you can
continue reading the documentation about the cgi module and writing your
own CGI scripts. The cgitb and Cookie modules should come in useful, too.

If you experience the limits of CGI one day, you can think about using a
full-blown appliation server like WebWare.

-- Gerhard


From gstaniak at inka.zagiel.pl  Fri Sep  6 08:45:57 2002
From: gstaniak at inka.zagiel.pl (Grzegorz Staniak)
Date: Fri, 6 Sep 2002 12:45:57 +0000 (UTC)
Subject: Access to shadow passwords?
Message-ID: <slrnanh90d.3k0.gstaniak@inka.zagiel.pl>

Hi all,

I didn't succeed in finding a Python interface to the /etc/shadow system
database - is anybody aware of a class/module that would allow me to
check account/password expiry dates etc.? I'm a Python beginner, but I
think there's no such thing in the standard library. I'll appreciate
your help.

Thanks,

-- 
Grzegorz Staniak <gstaniak at zagiel.pl>


From shalehperry at attbi.com  Wed Sep 11 12:10:44 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 11 Sep 2002 09:10:44 -0700
Subject: deleting the first and the last character of a string
In-Reply-To: <Xns9286ACD30F92Eduncanrcpcouk@127.0.0.1>
References: <20020911105818.GC2054@partagas.as.de.cw.net> <mailman.1031756234.22942.python-list@python.org> <Xns9286ACD30F92Eduncanrcpcouk@127.0.0.1>
Message-ID: <200209110910.44314.shalehperry@attbi.com>

On Wednesday 11 September 2002 09:03, Duncan Booth wrote:
>
> This doesn't work on Python 2.2.1 (or earlier), I think you need to be
> sitting astride the cutting edge of CVS development to do this.
>
> Specifically the change to implement this was April 22nd and 2.2.1 was
> released April 10th.

hmm, the 2.2.1 in Debian must have been updated then.  How annoying.  Thanks.

Python 2.2.1 (#2, Aug 30 2002, 10:27:17) 
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2



From maxm at mxm.dk  Fri Sep  6 09:24:30 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 06 Sep 2002 15:24:30 +0200
Subject: tree data structure
References: <FM0e9.2860$cP3.6165@news.iol.ie> <ala6c3$6i2$1@bob.news.rcn.net>
Message-ID: <3D78AC8E.4040900@mxm.dk>

Raymond Hettinger wrote:

>>>>tree = {}
>>>>for key, value in original:
>>>
>>>  tree.setdefault(key,[]).append(value)


That is a darn nice way to do it, even though it's a bit more readable to::

     for key, value in original:
         if map.has_key(key):
             map[key].append(value)
         else:
             map[key] = [value]

I just hate to see something I have done a lot of times in four line 
boiled down to one... :-)

I knew the setdefault() method, but would probably never have thought of 
using it like that on my own.


regards Max M



From shalehperry at attbi.com  Tue Sep 17 03:30:52 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 17 Sep 2002 00:30:52 -0700
Subject: problems getting a single item out of a list
In-Reply-To: <20020917071900.GB23629@partagas.as.de.cw.net>
References: <20020917071900.GB23629@partagas.as.de.cw.net>
Message-ID: <200209170030.52964.shalehperry@attbi.com>

On Tuesday 17 September 2002 00:19, Manuel Hendel wrote:
> I try to get a single item out of a list, but after this, the single
> item is stil a list item.
>
> domainfield = fields[3:4]
>
> I just want to get a new variable with a the value of the 3rd item of
> the list fields.
>
> Can anyone help?
>
> Thanks,
>     Manuel

just use fields[2] (you start counting at zero).



From max at alcyone.com  Mon Sep  2 23:40:56 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 02 Sep 2002 20:40:56 -0700
Subject: what about things like __*** ?
References: <7153651c.0209021927.16a3efaf@posting.google.com>
Message-ID: <3D742F48.9267B5B7@alcyone.com>

black wrote:

> That is I couldnt figure out what __*** is. Here is the description of
> that article but even confused me:
> 
> There is limited support for class-private identifiers. Any identifier
> of the form __spam (at least two leading underscores, at most one
> trailing underscore) is now textually replaced with _classname__spam,
> where classname is the current class name with leading underscore(s)
> stripped.
> 
> What's it trying to say please ???

The best way of finding out these kinds of things is playing with the
interactive interpreter:

max at oxygen:~% python
Python 2.2 (#1, Feb  6 2002, 19:31:45) 
[GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> class C:
...  def __init__(self):
...   self.__x = 1
... 
>>> c = C()
>>> dir(c)
['_C__x', '__doc__', '__init__', '__module__']
>>> c.__x
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: C instance has no attribute '__x'

When you create an attribute beginning with two underscores (and not
ending with two more underscores!), Python mangles the name and replaces
it with an encoding of the class name.  This is intended to obscure the
name so that in effect it acts as a private variable.  It doesn't
prevent malicious people from accessing the attribute in its mangled
form, it just prevents accidental accesses.

There is also a fairly common but unenforced convention that identifiers
starting with a single underscore are reserved and shouldn't be used
externally.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ There is nothing so subject to the inconstancy of fortune as war.
\__/ Miguel de Cervantes
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From syver-en+usenet at online.no  Mon Sep 16 12:38:34 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Mon, 16 Sep 2002 16:38:34 GMT
Subject: pyro & parallel processing
References: <rLOb1LAK5fh9EwRi@jessikat.demon.co.uk>
Message-ID: <uy9a1ritm.fsf@online.no>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> Has anyone used pyro to do symmetric parallel processing. The pyro
> framework is nice with master slave type applications, but the remote
> objects don't seem to be able to message back to their invoker.

They can, checkout the callback examples in the Pyro distro. Basically
what happens is that you pass a Pyro proxy as a parameter to the a
method on the Pyro object. 

> 
> I thought I'd ask if any one has got pyro to work in a more symmetric
> way?
> -- 
> Robin Becker

-- 

Vennlig hilsen 

Syver Enstad


From wyojustin at hotmail.com  Tue Sep 17 22:36:15 2002
From: wyojustin at hotmail.com (Justin Shaw)
Date: Wed, 18 Sep 2002 02:36:15 GMT
Subject: can wxpython do this?
References: <737aefee.0209151702.3f735252@posting.google.com> <spcaou804m5oo6gah31oqaeuqsuj60h0ip@4ax.com> <am3d0b$jj$1@news01bc.so-net.ne.jp> <am7hrt$9b3$1@news01bb.so-net.ne.jp> <20929D5514805227.48A29064FC554044.5053D887939EB774@lp.airnews.net>
Message-ID: <zERh9.105973$z91.4185575@bin3.nnrp.aus1.giganews.com>

> You might start with
>   from urllib import urlopen
>
>   URL = "http://www.python.org"
>   page = urlopen(URL).read()

Then use wxHtmlWindow to display it.
Justin




From jwbaxter at spamcop.net  Mon Sep 23 19:36:47 2002
From: jwbaxter at spamcop.net (John Baxter)
Date: Mon, 23 Sep 2002 16:36:47 -0700
Subject: Elementary confusion
References: <878z1s1kqf.fsf@smtp.blueyonder.co.uk>
Message-ID: <jwbaxter-31BCDD.16364723092002@corp.supernews.com>

In article <878z1s1kqf.fsf at smtp.blueyonder.co.uk>,
 Keith <kroc at blueyonder.co.uk> wrote:

> Hi,
> 
> I am working through O'Reilly's "Learning Python". I am up to page 33,
> and already I have a problem. The text outlines a cli entry about math
> operators, all really straight forward;
> 
> .>>>a=3
> .>>>b=4
> .>>>b/2+a
>  5
> .>>>b/(2.0+a)
>  0.8
> 
> This is so straight forward that I wasn't going to type it in, but I
> did anyway, and got instead of '0.8' something I cannot account for. I
> got '0.80000000000000004'.
> 
> So after 32 pages, I am already flumoxed. What is going on

In fact, the example is far from straight forward...it (now) exemplifies 
the fact that your computer can't represent 0.8 exactly in binary 
floating point, which Python uses for values like 0.8.  It comes close 
(supposedly as close as possible).  It's not a question of 
precision...double (or far more than double) the number of bits 
involved, and 0.8 still can't be represented exactly.

The 0.80000000000000004 is as close as your hardware can come (or it 
isn't as close as your hardware can come, and there are issues with the 
libraries or the chip).

Learning Python describes an older Python in which the interpreter lied 
about the value by rounding it to 0.8; the interpreter no longer does 
that, so you learn sooner about that problem.

You will encounter this issue over and over...for example on my machine:

$python
Python 2.2 (#1, 07/14/02, 23:25:09)
[GCC Apple cpp-precomp 6.14] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> a=0.1
>>> b=0.2
>>> c=a+b
>>> c==0.3
0
>>> c
0.30000000000000004
>>> d=0.3
>>> d
0.29999999999999999
>>>

We "know" that c is 0.3.  The computer has added two inexact values 
which are "too big" and produced a result which is therefore both "too 
big" and not the nearest representation of 0.3, which is "too small".

  --John


From richie at entrian.com  Fri Sep 20 10:31:40 2002
From: richie at entrian.com (Richie Hindle)
Date: 20 Sep 2002 07:31:40 -0700
Subject: Minimal debug/rep functionality
References: <mailman.1032406044.16132.python-list@python.org>
Message-ID: <8a6ba1da.0209200631.6f6e72bd@posting.google.com>

Tim,

> Having been writing a fairly comprehensive code and path coverage tool
> (incomplete - currently working on determining what *is* an executable line
> ... think I need to work with the AST)

I've done this recently as well, and I'm now on my third iteration of
the work-out-which-lines-are-executable code (!)

The first version used the AST, but there are several holes in that -
'else' lines, for instance, are identified by the parser as executable
lines, but the trace function doesn't always get called for them.  I
then wrote a system that found all the SET_LINENO instructions, and
that gave perfect results (as long as you weren't using -O) but then I
found out that SET_LINENO is going away in Python 2.3...

My current system (which I think works 8-) uses co_lnotab, which is a
delightful data structure that describes the relationship between
bytecodes and line numbers.  Here's the relevant piece of code, which
builds in 'lineNumbers' a list of the execuable line number in the
code object 'code':

# Derive the line numbers from co_lnotab; that's a list of pairs of
# increments, one for the bytecode address and one for the line
# number (see compile.c in the Python sources).
lineNumbers = []
lnotab = code.co_lnotab
previousAddress = -1
previousLineNo = -1
lineNo = code.co_firstlineno
byteCodeAddress = 0
for i in range( 0, len( lnotab ), 2 ):
    byteCodeAddress = byteCodeAddress + ord( lnotab[ i ] )
    lineNo = lineNo + ord( lnotab[ i+1 ] )
    
    # When the compiler wants to increment the line number by more
    # than 255 in one go, it increments it by 255 as many times as it
    # needs and then the remainder.  Fair enough, but there's a bug
    # whereby the byte code address gets incremented along
    # with the *first* 255, not the last.  Hence you get a line that
    # claims to have code on it when in fact it doesn't.  We detect
    # that here.
    if byteCodeAddress != previousAddress and not \
       ( lineNo - previousLineNo == 255 and 
         len( lnotab ) > i+2 and
         ord( lnotab[ i+2 ] ) == 0 ):
        lineNumbers.append( lineNo )
        previousAddress = byteCodeAddress
        previousLineNo = lineNo

Hope that helps.  The coverage tool that this belongs to is finished
but unreleased (mostly because of lack of thorough testing, but I'm
using it myself with no problems) - if you'd like a copy, drop me an
email.

-- 
Richie Hindle
richie at entrian.com


From cliechti at gmx.net  Sun Sep 29 18:55:58 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 30 Sep 2002 00:55:58 +0200
Subject: tkinter & serial port
References: <IhKl9.45766$bX.8904@sccrnsc02>
Message-ID: <Xns929998C1273Ccliechtigmxnet@62.2.16.82>

user at domain.invalid wrote in news:IhKl9.45766$bX.8904 at sccrnsc02:
> I'm new with Python/Tkinter, but have written LOTS of X windows code.
> This may be the root of my problem.
> 
> My program needs to talk over a serial port to a device AND update
> the display.
...
> How do I do this or its moral equivalent under Python so that it
> runs without modification under both BSD and Windows? 

you can hook in an "idle function" in most gui toolkits. that function 
could poll the serial port then. (select is out of question on wiondowze 
anyway)

i do prefer threads: start a worker thread that handled the serial IO and 
the main thread serves the GUI. that's the module "threading" and search 
for aahz page and the threding intro. (url found on all his messages' sig)

adv: for an easy serial port handling on both platforms i can sugest my 
lib: http://pyserial.sf.net.

chris

-- 
Chris <cliechti at gmx.net>



From marklists at mceahern.com  Mon Sep 23 22:11:39 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 23 Sep 2002 21:11:39 -0500
Subject: Three dumb questions (ordered by dumbness descending)
In-Reply-To: <pan.2002.09.23.20.43.59.461504.1216@io.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOIEOLEAAA.marklists@mceahern.com>

[Steven Feil]
> Humm, what is is zip()[...]

Steven, meet zip.  <wink>

$ python
Python 2.2.1 (#1, Jun 25 2002, 10:55:46)
[GCC 2.95.3-5 (cygwin special)] on cygwin
Type "help", "copyright", "credits" or "license" for more information.
>>> print zip.__doc__
zip(seq1 [, seq2 [...]]) -> [(seq1[0], seq2[0] ...), (...)]

Return a list of tuples, where each tuple contains the i-th element
from each of the argument sequences.  The returned list is truncated
in length to the length of the shortest argument sequence.
>>> l = range(10)
>>> zip(l, l)
[(0, 0), (1, 1), (2, 2), (3, 3), (4, 4), (5, 5), (6, 6), (7, 7), (8, 8), (9,
9)]

>>> d = dict(zip(l, l))
>>> d
{0: 0, 1: 1, 2: 2, 3: 3, 4: 4, 5: 5, 6: 6, 7: 7, 8: 8, 9: 9}
>>>

Neato, huh?

Cheers,

// m

-




From gerhard.haering at gmx.de  Tue Sep 17 03:41:07 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 17 Sep 2002 07:41:07 GMT
Subject: pypgsql -- not preserving transactions ?
References: <am3id4$1epu$1@nntp6.u.washington.edu> <mailman.1032151518.3079.python-list@python.org> <am6aga$21tu$1@nntp6.u.washington.edu>
Message-ID: <slrnaodn4i.a06.gerhard.haering@lilith.my-fqdn.de>

Frank Miles wrote in comp.lang.python:
> Gerhard H?ring <gerhard.haering at gmx.de> wrote:
>>* Frank Miles <fpm at u.washington.edu> [2002-09-16 03:15 +0000]:
>>> I've discovered that the 2.0 version of pyPgSQL packaged with Debian/woody
>>> doesn't preserve transactions...
> [snip]
> 
>>I can't reproduce that behaviour here. What do you do to ignore the
>>PgSQL.OperationalError? Or is it a different exception that gets thrown?
> 
> Yes, that exception gets thrown.  Perhaps this simply reveals a gap in my
> understanding -- I thought that it was Postgres that was handling the
> transactions.

Yeah, but pyPgSQL keeps track of wether we're in a transaction

> Presuming that the cursor wasn't starting a new transaction,

Ignoring the autocommit behaviour that's off by default: In reality,
the first cursor in a connection will automatically start a new
transaction. So do all other methods on the cursor object, if there is
no open transaction.

> even if the dumb user used a try/except block to bypass the exception (and
> continue with succeeding INSERT attempts), shouldn't Postgres prevent those
> later INSERTs?

An implicit rollback will be issued in case of an OperationalError.
The next cursor.execute() will open a new transaction, because there
is no open transaction now, which leads to the behaviour you describe.

>>This area of code hasn't changed up to 2.2, IIRC. Could you provide
>>a test case [...] 

> Sure, I've got some simple test code.  But at this point I'm now
> thinking that this is more likely a difference in conception -- what
> _should_ pypgsql do?

As far as I see, the current behaviour is what it should do.

> Should the application have to track if/when exceptions occur?  It's
> probably more efficient, though ISTM less clean to do it this way.

The problem as far as I can see is that you're using code like:

    for item in data:
        try:
            cursor.execute("INSERT INTO ...")
        except OperationalError:
            pass # better log error here
    conn.commit()

which leads to the behaviour I described above.

If you want either all or none of the inserts to succeed, you should
use code like this instead:

    try:
        for item in data:
            cursor.execute("INSERT INTO ...")
        conn.commit()
    except OperationalError:
        pass # better log error here

> Thanks for your reply, Gerhard.  Can you (or someone else) confirm
> that the "error" is a difference in conception?

Looks like so.

-- Gerhard


From webagency at simail.it  Tue Sep  3 08:38:33 2002
From: webagency at simail.it (Servizi Internet)
Date: Tue, 3 Sep 2002 14:38:33 +0200
Subject: Servizi Internet
Message-ID: <B0005002240@mailserver2.ede.it>

Salve.
Sono il titolare di "Web Ceating.it", una delle web agency piu' in vista
sul mercato europeo. Volevamo informarLa che la nostra azienda si 
mette a Sua completa disposizione per qualsiasi esigenza o servizio Lei desideri.
Ci occupiamo principalmente di:

* Servizio Hosting e registrazione domini .com/.net/.org/.biz/.it;
* Registrazione nei principali motori di ricerca;
* Posizionamento garantito in testa ai motori di ricerca;
* Creazione Siti Personali e Aziendali;
* Consulenza Web Marketing;
* Realizzazione Grafica;
* Realizzazione di siti E-Commerce;
* Virtual Store;
* Banner e Loghi Aziendali;
* Creazione di pagine HTML, ASP e FLASH;
* Creazione di Database per il Web;
* Manutenzione e gestione dei siti creati;
* Manutenzione e gestione di altri siti;
* Registrazione del Dominio .it .com .org .net;
* E-mail Target Advertise;

Per qualsiasi Sua esigenza o informazione La preghiamo di
contattarci compilando il form all'URL: http://www.webcreating.it/contatti.htm
I nostri consulenti saranno a Sua completa disposizione!
L'URL aziendale e': http://www.webcreating.it
_____________________________________________________________

L'87% degli utenti in internet utilizzano i motori di ricerca per trovare i
siti che vogliono!
Special Submitt Professional e' un servizio da noi elaborato che permette di
effettuare l' inserimento mirato su 500 e piu' motori di ricerca e portali
nazionali ed internazionali come: MSN - VIRGILIO - GOOGLE! - LYCOS - YAHOO -
LOOKSMART - GOTO ecc.ecc.

* PERCHE' ASPETTARE? In meno di tre giorni avverra' la sottoscrizione in
piu' di 500 motori di ricerca.

* IL TEMPO E' DANARO! Noi facciamo il lavoro per te! Basta passare notti in
bianco per poi non essere certi di vedere il proprio sito ben posizionato e
correttamente registrato sui principali motori di ricerca.

* REPORT! Periodicamente invieremo dei report e ti terremo aggiornato su quello
che succede sul tuo sito.

* SUPPORTO TECNICO! Potrai sempre contare su di un supporto tecnico esperto
nel settore e potrai porre tutte le tue domande.

E allora che aspetti? vai su http://www.webcreating.it/special_submitt.htm
e approfitta di questa vantaggiosa offerta.

Scusandoci per l'invadenza speriamo aver fatto cosa gradita e
cordialmente salutiamo.
________________________________________________________________
Questo messaggio non ? spam ma un semplice messaggio informativo che 
non verr? ripetuto.La ringraziamo della Sua cortese attenzione. Nel rispetto
della Sua privacy e' doveroso comunicarLe che il tuo indirizzo e-mail ? presente
su internet e nel nostro archivio e non verr? mai ceduto a terzi per nessun motivo.
________________________________________________________________

Web Creating.it - Soluzioni internet per le aziende.

python-list at python.org
www.python.it/faq/faq-1.html






From aleax at aleax.it  Fri Sep 20 14:56:30 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 18:56:30 GMT
Subject: Copy with __slots__
References: <mailman.1032419786.7187.python-list@python.org> <amd082$r5j$1@newsreader.mailgate.org> <cUoi9.134531$pX1.4784251@news2.tin.it> <amfj0b$lb2$1@newsreader.mailgate.org>
Message-ID: <ybKi9.140210$pX1.4975635@news2.tin.it>

Antonio Cuni wrote:

> Alex Martelli wrote:
> 
>> This doesn't deal with inherited __slots__.
> [cut]
> 
> you're right, I couldn't think of it :-(
> 
>> Some care
>> is needed, though (I can think of several possible anomalies).
> 
> what kind of anomalies?

For example, slots with the same name might be present
in more than one base, but it would be better to only
copy them once.


Alex



From tim.one at comcast.net  Mon Sep  9 14:22:49 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 09 Sep 2002 14:22:49 -0400
Subject: Can't exponentiate zero in NumPy
In-Reply-To: <alimef$lc9$1@naig.caltech.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEDBBDAB.tim.one@comcast.net>

[Rick Muller]
> Turns out it appears to be a more general problem than I thought.
>
>     1e-310 ** 2
>
> Works on RH linux/1.5.2, overflows on RH linux 2.2
>
>     pow(1e-310, 2)
>
> Works on linux 1.5.2, overflows on linux 2.2
>
>     import math
>     math.pow(1e-310, 2)
>
> Works on linux 1.5.2, works on linux 2.2 (??!)
>
>     1e-310 ** 2.1
>
> Works on linux 1.5.2, fails on linux 2.2
>
>     pow(1e-310, 2.1)
>
> Works on linux 1.5.2, fails on linux 2.2
>
>     math.pow(1e-310, 2.1)
>
> Works on linux 1.5.2, works on linux 2.2
> > In addition, exactly which version of Python are you using
> (2.0, 2.1, 2.1.1,
> > 2.1.2, 2.1.3, 2.2, 2.2.1, ...)?
>
> These are the standard python/python2 distributed with RedHat Linux 7.3.
>
> "Works" means returns 0.0
> "Fails" means raises an OverflowError
>
> Can someone suggest a decent workaround?

Upgrade to Python 2.2.1, or downgrade to an older version of glibc.  glibc
changed its errno behavior in a way that nobody noticed before 2.2 final was
released.  While I've never seen the problem myself, I fixed it <wink>, and
many people reported that this class of problem is indeed fixed in 2.2.1
without backing off to an older version of glibc.




From jubafre at brturbo.com  Mon Sep 16 13:41:12 2002
From: jubafre at brturbo.com (jubafre at brturbo.com)
Date: Mon, 16 Sep 2002 14:41:12 -0300 (GMT-03:00)
Subject: search re and put in a list??
Message-ID: <-763904540.1032198072820.JavaMail.nobody@webmail1>

i have 3 lists:

mem=['7','8', '9']
labels['D1','D2','D3']
S=['20 D1', '30 D2', '10 D3', 'F0', '  3', '  2', '  0']

I want to put mem in s trougth of labels.

#+++++++++++++++++++++++++++++++++++++++++++++
final=[]
for i in range(len(j)):
    for c in range(len(s)):
        p= re.compile(labels[i])
        t1=p.search(s[c])
        if t1:
            t=sub(p, j[i], s[c].strip())
            final.append(t)
print final
#+++++++++++++++++++++++++++++++++++++++++++++

but doing like this i have just de nodes replaced, 
['20 7', '30 8', '10 9']

and i want the replaced and the not replaced in order like this:
['20 7', '30 8', '10 9', 'F0', '  3', '  2', '  0']
	


Juliano Freitas
www.gebrasil.hpg.com.br 

From duncan at NOSPAMrcp.co.uk  Fri Sep 13 06:38:59 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Fri, 13 Sep 2002 10:38:59 +0000 (UTC)
Subject: wrapping yield ?
References: <alqehl$9qv$1@nntp0.reith.bbc.co.uk> <Xns92885F082A21Cduncanrcpcouk@127.0.0.1> <alsdpt$2nq$1@nntp0.reith.bbc.co.uk>
Message-ID: <Xns9288759449799duncanrcpcouk@127.0.0.1>

"Michael Sparks" <Michael.Sparks at rd.bbc.co.uk> wrote in
news:alsdpt$2nq$1 at nntp0.reith.bbc.co.uk: 

> Whereas if I could wrap up yield, I simply have to do:
> 
> item = waitForSomethingCheckingPeriodically()
> process(item)
> 
> Which if I have lots of this sort of thing going on, results in much
> clearer (and hence maintainable) code.
> 
> If it can't (putting yield in a function, and not turning the function
> into a
> generator) be done, and AFAICT it can't, and this is a suggested work
> around - many thanks :-)

The syntax wouldn't look quite so bad if your function wasn't returning a 
value. At present 'for x in gen(): yield x' is the only way to nest 
generators, but there were discussions here recently suggesting a new 
syntactic sugar:
   yield every gen()
as an equivalent for the for loop.

There definitely isn't any way to put a yield in a function and not have it 
turn into a generator. When you yield from a generator, only a single frame 
is saved, not a stack, so you can only yield from the outer loop. By 
yielding at each level you effectively simulate the stack since each frame 
chains to the others.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From strat_addict at yahoo.com  Tue Sep 24 16:33:56 2002
From: strat_addict at yahoo.com (strat_addict)
Date: 24 Sep 2002 13:33:56 -0700
Subject: COM, Word problem
References: <e00ba67.0209211751.1411ad7c@posting.google.com> <mailman.1032705750.12449.python-list@python.org>
Message-ID: <e00ba67.0209241233.5113ae84@posting.google.com>

"Emile van Sebille" <emile at fenx.com> wrote in message news:<mailman.1032705750.12449.python-list at python.org>...
> strat_addict:
> > com_error: (-2147417846, 'The message filter indicated that the
> > application is busy.', None, None)
> 
> Google is your friend... perhaps this helps:
> 
> http://groups.google.com/groups?selm=MPG.13553ec91f8cb3b98968e%40192.168
> .0.28

Thanks Emile, I never would have guessed to think it was my antivirus software.

Thanks again,
Glenn (strat addict)


From NineOfSix at gmx.de  Mon Sep  2 03:43:52 2002
From: NineOfSix at gmx.de (Axel Grune)
Date: Mon, 02 Sep 2002 09:43:52 +0200
Subject: ascii value
References: <akv4fn$eo2$03$1@news.t-online.com>
Message-ID: <akv4r1$qd7$04$1@news.t-online.com>

A sorry, I missed ord(), shoot me

Axel



From ark at research.att.com  Fri Sep 20 15:07:16 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 20 Sep 2002 19:07:16 GMT
Subject: Python 2.2.1 compatible with Tcl/Tk 8.4?
Message-ID: <yu9965x05vkr.fsf@europa.research.att.com>

Now that Tcl/Tk 8.4 is out, can anyone tell me whether Python 2.2.1
will work with it?  Or should I wait for 2.3?


-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From freddie at madcowdisease.org  Sat Sep 14 13:44:58 2002
From: freddie at madcowdisease.org (Freddie)
Date: Sat, 14 Sep 2002 17:44:58 GMT
Subject: Argh! Python 2.2 + threading + sockets = trouble
References: <R_ug9.753$Ee4.1406@news-server.bigpond.net.au> <b03e80d.0209140002.44c6b816@posting.google.com>
Message-ID: <uAKg9.3962$Ee4.10884@news-server.bigpond.net.au>

Sandy Norton wrote:
> A more precise description of your problem will often help others help
> you; more specifically, stating that an app 'just stops working' is
> not a great problem description.
> 
> Localize the problem and someone may be able to point you in the right
> direction.
> 
> cia,
> 
> Sandy

I would if I could, but I don't know HOW to localize the problem. If I 
put a 'print alive!' statement in the main cowbounce loop, it prints 
alive! for a while, but it will stop when a random client manges to 
connect to a server. Ctrl-C stops working, and the server stops spawning 
a new thread to handle an incoming connection. This is after the 
connect() call in BounceClient succeeds. Other times, runnings threads 
(already connected to a server) will just stop passing data back and 
forth. The Python profiler is useless in a threaded application, and 
Linux doesn't have ktrace.

MadCowOffer starts a thread for each 'task'. IRC, database, transfer 
queues, admin interface, etc. If I unroll the map(lambda x: x.start(), 
self.__Threads) loop, it goes something like this:

Starting PackMan
Starting HeadHoncho
Starting Chatterbox

Then it becomes unresponsive. Ctrl-C won't kill the process, I have to 
terminate it with kill. The problem is, the exact same code runs 
PERFECTLY FINE in Python 2.1.x. This is what has me stumped :(

If you need any more info, feel free to contact me.

Freddie



From gerhard.haering at gmx.de  Sat Sep 14 21:45:27 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sun, 15 Sep 2002 03:45:27 +0200
Subject: python + java for web development?
In-Reply-To: <am0nea$mhd$1@newsreader.mailgate.org>
References: <am0nea$mhd$1@newsreader.mailgate.org>
Message-ID: <20020915014527.GA7233@lilith.ghaering.test>

* eugene kim <eugene1977 at hotmail.com> [2002-09-15 08:36 +0000]:
> [...] since java passes objects , not just strings among servlets and
> jsps i'm wondering if it is possible to mix java and other language..

Did you check out Jython, yet? http://www.jython.org/

Jython is an alternative Python implementation that runs on a Java
Virtual Machine. Consequently, you can call Java class libraries from
Jython, and the other way round.

> how python can interact with these servlet/jsp..?

Jython can be used to write servlets and with some determination, you
can even use it as a language for JSPs [1].

Of course, CPython is very well suited for web development, too. I'd
even prefer it, since you don't have to touch Java, then ;-)

-- Gerhard

[1] That's a little hard, though, it involves using BSF and some tag
library from the Jakarta project, plus patching BSF to use Jython
instead of JPython, Jython's predecessor).



From dhein at acm.org  Fri Sep 20 19:49:36 2002
From: dhein at acm.org (Coot)
Date: 20 Sep 2002 16:49:36 -0700
Subject: Popup menu won't hide on Linux
References: <a3a96207.0209151220.5228a262@posting.google.com>
Message-ID: <a3a96207.0209201549.5a6b7e75@posting.google.com>

dhein at acm.org (Coot) wrote in message news:<a3a96207.0209151220.5228a262 at posting.google.com>...
> When a popup menu is displayed, I expect that it should goes away when
> one of the menu items is clicked (button <1>) or when I click anywhere
> else (such that the focus is no longer on the popup menu).
> 
> However, on Linux, the latter behavior doesn't happen.  The popup
> remains on top of the main app window and the only way to get rid of
> it is to click one of its menu command items or close the application.
> 
> This behavior isn't there on Windows.
> 
> Can anyone tell me how to solve, or work around, the problem???
> 
> Here is some example code you can use to demonstrate this problem for
> yourself (taken from
> http://www.pythonware.com/library/tkinter/introduction/x5819-patterns.htm):
> 
> # menu-example-4.py
> 
> from Tkinter import *
> 
> root = Tk()
> 
> def hello():
>     print "hello!"
> 
> # create a popup menu
> menu = Menu(root, tearoff=0)
> menu.add_command(label="Undo", command=hello)
> menu.add_command(label="Redo", command=hello)
> 
> # create a canvas
> frame = Frame(root, width=512, height=512)
> frame.pack()
> 
> def popup(event):
>     menu.post(event.x_root, event.y_root)
> 
> # attach popup to canvas
> frame.bind("<Button-3>", popup)
> 
> mainloop()


OK, on Linux I have to watch for a <FocusOut> event and "unpost" when
that event fires.  If I do that and click outside the window, then the
popup disappears.

Here is the new code:

from Tkinter import *

root = Tk()

def hello():
        print "hello!"

def popup(event):
        menu.post(event.x_root, event.y_root)
        menu.focus_set()

def popupFocusOut(self,event=None):
        menu.unpost()

# create a canvas
frame = Frame(root, width=512, height=512)
frame.pack()

# create a popup menu
menu = Menu(frame, tearoff=0)
menu.add_command(label="Undo", command=hello)
menu.add_command(label="Redo", command=hello)
menu.bind("<FocusOut>",popupFocusOut)

# attach popup to canvas
frame.bind("<Button-3>", popup)

mainloop()


From gerhard.haering at gmx.de  Sun Sep  1 21:18:18 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 2 Sep 2002 01:18:18 GMT
Subject: Keyerror problem
References: <_lyc9.7164$7N.29584@newsfep3-gui.server.ntli.net>
Message-ID: <slrnan5f2o.4s9.gerhard.haering@lilith.my-fqdn.de>

davbucko wrote in comp.lang.python:
> I have a class Flags as follows:
> 
> class Flags:
>     def __init__(self):
>         self.flags = {}
> 
>     def getInverseFlags(self):
>         for i in self.flags.keys():
>             #switch some keys/values around
> 
>     def setFlag(self,name,value):
>         self.flags[`name`] = value
                     ^^^^^^

Here's the bug. It should read: self.flags[name] = value.

`name` calls repr on name, which is most likely not what you want.

Btw. if you want to invert a dictionary, there are probably hundreds
of examples out there, here's a hopefully readable one:

def invertdict(orig):
    new = {}
    for k, v in orig.items():
        new[v] = k
    return new

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From sandorlevi at yahoo.com  Fri Sep  6 19:50:51 2002
From: sandorlevi at yahoo.com (Levente Sandor)
Date: 6 Sep 2002 16:50:51 -0700
Subject: Access to shadow passwords?
References: <slrnanh90d.3k0.gstaniak@inka.zagiel.pl>
Message-ID: <8b5e42a6.0209061550.1e786b7d@posting.google.com>

Just take a look at the pwd module.
levi

Grzegorz Staniak <gstaniak at inka.zagiel.pl> wrote in message news:<slrnanh90d.3k0.gstaniak at inka.zagiel.pl>...
[...]
> I didn't succeed in finding a Python interface to the /etc/shadow system
> database [...]


From lu_gio at hotmail.com  Thu Sep 26 18:37:11 2002
From: lu_gio at hotmail.com (Max)
Date: Thu, 26 Sep 2002 22:37:11 GMT
Subject: pygame.event.Event?
Message-ID: <r_Lk9.174774$pX1.6183293@news2.tin.it>

Hi!
I can't generate a QUIT event in pygame:
pygame.event.Event(type, dict, [keyword_args]) -> Event

What do I have to put in "dict"?




From skip at pobox.com  Thu Sep  5 16:24:00 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 5 Sep 2002 15:24:00 -0500
Subject: Why Python?
In-Reply-To: <al8amr$d3q$1@panix2.panix.com>
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com>
        <4f49d232.0209050418.7668c21@posting.google.com>
        <al7ub8$crj$1@panix2.panix.com>
        <3D77FB95.3070300@nospam.free.fr>
        <al8amr$d3q$1@panix2.panix.com>
Message-ID: <15735.48480.165861.793285@12-248-11-90.client.attbi.com>

    Roy> I wrote:
    >>> If you're learning languages to study computer
    >>> science, lisp is looking a lot better.

    Roy> laotseu  <bdesth at nospam.free.fr> wrote:
    >> I think once you know them both, you know *how* to program, whatever
    >> the language you have to use.

    Roy> I believe we actually both said the same thing, just worded
    Roy> differently :-)

Alas, to even get to a phone screen these days, it appears you need "at
least five years experience" with a buzzword language on your resume.  In
most cases, neither Lisp nor Python qualify.

-- 
Skip Montanaro
skip at pobox.com



From ckasso at sprynet.com  Tue Sep 24 21:57:22 2002
From: ckasso at sprynet.com (Chris Kassopulo)
Date: Tue, 24 Sep 2002 21:57:22 -0400
Subject: hex numbers??
References: <mailman.1032881550.14068.python-list@python.org>
Message-ID: <20020924215722.78b60f64.ckasso@sprynet.com>

On Tue, 24 Sep 2002 12:30:48 -0300 (GMT-03:00)
jubafre at brturbo.com wrote:

> 
> x = [12,13,14]
> y = ['0xc', '0xd', '0xe']
> 
> hex() create a list of string,  how i can write in a file just the
> real value of hexadecimal number?
> 

x = [12,13,14]
y = []
for char in x:
	y.append(str(hex(char)).replace ('0x','',1))
print y 	

['c', 'd', 'e']

Chris Kassopulo


From sludtke at bcm.tmc.edu  Mon Sep 30 13:11:09 2002
From: sludtke at bcm.tmc.edu (Steve Ludtke)
Date: Mon, 30 Sep 2002 12:11:09 -0500
Subject: XMLRPC calls don't work in threads, or rather, they don't play
 well with others ! (help)
References: <3D97D190.9030003@bcm.tmc.edu> <m38z1kqb4r.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3D9885AD.2030603@bcm.tmc.edu>


Martin v. Loewis wrote:
> Steve Ludtke <sludtke at bcm.tmc.edu> writes:
> 
> 
>>Hi. I've been trying to write a python app that makes somewhat
>>time-consuming XMLRPC calls. I planned to have several calls going at
>>once using threads, but it seems like something in the XMLRPC client
>>code isn't releasing the python interpreter lock properly. 
> 
> 
> This is unlikely. xmlrpclib does not use threads. I don't know what
> xmlrpclib_auth is, so all bets are off.
> 
> If it doesn't use any threading facilities, either, blocking calls
> will occur eventually only in the socket read/write calls, which do
> allow other threads to run, and do release the interpreter lock
> properly.
xmlrpclib_auth just adds a couple of header entries, and calls 
xmlrpclib. No actual functionality changes.  What you describe is what 
I'd expect, but not what I get.

>>Is there something I can do at the Python level to deal with this
>>issue ?
> 
> 
> I'd add print statements at all levels of the code, to see where
> exactly it is blocking.

Already did this. It definitely seems to be the xmlrpclib calls that are 
   preventing the thread switching.    I'll have to see if I can come up 
with a simple demonstration script to show the problem.



From tismer at tismer.com  Thu Sep 19 09:08:07 2002
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 19 Sep 2002 15:08:07 +0200
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <mailman.1032329366.6813.python-list@python.org> <j4ptvbaapm.fsf@informatik.hu-berlin.de> <3D890FB0.4030408@something.invalid> <m3admeo4u1.fsf@mira.informatik.hu-berlin.de> <ambso1$4g8vh$1@ID-59885.news.dfncis.de>
Message-ID: <3D89CC37.5080209@tismer.com>

Thomas Heller wrote:
> Martin v. Loewis wrote:
> 
>> Greg Ewing <see_reply_address at something.invalid> writes:
>>
>>
>>> So, the only way to find out whether a particular
>>> extension is safe to use in Stackless is to try
>>> it. 
>>
>>
>>
>> That is not true. It is more reliable to inspect the source code.
>>
>> In general, it is very easy to do so: if there are no callbacks to
>> Python, the module is stackless-safe.
> 
> 
> So win32all probably wxPython are incompatible with stackless?

PythonWin has a problem when locking Scintilla for printing,
and a tasklet prints then, too. I will prohibit this by
some magic which is already designed: I don't allow switching
for modules which are not registered. (registrry to be done)

wxPython: Nooooo, it works like a charm! I have a commercial
application which runs multiple tasks in wxPython.
There are some small imcompatibilities which I will
discuss with the wxPython developers.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From nospam at bigfoot.com  Fri Sep  6 19:39:11 2002
From: nospam at bigfoot.com (Gillou)
Date: Sat, 7 Sep 2002 01:39:11 +0200
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com>
Message-ID: <albdk2$2iro$1@norfair.nerim.net>

"Rod Stephenson" <viking_kiwi at yahoo.com> a ?crit dans le message de news:
wkr8g7rn4c.fsf at yahoo.com...
> Slashdot has a list of questions posed to Larry Wall (perl). When
> asked his thoughts on other scripting languages, he makes the
> following observation about python
>
> "Python is cool to look at small bits of, but I think the "outline"
> syntax breaks down with larger chunks of code. I'm with Aristotle on
> the structure of discourse--a story should have a beginning, and
> middle, and an end. So should blocks"
>
> I'm not quite sure what he's trying to get at here - I guess that for
> a long heavily indented chunk of code, you could lose track of the
> overall structure, but I don't write code this way.
>
> Any comments?

Long heavily indented chunks of code is a really bad programming style
(whatever's the language).
Modern python editors can find blocks, hide/show blocks, auto indent/dedent
blocks, and find start/end of blocks.
Tabnanny does a fine job on ugly indenting.

def longfunction():
    # 300 lines of code below
    ...
    return
# end longFunction()




From DennisR at dair.com  Tue Sep 17 16:33:48 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Tue, 17 Sep 2002 20:33:48 GMT
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de>
Message-ID: <MkMh9.1195$q52.53148952@newssvr21.news.prodigy.com>

> If I would like to write some code that would not be possible for others
to
> read, after I distribute the program, would Python be suitable?

I am working on an unreleased system that may work out for you.  I have an
http (web) server distributed as a single AutoRunning EXE which serves CGI
from memory.  That CGI can be Python scripts and the GUI interface would be
a web browser.  See sample at http://www.dair.com/hellopy.zip  Right now, I
have not documented this.

Import libraries are copied to a temp directory and erased when the browser
is dismissed.  Http pages are weakly encrypted in memory and only the
current active page is decrypted.

If the protection standard is "would not be possible for others to read"
then it may well be that no finite protection scheme is adequate.  The
protection outlined is adequate for some casual attacks (e.g. dumping the
exe).  It is not adequate against a determined, knowledgeable attack.

What is your time frame here?  I am looking for an early October 2002 first
release.  Beta testers welcome (write me at email below)

--

Dennis Reinhardt

DennisR at dair.com
http://www.dair.com




From tim.one at comcast.net  Sun Sep 22 14:45:34 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 22 Sep 2002 14:45:34 -0400
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
In-Reply-To: <3D8D3491.8010803@tismer.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEOJBFAB.tim.one@comcast.net>

[Christian Tismer]
> There is no reason to argue about that.
> Decimal arithmetic is exact for most of the operations
> bankers require. They have been demanding for it since
> decades, and they will not drop it. Having "no rounding
> issue at all" is better than any consideration about
> neglectible errors, it cannot be beaten, since *they*
> understand it.
> Cannot be discussed. They pay the bill.

Bless you, Christian!  That's all the truth.  It reminds me of my first job
in this business, where after a few months my boss took me aside and
reminded me that I was hired to be a prostitute, not a missionary <wink>.

otoh-the-bankers-need-stackless-but-just-don't-know-it-yet-ly y'rs  - tim




From skip at pobox.com  Thu Sep 12 13:59:32 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 12 Sep 2002 12:59:32 -0500
Subject: Coding Style: Quotes
In-Reply-To: <bt1g9.17429$Pf7.590136@news1.west.cox.net>
References: <bt1g9.17429$Pf7.590136@news1.west.cox.net>
Message-ID: <15744.54788.570744.899923@12-248-11-90.client.attbi.com>

    John> I'm curious if others have adopted any standards for choice of
    John> single vs.  double quotation marks to delimit strings. A look
    John> through the standard library reveals the choice of one over the
    John> other is rather arbitrary. I suspect it really makes no
    John> difference, but the question has come up during a code review.

In situations where it doesn't matter (one or the other of the quotes
embedded in the string) I sort of half-heartedly pretend I'm programming in
C.  Single-character strings are single-quoted while multi-character strings
are double-quoted.  It really makes no sense to do it this way, and I'm not
very consistent about it.  It's just a carryover from C.  Your code
reviewers might appreciate the consistency if they have a C/C++/Java
background. 

For multi-line strings I use triple quotes.  (Again, the choice of style is
a bit arbitrary.)  Generally I use """ because long chunks of English text
frequently contain contractions.  For long strings which don't contain
newlines (like embedded SQL statements) I tend to enclose multiple "-quoted
literals in parens of some sort, like so:

    rows = self.executesql("select cities.city, state, country"
                           "  from cities, venues, events, addresses"
                           "  where cities.city like %s"
                           "    and events.active = 1"
                           "    and venues.address = addresses.id"
                           "    and addresses.city = cities.id"
                           "    and events.venue = venues.id",
                           (city,))

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From peter at engcorp.com  Fri Sep  6 22:55:42 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 06 Sep 2002 22:55:42 -0400
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com>
Message-ID: <3d796a95$1@news.sentex.net>

Rod Stephenson wrote:
> [about Larry Wall on Python]
> 
> "Python is cool to look at small bits of, but I think the "outline"
> syntax breaks down with larger chunks of code. I'm with Aristotle on
> the structure of discourse--a story should have a beginning, and
> middle, and an end. So should blocks"
> 
> Any comments?

Anyone know how long is the largest chunk of code Larry Wall has
actually written in Python?

The fact he has to say "I think" about this issue implies he is
guessing.  Maybe if he actually tried it he'd conclude differently.

-Peter



From oren-py-l at hishome.net  Thu Sep 12 06:02:14 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Thu, 12 Sep 2002 06:02:14 -0400
Subject: [FEEDBACK] Is this script efficient...is there a better way?
In-Reply-To: <GCEDKONBLEFPPADDJCOECEPFEOAA.whisper@oz.net>
References: <20020912074400.GA22578@hishome.net> <GCEDKONBLEFPPADDJCOECEPFEOAA.whisper@oz.net>
Message-ID: <20020912100214.GA41211@hishome.net>

On Thu, Sep 12, 2002 at 02:31:44AM -0700, David LeBlanc wrote:
> <snip>
> > Have you actually tested this? I used to believe that larger buffers are
> > always better for performance.
> >
> > Wrong.
> >
> > I ran some test on Linux for the effect of buffer size on file reading
> > speed and the results were very interesting. I started with a buffer size
> > of 32 bytes, tested file I/O throughput and increased it
> > logarithmically by
> > about 2% for each step.  As expected, the time to read 1mb improved as the
> > buffer size increased until it hit a minimum for a buffer size of around
> > 4-8k (the graph is very noisy so it's hard to tell) and then rose back up
> > to a value that is 10-20% worse for buffer sizes of 32-64k and remained
> > more-or-less constant for anything higher.
> >
> > This performance curve may be a result of the CPU cache or the OS
> > architecture.
> >
> > The chunk size used by xreadlines is 8k which is about the optimum
> > value (at least for linux).
> >
> > 	Oren
> 
> Disk cluster size? Fragmentation?

Irrelevant. The file was entirely cached in memory. This means that the
test was affected only by the CPU, caches, main memory bank and the way the 
OS architecture interacts with them.

If the file isnt't already in memory there may be some advantage to using 
larger buffers - the disk readahead size used by the OS is a compromise 
between reading more to improve performance and reading less to avoid 
overhead in case the data wasn't needed after all. In this case the 
performance curve might not have the dip in the 4-8k area and keep improving 
with buffer size. Even so, it should rapidly come very close to its
asymptotic limit. I doubt that a buffer size over 32-64k would make any
difference. Slurping 30mb into memory only makes sense if you need to 
access it multiple times. For a single-pass operation there will always be 
a pretty small buffer size that will have equivalent or better performance. 

	Oren




From peter at engcorp.com  Tue Sep 10 11:44:57 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 10 Sep 2002 11:44:57 -0400
Subject: PyGTA: First Meeting notice (Toronto area Python/Zope group)
Message-ID: <3D7E1379.E0D2E231@engcorp.com>

As mentioned here previously, the newly formed Toronto area Python/Zope user 
group (http://web.engcorp.com/pygta) will hold its first meeting on Tuesday, 
September 17, from 7pm to 9pm, at the Willowdale United Church, 379 Kenneth Ave, 
North York (near the Finch subway station).  Doors will open at 7:00 sharp.

Some free parking is available in a lot and on the street.  More details 
and a map are available here: http://web.engcorp.com/pygta/wiki/FirstMeeting .

Among other activities, members will speak on interesting uses of Python
within their companies.  Note that, contrary to earlier reports, this will
not be merely an _organizational_ meeting, but a real first meeting.  

We will also be convening a post-meeting meeting at an area pub, the 
"Slug and Lettuce", apparently favoured by the local Smalltalk user group.

Please join us if you're in the area!

-Peter Hansen, for PyGTA


From mcherm at destiny.com  Wed Sep 11 09:29:56 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Wed, 11 Sep 2002 09:29:56 -0400
Subject: need Python and Zope experienced programers
References: <3d647529.27575997@news.earthlink.net> <15717.3618.449268.769068@gargle.gargle.HOWL>
Message-ID: <alnggc$vq5$2@main.gmane.org>

Skip Montanaro wrote:
>     Mike> I need experienced programmers for jobs in the NY metro area.
>     Mike> ETR Technology Center 
>     Mike> 180 Oser Ave Suite 0400
>     Mike> Hauppauge, NY 11788
> 
>     Mike> Call, e-mail, or fax
>     Mike> Attention: Mike Clemente
> 
>     Mike> Ph  (631)952-1300 Ext. 15
>     Mike> FAX (631)952-1248
>     Mike> E-mail: mike at etrtechcenter.com
>     Mike> Web: http://www.etrtechcenter.com
> 
> Mike,
> 
> You might want to post to the Python Jobs board:
> 
>     http://www.python.org/Jobs.html
> 

Unfortunately, it's for posting job LISTINGS only, not job REQUESTS. But 
it seems I've found something anyhow (though without Python :-( ).

-- Michael Chermside






From lopik at mail.telepac.pt  Sun Sep 22 15:16:32 2002
From: lopik at mail.telepac.pt (Robert A.M. van Lopik)
Date: Sun, 22 Sep 2002 20:16:32 +0100
Subject: Smalltalk vs. Python and COM was Re: Byte Magazine and Smalltalk
References: <MXFh9.23885$jF4.2037258@twister.southeast.rr.com> <3d87a686@tobjects.> <rd8fougqbgdivimn7vuest3ugoj57lvrr6@4ax.com> <Xns928CA588EF17Dtrippyparcplacecom@64.164.98.7> <3d87dd7c$1@tobjects.> <g9Vh9.313655$kp.1041470@rwcrnsc52.ops.asp.att.net> <Xns928D58DF09B6trippyparcplacecom@64.164.98.6> <20020918.115128.1605894428.1527@cybermesa.com> <amkpk0$or2$1@fe2.cs.interbusiness.it>
Message-ID: <1032791839.61940@spynews5>

When you ask this question on comp.lang.smalltalk.dolphin you will certainly
hear that Dolphin Smalltalk is the way to go if you have to do COM stuff.
And they are a very helpful (and knowledgable) bunch of people, those
readers
of clsd.

hth
rob van lopik



"Pino Gargiulo" <jk3380 at naida.org> wrote in message
news:amkpk0$or2$1 at fe2.cs.interbusiness.it...
>
> "Jay O'Connor" <joconnor at cybermesa.com> wrote in message >
>
> ... snip ...
>
> > I use Python in a lot of places where others would probably use Perl.
> > Pythons regex support is not a strong as Perl's, but its support for
data
> > abstraction(and hence, more complicated problems)  is much more
powerful.
>
> This brings up a question I was about to ask. I'm getting involved in a
> small project
> that requires a couple of COM servers, whatevere these are exactly... the
> nice thing
> is that any developmente language can be chose. My backgroung is Unix &
> CORBA,
> dabble a bit with VWNC (no serious work) and this is my second time I
write
> something on m$ platform.
>
> For this project I was considering Smalltalk and Python. The first so I
> could be
> some real stuff done and Python because apparently it permits quite easy
> access
> the the mysterious (to me) m$ APIs. I really would want to stay away from
> VC++,
> and I'm looking for some quite turnaround environment since I'll be
learning
> COM on the job.
>
> Anybody care to give some advice? I played around with DST but never
looked
> much at the existance of a COM wrapper/interface.
>
> TIA
>
> Pino
>
>






From martin at v.loewis.de  Fri Sep 13 01:57:44 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 13 Sep 2002 07:57:44 +0200
Subject: Detecting Russian and Ukrainian character sets
References: <mailman.1031865256.12992.python-list@python.org>
Message-ID: <m3u1kubfdz.fsf@mira.informatik.hu-berlin.de>

Tim Churches <tchur at optushome.com.au> writes:

> Is this correct?  

I would claim that the problem is, for many applications, slightly
different. For a number of applications, in particular in the
internet, there are well-established procedures for communicating the
encoding/charset of data as meta-information. The problem then is that
applications often ignore those declarations, when they should really
consider them.

For other applications, in particular with text files, it is true that
the text file does not carry a charset declaration. However, a good
estimate is that you find out what computer system is being used, and
follow the local conventions of that system.

Both approaches may go wrong: the charset declaration in an HTML file
may be missing, and the text file may have been copied from one system
to another. But so may auto-detection of encodings.

That is not to say that auto-detection isn't useful - but I'd like to
point out that it should be only one of many options.

In addition to the Perl module, it appears that there are a number of
other autodetection modules: Mozilla supports Cyrillic auto-detection,
and the ru-xcode package (available atleast as a FreeBSD port) does
autodetection as well. I'm not aware of a Python module, but it
shouldn't be difficult to port any of these algorithms to Python.

Regards,
Martin


From SBrunning at trisystems.co.uk  Tue Sep 17 04:42:34 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 17 Sep 2002 09:42:34 +0100
Subject: two or more whitespace regular expression?
Message-ID: <31575A892FF6D1118F5800600846864DCBD7EA@intrepid>

> From:	Manuel Hendel [SMTP:manuel at hendel.net]
> On Sun, Sep 15, 2002 at 07:33:50AM +0000, Alex Martelli wrote:
> > import re
> > twoormorewhites = re.compile(r'\s\s+')
>                                ^
> What is the "r" for? I tried to find something about this in the docs
> (Regular Expression HowTo , etc.) but I couldn't find anything.
 
It's a raw string - see
n<http://www.python.org/doc/current/ref/strings.html>.

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




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



From imbosol at vt.edu  Mon Sep 23 23:16:15 2002
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 23 Sep 2002 23:16:15 -0400
Subject: Three dumb questions (ordered by dumbness descending)
References: <amnvit$7kqdt$1@ID-77524.news.dfncis.de>
Message-ID: <amomot$knu$2@solaris.cc.vt.edu>

Thorsten Kampe wrote:
> Okay, here they are:
> 
> 1. Why is 'zip(zip(x)) != x' (should be the same because it's the 
> transposed)

Because zip doesn't transpose.  If zip were a transpose, it would take
a *single* argument, which is a sequence of sequences, and would
return the sequences transposed.  But zip doesn't do that.  Rather, it
takes *multiple* arguments, all of which must be sequences, and
returns a list of tuples of the nth item of the arguments.

Maybe you'll understand it this way, if you're a LISP programmer:

    zip = lambda x: apply(transpose,x)

where transpose is the function that does what you think zip does.



> 2. Peter Norvig mentions in "Python for Lisp Programmers" some 
> "don'ts": "[x] + y" and "x[1:]". Are there more things to avoid 
> (especially in a loop)?

A common one is "don't use 'in' in a loop".  The 'in' operator (for
regular lists) does a linear search of the list.  I did this recently
for an file sync'ing program and that part goes slow.

Really new versions of Python define an 'in' operation for dicts also,
but that's ok to use in a loop because it does a hash lookup.


-- 
CARL BANKS
http://www.aerojockey.com


From jpm at papercut.org  Fri Sep 20 10:27:16 2002
From: jpm at papercut.org (Joao Prado Maia)
Date: Fri, 20 Sep 2002 10:27:16 -0400 (EDT)
Subject: Overriding all methods in a class
Message-ID: <Pine.LNX.4.44.0209201021580.13541-100000@mercury.sabren.com>

Hi,

I have been looking for an answer for this question for a while, but 
couldn't find anything that would work for what I want to do.

I know that you can use __getattr__ to overload methods that your class 
don't have, but what about overloading all methods, so I could in a way 
'intercept' calls to my methods and do something different if I wanted, is 
that possible ?

What I want to do basically is create a caching system for a specific 
class, and have __getattr__ (or any other magic method name) handle the 
decision to sue the cached return value or to actually call the method.

Maybe it would be just easier to rename my current class to _Classname and 
create a new Classname with just __getattr__.

Any suggestions ?

Cheers,
Joao




From mr.random at gmx.net  Sun Sep 29 14:25:37 2002
From: mr.random at gmx.net (Dave Richards)
Date: Sun, 29 Sep 2002 20:25:37 +0200
Subject: Python and GUI question
References: <3d96b808_3@news.bluewin.ch>
Message-ID: <3d97459e_2@news.arcor-ip.de>

There are loads of GUIs for Python - Gtk, Qt, Tk, WxWindows, Fox etc.
Which you use depends on what you need, and which platform you are
developing for. Tk is simple to use, but doesn't have list and tree
widgets. It also looks good on Win32, but crappy in Linux. I mostly use
Gtk, because it's relatively simple, but powerful - though I've never
tried it on Win32. The main reason to use Gtk is Glade - it makes
building GUIs pretty quick work. WxPython looks good on Win32 and Linux,
and it pretty powerful. The downside is it doesn't have a (free) GUI
designer like Gtk or Qt.
I would say use Tk, if it's a simple app, Wx if it's more complex. If
it's a Linux app, use Gtk. I don't have much experience of Qt myself.

Dave

On Sun, 29 Sep 2002 11:21:36 +0200, Josef B?rgi wrote:

> Dear NG
> 
> I designed a CMS using mySQL and PHP. Now I like to write an
> administrativ GUI Client Tool using Python. So I bought a few books
> about it, saying that there is more then one GUI interface (Tk, Qt,
> xvWindow). So can anyone tell me which one you use and why ? And how do
> you Design GUI interfaces (IDE) ?
> 
> Best wishes fromSwitzerland
> 
> Joe


From notvalid at hotmail.com  Tue Sep 24 16:01:30 2002
From: notvalid at hotmail.com (jano)
Date: Tue, 24 Sep 2002 13:01:30 -0700
Subject: cgi post problem
References: <FVRj9.192186$Jo.71059@rwcrnsc53> <9aWj9.39314$V7.10439005@twister.socal.rr.com> <amq8av$2cum$1@agate.berkeley.edu>
Message-ID: <amqgaq$2g05$1@agate.berkeley.edu>

There was an error in my prior post.  See ^^^^


"jano" <notvalid at hotmail.com> wrote in message
news:amq8av$2cum$1 at agate.berkeley.edu...
>
>
> "Jeff Davis" <jdavis at empires.org> wrote in message
> news:9aWj9.39314$V7.10439005 at twister.socal.rr.com...
> > Hi,
> >
> > It looks as though you have a lot of code. It might be helpful to try to
> > narrow the problem further by removing a lot of the code unrelated to
your
> > problem, which appears to be that "keys" does not contain any items.
After
> > you work that out, then put the other code back in.
> >
> > My first impression is that you have some "weird" form element names
> > because they contain the "/" character. That might be perfectly legal,
but
> > perhaps the cgi.FieldStorage class does not properly handle such element
> > names, or perhaps it considers such names erroneous.
> >
> > Next, it seems that the line "keys = form.keys()" might be out of
> alignment
> > with the rest of the code in it's block. Perhaps it's just the way it
> > looks after you copied your code into the post.
> >
> > Regards,
> >         Jeff
> >
> >
>
>
>
> Yes, the problem is that keys does not contain any items.  The 'weird'
names
> are not the problem, because i tried it also with regular 1-word names (no
> slashes), with the same results.  Also, the keys=form.keys() line is
> properly indented in the code -- it was just a copy/paste problem.  Here
is
> the code again, with the extraneous stuff stripped out:
>
> #!/usr/local/bin/python
>
> import cgitb; cgitb.enable()
> import sys, traceback, cgi
>
> def main():
>     try:
>         form = cgi.FieldStorage()
>         input = form.keys()
>
>         print "Content-type: text/html\n\n"
>         print("""<?xml version='1.0' encoding='UTF-8'?>
>         <!DOCTYPE html PUBLIC '-//W3C//DTD XHTML 1.1//EN'
> 'http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd'>
>         <html xmlns='http://www.w3.org/1999/xhtml' xml:lang='en'>
>             <head>
>                 <title>test</title>
>             </head>
>             <body>""")
>         print "<p>",keys,"</p>"

#                              ^^^^ should be 'input' not 'keys', as it is
in my script.

>         print "</body></html>"
>    except:
>         pass
>
> if __name__ == "__main__":
>      main()
>
> And the form is:
>
> <form action="http://someurl.com/cgi-bin/updatexml.py" method="post">
>      <input type="text" name="Name" size="20" />
>      <input type="text" name="PartNumber" size="20" />
>      <input type="text" name="Inventory" size="20" />
>      <input type="text" name="Cost" size="20" />
>      <input type="text" name="WholesalePrice" size="20" />
>     <input type="submit" value="Update Records" />
> </form>
>
>
> The script, again, is currently just supposed to return the keys, which I
am
> expecting to be Name, PartNumber, etc., but it returns only an empty list.
> Any ideas.
>
> thanks,
>
> jano
>
>
>




From mstenner at phy.duke.edu  Tue Sep  3 21:54:01 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Tue, 3 Sep 2002 21:54:01 -0400
Subject: simple xml pickle
Message-ID: <20020903215401.A6097@phy.duke.edu>

I would like to find a simple xml pickle module.  I have very simple
objects and so very simple needs.  I've found two things:

  the "old" xml_pickle.py, which appears to not work at all with
  python2.2, but is otherwise appealing.  It is a single file and
  could be easily included with a program.

    and

  the "new" xml.pickle as it appears in the Gnosis tools.  This is
  less appealing because it's very large and distributed over several
  files, and would almost necessitate full installation, and just one
  more prerequisite.  Besides, I can't get it to build (into an rpm,
  which is the only way that I'm really interested in installing
  software here).

Does anyone know of anything else?  Again, all I need is the ability
to handle simple objects & types.  I'd prefer something that I can
include in my program/package, rather than something that would need
to be installed separately.
				Thanks,
				 Michael

(I've looked through the vaults and put in lots of google time.  If
there are other resources, I'd love to hear about them.)
-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From cotabas at earthlink.net  Sun Sep  1 14:58:05 2002
From: cotabas at earthlink.net (CotaBas)
Date: Sun, 01 Sep 2002 18:58:05 GMT
Subject: Executable Python Programs
Message-ID: <1rtc9.108$LI2.7238@newsread2.prod.itd.earthlink.net>

This is probly a stupid question......

how do you make your python programs executable (without python IDLE).

For example:

I write a program that I want to give to a friend to run on his computer. He
doesn't have Python. What can I do to make it so he can run it without
Python?




---------------------------
-CotaBas




From gerhard.haering at gmx.de  Tue Sep  3 16:29:19 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 3 Sep 2002 20:29:19 GMT
Subject: basic cgi question
References: <67f6bbe5.0209031225.56b1b7bd@posting.google.com>
Message-ID: <slrnana6sv.t8.gerhard.haering@lilith.my-fqdn.de>

Jeffrey Keil wrote in comp.lang.python:
> Greetings:
> 
> I've used Perl in the past for cgi programs. I thought I'd play around
> a little with Python. I've tried a few experiments.
> 
> I have programs that prints out to the browser. Unfortunately, it
> prints out everything, including the "import" statements, "print",
> etc. I'm not quite sure what I'm doing wrong.

Your web server doesn't handle the Python script as a CGI script, but
as a normal document instead.

So the answer is: configure your web server appropriately.

Depending on your webserver, one of these groups is appropriate for
questions about configuration:

comp.infosystems.www.servers.mac
comp.infosystems.www.servers.misc
comp.infosystems.www.servers.ms-windows
comp.infosystems.www.servers.unix

You can also find some Python specific answers for at least IIS,
Apache on http://python.faqts.com and IIRC also in the Python FAQ.

-- Gerhard


From pereira at cis.upenn.edu  Sat Sep  7 12:06:07 2002
From: pereira at cis.upenn.edu (Fernando Pereira)
Date: Sat, 07 Sep 2002 12:06:07 -0400
Subject: Smoothing a discrete set of data
References: <3csmdso5.fsf@morpheus.demon.co.uk>
Message-ID: <B99F9C2F.1238E%pereira@cis.upenn.edu>

On 9/7/02 6:02 AM, in article 3csmdso5.fsf at morpheus.demon.co.uk, "Paul
Moore" <gustav at morpheus.demon.co.uk> wrote:
> I have a set of data, basically a histogram. The data is pretty
> variable, and I'd like to "smooth" it to find trends. [...] When I've tried to
formalise this, I've always stumbled over the fact
> that it's a trade-off exercise (there most be some sort of "tolerance"
> parameter in the algorithm) - on the one hand, the original data is
> "best" (in the sense that it matches exactly) whereas on the other
> hand, a single horizontal line at the average is "best" (fewest number
> of steps).
> 
> My instinct is that you're looking for something that balances number
> of steps vs difference from the original data. But for the life of me
> I can't see how to make this intuition precise.
> 
> Can anyone help me? I can't believe that this problem has never come
> up before, but I can't find any literature on it.
You must have been using the wrong search engine <wink> This is the problem
of *regression*, extensively studied in statistics and machine learning. The
tradeoff you mention is discussed in the literature under headings like
"bias-variance tradeoff", "generalization bounds", "structural risk
minimization"... If you want a single good book that covers this and related
topics with a minimum of prerequisites (elementary calculus and linear
algebra, a teeny bit of probability), I recommend

<http://www-stat.stanford.edu/~tibs/ElemStatLearn/>

-- F



From DennisR at dair.com  Wed Sep 18 16:54:24 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Wed, 18 Sep 2002 20:54:24 GMT
Subject: Python comment stripper
References: <mailman.1032369743.18855.python-list@python.org>
Message-ID: <4K5i9.674$HD4.41010424@newssvr13.news.prodigy.com>

> One question: How big is your disk page size?

My disk page size is 512.  However, the object of the compression here has
nothing to do with saving disk space.  The intent here is speeding up
internet download of an AutoRunning EXE that has Python DLLs and support
libraries included.

Perhaps, I should have stated my reasons for wanting to strip comments out
up front.  I apparently caused everyone who responded to question whether my
objective made any sense at all.  In retrospect, I should have stated my
larger objective (speeding internet upload/download) in addition to my
narrower objective (stripping comments for maximal compression).  I am sorry
for an misunderstanding.

--

Dennis Reinhardt

http://www.dair.com




From paul at boddie.net  Wed Sep  4 04:26:20 2002
From: paul at boddie.net (Paul Boddie)
Date: 4 Sep 2002 01:26:20 -0700
Subject: RedHat 7.3 and Python2.2.1
References: <mailman.1031020923.22683.python-list@python.org> <88bc63c6.0209030640.71e074ff@posting.google.com>
Message-ID: <23891c90.0209040026.3b9a0691@posting.google.com>

writeson at earthlink.net (Doug Farrell) wrote in message news:<88bc63c6.0209030640.71e074ff at posting.google.com>...
> To everyone who responded,
> 
> Thanks you for all your help, much appreciated. I ended up solving the
> problem (at least I hope so, and didn't create more for myself! <g>)
> by doing the following:
> 
> rpm -ivh --replacefiles python-2.2*

I do wonder if that should have been "-Uvh", but then my strategy with
RPM installation tends to involve stripping away the conflicting
packages (with "--erase") and installing the new ones in their place.
Indeed, I did this recently with KDE and found Python very useful in
the automation of the process - popen2 and os were most helpful in
identifying and removing the 30 or so old packages. ;-)

> The --replacefiles option seemed to eliminate the file conflicts and
> the RedHat tools (like up2date) still are working.

I suppose that Python 2.2.1 attempts to act just like Python 2.2 and
that various files/directories are identically named. Whilst this is
likely to confuse the RPM system, there's a reasonable chance that the
replacement files will do more or less what the old ones did. There's
still a risk that various extension modules won't exist in the
newly-installed package, but then they probably won't have been
deleted from the filesystem if you do a "package overwrite" in the
manner illustrated, and the extension API probably won't have changed
enough to cause severe problems when loading old extension modules
from Python 2.2 into Python 2.2.1.

Even though the concepts behind package management are sound, the RPM
implementation seems to be a solid justification for building packages
from source, however.

Paul


From maxm at mxm.dk  Tue Sep 24 04:47:16 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 24 Sep 2002 10:47:16 +0200
Subject: confusion about opening files
References: <mailman.1032835770.5997.python-list@python.org>
Message-ID: <3D902694.3090508@mxm.dk>

Mark McEahern wrote:

> I should have read the post I was replying to more carefully.  I saw the
> "integer required" error and didn't notice that the OP was using os.open.
> So the 'from x import *' was a red herring.  As Erik Max Francis pointed
> out, the OP should use open instead of os.open.


Actually he should use file() as open() is deprecated (well sort of ;-) )

"2.1 Built-in Functions
The file() constructor is new in Python 2.2. The previous spelling, 
open(), is retained for compatibility, and is an alias for file(). "


regards Max M



From fb at frank-buss.de  Sun Sep  1 04:58:32 2002
From: fb at frank-buss.de (Frank Buss)
Date: Sun, 1 Sep 2002 08:58:32 +0000 (UTC)
Subject: Crossword generator
References: <pan.2002.08.05.23.29.50.205338.718@thomas-guettler.de> <aknmta$h7f$3@newsreader2.netcologne.de> <a873nukqsjots9id3dgbnoomoajs3ieg8v@4ax.com> <akskq6$8c3$1@newsreader2.netcologne.de>
Message-ID: <akskro$8c3$2@newsreader2.netcologne.de>

Frank Buss <fb at frank-buss.de> wrote:

Sorry, this is the online link:
http://www.pdos.lcs.mit.edu/cgi-bin/theme-cword

-- 
Frank Bu?, fb at frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de


From see_reply_address at something.invalid  Mon Sep  9 00:18:23 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Mon, 09 Sep 2002 16:18:23 +1200
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com>
Message-ID: <3D7C210F.7090708@something.invalid>

Stephen Ferg wrote:

> "Joseph A. Knapka" <jknapka at earthlink.net> wrote in message news:<3D798461.B3FE0E06 at earthlink.net>...

 >

>>Also, you should be aware of the "command" option
>
> I know about it, but I think that for beginners, it is very confusing.
< ...  So I think that
> beginners should consistently use the more plodding technique of
> explicit binding.


I hope you're aware that what you're doing in your buttons
example is NOT equivalent to using the "command" option.
The latter does a number of extra things: it waits until
the mouse is released, and triggers the handler only if
the release is within the button.

So if you insist on presenting things the way you are, you'd
better include a clear warning that you're taking a big
short-cut and that you'll explain how to do it correctly
later, otherwise you'll lead people into a very bad habit.

But I wonder whether this isn't going to be more confusing
than just showing the use of the command option from the
beginning.

 >  For one thing, it leads down the path to lambda,

Not if you present your examples using one of the several
reasonable alternatives to lambda which exist.

 > For another, it doesn't automatically pass an event
 > object, which makes it even more difficult to use.

In the case of buttons, this isn't usually a problem,
since all you want to know is that the button was clicked.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From andrewm at object-craft.com.au  Mon Sep 16 02:40:37 2002
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Mon, 16 Sep 2002 16:40:37 +1000
Subject: Pickling slotted instances 
In-Reply-To: Message from martin@v.loewis.de (Martin v. Loewis) 
   of "16 Sep 2002 07:37:23 +0200." <m3it16ze98.fsf@mira.informatik.hu-berlin.de> 
References: <mailman.1031805375.9030.python-list@python.org> <j4fzwf349u.fsf@informatik.hu-berlin.de> <mailman.1032144560.28257.python-list@python.org>  <m3it16ze98.fsf@mira.informatik.hu-berlin.de> 
Message-ID: <20020916064037.585853C8FF@coffee.object-craft.com.au>

>I don't know what issues Guido had in mind, but here are a few for
>consideration:
>
>1. How to determine all slots? It has been proposed that __slots__
>   should return a tuple that includes base slots as well;
>   alternatively, iterate over base types yourself. Alternatively,
>   return slots in __dict__.

How is this handled in an instance now? Does an instance have slots
for all it's direct slots as well as the slots defined by it's base
classes? If so, it should probably be serialised in exactly this manner?

This is probably the nub of the matter - I'll need to look into the
implementation to learn more.

>2. Should all slots be pickled unconditionally? You seem to expect
>   this, but Samuele Pedroni points out that, in Java, some slots can
>   be marked transient, to avoid pickling them.

That would be nice, but we already have this problem with traditional
classes (I would guess this is one of the more common reasons to code
your own __getstate__ machinery with traditional classes)?

>3. How to pickle slots? You are suggesting that the list of slot names
>   is included, but that appears to be redundant - if you know the type,
>   you also know what slots it has.

Very true (what happens if the class implementation is changed? Some sort
of signature, at least, might be required, I guess).

>4. How to pickle unset slots? If a type has a slot but the instance
>   has not assigned the slot, requesting the slot value raises an
>   AttributeError.

It would seem simple enough to include a "not set" token in the serial
stream.

>If you are in doubt that this is really as tricky as presented, I'd
>encourage you to come up with a patch. In the process of implementing
>the patch, you may either find that all issues have a straight-forward
>solution, or that there are really serious issues remaining.

"Put your code where your mouth is"... 8-)

I'm sort of surprised that slots were added without having all the
bases covered, but I'm presuming that they are really a manifestation
of the internal machinery that was added for class/type unification,
and have been exposed in the language itself so people can see if it's
a useful feature.

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/



From benji_75 at hotmail.com  Wed Sep 18 05:31:05 2002
From: benji_75 at hotmail.com (Ben Gerblich)
Date: Wed, 18 Sep 2002 19:01:05 +0930
Subject: try except ImportError failing to catch error!
Message-ID: <F304IwPARlZ7bneA9zp0000db23@hotmail.com>

I am importing all module files in the CWD. This part works fine.

I want to add the functionality to catch the error if there is a
typo in the imported module file. The following does not catch
the error!

try:
    self.userDefs[file[:-3]]['class'] = __import__(file[:-3])
except ImportError, e:
    print 'cant do it', e

I still get the runtime error:

Traceback (most recent call last):
  File "./main.py", line 8, in ?
    application = testApp()
  File "/usr/lib/python2.1/site-packages/wxPython/wx.py", line 1581, in 
__init__    _wxStart(self.OnInit)
  File "./mainScript.py", line 3444, in OnInit
    self.frame = MainFrame(None)
  File "./mainScript.py", line 2836, in __init__
    exec code
  File "<string>", line 1, in ?
  File "./Display.py", line 33, in __init__
    self.line(5, 60, 15, 60,999)
TypeError: line() takes exactly 5 arguments (6 given)

I expect this error! as the '999' should not be there.
Why does not the 'cant to it' line print instead?

And is there a way for my handler routine to return the line-number of the 
imported module file that caused the error?

Thanks,
Ben.
Im using:
Red Hat Linux 7.1
python2-2.1-5.i386.rpm
wxPython-2.3.2.1-1-Py21.i386.rpm
wxGTK-2.3.2-1.i386.rpm


_________________________________________________________________
MSN Photos is the easiest way to share and print your photos: 
http://photos.msn.com/support/worldwide.aspx




From marklists at mceahern.com  Wed Sep 18 18:41:31 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 18 Sep 2002 17:41:31 -0500
Subject: When we say 'One obvious way...',  do we mean 'generally obvious' or does that matter?
In-Reply-To: <8d148763.0209181359.68cd789f@posting.google.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOCELLDPAA.marklists@mceahern.com>

[yaipa]
> I LIKE the below code snippet. While it might be Pythonic, it will
> only be 'obvious' to a Python programmer with some experience under
> his/her belt.

Actually (as the author of the code snippet), I would say it's decidedly
unpythonic.  Why go through contortions to restrict access to base class
methods?  Nonetheless, that's what the original poster wanted.

Besides, I needed to fill in gaps in my understanding of metaclasses and
this seemed like the perfect opportunity.

Cheers,

// mark

-




From m.faassen at vet.uu.nl  Wed Sep  4 06:15:27 2002
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 4 Sep 2002 10:15:27 GMT
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com>
Message-ID: <al4mfv$hd0$1@newshost.accu.uu.nl>

Benjamin <phncontact.libero.it@> wrote:
> I just begun to learn how to program. And I choosed Python for
> doing that. surfing around the web i found an article about this
> language, they said it's the ideal language for beginners to the
> world of programming. and i admit it, the syntax is very clear making
> the code very readable, i learn easily the concepts of programming
> and i soon wrote some nice programs. but since i finished my
> newbie-tutorial, i am wondering about one question: is it worth to go
> deeper into python? i talked to a friend of mine, he told me that i
> should try php..
> what should i do?? ^_^ in other words: what is python good for?
> why should i continue with python?

Python's good for everything! Well, perhaps not everything, but it's
a general programming language and it can take quite a while until you
run into its limitations. And every language has its own limitations
and advantages.

Python's main limitation in my mind is that it is relatively
slow compared to code written in C, say. However, don't let that scare
you away -- it's fast enough for many purposes, there's a lot that you
can do to improve performance without leaving Python, and performance
is a subtle issue; as Python is a higher level language sometimes it is
easier to write a performing but more complicated algorithm in
Python, compared to a lot of time and struggle in C. Don't worry too
much about performance when you're starting out.
Besides, if you compare to PHP I would think Python wouldn't do so
badly performancewise anyway.

I wouldn't recommend learning PHP at this stage; from what I've seen and
heard about it is it not very good at showing beginning programmers
how to code cleanly -- Python is a lot better at that (though there are
certainly other languages that would help there as well). PHP might
obscure the concepts from you or expose you to bad practices. Then again,
it might help to deliver quick results and thus stimulate you to
program more, so I don't imagine it would do a lot of harm either as long
as you stay aware of good programming practices.

PHP was developed to create dynamic web pages on the server side, and
grew from that. People still mostly use it to create web applications.
It is certainly possible to write web applications with Python; I do
that for a living using Zope (http://www.zope.org). Here's a page
describing a whole range of options if you want to do web programming
with Python:

http://www.python.org/topics/web/

But Python can be used for many other purposes besides web programming. 
It depends entirely on what you want to do. Best way to learn is to set
yourself a goal that's not too ambitious and try implementing it.
Even if you fail you'll have learned new things.
 
Regards,

Martijn 
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From jim at dsdd.org  Thu Sep 19 22:06:37 2002
From: jim at dsdd.org (Jim Meier)
Date: Fri, 20 Sep 2002 02:06:37 GMT
Subject: Smalltalk style inspector
References: <mailman.1032433164.17855.python-list@python.org>
Message-ID: <Novi9.375064$f05.18290942@news1.calgary.shaw.ca>

On Thu, 19 Sep 2002 12:00:08 +0100, Andrew Thompson wrote:
> I don't even know the process for contributing the module...

1. Make it available for download.
2. ???
3. Profit!

-Jim



From bkc at Murkworks.com  Wed Sep 11 12:51:07 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Wed, 11 Sep 2002 12:51:07 -0400
Subject: UNIX processes
References: <mailman.1031654172.15956.python-list@python.org> <kuhegycb8r.fsf@lasipalatsi.fi> <mailman.1031755292.20295.python-list@python.org>
Message-ID: <3d7f71e4$1_5@goliath.newsgroups.com>

Can you describe your module more?

I want to implement something like top/gtop on Linux, but using XWT as the
client via Zope or a CGI.

I also want to emulate gkrellm too.

So, I need load averages (I can get via /proc) but I also want top style
output, memory usage and current CPU usage by process.

Folding threads together into a "single process object" would also be nice
for presentation purposes.


--
Novell DeveloperNet Sysop #5

_
"Bo M. Maryniuck" <b.maryniuk at forbis.lt> wrote in message
news:mailman.1031755292.20295.python-list at python.org...
On Tuesday 10 September 2002 13:52, Erno Kuusela wrote:
> use the ps program (ps -e works on linux). there is no other
> standard api for it.

I already made native Python module for this, because I don't like 'ps' in
this case. Well, module works even faster. ;-)

--
Regards, Bogdan

I doubt, therefore I might be.






-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From dsavitsk at e-coli.net  Mon Sep  9 11:13:31 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Mon, 09 Sep 2002 15:13:31 GMT
Subject: PythonWin bug [very weird]
Message-ID: <vU2f9.7489$yt3.3820112@newssrv26.news.prodigy.com>

This is PythonWin 145 (patched for asp), Python 2.1.3, Win2k.  I have no
means for testing other versions.

The bug is that pressing the wrong sequence of buttons can render a file
unsaveable. To reproduce it:
1. open a .py file
2. make some changes
3. press the X to close the file (not to close PythonWin)
4. when asked to save (Yes, No, Cancel) press Cancel
5. Save the file (File | Save)
6. Any new edits to the file are unsaveable (the buton is grayed out). To do
so, you must copy the entire contents, close the file, re-open it, and paste
the clipboard.

-doug




From gerhard.haering at opus-gmbh.net  Mon Sep  2 07:51:27 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 2 Sep 2002 11:51:27 GMT
Subject: Must have Python modules?
References: <_THc9.42153$vY2.1039887@e3500-atl2.usenetserver.com>
Message-ID: <slrnan6k4s.eg.gerhard.haering@haering.opus-gmbh.net>

Robert Oschler wrote:
> I've scanned Vaults of Parnassus and several other hot Python sites and
> was overwhelmed by the amount of stuff that is available.  So I'm asking
> you, what is your favorite, absolute must-have Python module?

What do you want to hear? Everyone's list of favourite Python extensions?
There's likely is more than one for most of us. My favourite is the "this"
module.

Don't know what you expect :-) Maybe better ask a specific question as to
what is the recommended module for a certain task, like parser generation,
GUI, database access, etc.

Previous similar discussion: http://tinyurl.com/19a9

-- Gerhard


From tismer at tismer.com  Sun Sep 15 14:24:18 2002
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 15 Sep 2002 20:24:18 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <j43csbruq7.fsf@informatik.hu-berlin.de>
Message-ID: <3D84D052.1080600@tismer.com>

Martin v. L?wis wrote:
> "David Abrahams" <david.abrahams at rcn.com> writes:
> 
> 
>>I think it still matters. You don't want people who need stackless'
>>capabilities to have to use a non-standard Python forever, do you?
> 
> 
> Unfortunately, the current implementation is processor-dependent; it
> supports x86, SPARC, and PPC only (not really sure about PPC).

Since today it supports X86, Sparc, PPC and Cygwin.

> It also
> limits extensions: if they call back to Python, they must not keep
> "global" variables on the stack (global in the sense that their
> address is known outside their function).
> 
> These limitations require Stackless users to be more aware of the
> inner workings of Python, or they risk crashes - something that would
> be considered a bug in standard Python.

Yes, but :-)
Stackless is built in a way that by default there is no difference
to the standard. No stack switching occours, all is identical.
If you want stackless behavior, you must switch it on.
And if you want to use task switching, you need to import the
stackless extension (which is now apart).

Therefore I'd claim that putting Stackless into the core trunk
would not do any harm. Users who want to use it can use it,
others may ignore it (and it is very tiny, about 50k).

This could be something like the thread package, which also
does not work on all platforms.

ciao - chris
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From fperez528 at yahoo.com  Thu Sep 26 14:14:20 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 26 Sep 2002 12:14:20 -0600
Subject: Creating daily/monthly averages from datafiles
References: <mailman.1033011994.21740.python-list@python.org>
Message-ID: <amvips$9t0$1@peabody.colorado.edu>

At the risk of sounding stupid, may I suggest you look into Numeric for this? 
It has most of what you want already in it, and it's fast to boot. Multi-year 
data every 15 minutes is bound to be a lot, I can't imagine how slow doing 
that in raw python will be. Unless you are going to throw a 1024 node cluster 
at this simple problem :)

Now, the trick with Numeric is that it wants homogeneous data, so none of that 
date/data mixing you want. But with a bit of helper tables, at some extra 
storage cost, I'm willing to bet you can fix this very easily. Store the 
numerical data in one or a few Numeric arrays, and have a few auxiliary 
python structures (dicts, probably) to help along with the indexing. This way 
you can quickly find where in the Numarray the data you want is, and then you 
can efficiently process it using Numeric.

Just my $0.02.

Cheers,

f.


From michael at stroeder.com  Tue Sep 17 07:14:51 2002
From: michael at stroeder.com (=?ISO-8859-1?Q?Michael_Str=F6der?=)
Date: Tue, 17 Sep 2002 13:14:51 +0200
Subject: More on Protecting Source Code
References: <mailman.1032247760.982.python-list@python.org>
Message-ID: <defj5-8e3.ln1@junker.stroeder.com>

David LeBlanc wrote:
> Well, I think of it this way: machine coded binaries are more like a 128 bit
> key and Python is more like a 40 bit key. I agree that nothing is safe from
> reverse engineering; it's a matter of how much pain and money it takes to
> unscrut the inscrutable. Python could do better - how I'm not totally sure.

Dave, without knowing the business model for the software you write 
nobody can give any reasonable answer.

As usual: Define your business model first and derive the technical 
solution from that afterwards.

Ciao, Michael.



From rajarshi at presidency.com  Wed Sep 18 13:21:26 2002
From: rajarshi at presidency.com (Rajarshi Guha)
Date: Wed, 18 Sep 2002 13:21:26 -0400
Subject: isFloat: Without Exception-Handling
References: <ama6nt$n6i$06$1@news.t-online.com> <pan.2002.09.18.11.40.50.766802.29524@presidency.com> <%82i9.1366$bh1.943@fe03>
Message-ID: <pan.2002.09.18.13.21.24.660912.29524@presidency.com>

On Wed, 18 Sep 2002 12:47:40 -0400, Steve Holden wrote:


>>
>>
>> def isFloat(string):
>>   if type(string) == type(1.0):
>>     return 1
>>   else
>>     return 0
>>
>> This should do what you want without exceptions
> 
> I doubt that VERY much. Please test your code before submitting, or
> ensure you note it wasn't tested ...

[snip]

>>>> isFloat("banana")
> 0
>>>> isFloat("1.2")
> 0

If you call it as isFloat(1.2) it works


From sross at connectmail.carleton.ca  Tue Sep 24 22:57:53 2002
From: sross at connectmail.carleton.ca (Sean Ross)
Date: 25 Sep 2002 02:57:53 GMT
Subject: Removing all occurences of a character from a string.
Message-ID: <amr8nh$rd8$1@driftwood.ccs.carleton.ca>

Hi.
I have a string, say "'Hello World'", where Hello World is in single
quotes(i.e., 'Hello World'). How can I remove the single quotes so that
"'Hello World'" becomes "Hello World"?
Thanks
Sean Ross




From cliechti at gmx.net  Fri Sep 27 18:17:03 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 28 Sep 2002 00:17:03 +0200
Subject: Better terminal for windows?
References: <3D94B037.2020202@motorola.com> <mailman.1033158756.25404.python-list@python.org> <3D94D1DD.40407@motorola.com>
Message-ID: <Xns92972F21C889cliechtigmxnet@62.2.16.82>

Stephen Boulet <stephen.boulet at motorola.com> wrote in 
news:3D94D1DD.40407 at motorola.com:
> Gerhard H?ring wrote:
>> cmd.exe isn't *that* bad once you activate tab completion in the
>> registry. Have you done that?
> 
> Nope. How do you do that? I couldn't figure it out just looking through 
> regedit.

had to google for that too:
http://www.winnetmag.com/Articles/Index.cfm?ArticleID=8746

chris

-- 
Chris <cliechti at gmx.net>



From aleax at aleax.it  Mon Sep 23 02:39:23 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 23 Sep 2002 06:39:23 GMT
Subject: Universally unpacking tuples
References: <158d3913.0209222233.123a9675@posting.google.com>
Message-ID: <vGyj9.150476$pX1.5396552@news2.tin.it>

Peter Bittner wrote:
        ...
> arbitraryTuple = ('abc', '123', 'qwert', ...)
> while len(arbitraryTuple) > 0:
>   elem, arbitraryTuple = arbitraryTuple
>   print elem
> 
> Well, I know this doesn't work since unpacking requires the correct
> number of elements on the left side... :-(
        ...
> mylist = ['abc', '123', 'qwert', ...]
> for i in range(len(mylist)):
>   elem = mylist[i]
>   print elem

The second snippet works just as well whether name mylist refers to
a list, or to a tuple.  Tuples are indexable just like lists are --
they're not _modifiable_, but that's a different issue.

However, the simplest solution of all, also universally applicable
to all sequences (lists, tuples, strings, ...) and other iterables
(directories, files, ...) is:

for elem in whatever:
    print elem


Back to the problem of what you try to accomplish in the first
statement of the first loop above:

elem, whatever = whatever[0], whatever[1:]

this sets elem to the first argument of 'whatever' (can be a
tuple, list, string, ...) and re-binds name 'whatever' to all
elements except the first (keeping the same type as was in use
before -- again, could be tuple, list, string, etc, etc).  You
do not need it here, and it would be a substantial waste of
both programming and computational effort to use it, but in
some other cases it might indeed come in handy.


Alex



From aleax at aleax.it  Tue Sep 17 07:31:40 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 17 Sep 2002 11:31:40 GMT
Subject: Regular expression help needed
References: <mailman.1032256460.12295.python-list@python.org> <3D870A74.5D2D7595@fast.no>
Message-ID: <woEh9.119795$pX1.4310959@news2.tin.it>

Torkil Grindstein wrote:
        ...
> Btw, I have no knowledge of SAX. Perhaps I should dig into that?

Yes, I WOULD suggest looking into Sax (and minidom, and Prescod's
pulldom) for any kind of XML processing.  Jones and Drake, "Python
& XML", O'Reilly, is a good book dedicated to the subject (very
little on pulldom -- i'm learning to appreciate pulldom myself just
quite recently, as it's saving my bacon on a complex XML transcript-
with-mods project, one that needs to work on large enough files that
minidom is not an option, but also needs enough localized "dig into
this node and change it like this and like that" that SAX would be
rather a PITA).


Alex



From jp at NOSPAMdemonseed.net  Wed Sep 11 11:07:33 2002
From: jp at NOSPAMdemonseed.net (jason petrone)
Date: Wed, 11 Sep 2002 15:07:33 +0000
Subject: Python iCalendar API?
References: <mailman.1031673134.15031.python-list@python.org>
Message-ID: <l7mnla.rii.ln@demonseed.net>

Erik Price <erikprice at mac.com> wrote:
> Is there a Python module that provides capability of processing files 
> in .ics (iCalendar) format?  I found nothing concrete on Google or at 
> python.org ...

I have a horrible, messy, rats nest of an ical module.  It converts my
gnomecal/evolution calendars into html fragments.  

If you promise to not tell anyone I'm the one that wrote it I can send it 
to you.

jason


From Pekka.Pessi at nokia.com  Tue Sep 17 06:56:39 2002
From: Pekka.Pessi at nokia.com (Pekka Pessi)
Date: 17 Sep 2002 13:56:39 +0300
Subject: Detecting IPv6 support 
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032245506.8580.python-list@python.org> <j4n0qhjbkf.fsf@informatik.hu-berlin.de>
Message-ID: <pvsn08lw9k.fsf_-_@agni.research.nokia.com>

loewis at informatik.hu-berlin.de (Martin v. L?wis) writes:
>- the IPv6 patches took about 15 months to integrate. For many months,
>  the patches were unreviewed, because nobody thought to have the
>  expertise to review them. I eventually tought myself the necessary
>  background to review the patch, after which about 4 further
>  iterations where necessary before the patch was installed, and about
>  10 changes after the patch was installed. 

	Speaking of IPv6: is the IPv6 enabled on the Pythonlabs Windows
	distribution by default (sorry, I don't have a Windows box in my
	use at the moment)? Does Python support the older IPv6 stacks
	for NT4 and W2K?

	What is the recommended method of finding out if Python and the
	underlying OS support IPv6? I can, for instance, create an IPv6
	socket with a Python version that does *not* support IPv6.

					Pekka


From erikprice at mac.com  Wed Sep 18 08:57:59 2002
From: erikprice at mac.com (Erik Price)
Date: Wed, 18 Sep 2002 08:57:59 -0400
Subject: Will "Python in a Nutshell" be too short?
In-Reply-To: <02091723502402.21243@arthur>
Message-ID: <4278C014-CB06-11D6-B699-00039351FE6A@mac.com>

On Tuesday, September 17, 2002, at 05:50  PM, Alex Martelli wrote:

> The number of pages I had to devote, in particular, to tkinter and
> xml, still boggles my mind -- yet, I do see that they add to the book's
> usefulness more than they take away from it.

I just hope that the "Sample Chapter" that ORA offers from its website 
isn't the XML chapter.  It's not that I'm not interested in XML or that 
I don't think it will be worthwhile, but rather because it seems like 
almost every "Sample Chapter" offered on their site is "the XML 
chapter" for the given book!!

;)


Erik

PS: I know this is an overstatement.




--
Erik Price                                   (zombies roam)

email: erikprice at mac.com
jabber: erikprice at jabber.org




From sholden at holdenweb.com  Mon Sep 30 13:35:47 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 30 Sep 2002 13:35:47 -0400
Subject: rank-0 arrays?
References: <3d988160.79509684@news>
Message-ID: <U00m9.6260$_32.1554@fe04>

<brobbins333 at shaw.ca> wrote in message news:3d988160.79509684 at news...
> In Numeric Python, what are rank-0 arrays?
> Must not arrays have at least 1 dimension?
>

Only if they must contain data. A xero-rank array will (sort of) be the None
of Numeric arrays.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From tdelaney at avaya.com  Thu Sep 19 21:18:35 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 20 Sep 2002 11:18:35 +1000
Subject: Minimal debug/rep functionality
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5BA@natasha.auslabs.avaya.com>

> From: holger krekel [mailto:pyth at devel.trillke.net]
> 
> Delaney, Timothy wrote:
> > > From: Peter Hansen [mailto:peter at engcorp.com]
> > > 
> > As I said, I *expect* to be able to release (there's no 
> commercial reason
> > not to, and we have done so with such things in the past).
> 
> Maybe we are turning circles [1] but why don't you do it, then? 
> 
> Maybe you need to invest some time in obfuscating your
> source code and you don't want to do it? 
> 
> Why do you *expect* others to solve this problem for you?

Holger, what the hell are you talking about?

I'm not releasing yet because:

1. The tool is not feature-complete.

2. The code is not in a fit state for release - needs more documentation,
etc - it's in a state of high flux, and it's a personal (but work-related)
project so I've been a bit lax ;)

3. I have not even approached my employer as to whether I am allowed to
release it (but I expect that when I do so I will be given permission).

As I stated, I would like this to get into the core library eventually - how
does that in any way suggest that I want to:

1. Obfuscate the code

or

2. Expect others to do the work for me

???

Tim Delaney



From max at alcyone.com  Fri Sep  6 16:06:51 2002
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 06 Sep 2002 13:06:51 -0700
Subject: Interrupting Python
References: <yz5e9.4237$rU2.145619@news4.srv.hcvlny.cv.net>
Message-ID: <3D790ADB.912B9E3E@alcyone.com>

Bob Easton wrote:

> I have a script that can run, accessing network resources, for several
> days.
> Since the script does not normally need keyed input, exception
> processing
> does not raise the keyboard exception until after the program ends
> normally.
> I would like to be able to interrupt if from the keyboard, but have
> not
> learned the trick.  How can I do this?

If you're under a UNIX-like operating system, kill its process ID with
the -INT signal.

	kill -INT processID

You can find the process ID with ps if you don't already know it.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ There is nothing so subject to the inconstancy of fortune as war.
\__/ Miguel de Cervantes
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From Gareth.McCaughan at pobox.com  Sun Sep 15 10:31:42 2002
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: Sun, 15 Sep 2002 15:31:42 +0100
Subject: Extending a C++ App
References: <996eade8.0209120912.3a653d7@posting.google.com> <n0qnhvfn.fsf@boeing.com> <996eade8.0209122352.45721938@posting.google.com> <altqiu$fjr$1@bob.news.rcn.net>
Message-ID: <slrnao96ee.52u.Gareth.McCaughan@g.local>

David Abrahams wrote:
> 
> "Bwuce_Wee" <bwucew at yahoo.com> wrote in message
> news:996eade8.0209122352.45721938 at posting.google.com...
> 
> >     I'm busy investigating Boost.Python...
> 
> In that case, get Boost.Python v2 using the instructions at
> 
> http://mail.python.org/pipermail/c++-sig/2002-May/001100.html
> 
> It's nearing release, and v1 isn't long for this world.

If you recommend new users to start with v2 rather than v1,
then it's already time for a (pre-)release, no?

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
.sig under construc


From brian at sweetapp.com  Mon Sep 23 13:41:33 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Mon, 23 Sep 2002 10:41:33 -0700
Subject: WSDL for XML-RPC ??
In-Reply-To: <_YAj9.21316$0p1.377838@news2.ulv.nextra.no>
Message-ID: <001001c26328$75194ff0$df7e4e18@brianspiv1700>

> SOAP has WSDL to describe a webservice. Doeas XML-RPC have something
> similar?

XML-RPC has an optional set of methods that allow introspection. They
are:

system.listMethods()
system.methodSignature(method_name)
system.methodHelp(method_name)

Here is a usage example:

>>> from xmlrpclib import Server
>>> s = Server('http://www.sweetapp.com/cgi-bin/xmlrpc-test/rpc1.py')
>>> s.system.listMethods()
['add', 'get_source_code', 'mul', 'pow', 'system.listMethods',
'system.methodHelp', 'system.methodSignature', 'system.multicall']
>>> print s.system.methodHelp('pow')
pow(x, y[, z]) -> number

With two arguments, equivalent to x**y.  With three arguments,
equivalent to (x**y) % z, but may be more efficient (e.g. for longs).
>>>

For information on how to create servers that automatically support
these methods, see:
http://www.sweetapp.com/xmlrpc/

Cheers,
Brian




From hwlgw at hotmail.com  Sat Sep 28 06:41:52 2002
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 28 Sep 2002 03:41:52 -0700
Subject: comp.lang.python...
References: <mailman.1033054480.24506.python-list@python.org>
Message-ID: <cb035744.0209280241.53ffc327@posting.google.com>

"Vincent A. Primavera" <vincent_a_primavera at netzero.net> wrote in message news:<mailman.1033054480.24506.python-list at python.org>...
> Hello,
> 	Stupid question...  What information do I need to set up
> comp.lang.python with my news client(Pan)?
> 
> 	Thank you,
> 
> 	Vincent A. Primavera.

Perhaps stupid answer: did you know that you can read news using only
a web browser?  And sent messages too.  Have a look at
http://groups.google.com/groups?q=comp.lang.python

'''
Alexander Graham Bell is alive and well
in New York, and still waiting for a dial tone.
'''


From erikprice at mac.com  Tue Sep 17 10:26:09 2002
From: erikprice at mac.com (Erik Price)
Date: Tue, 17 Sep 2002 10:26:09 -0400
Subject: OT: Re: Python on MacOS X 10.2: a warning and instructions
In-Reply-To: <1F30F09A-CA06-11D6-8521-00050287C0A6@earthlink.net>
Message-ID: <691A8B2C-CA49-11D6-888F-00039351FE6A@mac.com>

On Tuesday, September 17, 2002, at 02:24  AM, Kow K wrote:

> I probaby won't migrate to Jaguar for a while.
> I can't trust Apple these days. Now, Apple is such a
> shameless company that they don't care privileges
> of previous users. Why do 10.1.x users have to pay as
> much as $129??? No matter how tight their budget is,
> that sucks.

Well, 10.0.x came with my computer, so I guess I paid for that version. 
  And it sucked.
Then 10.1.x came out and I got an upgrade package for free from a 
nearby Apple store.  You could also get one mailed to you for $20 
directly from Apple.  That version was a much-needed improvement.
Now 10.2.x is out and costs $129, but if I only have to pay for every 
even-numbered point release, it doesn't seem so bad.  Let's see if they 
offer 10.3.x for free (or $20).

It's the other things Apple is doing that worry me: stealing from 3rd 
party developers putting Watson into Sherlock.  [Though I do appreciate 
the decision to move "Find" back into the "Finder"]

To the OP: Did you say that you were able to get a Mac OS X version of 
Python working with GNU readline without using Fink??


Erik






--
Erik Price                                                  (zombies 
roam)

email: erikprice at mac.com
jabber: erikprice at jabber.org




From jb at cascade-sys.com  Tue Sep 17 06:54:36 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 17 Sep 2002 03:54:36 -0700
Subject: threads or queue for this task
References: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com> <mailman.1032202279.8360.python-list@python.org> <Qgqh9.117134$pX1.4203010@news2.tin.it> <mailman.1032241763.6933.python-list@python.org> <d7Ah9.103593$ub2.2156613@news1.tin.it>
Message-ID: <3D8709EC.8020208@cascade-sys.com>

Alex Martelli wrote:

>James J. Besemer wrote:
>        ...
>  
>
>>>>Generally, Queues never get "full" and you generally don't care if
>>>>they're empty.
>>>>        
>>>>
>>>You can define Queues that can hold a maximum number of items; [...]
>>>      
>>>
>
>And I proceed to give a very detailed and rather realistic example.
>How come you don't address it at all?
>

I certainly meant no disrespect. Nor did I mean to contradict anything 
material that you said.

I took your point to be that a bounded queue also is sometimes useful, 
which I do not and did not dispute.

Your example and discussion seemed complete and thorough.  I read it as 
an extension to what I had offered and I had nothing to add or to 
contradict.

In summary, you expressly classed it as an "optional embellishment" 
which seemed consistent with my own view.

It goes without saying that if the designer knows in advance that he 
doesn't have the throughput to handle incoming requests, then a 
significantly more complicated solution is called for.  I alluded to 
more complex topologies and scenarios but I lacked the time and patience 
to spell out the solutions.  Attempting to do so seemed doubly pointless 
as you seemed to be taking the lead on more elaborate scenarios.

So at a high level I am at a loss, as you seem to be picking a fight I 
did not intend to participate in.  Like I said, I thought we were 
largely in agreement, while expanding upon different scenarios.

>>Generally," NOT having a limit is by far the more elegant solution.
>>    
>>
>
>I dispute the generality you assert, 
>

I agree there are situations where unbounded queues might be 
inappropriate.  I only meant to suggest to this particular novice that 
it was a complication that I would avoid in a first cut solution.  

In this, I am "coming from" the standpoint of Extreme Programming -- you 
may recall -- that dictates you start by writing the simplest solution 
you don't know won't work.  Performance optimization and other 
"embellishments" are always deferred unless/until you can't avoid it.

That simplest, initial solution IMHO, is an unbounded queue.  I further 
characterize it as "elegant" as it's fully functional for many 
applications and it's as simple as it gets.    To me, real world 
complications, however realistic, tend to result in less "elegant" 
solutions but perhaps you disagree.

>and, again, I gave a
>specific example: where the time to produce work requests
>(posted to a queue) is always far less than the time to
>consume/process a work-request taken from the queue.  
>
>This
>is hardly a RARE situation: on the contrary, it's quite a
>normal state of affairs that processing a work-request is
>much more time-consuming than producing it.
>

Incidentally, the ratio of the "time to produce work requests" vs. the 
time to process them is immaterial (since the work requests may be 
created on a different machine).  Queuing theory usually refers to "the 
arrival rate" of the requests vs. the the average processing time that 
matters.  Little's result tells us that in a steady state system, if the 
average arrival rate is R (arrivals per unit) and the average processing 
time is T (units) then the average queue length will be N:

    N = R *  T

Of course, Little's result only holds for steady state systems, where 
the arrival rates are relatively low or the processing times are 
relatively short.  It may be more useful to consider inter-arrival times 
(the inverse of arrival rates), I:

    N = T / I.

Then it's easy to see that as the inter-arrival times shrink then the 
queue length can become arbitrarily large.  Of course, this all is 
assuming stochastic processes, with random variable arrival rates and 
processing times.

If you're looking at a static system with a fixed processing time and 
known arrival rates then it's a matter of simple arithmetic  Bottom line 
is if the requests exceed the capacity of the machine then your queue 
length grows without bounds and eventually you're hosed.  

Frankly, any time you do NOT have enough capacity to handle the expected 
traffic then your system design is flawed on a fundamental level.  You 
can perhaps mitigate the problem with bounded queues or what not but 
that still leaves you with possible data overruns or other problems 
upstream from the writer end of the queue.  Bounded queues only push the 
problem of to some other point in the system, since the writer is 
suspended and not responding to the requests.  Ultimately, at some 
level, the designer is going to have to deal with the problem of 
requests that cannot be fulfilled.  Conversely, if it is given that 
there is ample CPU power then he can dispense with all the complexity 
and use unbounded queues or something even simpler.

In my experience, computer queuing systems first and foremost have to be 
designed so that they can handle the max throughput without fail.  In a 
few embedded systems, throughput considerations drove the choice and 
speed of the CPU(s) and other hardware "assist."  More commonly, the CPU 
speed was fixed and max throughput drove constraints on request rates or 
processing times, usually both.  In a number of systems I worked on, 
queues always only had 0 or 1 items in them.  In one system like this, 
the queues actually were implemented in hardware and writer CPUs 
actually locked up until the reader emptied the previous request.  In 
another, the CPUs themselves did not lock but all further communication 
was effectively blocked until the "queue" cleared.  Blocking for any 
length of time was considered a bug and the system was designed never to 
do that.  In other systems, queues had fixed sizes but this was in cases 
where the RTOS required the developer to declare a fixed amount of 
space.  In this case, the queue size was determined by measuring the 
average arrival rates and processing times (plus a fudge factor), but 
the limits were chosen to conserve RAM; like the CPU's stack they were 
designed never to be exceeded in normal operation.  Then there were 
many, many cases where throughput was known to be sufficiently low that 
queue length was never a consideration.  Usually arrivals were low 
enough and processing time fast enough that the average queue length was 
virtually zero.

As far as producers outperforming consumers being the "normal state of 
affairs" -- while I agree once it was a common problem, I find it to be 
increasingly not to be the case, now that we have computers with 
multiple Giga Hz and hundreds of Mega Bytes of RAM.  Hell, my present 
desktop machine's RAM is 50X larger than my first hard drive. This is 
results in a qualitative difference in programming.  One of the really 
cool things about Python in these modern machines is you can do all 
sorts of things that previously were virtually impossible.  E.g., 
reading and writing multi megabyte files as a unit and treating them 
internally as a big string, where previously only line at a time 
processing was possible.  Used to be you'd never think of using 
"unbounded" strings -- or lists or other large, dynamic structures.  Now 
they're the norm.  Or -- in this case -- Python's unbounded queues are 
useful over a much wider range of conditions than previously thought. 
 Used to be all your elegant little CS algorithms never worked on a real 
computer because you had to translate them from their ideal world to the 
real world of extremely limited memory, disk and CPU power.  While we 
may always be machine limited for some applications, I find it 
delightful that increasingly we can ignore those nasty details, merely 
write cosmic-simple Python code and have it all work.

Really and truly, unbounded Queues seem to me to be the Pythonetic way 
to go.  ;o)

>Therefore, I consider your assertion "Generally, Queues never
>get *full*" a dangerous over-simplification: they get full
>if you specifically define them to have a maximum size (and
>post more items than you peel off), 
>

Yes.  If you make them too small they will overflow.  

I was saying DO NOT do this.

>and it's not a rare case
>to want that ("an optional embellishment", as I said, but one
>that does happen quite normally in general cases).
>

I don't know how to react to your shift in emphasis on this point from 
mere "optional embellishment" to "dangerous over-simplification".   I 
think you're exaggerating the "danger."  What is the "dangerous" outcome 
you're striving to avoid?

I think you were right the first time.

>If the number of work requests is not unbounded, 
>

The total number of requests is immaterial.  You mean if the arrival 
rate is sufficiently high or the processing time too long.

>you may be
>thrashing your computer into disgraceful performance rather
>than having your program fail with an exception -- wasting
>unbounded amounts of memory on an unbounded queue, 
>

Yes, I acknowledged that your system can thrash or fail some other way 
if you don't have enough power to process the requests.

>where a
>single, simple call to make it bounded would solve it all
>like magic.  
>

Magic sounds good but what about the time-critical requests that don't 
get serviced in time?

Bounded queues are A solution to some problems but not The solution for 
all situations.  

And as I said, they push part of the problem to some other place in the 
design, which may or may not be a good thing.

>>Anyway, I felt the "general" case was more applicable to this novice's
>>line of questioning.  He/she already was confounded by some
>>non-essential details.
>>    
>>
>
>We disagree on didactics.  I think over-simplification is a
>bad way to teach: everything must be kept as simple as possible,
>*but no simpler than that*.  
>

Is it my turn to retort with a Monsieur de La Palice allusion?

Actually, it sounds like we agree on didactics but disagree on exactly 
where to draw the line.  And I don't see that we're all that far apart.

I'd be curious to hear Robin's opinion on the matter.

I also sense that I've somehow done something to piss you off.  I'm 
sorry if I did but I can't think what it would have been.

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From maciej at maciejsobczak.com  Tue Sep 17 13:20:50 2002
From: maciej at maciejsobczak.com (Maciej Sobczak)
Date: Tue, 17 Sep 2002 19:20:50 +0200
Subject: Socket server and threads
References: <1032217212.2735.21.camel@caspian.everglade> <mailman.1032223099.23141.python-list@python.org>
Message-ID: <am7oeh$cdd$1@sunsite.icm.edu.pl>

Hi,

"Rob Brown-Bayliss" <rob at zoism.org> wrote in message
news:mailman.1032223099.23141.python-list at python.org...
>
> Hi, I am trying to write an app. it Has a gui, and I am using xmlrpc via
> socket server to communicate with other apps on other machines.

I envite you to try this library:

http://www.maciejsobczak.com/prog/yami/

You can go directly to:
http://www.maciejsobczak.com/prog/yami/impl/index.html#python

It uses sockets to provide object-based communication abstraction for C,
C++, Tcl and Python.
It can greatly simplify your code (including threading issues), unless you
need to use some specific protocol for some reason.

Note that there is work in progress on 2.0 version.

Do not hesitate to contact me if you have any questions.

Cheers,

--
Maciej Sobczak
http://www.maciejsobczak.com/





From mhammond at skippinet.com.au  Mon Sep 23 20:12:41 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 24 Sep 2002 00:12:41 GMT
Subject: COM: Returning IDispatch or NULL
References: <a230569c.0209230700.14116a1d@posting.google.com>
Message-ID: <Z5Oj9.12374$6g7.37565@news-server.bigpond.net.au>

Juan Carlos CORU?A wrote:
> Hello,
> 
> I'm trying to develop a COM Server in python, that has a method that
> returns an IDispatch object but sometimes it must return nothing. See
> the following example programmed in C/SIDE (Microsoft Navision
> Financials ERP):
> 
> --->
> otherCOMServer := COMServer.returnOtherCOMServer();
> IF ISCLEAR(otherCOMServer) THEN MESSAGE('otherCOMServer is clear');
> <----
> 
> Definition of ISCLEAR (extracted from MS Navision help):
>   The return value is TRUE if:
>     -the automation variable has not been created. See CREATE.
>     -the automation variable has been cleared. See CLEAR.
>     -the automation variable has been assigned NULL by an automation
> server.
> 
> I thing that in this case I must assign NULL to the otherCOMServer
> variable in order to keep the otherCOMServer variable CLEARED.
> What must I return from "COMServer.returnOtherCOMServer()" method to
> assign NULL?
> I have tried:
>   return None
>   return wrap(None)
>   return 1==0
>   return
> 
> but without success. The environment responds "data type not
> supported".

None is the way to do this.  What responds "data type not supported"? 
Have you registered your object for debugging and seen the output? 
There may be a traceback printed.

Mark.



From peter at engcorp.com  Thu Sep 19 07:36:09 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 19 Sep 2002 07:36:09 -0400
Subject: StructuredText -> HTML module??
References: <wbii9.19869$0p1.359879@news2.ulv.nextra.no>
Message-ID: <3d89b69d$1@news.sentex.net>

Thomas Weholt wrote:
> Need a module to parse StructuredText/Wiki into HTML. Looked at PyWiki, but
> the generated code was not what I expected. Does anybody have module or lib
> to do this?

Maybe you could help us by telling us *what* you expected.
Otherwise we're likely to waste our time making suggestions
that are just as inadequate from your point of view.

(For example, ZWiki... or the Zope structured text stuff.)

-Peter



From peter at engcorp.com  Wed Sep  4 22:37:05 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 04 Sep 2002 22:37:05 -0400
Subject: default tuple unpacking?
References: <slrnanaknb.1gh.huaiyu@gauss.almadan.ibm.com> <3d758180$1@news.sentex.net> <3d7584d2$1@news.sentex.net> <6qd6rtdg0e.fsf@thetis.intevation.de>
Message-ID: <3d76c336$1@news.sentex.net>

Bernhard Herzog wrote:
> Peter Hansen <peter at engcorp.com> writes:
> 
> 
>>Hmm... I think there might be an inconsistency or an unexpected
>>result in this.  What if e already contains a tuple?
>>
>>   e = (1, 2, 3)
>>   *a = e        # a is ((1, 2, 3), )
>>
>>That seems sensible, and it's just what happens if you pass
>>"e" in to a function defined as "def func(*a)".
> 
> No it isn't. If you call func as func(e) the tuple being unpacked into
> the function's parameters is (e,) not e. So the situation is more like
> calling func as func(*e) so that *a = e would be the same as a = e or
> perhaps a = tuple(e)

Quite right... my mistake.

>>Now what about this:
>>
>>   a, *b = e     # a is 1, b is (2, 3)
>>
>>That's one possible result, and presumably the one you want with
>>this syntax, but it's not what you'd get with the function syntax:
>>
>>   def func(a, *b):
>>     print a
>>     print b
>>
>>   func(e)
> 
> 
> Again the tuple being upacked to a, *b is (e,)...
> 
>>Here of course you get (1, 2, 3) in "a" and () in b. 
> 
> ... so this is obviously what you should get.

A sign that even when one can use these effectively in programs, one
can be confused by the arbitrary choice of syntax when talking about
it.  Thanks for the corrections.

-Peter



From peter at engcorp.com  Wed Sep 18 00:07:37 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 18 Sep 2002 00:07:37 -0400
Subject: can wxpython do this?
References: <737aefee.0209151702.3f735252@posting.google.com> <spcaou804m5oo6gah31oqaeuqsuj60h0ip@4ax.com> <3d854b44@news.sentex.net> <am7hv3$623$1@news01be.so-net.ne.jp>
Message-ID: <3d87fbfc@news.sentex.net>

Erick wrote:
>     > A better option, although I haven't tried it myself, would be
>     > to use py2exe to build an executable, then package that up
> 
> thanks Peter, but I was talking about something cross-platform. As py2exe's
> page says:
> 
>      py2exe is a distutils extension to convert python scripts into
>     executable windows programs, which are able to run without
>     requiring a python installation.
> 
> 
> do you know if it works across platforms?

I don't believe the current solutions are cross-platform, but
it could be feasible to build a meta-packager which uses the
appropriate package to build executables for different platforms
by calling on the appropriate platform-dependent packager.

Anyway, do you really want all platforms, or a specific subset?

-Peter



From jacek.generowicz at cern.ch  Mon Sep 23 08:00:45 2002
From: jacek.generowicz at cern.ch (Jacek Generowicz)
Date: 23 Sep 2002 14:00:45 +0200
Subject: Python Compiling
References: <amfveg$5g64k$1@ID-53055.news.dfncis.de> <j4lm5w9zft.fsf@informatik.hu-berlin.de>
Message-ID: <tyf7khczziq.fsf@pcitapi22.cern.ch>

loewis at informatik.hu-berlin.de (Martin v. L?wis) writes:

> The question is: why do you want to have a compiler? Using it is much
> more difficult than an interpreter, and more time-consuming during
> development of an application. People often expect significant speed
> increases from compilation, but it turns out that those don't
> materialize. 

How would you reconcile the above statements in context of Common Lisp
compilers ?  For example, working with the CMU Common Lisp compiler
(which, funnily enough, is called Python, and compiles to native
code), should provide a practical demonstration that your statements
are, how shall I put it ... not entirely true.

Specifically:

1) Compiling your code with CMUCL is no more difficult than leaving
   the code uncompiled.

2) It is not more time consuming (on any significant scale - by which
   I mean that a human rarely notices the time-taken-to-compile during
   a normal coding-testing-debugging session).

3) Significant speed increases (wrt the uncompiled code) are usually
   observed.


From alanmk at hotmail.com  Mon Sep 23 17:03:02 2002
From: alanmk at hotmail.com (Alan Kennedy)
Date: 23 Sep 2002 14:03:02 -0700
Subject: Strange values in Accept-Language header?
Message-ID: <f72dac0d.0209231303.12349f2e@posting.google.com>

Dear all,

I know that this is not quite the correct place to post this question,
but I'm not sure where else to go with it (suggestions? nice ones :-).
And I know that there are plenty of web developers/masters here, and
they might be interested.....

On looking at my server logs for the Accept-Language headers (to
answer another post in this group), I came across some strange values
that are definitely not ISO language codes. For example,

Accept-Language: en-us,x-ns1rDGeT4e2FpA,x-ns2f67971hgDw1
Accept-Language: x-ns13V8cN8S1Xz9,x-ns2T329gfxKa7d

I have no idea what these could mean (I've changed the actual values
BTW, on the remote off-chance that somebody's highly personal
information was coded in there)

They are obviously not meant to convey language preferences, in a
general sense.

Perhaps they are

A: A method of communicating language preference between bespoke
clients/servers coded to recognise them?

B: An experimental method of communicating language preferences which
has yet to be standardised?

C: The result of header anonymization through some form of anonymising
proxy? (why bother?)

D: Some form of cracking attempt involving a buffer-overrun
vulnerability?

E: A secret "digital handshake" for the Python Secret Underground ;-)

None of the search engines turned up anything relevant.

Stumped,

Alan.


From timh at zute.net  Tue Sep 24 09:10:13 2002
From: timh at zute.net (Tim Hoffman)
Date: Tue, 24 Sep 2002 21:10:13 +0800
Subject: Saving & Restoring Python objects in XML?
References: <3D8F7800.F757A72D@wyeth.gsfc.nasa.gov>
Message-ID: <3D906435.2030304@zute.net>

Hi

Zope does this, you may like to have a look at the export code in zope, 
to get some ideas.

Tim

Tom Bridgman wrote:
> Hello,
> 
> I'm setting up some codes where I want to save the state of the code in
> an XML file at some time and reload it later to continue the processing.
> 
> It's relatively easy to save out the innards of Python objects into XML
> regardless of how the objects are nested.  But what about loading this
> state information back in from the XML file?  
> 
> Is there a way to write the parser handlers and/or other components (I'm
> using PyXML 0.7 and higher) such that they can assemble the required
> objects as they read in the file?  I'm wondering if there is an elegant
> solution that might not be immediately obvious from the 'brute force'
> method I keep coming up with.  I'd prefer to do this using SAX (since
> I've worked with it some), but using DOM is not out of the question if
> it solves the problem in a straightforward way.
> 
> Thanks,
> Tom
> --
> Dr. William T."Tom" Bridgman           Scientific Visualization Studio
> Global Science & Technology, Inc.      NASA/Goddard Space Flight Center
> Email: bridgman at wyeth.gsfc.nasa.gov    Code 935
> Phone: 301-286-1346                    Greenbelt, MD 20771
> FAX:   TBD                             http://svs.gsfc.nasa.gov/




From Padraig at Linux.ie  Mon Sep 16 11:24:09 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Mon, 16 Sep 2002 15:24:09 GMT
Subject: python performance
References: <3D84E66F.6020701@linux.ie> <zm7h9.51182$AR1.2015178@bin2.nnrp.aus1.giganews.com> <IThh9.4238$cP3.10401@news.iol.ie> <j4r8fumezm.fsf@informatik.hu-berlin.de> <uk7lmm0sj.fsf@online.no>
Message-ID: <tImh9.4274$cP3.10169@news.iol.ie>

Syver Enstad wrote:
> loewis at informatik.hu-berlin.de (Martin v. L?wis) writes:
> 
> 
>>Padraig Brady <Padraig at Linux.ie> writes:
>>
>>
>>>>Calling functions adds overhead.  Running code within functions
>>>>reduces overhead for object access.
>>>
>>>Ah OK, this because it has to search a smaller namespace?
>>
>>No. Inside a function, local variables are accessed with integer
>>indices, instead of by name. This is possible because there is static
>>knowledge about the local variables in a function.
>>
>>
>>>Wouldn't compiling overcome this? 
>>
>>Not if you want to preserve Python semantics; the compiler would need
>>to emit code that performs the same actions that the interpreter
>>currently does.
> 
> What about Psyco?

Interesting. I wonder why something like this isn't done
when creating pyc files, hmm I suppose it would be non portable.
Anyway it didn't help for my very simple program. Well it seemed
to make the timings more volatile (a few microseconds each side of
the previous measurements).

P?draig.



From dsavitsk at e-coli.net  Tue Sep 17 00:46:44 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Tue, 17 Sep 2002 04:46:44 GMT
Subject: More on Protecting Source Code
References: <mailman.1032232999.22661.python-list@python.org>
Message-ID: <Usyh9.9807$yt3.4735811@newssrv26.news.prodigy.com>

"David LeBlanc" <whisper at oz.net> wrote in message
news:mailman.1032232999.22661.python-list at python.org...
> I have read what I consider to be the narrow minded and short sighted
views
> wrt to protecting IP in the form of Python source code, and I would like
to
> make a few points.
>
> * Not everyone believes that all or most software ought to be free. I for
> one, don't.
>
> * Owners of property have the right to control it. If you disagree and own
a
> nice car or house, I'll be right over :->
>
> * Why invest a substantial amount of time and money developing in a
language
> that makes it trivial to gain access to the work product?
>
> Python may be easy to use, but if it's also easy to steal software written
> in it, that's an impediment to it's being widely used in commercial
> products, especially shrink-wrapped products. Perhaps that's a reason why
> there seems to be so few commercial products in Python, beyond those that
> are fairly closely licensed and/or have substantial parts of the app
written
> in C/C++.
>
> I suppose one solution is to modify the Python interpreter with different
> op-codes and that ought to make it somewhat painful for the average
hacker.
> A better solution is to make a .pyc file approximately as hard as a binary
> .exe file to decompile - however that could be done.
>
> David LeBlanc
> Seattle, WA USA
>

I agree that some things should not be open source, but at some point,
however obfuscating the source seems like the wrong way.  In truth most
homes, and all cars, are trivial to break into. Yet we still park on the
street relying on our neighbors to respect the law. In the same way, a
software license should be enough to stop theft.

Further, there are daily calls in newsgroups requesting decompilers to
recover lost source code.  I am comforted knowing that as long as it runs I
can get it back.

I trust my users to respect my rules for software I develop.  Perhaps for
widely distributed projects Python is not the proper choice.  But no single
tool is always appropriate, and for times that one can expect some level or
honesty, Python works.

-doug




From LogiplexSoftware at earthlink.net  Fri Sep 20 08:17:14 2002
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 20 Sep 2002 05:17:14 -0700
Subject: Source code
In-Reply-To: <3D88B0E6.AC08EC27@crary.com>
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> 
	<3D88B0E6.AC08EC27@crary.com>
Message-ID: <1032524234.2820.26.camel@localhost.localdomain>

On Wed, 2002-09-18 at 09:59, Bradley D. Larson wrote:
> I might have missed it... (I read most of the posts on this subject).
> Why not just use a significant info registration key. The registration
> key would be sensative to date, registered name and/or some other
> piece of information that would be required to get the software to work.

The problem with this approach is that, by decompiling the software,
these protections can be easily found and removed.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From jolsen at mailme.dk  Tue Sep 10 03:01:39 2002
From: jolsen at mailme.dk (Jesper Olsen)
Date: 10 Sep 2002 00:01:39 -0700
Subject: Graph algorithms - minimisation
Message-ID: <cf0ad9fb.0209092301.44288bc7@posting.google.com>

Do you know of python extentions that implement efficient graph
algorithms?

In particular I am interested in minimising the number of nodes & arcs
used for representing directed graphs.

If you know of implementations in other languages than python - e.g.
C/C++, I am interested as well.

Cheers
Jesper


From sandorlevi at yahoo.com  Thu Sep 26 07:54:17 2002
From: sandorlevi at yahoo.com (Levente Sandor)
Date: 26 Sep 2002 04:54:17 -0700
Subject: Dynamic Event Handling in wxPython
References: <amt348$2pl$1@news-int.gatech.edu>
Message-ID: <8b5e42a6.0209260354.3a419f40@posting.google.com>

Try this:

def OnMenuSelection(event):
    id = event.GetId()
    # ...your code here...

----
levi

"Mikey at Work" <itsmikeytime at hotmail.com> wrote in message news:<amt348$2pl$1 at news-int.gatech.edu>...
> I'm trying to design a menu in wxPython that connects to an SQL server to
> get its menu commands.  So far, I've been able to display the records from
> the SQL server in the menu.  The problem, however, occurs when I try to
> connect event handlers to the menu items.  I know how to connect the event
> handlers (I'm either going to use EVT_MENU_RANGE or assign a EVT_MENU to
> each menu item when the items refresh from the server), but the function
> called by the event handler has no way of determining the id of the menu
> item (at least as far as I know).
> 
> If I use...
> 
> EVT_MENU_RANGE(lowID, highID, OnMenuSelection)
> 
> ... OnMenuSelection doesn't know which menu item was selected.  Is there a
> way to pass the menu item's id to the function OnMenuSelection or retrieve
> the id in the function?  Thanks for your help.  I'm sorry if some of my
> terminology is off -- I'm kinda new to Python and programming in general.


From steve at ferg.org  Fri Sep 13 09:03:37 2002
From: steve at ferg.org (Stephen Ferg)
Date: 13 Sep 2002 06:03:37 -0700
Subject: [newbie] Buying an introductory/reference text
References: <Mu5g9.1702$U .1021@news04.bloor.is.net.cable.rogers.com> <mailman.1031857662.1669.python-list@python.org>
Message-ID: <b16e4ef7.0209130503.3eb167b0@posting.google.com>

> I believe that "Python: Essential Reference" is about all you need if you
> have programmed in at least one other language.  Top notch book.

Absolutely!

For a very clear and readable introduction, rather than a reference
manual, I suggest "The Quick Python Book" by Daryl Harms.


From Janssen at rz.uni-frankfurt.de  Mon Sep 16 18:39:40 2002
From: Janssen at rz.uni-frankfurt.de (Michael Janssen)
Date: Tue, 17 Sep 2002 00:39:40 +0200
Subject: pickle magically not working
Message-ID: <3d865dac@nntp.server.uni-frankfurt.de>

Hello

at least i've solved it (as a workaround), but it's pretty annoying. 
Maybe someone has an answere :-)

i wanted to pickle a dictionary. The code was fine on my own maschine 
but throwing RuntimeError: "maximum recursion depth exceeded" on other 
maschines (same Linux, same Python-Version 2.2).

The mindthrilling thing was (and is), that i can copy the dictionary 
content with mouse2 into a working dictionary. Original dic still 
refused working. Have a look:

m is a single-key version of the bad dic. mm is hand-made

>>> m == mm
1
>>> m
{(2002, 9, 16, 0, 0, 0, 0, 259, 1): {'virus': 1, 'batches': 186, 
'scan': 152, 'mail': 307, 'spam': 0, 'bytes': 6213773}}
>>> mm
{(2002, 9, 16, 0, 0, 0, 0, 259, 1): {'virus': 1, 'batches': 186, 
'spam': 0, 'mail': 307, 'scan': 152, 'bytes': 6213773}}
>>> try: pickle.dump(m, fo); print "functional"
... except RuntimeError: print "not functional"
...
not functional
>>> try: pickle.dump(mm, fo); print "functional"
... except RuntimeError: print "not functional"
...
functional

copy.copy or copiing via:

    d = {}
    for key in dump_dic.keys():
        d[key] = dump_dic[key]

doesn't help. finally i've improved ;-) the copiing-code to:
    d = {}
    for key in dump_dic.keys():
        d[key[:]] = dump_dic[key]

i.e. making copies of the keys (which are tupel as returned by 
time.localtime(), by the way).

Pickling is now functional, but the question remains: 
  *what can cause this???*

Furthermore: is there a way to examine somewhat deep data-structure, or 
is it more reliable to blame voodoo and "computer-logik" for this ;-)

Michael

PS: Don't ask me, how i've build those nasty dics. Hundreds lines of 
slightly confused code have done it ;-)





From sismex01 at hebmex.com  Fri Sep 13 09:34:18 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Fri, 13 Sep 2002 08:34:18 -0500
Subject: [ANN] PySQLite 0.3.0 released
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F0623302425601@mail.hebmex.com>


> -----Original Message-----
> From: Bob X [mailto:bobx at linuxmail.org]
> Sent: Thursday, September 12, 2002 8:19 PM
> To: python-list at python.org
> Subject: Re: [ANN] PySQLite 0.3.0 released
> 
> 
> Gerhard H?ring wrote:
> > NAME:
> >     pysqlite - Extension module for SQLite databases.
> > 
> > ...[snip]
>
> Windows binary?
> 

Yeesh! You had to quote the whole message just to
add two words?

:-(

-gus
--
-- Please disregard offensive and arrogant corporate
-- legalese at end of this message.
--

Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From henrik.motakef at web.de  Fri Sep 27 17:54:42 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 27 Sep 2002 23:54:42 +0200
Subject: How do you pronounce "__" (double underscore)?
References: <525eaba.0209260956.72bf6786@posting.google.com> <3D934C5C.4000603@mxm.dk> <20020926.130938.1789366143.1524@cybermesa.com> <CnJk9.120675$Pf7.4559925@news1.west.cox.net> <3D94BFAB.5060506@sschwarzer.net>
Message-ID: <87znu3t7x9.fsf@pokey.henrik-motakef.de>

Stefan Schwarzer <sschwarzer at sschwarzer.net> writes:

> Tim Hochberg wrote:
>   > "Jay O'Connor" <joconnor at cybermesa.com> wrote:
>   >>How about "underder"??
>   >
>   > Or maybe "dunder"?
> 
> I think, you shouldn't use two words which sound very similar when
> spoken.

IRL, I tend not to speak __ at all, I just say "def init self...". For
the rare cases when I have to make a distinction between, say, __foo__
and foo, I use stuff like "special foo" for "__foo__", "private foo"
for "__foo" or "protected foo" for "_foo".

But that is usually followed by a remark that one should really invent
better names ASAP.


From fredp at mygale.org.nospam  Wed Sep 18 08:53:54 2002
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 18 Sep 2002 12:53:54 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <Xns928CED692541cliechtigmxnet@62.2.16.82> <6_Ph9.479711$q53.16315681@twister.austin.rr.com> <am9jj9$c2m$1@wanadoo.fr> <3d8868b1$0$28683@echo-01.iinet.net.au>
Message-ID: <Xns928D9818E8E85PaCmAnRDLM@161.48.128.20>

"Rob Hall" <bloke at ii.net> said :

> I settled on wx for the following reasons...
>  - I needed to settle on something, and pretty soon
>  - wx has a native look and feel
>  - wx is fast - compiled C++, whereas tkinter is a script.  I found it
>  - to be 
> almost unacceptably slow in some instances.
>  - PythonCardPrototype would simplify things even further.

not "would" -- "does" :-)

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


From PoulsenL at capanalysis.com  Thu Sep  5 14:22:43 2002
From: PoulsenL at capanalysis.com (PoulsenL at capanalysis.com)
Date: Thu, 5 Sep 2002 14:22:43 -0400 
Subject: Psycos and Pyros don't seem to play well together
Message-ID: <72F73B808A78D511A73600034771D9FF717EB7@dc_exchange2.howrey.com>

I am working on a distributed application using Pyro that I thought would
benefit from Psyco.  All is well until I add "from psyco.classes import *"
and then it gives me some Pyro based error messages (be happy to send them
along if requested) when I launch the client.  I am using the "quickstart"
method to get my objects up and running on Pyro quickly.  Anyone try to use
these two modules together before?  What were your results?

Thanks,

Loren



From pinard at iro.umontreal.ca  Wed Sep  4 18:04:15 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Wed, 04 Sep 2002 18:04:15 -0400
Subject: Garbage collection slowdown? [was: Re: efficient list merging]
In-Reply-To: <al5a1g$hat$1@wanadoo.fr> (Eric Brunel's message of "Wed, 4 Sep
 2002 17:56:38 +0200")
References: <mailman.1031150946.17547.python-list@python.org>
	<al5a1g$hat$1@wanadoo.fr>
Message-ID: <oqvg5lif8g.fsf_-_@carouge.sram.qc.ca>

Hi!  Curious about timings, I played with some of the suggestions from the
"efficient list merging" thread, but not considering lists of lists.

I am surprised by the impressive slowdown coming from leaving the garbage
collector enabled.  Here are a few results:

   0.51 gc-on    0.34 gc-off    1.50 ratio    test_comprehension
   0.52 gc-on    0.37 gc-off    1.41 ratio    test_zip_and_update
   0.43 gc-on    0.29 gc-off    1.48 ratio    test_zip_with_none
   0.51 gc-on    0.28 gc-off    1.82 ratio    test_zip_with_self

Here is the testing code.
   
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: xx.py
URL: <http://mail.python.org/pipermail/python-list/attachments/20020904/123ea8c7/attachment.ksh>
-------------- next part --------------

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard

From dwelch91 at nospam.attbi.com  Mon Sep  2 18:04:13 2002
From: dwelch91 at nospam.attbi.com (djw)
Date: Mon, 02 Sep 2002 22:04:13 GMT
Subject: Why it is so dramatical?
References: <200209021107.25761.b.maryniuk@forbis.lt> <mailman.1030961043.19190.python-list@python.org>
Message-ID: <3D73E035.2040300@nospam.attbi.com>

holger krekel wrote:
> Bo M. Maryniuck wrote:
> 
>>The task:
>>	Adding a text to a text.
>>
>>The problem:
>>	It is tragically slow to add string to huge string.
> 
> 
> That's because strings are immutable objects. When you issue
> 
>     somestring += otherstring
> 
> a new string object is allocated where the contents of
> somestring and otherstring are copied in.  Then the
> name 'somestring' is bound to this new string.
> 
> If you have something like this in a loop it's often
> *much* better to work with a list and later do a 
> 
>     "".join(stringlist)
> 
> As you observed yourself this is around 100-1000 times faster.  
> 
> And consider: if strings were not immutable but modified in-place
> (like lists) you couldn't use them as dictionary keys. 
> 
> if-it-hurts-don't-do-it-ly, yours Holger
> 

Stupid question:

If ""join() is 100-1000 X faster, why doesn't the Python interperter 
translate s = s1 + s2 + ... into s = "".join( [ s1, s2, ... ] ) 
automatically?

Just curious...

D



From eric.brunel at pragmadev.com  Fri Sep 20 10:43:45 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Fri, 20 Sep 2002 16:43:45 +0200
Subject: Setting cursor colors in tkinter
References: <rvFi9.926$fO.166520@kent.svc.tds.net>
Message-ID: <amfbo5$e2r$1@wanadoo.fr>

Edward K. Ream wrote:
> Hi,
> 
> I would like to set the foreground an background colors of a cursor.
> 
> Following:
> 
> http://www.tcl.tk/man/tcl8.3/TkLib/GetCursor.htm#M4
> 
> I am synthesizing a cursor description with
> 
> cursor="xterm " + fg + " " + bg
> 
> yielding, e.g.,
> 
> cursor = "xterm red green"
> 
> I then call:
> 
> body.configure(cursor=cursor)
> 
> However, Tcl complains:
> 
>     self.tk.call((self._w, 'configure')
> TclError: bad cursor spec "xterm red green"
> 
> Can anyone say how this should properly be done?  Put another way, I may
> be asking what the proper format of
> 
> "name [fgColor [bgColor]]"
> 
> is in the URL above.

What platform are you on? On Linux, doing:

from Tkinter import *
root = Tk()
root.configure(cursor='xterm red green')

doesn't produce an error. The colours seem to to be taken into account, but 
the xterm cursor isn't really a good example. I tried 
root.configure(cursor='gumby red green') and the result was what I expected.

Maybe there are some issues on Windows?
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From marklists at mceahern.com  Fri Sep  6 18:11:37 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 6 Sep 2002 17:11:37 -0500
Subject: Larry Wall's comment on python...
In-Reply-To: <200209062206.g86M6UVm092251@mail.hcsmail.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOCEAMDNAA.marklists@mceahern.com>

[mmaddox at hcsmail.com]
> Although this layout may be considered more elegant from a
> readability standpoint, it's the enforcement of the layout that is
> irritating to the more Libertarian of us. It's a mindset difference
> similar to that between Pascal and C++.

So does the fact that C++ forces you to use { } as block delimiters irritate
the Libertarian in you?

// m

-




From peter at engcorp.com  Mon Sep 16 22:00:59 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 16 Sep 2002 22:00:59 -0400
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de>	<pUdh9.9377$yt3.4604908@newssrv26.news.prodigy.com>	<mailman.1032184821.5180.python-list@python.org> 	<c%qh9.102794$ub2.2120685@news1.tin.it> <mailman.1032212180.11973.python-list@python.org>
Message-ID: <3d868ccc$1@news.sentex.net>

Cliff Wells wrote:
> On Mon, 2002-09-16 at 13:17, BT wrote:
> 
>>Ok, let's say I make a software in two years and I market it on internet.
>>The software is good and sell some copy.
>>You find the software, like it, change the name, add some feature and resell
>>it. My time: 2 years, yours: 1 month.
> 
> And that would be illegal, hence the suggestion to *copyright* your
> work.

Just a note about the copyright issue, as I agree with most of the
comments about the futility of spending much, and often any, effort
obscuring the source.  (There are disassemblers for almost any
processor on the planet, and most of the value is not in the code
anyway, but in the more concrete representations of the requirements
behind the program, such as the user interface designs and the
overall feature set.)

Copyright is actually a passive thing now, as opposed to active.
That is, you don't have "to copyright your work", because unless
you explicitly disclaim the copyright, you automatically get it.

As I recall, this was not always the case in at least the U.S.,
but has been the case for a number of years now.  In other words,
if you don't write something like "placed in the public domain"
then you still have copyright on it and could take steps in
courts of law to recover lost income if someone else attempts
to profit from your work without permission.

There *are* steps you can take that are considered by some to
increase the strength of the copyright (registering it officially
and so forth), but others believe this is mostly a waste of time.

The usual disclaimers about lawyers and advice on Usenet apply...

-Peter



From kowald at molgen.mpg.de  Tue Sep  3 05:22:26 2002
From: kowald at molgen.mpg.de (Axel Kowald)
Date: Tue, 03 Sep 2002 11:22:26 +0200
Subject: string reading bug in ActiveState Python 2.2 ?
References: <3D74757D.1010405@molgen.mpg.de>
Message-ID: <3D747F52.4080101@molgen.mpg.de>

Of course, binary mode !!
How stupid of me.

Many thanks to all,

                          Axel



From ark at research.att.com  Tue Sep  3 23:21:36 2002
From: ark at research.att.com (Andrew Koenig)
Date: Wed, 4 Sep 2002 03:21:36 GMT
Subject: Memory fault - core dumped   while compiling 2.2.1 on SCO
References: <3D757EFA.9050401@nospam.yahoo.com>
Message-ID: <yu99sn0qqw1r.fsf@europa.research.att.com>

Anton> Any ideas?

Looks like the same symptom I had.

If you are using binutils 2.13, try going back to 2.12.1 and
see if that fixes the problem.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From aleax at aleax.it  Thu Sep 26 08:31:10 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 26 Sep 2002 12:31:10 GMT
Subject: How to send Email using Python?
References: <amu8qh$93n60$1@ID-49758.news.dfncis.de>
Message-ID: <i6Dk9.170431$pX1.6070916@news2.tin.it>

Ken wrote:

> Hi all, I like to know how to send a HTML featured Email from Python? I
> know how to do a text based version using SMTPLib, but not sure how to do
> it with a HTML featured Email. Can someone help me? Can someone provide me
> with sample code?

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/67083

or, of course, the edited and enhanced version thereof we prepared for
O'Reilly's paper-edition of the Python Cookbook (I know you can download
all of the printed coobook's recipes' code from O'Reilly's site, but don't 
have the exact URL for that at hand).


Alex



From brendon.co at asia.bigfoot.com  Thu Sep 19 05:07:48 2002
From: brendon.co at asia.bigfoot.com (dd)
Date: 19 Sep 2002 02:07:48 -0700
Subject: execfile syntax & zope error
Message-ID: <a65110ed.0209190107.54b4ae46@posting.google.com>

Hi,

Im having problem using execfile.

Using execfile('filename.exe') on zope for web purpose, generate an error:

An error has occurred. The Site Admistrator has been notified. 
global name 'execfile' is not defined and Traceback (innermost last)

should I import something on python like:

import sys
execfile('filename.exe')


but this gives me an error, refer above.....


Any help...thanks

dd


From aleax at aleax.it  Wed Sep 18 12:21:19 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 18 Sep 2002 16:21:19 GMT
Subject: Copy with __slots__
References: <mailman.1032358481.7625.python-list@python.org>
Message-ID: <3K1i9.110352$ub2.2335897@news1.tin.it>

Griebel, Peer wrote:
        ...
> I'm using Python 2.2 (Cygwin). There I don't get the exception you
> mentioned.
>  
>> This is with Python 2.2.1 -- as you see, it doesn't get anywhere

I suggest you upgrade to Python 2.2.1 -- it has enhancements such
as this one, which help a lot.

>> class C2(C1):
>>     __slots__ = "s2"
>>     def __copy__(self):
>>         x = self.__class__()
>>         x.s1 = self.s1
>>         x.s2 = self.s2
>>         return x
> 
> Yes I know this works. But this hat two drawbacks (at least in my eyes):
> 
> 1. Formerly I used the approach of copy: I simply updated the dict of a
> newly created empty object. This is fast and works well - but not with the
> new classes. I think your approach runs much more slowly.

It's of course not a matter of new vs old classes -- new classes
would let you use the dictionary-update method just as well... as
long as you didn't use __slots__.

Regarding performance -- *measure* things, man!  Intuition about
what runs slowly, what rapidly, is notoriously misleading.  E.g.:

import time, copy

class noslots(object):
    pass

class withslots(object):
    __slots__ = 's1', 's2'
    def __copy__(self):
        result = self.__class__.__new__(self.__class__)
        result.s1 = self.s1
        result.s2 = self.s2
        return result

x1 = noslots()
x1.s1 = 23
x1.s2 = 'bebop'

x2 = withslots()
x2.s1 = 23
x2.s2 = 'bebop'

repetitions = [None] * 10000
def copyalot(x):
    start = time.clock()
    for i in repetitions: copy.copy(x)
    stend = time.clock()
    return '%.2f %s' % (stend-start, x.__class__.__name__)

for x in x1, x2:
    print copyalot(x)

[alex at lancelot ~]$ python -O slorno.py
1.04 noslots
0.29 withslots

"much more slowly", hmm...?  Looks over 3 times faster to *ME*...


> 2. The approach is error prone. When simply updating a dict I'm not going
> to miss a newly added attribute. But using your method I have to remember
> the method whenever I add an attribute.

Yes, that IS part of the price you pay for getting the optimization
that __slots__ affords: you do have to know exactly what you're doing.
Thus, __slots__ should be used when the program is working, in the
optimization phase, for classes such that saving a __dict__ per
instance IS an important optimization, only (classes of which MANY
instances exist, basically).

At the price of a small slow-down, if the __slots__ are all defined
in the current class, you can safeguard against this, e.g.:

class withslots(object):
    __slots__ = 's1', 's2'
    def __copy__(self):
        result = self.__class__.__new__(self.__class__)
        for s in self.__slots__:
            setattr(result, s, getattr(self, s))
        return result

This slows my timing test down from 0.29 to about 0.36.  If you
must also consider _inherited_ slots, e.g;

class withslots_base(object):
    __slots__ = ('s1',)

class yetanother(withslots_base):
    pass

class withslots(yetanother):
    __slots__ = ('s2',)
    def __copy__(self):
        result = self.__class__.__new__(self.__class__)
        for base in self.__class__.__mro__:
            for s in getattr(base, '__slots__', []):
                setattr(result, s, getattr(self, s))
        return result

the slowdown is even more severe -- to 0.62.  Still nothing
like the 1.04 you pay for NOT having the slots at all, of
course, but it's starting to undermine performance improvement,
the only real rationale for having __slots__ at all.  Still,
you pays your money, and you makes your choices.


>> The way to cover copy.copy, copy.deepcopy AND pickle is to have the
>> class supply a special method __reduce__.  __reduce__ can return a
>> pair where the first item is callable, the second item is the tuple
>> to pass as arguments to the first.  Thus, for example:
> 
> Here once again I have the two disadvantages mentioned above.

You mis-count, I think.  You have only one disadvantage: you
have to write more code and decide how fail-safe you want it
to be wrt changes to __slots__ in the inheritance hierarchy,
depending on whether you need just a LITTLE optimization wrt
not using __slots__, or really crave for all you can get.

The advantage is the optimization.  The disadvantage is that
you have to write more code, with some care, to get the
optimization, particularly if you also want flexibility (more
optimization, of course, is gained by foregoing flexibility
and nailing things down).  That's how MOST optimization
activities work, in my experience.  Is yours so different?

> Yes this would work. But I'm looking for speed (as you already abserved).
> And my constructor is a little more heavy weight. So it is much faster to
> use the old mechanism (update dict) than always calling the constructor.

So, write that little extra factory function -- that's hardly
the problem!  The ability to reuse the constructor is just a
handy work-saver when feasible, as it often is.

> Thanks Alex!

You're welcome.

> I'm still in hope to find a fast and secure solution...

I think you need to choose your preferred tradeoff point between
'fast' and 'handy'.  It's not a matter of security -- none of
these solutions run higher security risks than others -- it's
that handiness, flexibility and other admirable qualities 
(including ease of changing the software later) ARE more often
than not things you can sacrifice if and when you really need
to increase performance.


Alex



From s3v3n3v3s73 at yahoo.com  Mon Sep 16 02:17:51 2002
From: s3v3n3v3s73 at yahoo.com (7)
Date: 15 Sep 2002 23:17:51 -0700
Subject: basic question
Message-ID: <ad3fe3.0209152217.1d7d4468@posting.google.com>

I am using Pythonwin. How can I always make sure when I make a change
to a script and save it, that when I go to run it, that I run the
changed version. I know there is a very basic answer, this has been
getting to me.

Thanks in advance.


From whisper at oz.net  Fri Sep  6 16:56:07 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 6 Sep 2002 13:56:07 -0700
Subject: Interrupting Python
In-Reply-To: <yz5e9.4237$rU2.145619@news4.srv.hcvlny.cv.net>
Message-ID: <GCEDKONBLEFPPADDJCOEKEHOENAA.whisper@oz.net>

> I have a script that can run, accessing network resources, for
> several days.
> Since the script does not normally need keyed input, exception processing
> does not raise the keyboard exception until after the program
> ends normally.
> I would like to be able to interrupt if from the keyboard, but have not
> learned the trick.  How can I do this?
>
> --
> Bob Easton

I have had the experience of Python progs not responding to ctrl-c or ctrl-d
on Windows, but I found that ctrl-break works every time.

Dave LeBlanc




From lbrannma at cablespeed.com  Sun Sep 22 15:32:47 2002
From: lbrannma at cablespeed.com (Lance)
Date: Sun, 22 Sep 2002 12:32:47 -0700
Subject: send to Mitya
Message-ID: <uos6ip92qmib4e@corp.supernews.com>

9.2 Python Scopes and Name Spaces
Before introducing classes, I first have to tell you something about
Python's scope rules. Class definitions play some neat tricks with
namespaces, and you need to know how scopes and namespaces work to fully
understand what's going on. Incidentally, knowledge about this subject is
useful for any advanced Python programmer.

Let's begin with some definitions.

A namespace is a mapping from names to objects. Most namespaces are
currently implemented as Python dictionaries, but that's normally not
noticeable in any way (except for performance), and it may change in the
future. Examples of namespaces are: the set of built-in names (functions
such as abs(), and built-in exception names); the global names in a module;
and the local names in a function invocation. In a sense the set of
attributes of an object also form a namespace. The important thing to know
about namespaces is that there is absolutely no relation between names in
different namespaces; for instance, two different modules may both define a
function ``maximize'' without confusion -- users of the modules must prefix
it with the module name.

By the way, I use the word attribute for any name following a dot -- for
example, in the expression z.real, real is an attribute of the object z.
Strictly speaking, references to names in modules are attribute references:
in the expression modname.funcname, modname is a module object and funcname
is an attribute of it. In this case there happens to be a straightforward
mapping between the module's attributes and the global names defined in the
module: they share the same namespace! 9.1

Attributes may be read-only or writable. In the latter case, assignment to
attributes is possible. Module attributes are writable: you can write
"modname.the_answer = 42". Writable attributes may also be deleted with the
del statement. For example, "del modname.the_answer" will remove the
attribute the_answer from the object named by modname.

Name spaces are created at different moments and have different lifetimes.
The namespace containing the built-in names is created when the Python
interpreter starts up, and is never deleted. The global namespace for a
module is created when the module definition is read in; normally, module
namespaces also last until the interpreter quits. The statements executed by
the top-level invocation of the interpreter, either read from a script file
or interactively, are considered part of a module called __main__, so they
have their own global namespace. (The built-in names actually also live in a
module; this is called __builtin__.)

The local namespace for a function is created when the function is called,
and deleted when the function returns or raises an exception that is not
handled within the function. (Actually, forgetting would be a better way to
describe what actually happens.) Of course, recursive invocations each have
their own local namespace.

A scope is a textual region of a Python program where a namespace is
directly accessible. ``Directly accessible'' here means that an unqualified
reference to a name attempts to find the name in the namespace.

Although scopes are determined statically, they are used dynamically. At any
time during execution, exactly three nested scopes are in use (exactly three
namespaces are directly accessible): the innermost scope, which is searched
first, contains the local names, the middle scope, searched next, contains
the current module's global names, and the outermost scope (searched last)
is the namespace containing built-in names.

Usually, the local scope references the local names of the (textually)
current function. Outside of functions, the local scope references the same
namespace as the global scope: the module's namespace. Class definitions
place yet another namespace in the local scope.

It is important to realize that scopes are determined textually: the global
scope of a function defined in a module is that module's namespace, no
matter from where or by what alias the function is called. On the other
hand, the actual search for names is done dynamically, at run time --
however, the language definition is evolving towards static name resolution,
at ``compile'' time, so don't rely on dynamic name resolution! (In fact,
local variables are already determined statically.)

A special quirk of Python is that assignments always go into the innermost
scope. Assignments do not copy data -- they just bind names to objects. The
same is true for deletions: the statement "del x" removes the binding of x
from the namespace referenced by the local scope. In fact, all operations
that introduce new names use the local scope: in particular, import
statements and function definitions bind the module or function name in the
local scope. (The global statement can be used to indicate that particular
variables live in the global scope.)



9.3 A First Look at Classes
Classes introduce a little bit of new syntax, three new object types, and
some new semantics.



9.3.1 Class Definition Syntax
The simplest form of class definition looks like this:


class ClassName:
    <statement-1>
    .
    .
    .
    <statement-N>
Class definitions, like function definitions (def statements) must be
executed before they have any effect. (You could conceivably place a class
definition in a branch of an if statement, or inside a function.)

In practice, the statements inside a class definition will usually be
function definitions, but other statements are allowed, and sometimes
useful -- we'll come back to this later. The function definitions inside a
class normally have a peculiar form of argument list, dictated by the
calling conventions for methods -- again, this is explained later.

When a class definition is entered, a new namespace is created, and used as
the local scope -- thus, all assignments to local variables go into this new
namespace. In particular, function definitions bind the name of the new
function here.

When a class definition is left normally (via the end), a class object is
created. This is basically a wrapper around the contents of the namespace
created by the class definition; we'll learn more about class objects in the
next section. The original local scope (the one in effect just before the
class definitions was entered) is reinstated, and the class object is bound
here to the class name given in the class definition header (ClassName in
the example).







From ianholmes01 at lycos.com  Fri Sep  6 08:10:00 2002
From: ianholmes01 at lycos.com (Ian Holmes)
Date: 6 Sep 2002 05:10:00 -0700
Subject: Python "compiler" is  too slow for processing large data files???
References: <mailman.1030614584.8635.python-list@python.org> <Fpub9.22728$Ic7.1819173@news2.west.cox.net>
Message-ID: <b31eb63a.0209060410.7257ee8d@posting.google.com>

"Ron Horn" <rchii at lycos.com> wrote in message news:<Fpub9.22728$Ic7.1819173 at news2.west.cox.net>...
> test on my Windows 2000 machine, 1.5GHz, Pentium 4:
> 
> C:\test>python compiletest.py 8000 10 comp.py
> C:\test>timer r "python -c \"import comp\""
> 8000
> Time passed: 00:00:09.39
> 
> C:\test>python compiletest.py 8000 20 comp2.py
> C:\test>timer r "python -c \"import comp2\""
> 8000
> Time passed: 00:00:26.47

What else is your p4 running? - it seems to running a tad slow

on my celeron 500Mhz win2k 384meg py 2.2.1
runs at 15-17s for 8000x20 and python never got above 2meg (though cpu
was at 80-98% through out)

Also I imported psyco and proxied the main method which reduced the
time to ~6s

see below. psyco is available from psyco.sourceforge.net - I imagine
that if the other functions are proxied to then it may speed up some
more.


#modified with psyco and in script timing
# gen.py
# create a script to generate modules that will stress
# out the python compiler a bit?

import sys, random, time, psyco

def GenLine(numArgs, outfile):
    outfile.write("    (")
    for i in range(numArgs):
        outfile.write("%d, " % random.randint(0,1000))
    outfile.write("),\n")

def Generate(numLines, numArgs, outfile):
    outfile.write("list1 = [\n")
    for i in range(numLines):
        GenLine(numArgs, outfile)
    outfile.write("]\n")
    outfile.write("print len(list1)\n")


def main():
    outfile = None
    try:
        if len(sys.argv) != 4:
            print "Usage: gen <lines> <args> <outfile>"
            return

        numLines = int(sys.argv[1])
        numArgs = int(sys.argv[2])
        outfile = open(sys.argv[3], 'wt')

        Generate(numLines, numArgs, outfile)

    finally:
        if outfile: outfile.close()

if __name__ == "__main__":
    START_TIME = time.time()
    mainx = psyco.proxy(main)
    mainx()
    print "end",time.time() - START_TIME


From oren-py-l at hishome.net  Wed Sep 11 01:14:00 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Wed, 11 Sep 2002 01:14:00 -0400
Subject: Reset a generator function?
In-Reply-To: <%Lrf9.30107$w51.7961593@news2.news.adelphia.net>
References: <%Lrf9.30107$w51.7961593@news2.news.adelphia.net>
Message-ID: <20020911051359.GA664@hishome.net>

On Tue, Sep 10, 2002 at 07:31:07PM +0000, Robert Oschler wrote:
> How can I reset a generator function so that it begins yielding values as
> it did when first called?

You need to call the function again with the same arguments. Assuming that
its result depends only on the arguments and that it does not modify them
it should regenerate the same sequence.

Look at the regen function in http://tothink.com/python/dataflow/xlazy.py
It wraps a generator function with a wrapper class that behaves just like
the generator function except that its return value is an iterable object
instead of an iterator. Each time iter() is called on this object it calls 
the generator function again with the original arguments.

example:

>>> from __future__ import generators
>>> def f(n):
...   for i in xrange(n):
...     yield i*10
...
>>> x=f(10)
>>> list(x)
[0, 10, 20, 30, 40, 50, 60, 70, 80, 90]
>>> list(x)
[]
>>>
>>> from xlazy import *
>>> f = regen(f)
>>>
>>> x = f(10)
>>> list(x)
[0, 10, 20, 30, 40, 50, 60, 70, 80, 90]
>>> list(x)
[0, 10, 20, 30, 40, 50, 60, 70, 80, 90]
>>>

	Oren



From marklists at mceahern.com  Tue Sep 17 23:31:36 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 17 Sep 2002 22:31:36 -0500
Subject: Splitting a list of strings
In-Reply-To: <am8pvo$dcn$1@driftwood.ccs.carleton.ca>
Message-ID: <JHEOKEOOLIGLDHCMAHMOAEGNDPAA.marklists@mceahern.com>

> Okay. Here is some representative sample data:

This does everything except ignore the last attribute--easy enough to add
either here or have the caller do it.

def getDataAndAttributes2(filename):
    attrib_prefix = "@"
    comment_prefix = "%"
    attributes = []
    data = []
    f = file(filename)
    lines = f.readlines()
    f.close()
    for line in lines:
        # Ignore empty lines and comments.
        line = line.strip()
        if not line or line.startswith(comment_prefix):
            continue
        # Put attributes and data into separate lists.
        if line.startswith(attrib_prefix):
            attributes.append(line)
        else:
            data.append(line)
    return data, attributes

// m

-




From kseehof at neuralintegrator.com  Tue Sep  3 16:27:53 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Tue, 3 Sep 2002 13:27:53 -0700
Subject: efficient list merging
In-Reply-To: <oqu1l66fu7.fsf@titan.progiciels-bpi.ca>
Message-ID: <EFEFIHAINIINNELLHFAIKEFCDFAA.kseehof@neuralintegrator.com>

> [Peter Saffrey]
>
> > I have two lists of lists of strings which I would like to merge
> > efficiently without repeats.  [...] Or is there a better way?
>
> I am not sure about the relative speed of everything, but I would be
> tempted to try:
>
>    dict(zip(list1 + list2, [None] * (len(list1) + len(list2)))).keys()
>
> It seems to work:
>
> >>> list1 = list('abcdebxyz')
> >>> list2 = list('pqqr')
> >>> list1
> ['a', 'b', 'c', 'd', 'e', 'b', 'x', 'y', 'z']
> >>> list2
> ['p', 'q', 'q', 'r']
> >>> dict(zip(list1 + list2, [None] * (len(list1) + len(list2)))).keys()
> ['a', 'c', 'b', 'e', 'd', 'q', 'p', 'r', 'y', 'x', 'z']
>
> --
> Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard

This is a bit faster, but assumes that the first list does not initially
contain duplicates.  It eliminates duplicates that would be introduced by
merging items from list2.  Also, this maintains the original list order
(i.e. list1+list2 without adding duplicates).

def merge(list1,list2):
    d1 = dict(zip(list1,list1))
    return list1+[s for s in list2 if s not in d1]

- Ken Seehof




From anthony_barker at hotmail.com  Mon Sep 16 14:34:55 2002
From: anthony_barker at hotmail.com (Anthony_Barker)
Date: 16 Sep 2002 11:34:55 -0700
Subject: FYI: Oreilly Programming Python book $6
Message-ID: <899f842.0209161034.1e587611@posting.google.com>

For those on a budget - I just picked up Programming Python (note -
1st Edition so parts of it are dated) for $5.99 at
http://www.bookcloseouts.com. They have 37 more copies. They also have
the pocket reference listed for $3.

I don't have anything to do with that company. 

The order arrived fairly promptly.

Anthony
http://xminc.com/anthony/


From see_reply_address at something.invalid  Mon Sep  2 21:18:52 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Tue, 03 Sep 2002 13:18:52 +1200
Subject: Would you like a Snobol pattern matcher? (Re: Making regex suck less)
References: <3d725881.345921@news.t-online.de> <un7n06qrl1gm93@news.supernews.com>
Message-ID: <3D740DFC.1040009@something.invalid>

John Roth wrote:

> http://www.perl.com/pub/a/2002/06/04/apo5.html
> 
> http://www.perl.com/pub/a/2002/08/22/exegesis5.html
 

I just had a brief look at this, and the underlying
ideas seem to be a lot like the way Snobol patterns
work.

Maybe it's time for me to resurrect the Snobol-style
pattern matching module that I started on a while
back and never got around to releasing.

Would anyone be interested in this? Its
syntax is similar to that of Plex REs, except that
the primitives are Snobol-like, and it uses a
backtracking matching algorithm that's much more
powerful than a DFA (you can write entire parsers
in it, for example).

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From klaus.hoeppner at bigfoot.de  Mon Sep 30 08:26:47 2002
From: klaus.hoeppner at bigfoot.de (Klaus Hoeppner)
Date: Mon, 30 Sep 2002 14:26:47 +0200
Subject: Find out link location
Message-ID: <7uf9na-vd9.ln@lbdb01.gsi.de>

Hi,

with the os.path module I can use the islink function to check, wether
a file is symbolic link. But how can I find out, where the link points
to. E.g. if I have readme.txt --> ../packages/mypackage/readme.txt how
do I get the "../packages/mypackage/readme.txt" with python?

Regards
Klaus


From peter at engcorp.com  Thu Sep 19 20:08:23 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 19 Sep 2002 20:08:23 -0400
Subject: Minimal debug/rep functionality
References: <mailman.1032479306.27647.python-list@python.org>
Message-ID: <3d8a66ec$1@news.sentex.net>

Delaney, Timothy wrote:
>>From: Peter Hansen [mailto:peter at engcorp.com]
> 
>>I'm looking forward to seeing your tool though, Timothy. :)
> 
> I'd love it to eventually become part of the core - I *expect* I'll be able
> to release it (but of course, it's up to my employer ...). It's doing
> everything I want now except for determining exactly which zero-count lines
> can be safely ignored (comments, multiline expressions). Oh - and I haven't
> got an actual "trace as you go" mode (which is the topic of this thread ;)
> but that will be easy to add. And I haven't got any path coverage analysis
> yet - the data is captured, but no analysis is performed. And it currently
> does a relatively primitive text-only coverage dump. And ... 

Heh. :-)

I'm a commercial Python user as well, so I can imagine the kinds of
things your employer think about whether to release or not.  We have
some items which we simply wouldn't consider releasing, but there
are a number of tools which we hope to release at some point.  They
have similar limitations to yours (well, okay, they're actually much
more primitive even than that :-) yet with a little more work they
would be useful to others.

More helpfully in convincing my own employer that we should release
back to the community is pointing out how with our limited resources
these tools will always remain primitive and simplistic.  If we can
do a little more work *and* release them, we stand a chance of seeing
them improved by others, and taking advantage of the improvements
ourselves.  See, it's easy to convince those conservative, selfish
minds that they can profit by giving things away. :-)

Well, if you can't release it (I guess I just assume from all your
mentions of it that it was going to be open), we'll probably be
building some of the same types of things ourselves.  It would
be a real shame if we both (and others) went to all that effort
in parallel without sharing. :(

-Peter



From tim.one at comcast.net  Sat Sep 21 01:25:59 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 21 Sep 2002 01:25:59 -0400
Subject: e vs exp()?
In-Reply-To: <iiOi9.173334$5r1.7100346@bin5.nnrp.aus1.giganews.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEKKBFAB.tim.one@comcast.net>

[Tim]
>> ... The implementation of a decent x**y function is one
>> of the> most difficult tasks in a platform's math library....
>> I've written such a beast without benefit of an extended hardware
>> precision; I wouldn't want to do it again <wink>.

[Terry Reedy]
> What was your basic approach?  Power series?  Rational function?
> Continued fraction? ???

exp(log(x)*y) -- really <wink>.  It was a combination of table-driven and
economized power series approaches, like the Tang algorithm I mentioned last
time, carefully faking just enough extra precision to guarantee < 1 ulp
worst-case error in the end.

> http://functions.wolfram.com/ElementaryFunctions/Power/
> gives 343 formulas (many only relevant to complex domain) but doesn't
> seem to indicate which Mathematica actually uses (trade secret I
> presume).

Those kinds of references won't do you any good at this level:
good-to-the-last-bit library functions are a whole different game.  Check
out Peter Tang's celebrated paper on the implementation of exp (sorry, it's
not free):

    http://portal.acm.org/citation.cfm?doid=63522.214389

That's still much the state of modern art for native-precision
transcendentals, and, with enough pain (in x**y's case, transcendant pain),
all can be approached in a similar way.

Dynamic-precision calculations are yet another entirely different kind of
game.  David Bailey's MPFUN is freely available, and so is Dave Gillespie's
Emacs calc mode; you can get as much precision out of those as you want.
Jurjen N.E. Bos's real.py for Python is yet another entirely different game:
an implementation of the constructive reals, where the result of an
arbitrarily complex chain of calculations (not just one!) can be obtained to
any desired accuracy.

As you climb up this chain of wilder ambitions, runtime increases
accordingly, of course.  For best speed, exp(x) can be approximated by 1
<wink>.




From aleax at aleax.it  Sat Sep 14 15:51:46 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 14 Sep 2002 19:51:46 GMT
Subject: Skipping Parts of a For Loop
References: <pan.2002.09.14.12.23.44.103401.723@thomas-guettler.de>
Message-ID: <mrMg9.95802$ub2.1909231@news1.tin.it>

Thomas Guettler wrote:

> Hi!
> 
> Is there a way to skip some parts in a for loop?
> 
> 
> for i in range(3):
> print "i before", i
> i=2
> print "i after", i
> 
> #Results in:
> 
> i before 0
> i after 2
> i before 1
> i after 2
> i before 2
> i after 2
> 
> Is there a way to skip some parts?
> 
> "continue" skipps one element. But how can
> I skip several?

In Python 2.2, a for loop like:
    for x in something:
        "any statements here"
is equivalent to:
    _aux = iter.something()
    while True:
        try: x = _aux.next()
        except StopIteration: break
        "any statements here"

You can take advantage of this as long as you
bind the iterator to a name, exploiting also the fact
that iter(iter(x)) is iter(x) [i.e., iter is idempotent
on iterators]:

aux = iter(range(3))
for i in aux:
  print "i before", i
  for i in range(i,3): aux.next()
  print "i after", i

Each calls to the next method of the iterator object
"skips" a whole coming leg of the for loop.


Alex



From NWMoriarty at lbl.gov  Thu Sep 19 12:58:47 2002
From: NWMoriarty at lbl.gov (Nigel W. Moriarty)
Date: Thu, 19 Sep 2002 09:58:47 -0700
Subject: thread error
Message-ID: <3D8A0247.B87A7CD@lbl.gov>

Folks

I have a app using a threading TCP server which ocassionally spits out the following error when it exits.

Unhandled exception in thread:
Traceback (most recent call last):
  File "/usr/local_cci/Python-2.2.1t/lib/python2.2/threading.py", line 415, in __bootstrap
    s = _StringIO()
  File "/usr/local_cci/Python-2.2.1t/lib/python2.2/StringIO.py", line 42, in __init__
    if type(buf) not in types.StringTypes:
AttributeError: 'NoneType' object has no attribute 'StringTypes'

A look at the code shows that the object types is the standard module types.  Any ideas?

Nigel

--
Nigel W. Moriarty
Building 4R0230
Physical Biosciences Division
Lawrence Berkeley National Laboratory
Berkeley, CA 94720-8235
Phone : 510-486-5709
Fax   : 510-486-5909
Email : NWMoriarty at LBL.gov
Web   : CCI.LBL.gov




From maxm at mxm.dk  Sat Sep 14 04:59:07 2002
From: maxm at mxm.dk (Max M)
Date: Sat, 14 Sep 2002 10:59:07 +0200
Subject: Calling Python from C# - please help
References: <6de16ac9.0209130508.5d1b56c6@posting.google.com> <j4elbyt3yw.fsf@informatik.hu-berlin.de> <ur8fyxau0.fsf@online.no> <0iCg9.2350$Ee4.6516@news-server.bigpond.net.au>
Message-ID: <3D82FA5B.3050003@mxm.dk>

Mark Hammond wrote:

> I have no examples of this, but I should.  I'm co-authoring a .NET book 
> which is nearly out :) - 


That sounds great. I enjoyed the win32 book, and still use it as a 
reference.

Are you by any chance still developing Pyhton .NET, or has it completely 
stopped?

regards Max M



From =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?=  Fri Sep 13 10:37:14 2002
From: =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?= (=?iso-8859-5?Q?=B4=D0=DC=F8=D0=DD_=B3=2E?=)
Date: Fri, 13 Sep 2002 16:37:14 +0200
Subject: calling Oracle functions
Message-ID: <3d81f811$1@news.mt.net.mk>

Does anyone know, how can I call Oracle stored functions from Python.
I'm using cx_Oracle.

The function takes three parameters and returns one result.

I tried:

	params = ['param1', 'param2', 'param3']
	v_Cursor.callproc("SCH.PKG.MyFunction", params)

and I got the error:

	cx_Oracle.DatabaseError: ORA-06550: line 1, column 8:
	PLS-00221: 'MyFunction' is not a procedure or is undefined

But I can execute the function from sqlplus

	declare result number;
    
	BEGIN
	result := SCH.PKG.MyFunction('param1', 'param2', 'param3');
	END;
		  
Any hints?


-- 
??????

            Yes, I've heard of "decaf." What's your point?



From bbin__ at hotmail.com  Tue Sep 24 09:36:21 2002
From: bbin__ at hotmail.com (Benjamin Binford)
Date: Tue, 24 Sep 2002 09:36:21 -0400
Subject: Three dumb questions (ordered by dumbness descending)
References: <amnvit$7kqdt$1@ID-77524.news.dfncis.de> <pan.2002.09.23.20.43.59.461504.1216@io.com>
Message-ID: <pan.2002.09.24.13.36.21.366582@hotmail.com>

On Tue, 24 Sep 2002 01:45:29 +0000, Steven Feil wrote:

> On Mon, 23 Sep 2002 16:03:21 -0500, Thorsten Kampe wrote:
> 
>> Okay, here they are:
> 
> Ok, I'll take a stab at the first two questions. I do think that a
> little more clarity on your part could make it a bit easer.
> 
> The first thing to remember is that Python is a dynamically typed
> language. It is typed and each object has a type, It would be a little
> easer if you would disclose the type for each object you are referring
> to.
> 
> 
>> 1. Why is 'zip(zip(x)) != x' (should be the same because it's the
>> transposed)
> 
> Humm, what is is zip(), a function in the standard Python library or is
> it something you or someone else made up? I did a quick search in the
> Python docs and the only zip I found referred to zip compression. If
> this is something from the standard Python libraries please state the
> library it comes from.
> 
> If zip is some general function, then in general, zip(zip(x)) is not
> equal to x. This comes from fundamental abstract algebra. In trig
> sin(sin(x)) is not x. There do exist functions who are there own inverse
> and if zip is one of them then zip(zip(x)) is equal to x.
> 
> 

Zip is a builtin function in Python versions >= 2
zip(seq1, ...)
This function returns a list of tuples, where the i-th tuple contains the
i-th element from each of the argument sequences. At least one sequence is
required, otherwise a TypeError is raised. The returned list is truncated
in length to the length of the shortest argument sequence. When there are
multiple argument sequences which are all of the same length, zip() is
similar to map() with an initial argument of None. With a single sequence
argument, it returns a list of 1-tuples. 


>From the definition you can see that zip is not the transpose of itself.
zip(zip((1,2,3))) = [((1,),), ((2,),), ((3,),)].
If the original sequence is represented by a list like (1,2,3), the result
of a double zip is a doubly nested set of lists or a tree. Thus the result
could also be represented as:

		1
 	       /
	list  
       / 
      /  	2
     /       /
	list  
      /
root
      \		3
       \      /
	list  



From oren-py-l at hishome.net  Wed Sep  4 14:57:04 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Wed, 4 Sep 2002 14:57:04 -0400
Subject: Daemonize?
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F06233022B66EC@mail.hebmex.com>
References: <F7DB8D13DB61D511B6FF00B0D0F06233022B66EC@mail.hebmex.com>
Message-ID: <20020904185704.GA28154@hishome.net>

On Wed, Sep 04, 2002 at 08:36:57AM -0500, sismex01 at hebmex.com wrote:
> Sorry to ask here, but I don't have any "unixy" python
> documentation at hand.
> 
> Normally, I put an app into "daemon" mode, by first
> forking, then reopening stdin/out/err to /dev/null,
> and then forking again. Works good enough for my
> purposes.
> 
> But, am I right? Is this the correct way to do this?
> Also, I can't find any "reopen" in python to apply
> to sys.stdin/out/err.

No need to do anything to the sys.std* objects - you can manipulate
the underlying file handles directly with os.close, os.dup2, etc. 

	Oren




From marklists at mceahern.com  Thu Sep 26 10:10:56 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 26 Sep 2002 09:10:56 -0500
Subject: Tabs vs. Spaces, VIM vs. EMACS (Was Re: Larry Wall's comment on python...)
In-Reply-To: <006101c2651b$0df0c340$0101010a@local>
Message-ID: <JHEOKEOOLIGLDHCMAHMOOELPEBAA.marklists@mceahern.com>

[Chris Gonnerman]
> Surely emacs can do as well as VIM...  (and I expect a rabid
> emacs user to fill me in posthaste).

Yup.  Here are the relevant portions of my .emacs file:

;; Turn tabs off by default.
(setq-default intend-tabs-mode nil)

;; Set tab width to 4 by default.
(setq-default tab-width 4)

;; Use Python mode for .py files.
(setq auto-mode-alist
      (cons '("\\.py$" . python-mode)
            auto-mode-alist))

(setq interpreter-mode-alist
      (cons '("python" . python-mode)
            interpreter-mode-alist))

;; Load Python mode.
(autoload 'python-mode "python-mode" "Python editing mode." t)

;; Set up initialization parameters for python mode:
(setq python-mode-hook
      '(lambda () (progn
                    (set-variable 'py-indent-offset 4)
                    (set-variable 'py-smart-indentation nil)
                    (set-variable 'indent-tabs-mode nil) )))

;; Define untabify-buffer.
(defun untabify-buffer()
  (interactive)
  "Untabify the current buffer.  Return nil.

  Must return nil, if added to write-file-hooks."
  (save-excursion
    (goto-char (point-min))
    (while (re-search-forward "[ \t]+$" nil t)
      (delete-region (match-beginning 0) (match-end 0)))
    (goto-char (point-min))
    (if (search-forward "\t" nil t)
        (untabify (1- (point)) (point-max))))
  nil)

;; Untabify buffers on write file.
(defun do-on-write-file()
  (untabify-buffer))

;; Add the hook.
(add-hook 'write-file-hooks 'do-on-write-file)

// m



From martin at v.loewis.de  Mon Sep  2 01:11:13 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 02 Sep 2002 07:11:13 +0200
Subject: Tix/Tk Module frustrations HELP!
References: <pan.2002.08.17.23.47.25.28221.3534@example.com> <m3it288zgs.fsf@mira.informatik.hu-berlin.de> <pan.2002.08.18.14.13.07.553600.18030@example.com>
Message-ID: <m3sn0t7z7y.fsf@mira.informatik.hu-berlin.de>

Jonathan S <python_hacker at example.com> writes:

> I tried rebuilding Tix from sources, yet it still has the same trouble. I
> guess I'll try rebuilding my X11 libs...

Can you please report the linker line used to link libtix.so? It
appears that it sometimes fails to link -lX11, which would also
explain this problem.

HTH,
Martin


From jon at csh.rit.edu  Mon Sep 30 15:32:24 2002
From: jon at csh.rit.edu (Jon Parise)
Date: Mon, 30 Sep 2002 15:32:24 -0400
Subject: About data structure
In-Reply-To: <3d98a5d8$0$496$7a628cd7@news.club-internet.fr>
References: <3d98a5d8$0$496$7a628cd7@news.club-internet.fr>
Message-ID: <20020930193223.GB19636@csh.rit.edu>

On Mon, Sep 30, 2002 at 09:29:52PM +0200, malan wrote:

> I need to build a data structure like this :
> 
> size	4 bytes ( example 0x0000000A )
> flag	1 byte  ( example 0x87 )
> trans   8 bytes ( example 'URTVMI1' )
> F5	3 bytes ( example 0x00125C )
> 
> How can I define that structure, format it and send it via a socket ?
> My problem is to code / decode the message to and from the server.
 
You want to use the 'struct' module:

    http://www.python.org/doc/current/lib/module-struct.html

-- 
Jon Parise (jon at csh.rit.edu)  .  Information Technology (2001)
http://www.csh.rit.edu/~jon/  :  Computer Science House Member



From keith_s at ntlworld.nospam.com  Thu Sep  5 04:45:29 2002
From: keith_s at ntlworld.nospam.com (Keith S.)
Date: Thu, 05 Sep 2002 09:45:29 +0100
Subject: linking embedded python
Message-ID: <3D7719A9.8050804@ntlworld.nospam.com>

Hi,

This might be a dumb question, but how does one link
a C++ application in Unix with embedded python?

Under Solaris 2.8 I found what appears to be the
link library /usr/local/lib/python2.1/config/libpython2.1.a
(the python installation came from the Sun freeware site).

However, when I try and link, I get errors like:

Undefined                       first referenced
  symbol                             in file
dlsym 
/usr/local/lib/python2.1/config/libpython2.1.a(dynload_shlib.o)  (symbol 
belongs to implicit dependency /usr/lib/libdl.so.1)
dlopen 
/usr/local/lib/python2.1/config/libpython2.1.a(dynload_shlib.o)  (symbol 
belongs to implicit dependency /usr/lib/libdl.so.1)
dlerror 
/usr/local/lib/python2.1/config/libpython2.1.a(dynload_shlib.o)  (symbol 
belongs to implicit dependency /usr/lib/libdl.so.1)

Anyone shed any light on what I'm doing wrong?

- Keith



From mfranklin1 at gatwick.westerngeco.slb.com  Wed Sep 25 04:53:06 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Wed, 25 Sep 2002 08:53:06 +0000
Subject: how to change font and its size in IDLE (in UNIX)?
Message-ID: <200209250853.06075.mfranklin1@gatwick.westerngeco.slb.com>

On Wednesday 25 Sep 2002 5:22 am, Kow K wrote:
> Hi all,
>
> I'm trying to figure out how to change the font and its size in IDLE.
> It's Courier in 10pt. I don't like Courier, and what's worse,
> 10pt is too small for me. I wish I could change it to Fixed 7x13.
>
> Any input will be appreciated.
>
> Kow

Take a look here (from http://groups.google.com/advanced_group_search)


http://groups.google.com/groups?q=font+size+idle&hl=en&lr=&ie=UTF-8&scoring=r
&selm=3a942c61_4%40corp.newsfeeds.com&rnum=1


Cheers
Martin

--
### Python Powered Signature
I started writting this email on
Wed Sep 25 08:51:33 2002




From Minews at ms6.de  Mon Sep  9 09:01:16 2002
From: Minews at ms6.de (Michael Schwarz)
Date: Mon, 09 Sep 2002 15:01:16 +0200
Subject: Tkinter: How to "break" class-event-bindings from within tag-bindings?
Message-ID: <3D7C9B9C.60603@ms6.de>

A "break" as the return value of a tag-event-binding for the 
Text-widget seems not to prevent the standard event-handling of the 
widget to occur.

This seems to be an error!

How can it be done another (elegant) way?

See the following example:

===

# How can tag-events "break" class-bindings?
from Tkinter import *

def Jump(e):
    # Sprung:
    targetFrom, targetTo = "3.5", "3.9"
    F.Text.tag_remove("sel", "0.0", "end")
    F.Text.tag_add("sel", targetFrom, targetTo)
    F.Text.mark_set("insert", targetFrom)
    return "break" # Doesn't do it !!!

F = Tk()
F.Text = Text(F) # , width=120, height=35, bg="white")
F.Text.pack() # (side="right", fill="both", expand=1)
F.Text.focus_set()
F.Text.insert(INSERT, """\
Link

Jump HERE
""")
F.Text.tag_bind("link", "<1>", Jump)
F.Text.tag_config("link", underline = "on")
F.Text.tag_add("link", "1.0", "1.4")

# Comment the following line and the jump won't do it anymore
# due to the standard class-bindings of the text-widget
F.Text.bind("<1>",lambda e: "break")

F.mainloop()



From richie at entrian.com  Mon Sep 16 12:33:27 2002
From: richie at entrian.com (Richie Hindle)
Date: 16 Sep 2002 09:33:27 -0700
Subject: possibly OT - delete char?
References: <3pch9.10253$Ee4.26056@news-server.bigpond.net.au>
Message-ID: <8a6ba1da.0209160833.154251f5@posting.google.com>

> - is there a character which deletes the previous printed character when
> printed?

Yes, \b, in a reasonably portable manner:

$ uname -sr
Linux 2.4.7-10smp
$ python
Python 2.1.3 (#1, May 15 2002, 14:56:49) 
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> print "12\b3"
13
>>> 

C:\>ver

Microsoft Windows XP [Version 5.1.2600]

C:\>python
Python 1.5.2 (#0, Feb  4 2002, 16:19:07) [MSC 32 bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> print "12\b3"
13
>>>

-- 
Richie Hindle
richie at entrian.com


From sross at connectmail.carleton.ca  Tue Sep 17 22:55:52 2002
From: sross at connectmail.carleton.ca (Sean Ross)
Date: 18 Sep 2002 02:55:52 GMT
Subject: Splitting a list of strings
References: <mailman.1032309381.3626.python-list@python.org>
Message-ID: <am8pvo$dcn$1@driftwood.ccs.carleton.ca>

Okay. Here is some representative sample data:

%  Relevant Information:
%     This data set includes descriptions of hypothetical samples
%     corresponding to 23 species of gilled mushrooms in the Agaricus and
%     Lepiota Family (pp. 500-525).
[snip]
%
%  Number of Instances: 8124
%
%  Number of Attributes: 22 (all nominally valued)
%
[snip]
@attribute 'cap-shape' { 'b', 'c', 'f', 'k', 's', 'x'}
@attribute 'cap-surface' { 'f', 'g', 's', 'y'}
@attribute 'cap-color' { 'b', 'c', 'e', 'g', 'n', 'p', 'r', 'u', 'w', 'y'}
[snip]
'x','s','n','t','p','f','c','n','k','e','e','s','s','w','w','p','w','o','p',
'k','s','u','p'
'x','s','y','t','a','f','c','b','k','e','c','s','s','w','w','p','w','o','p',
'n','n','g','e'
'b','s','w','t','l','f','c','b','n','e','c','s','s','w','w','p','w','o','p',
'n','n','m','e'
'x','y','w','t','p','f','c','n','n','e','e','s','s','w','w','p','w','o','p',
'k','s','u','p'
[snip]

So, essentially, there are three types of strings in the file: comments,
attribute descriptors, and data. I'm pulling out the attribute descriptor
and data strings and storing them in separate lists. I'll be doing stuff
with those latter; making a decision tree, for instance. Anyway, the code
I've written does what I need, but I'm thinking there may be a more
efficient method. Hence, the post.

As for str, I agree and the name has been changed. Thank you.

"Mark McEahern" <marklists at mceahern.com> wrote in message
news:mailman.1032309381.3626.python-list at python.org...
> > I need to read a list of strings from a file[...]
>
> For those of us who are lazy and lacking in imagination, why not share a
few
> lines of representative sample data?
>
> // m
>
> p.s.  Don't use str as the name of a variable, you'll clobber the builtin
> str.
>
> -
>
>




From emile at fenx.com  Wed Sep 18 10:39:51 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 18 Sep 2002 07:39:51 -0700
Subject: Splitting a list of strings
References: <mailman.1032309381.3626.python-list@python.org> <am8pvo$dcn$1@driftwood.ccs.carleton.ca> <w0Uh9.116763$5r1.4544705@bin5.nnrp.aus1.giganews.com> <am9tpv$9gp$1@driftwood.ccs.carleton.ca>
Message-ID: <ama34r$nga$1@main.gmane.org>

Sean Ross:
> However, I guess what I was looking for was a
> side-effect, or some such mechanism where, by extracting the nickels,
I get
> the two piles: one of nickels, one of pennies and dimes.

lines = ["sdflkj","@;lkjdsa","#lkjasdf",
        "sdflkj","@;lkjdsa","#lkjasdf",
        "sdflkj","@;lkjdsa","#lkjasdf",
        "sdflkj","@;lkjdsa","#lkjasdf",]
vals = [[],[],[]]
for line in lines:
    vals["@#".find(line[:1])+1].append(line)
data, attrs, comments = vals
print data, attrs, comments


Or, if you want to use an lc:

[ vals["@#".find(line[:1])+1].append(line) for line in lines]


--

Emile van Sebille
emile at fenx.com

---------






From member at dbforums.com  Sun Sep 15 06:50:13 2002
From: member at dbforums.com (hpyhpy)
Date: Sun, 15 Sep 2002 10:50:13 +0000
Subject: Anyone can tell me why?
References: <1815200.1031973626@dbforums.com>
Message-ID: <1817360.1032087013@dbforums.com>

But I try,change the filename as "example.so","_example.so",
"examplemodule.so",#_#,same error......have other advise?^_^

--
Posted via http://dbforums.com



From aleax at aleax.it  Fri Sep 20 10:18:19 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 14:18:19 GMT
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com> <amet8v$e50$1@scotsman.ed.ac.uk>
Message-ID: <L6Gi9.119475$ub2.2533804@news1.tin.it>

Eddie Corns wrote:

> If you want to learn "how it (computer languages) works" I suggest reading
> "Structure and Interpretation of Computer Languages" by Abelson and
> Sussman. This is about understanding the principles rather than how to do
> things in a particular language and is probably the best book on the
> subject.

I'll second this (well, third this, I guess, since P. Rubin recommended
the very same book independently in another post a while ago:-).  SICP
is the best book I know about the "deep principles" of programming.

I think you need substantial capacity and appetite for abstract thought
to really _enjoy_ SICP -- it doesn't spoon-feed you, nor does it use
baby-talk, lots of somewhat-repetitive examples, and other standard
didactical techniques.  But if the book's style does match your own
character and inclination, you're likely to fall in love with it.


Alex



From bokr at oz.net  Wed Sep  4 16:38:39 2002
From: bokr at oz.net (Bengt Richter)
Date: 4 Sep 2002 20:38:39 GMT
Subject: Regexp Over Multiple Lines
References: <565e316f.0209040740.36573b56@posting.google.com>
Message-ID: <al5r0f$p09$0@216.39.172.122>

On 4 Sep 2002 08:40:22 -0700, vetter at lincom-asg.com (Keith) wrote:

>In Python can I have a regular expression that spans multiple lines...
Yes, see X flag on

    http://www.python.org/doc/current/lib/node99.html

and read about the re module at

    http://www.python.org/doc/current/lib/module-re.html

wherefrom you can find the syntax at

    http://www.python.org/doc/current/lib/re-syntax.html

etc. etc. It should look familiar.

[...]

Regards,
Bengt Richter


From torkil at fast.no  Tue Sep 17 06:56:52 2002
From: torkil at fast.no (Torkil Grindstein)
Date: Tue, 17 Sep 2002 12:56:52 +0200
Subject: Regular expression help needed
References: <mailman.1032256460.12295.python-list@python.org>
Message-ID: <3D870A74.5D2D7595@fast.no>

Harvey Thomas wrote:
> 1) There are no comments or processing instructions that could
> contain similar data
> 2) The attributes always come in the order specified i.e. name
> first, content second, any others following
> 3) The attributes are always double quoted and there are no
> spaces surrounding the =

Thank you very much!
I have set up a little extended version:

re.compile("""<meta\s+name\s*=\s*[\"\']{0,1}michael[\"\']{0,1}\s+content\s*=\s*[\"\']{0,1}(\w*)[\"\']{0,1}\s*\/\s*>""",
re.IGNORECASE)

Which should rule out point 3) (except that it accepts "michael'
and michael' etc). 1) and 2) I think I can live with for my application
(due to knowledge of the documents).

Btw, I have no knowledge of SAX. Perhaps I should dig into that?

Torkil.


From fgeiger at datec.at  Wed Sep 25 03:01:11 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Wed, 25 Sep 2002 09:01:11 +0200
Subject: Saving & Restoring Python objects in XML?
References: <3D8F7800.F757A72D@wyeth.gsfc.nasa.gov>
Message-ID: <3d915f2b@news.swissonline.ch>

IIRC the Gnosis XML utilities offer XML/object mapping.

Cheers
Franz


"Tom Bridgman" <bridgman at wyeth.gsfc.nasa.gov> schrieb im Newsbeitrag
news:3D8F7800.F757A72D at wyeth.gsfc.nasa.gov...
> Hello,
>
> I'm setting up some codes where I want to save the state of the code in
> an XML file at some time and reload it later to continue the processing.
>
> It's relatively easy to save out the innards of Python objects into XML
> regardless of how the objects are nested.  But what about loading this
> state information back in from the XML file?
>
> Is there a way to write the parser handlers and/or other components (I'm
> using PyXML 0.7 and higher) such that they can assemble the required
> objects as they read in the file?  I'm wondering if there is an elegant
> solution that might not be immediately obvious from the 'brute force'
> method I keep coming up with.  I'd prefer to do this using SAX (since
> I've worked with it some), but using DOM is not out of the question if
> it solves the problem in a straightforward way.
>
> Thanks,
> Tom
> --
> Dr. William T."Tom" Bridgman           Scientific Visualization Studio
> Global Science & Technology, Inc.      NASA/Goddard Space Flight Center
> Email: bridgman at wyeth.gsfc.nasa.gov    Code 935
> Phone: 301-286-1346                    Greenbelt, MD 20771
> FAX:   TBD                             http://svs.gsfc.nasa.gov/




From fredrik at pythonware.com  Wed Sep  4 03:14:09 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 04 Sep 2002 07:14:09 GMT
Subject: Max line coords canvas.create_line() can display?
References: <fa06e058.0208291055.4b0dd338@posting.google.com>
Message-ID: <5pid9.5885$e5.1033551@newsb.telia.net>

"revyakin" wrote:

> Is there a limit of line coords (x1,y1,...xn,yn) which a Tk Canvas
> create_line method can display? I found that I can display a line
> consisting of 5000 x,y points, but a line of 20 000 points does not
> get displayed.

there shouldn't be a limit.

can you perhaps post a (short) code snippet that illustrates
the problem?

</F>




From dance_code at hotmail.com  Tue Sep  3 03:04:26 2002
From: dance_code at hotmail.com (lion)
Date: 3 Sep 2002 00:04:26 -0700
Subject: syntax error in Pyton shell
Message-ID: <e895ad50.0209022304.44c32987@posting.google.com>

I'm learning exception handling, I type these code in Python shell:

>>> try: f=open("nofile")
... except IOError: pass
... print "The exception is just passed!"
  File "<stdin>", line 3
    print "The exception is just passed!"
        ^
SyntaxError: invalid syntax

But these code run without any error as a program(or script). Could
anyone tell me why?

Thanks in advance,
lion


From ianb at colorstudy.com  Mon Sep 23 23:58:55 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 23 Sep 2002 22:58:55 -0500
Subject: Three dumb questions (ordered by dumbness descending)
In-Reply-To: <amnvit$7kqdt$1@ID-77524.news.dfncis.de>
References: <amnvit$7kqdt$1@ID-77524.news.dfncis.de>
Message-ID: <1032839942.11325.106.camel@dsl-65-184-205-5.telocity.com>

On Mon, 2002-09-23 at 16:03, Thorsten Kampe wrote:
> 2. Peter Norvig mentions in "Python for Lisp Programmers" some 
> "don'ts": "[x] + y" and "x[1:]". Are there more things to avoid 
> (especially in a loop)?

He's probably thinking about x[1:] because it is slower than (cdr x) --
in Python you are creating a new list.  You can use "del x[0]" instead.

The same is true of "[x] + y", as compared to (cons x y) -- you aren't
reusing the y list.  y.insert(x, 0) is considerably faster (as long as
you don't mind mutating y).

I haven't read what he said, but I think rather you shouldn't use:

x = x[1:]
# or
y = [x] + y

both these cases are suboptimal, even if they seem similar to some Lisp
cognate.

> 3. random.shuffle: the documentation says: "Note that for even rather 
> small len(x), the total number of permutations of x is larger than the 
> period of most random number generators; this implies that most 
> permutations of a long sequence can never be generated." I translate 
> this to "don't shuffle lists with more than 15 elements if you want a 
> randomly ordered list".
> 
> Alex Martelli says the same: "For example, Python users who are 
> calling random.shuffle on a sequence of substantial length have a need 
> for an underlying random generator with a *HUGE* period, as a 
> necessary although not sufficient condition towards ensuring that all 
> permutations of the shuffled sequence get generated with equal 
> likelihood.
> 
> Wichman-Hill has a period of less than 7,000 billions (7e12),
> insufficient to account even for the permutations of a sequence
> whose length is just 16 (16! > 2e13)."
> 
> Otherwise in the "Python Cookbook" (Selecting Random Elements from a 
> List Without Repetition) the shuffle version is called "the winner".
> 
> So: can I use random.shuffle or do I have to write my own - slower - 
> version to get a random list?

Well, it depends what you require of shuffle.  If you want to do
modeling on large sets with many permutations, maybe so.  The only way I
see to get around this is to use a true random number source -- like
/dev/random on Linux.  All pseudo-random sources will have a period
based on their potential seed size and amount of internal state.  For
long lists, it seems unreasonable to create a better pseudo-random
source -- though I haven't thought about it that hard.

random.shuffle is not optimized -- it's written in Python, and is very
short, and I don't think well implemented:

def shuffle(x):
    for i in xrange(len(x)-1, 0, -1):
        # pick an element in x[:i+1] with which to exchange x[i]
        j = int(random() * (i+1))
        x[i], x[j] = x[j], x[i]

I would implement it like this:

def shuffle(x):
    newList = [(random(), i) for i in x]
    newList.sort()
    return [i[1] for i in x]

Anyway, in either case you could create your own random number source
and copy and paste that algorithm using that new source.

  Ian





From mike at frcatel.fri.utc.sk  Thu Sep 19 09:54:02 2002
From: mike at frcatel.fri.utc.sk (Michal Kaukic)
Date: Thu, 19 Sep 2002 15:54:02 +0200 (CEST)
Subject: Python sequences reference - problem?
In-Reply-To: <pan.2002.09.19.13.14.36.452963.1508@metaslash.com>
Message-ID: <E17s1ko-0002Zc-00@frcatel.fri.utc.sk>

Hi,
   
   we like Python - I use it (with Numeric and SciPy modules) 
for teaching of Numerical analysis and my colleague is doing 
some research in discrete optimization in Python. And he often
surprises me with non-conformal use of language :-)

   We are mathematicians, not programmers. I suppose, he believes 
that Python "thinks" like Mathematician. Yesterday, he was struggling 
with the code shown below (somewhat simplified):
    

L_L=[[1,2,3], [2,3,4], [3,4,5]]

def do_process(LL):
    n=len(LL[0])
    rec = [0]*n
    Res=[]
    
    for row in LL:
        rec[0] = row[0]
        # further code - modifying rec 
        # lots of conditional processing and such...
        # ...
        Res += [rec]
                
    return Res

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

After calling do_process(L_L), the "expected" result should be

   [[1, 0, 0], [2, 0, 0], [3, 0, 0]]

but the true result is

   [[3, 0, 0], [3, 0, 0], [3, 0, 0]].


   Yes, this is fully in accordance with how the Python language should
behave:
       Res += [rec] inserts references to list object rec,
       which are further modified... (he should use copy(rec) instead).

   But there is nothing to make this behaviour clearly VISIBLE in code.
If I work with pointers in C/C++ I know and see they are pointers.
You can say - we also know that rec is list object and so be careful
with it.  Yes, but consider the complex code where the similar constructs
are very easy to overlook. And debugging of such code can be frustrating.

   My colleague was in state of despair and made thoughtful remarks
about FORTRAN and Python similarity. He believes that Python is corrupting 
his (computer) memory...

   So what is the point? I wrote this message in hope that there are
more people with similar experience. My question is - how to explain
to novice non-programmer users (maybe mathematically "infected")     
the quirks of Python sequence objects? Which methodology to use
in programs so we can clearly see we work with "pointers"?
Or maybe someone can propose the changes to the Language to overcome
this (psychological) barrier? I feel this as a serious obstacle
in using Python (my students experienced it too).

Thanks,
                  Mike





From thehaas at binary.net  Thu Sep 12 08:55:11 2002
From: thehaas at binary.net (thehaas at binary.net)
Date: Thu, 12 Sep 2002 12:55:11 GMT
Subject: Regexp Over Multiple Lines
References: <565e316f.0209040740.36573b56@posting.google.com> <565e316f.0209110529.30e61ea4@posting.google.com> <Xns928696BCE741duncanrcpcouk@127.0.0.1>
Message-ID: <P80g9.15253$TX5.453776@news1.east.cox.net>

Duncan Booth <duncan at nospamrcp.co.uk> wrote:
> Once you learn the idioms you should do fine.

> BTW, one of the idioms is not to use regular expressions as often as in 
> Perl.

I have done a lot of text processing using just the string module and the
built-in string methods (Py 2.x) and never touching regexes.  If you read
the regex how-to, you will remember it saying that for a lot of tasks
(like looking for an exact string) it's best to use string.find, etc.,
mostly because it's faster, and (to me) seems easier.

Just throwing it out there . . .

-- mikeh

-- 
Mike Hostetler          
thehaas at binary.net 
http://www.binary.net/thehaas 
GnuPG key: http://www.binary.net/thehaas/mikeh.gpg 


From tismer at tismer.com  Wed Sep 18 09:27:12 2002
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 18 Sep 2002 15:27:12 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <3D881B66.1080200@something.invalid>
Message-ID: <3D887F30.5000601@tismer.com>

Greg Ewing wrote:
> Martin v. Loewis wrote:
> 
>>
>> They aren't really dangerous, just unconventional.
> 
> 
> 
> Last I heard, Christian was moving pieces of the C
> stack while there could be pointers to it in use.
> That's definitely dangerous, as he discovered when
> it broke Tk.

And I fixed it by adding some small locking code
to _tkinter that prevends stack slicing in that
context.

> I'd be very nervous about using Stackless until
> it stops doing that.
> 
>  > And that cannot be
>  > fixed, as it is essential to operation of Stackless.
> 
> Seems to me it could be fixed by just copying the
> C stack as a whole, instead of piecemeal.

See my other posting. Everything is there already.
Everything is under your control and optional.

hard-to-discuss-if-people-haven't-even-looked-at-it - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From thorsten at thorstenkampe.de  Tue Sep 17 14:36:24 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Tue, 17 Sep 2002 20:36:24 +0200
Subject: Python equivalent to perldoc?
References: <Olvh9.7898$Lg2.1389903@news2.news.adelphia.net> <%Avh9.7933$Lg2.1392332@news2.news.adelphia.net>
Message-ID: <am7sna$3emja$2@ID-77524.news.dfncis.de>

* Bob X
> Bob X wrote:
>> In Perl you can run "perldoc warnings" and get the definition of what 
>> "warnings" does. Is there a equivalent in Python?
>> 
> Well...I found "pydoc". However I have to explicitly tell python to 
> execute it on Windows XP. I cannot type "pydoc sys" nor "python pydoc 
> sys" but I have to type "c:\python22\libs\pydoc sys" to get it to work.

Tough
 
> Shouldn't Python know its own libs?

It does:
,---
| Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
| Type "help", "copyright", "credits" or "license" for more information.
| >>> import pydoc
| >>>
`---

> Or at least put the "libs" in the ENV variable?

Windows XP doesn't have a ENV variable.


Thorsten


From marklists at mceahern.com  Wed Sep 25 11:04:49 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 25 Sep 2002 10:04:49 -0500
Subject: What function called this function ??
In-Reply-To: <98jk9.22340$0p1.391334@news2.ulv.nextra.no>
Message-ID: <JHEOKEOOLIGLDHCMAHMOEEGKEBAA.marklists@mceahern.com>

> Is it possible, by hacking the traceback/stack/etc, to get
> information about the function, that functions class and in what
> module that function and class are defined, from inside a function
> located elsewhere?  Something like a way of answering the question
> of "who called me?" without sending the answer along as a parameter.

Some variation on this theme:

**

#!/usr/bin/env python

def mycaller():
    """mycaller()

    Return the name of the caller of the caller of this function.
    """
    import sys
    steps_back = 2
    return sys._getframe(steps_back).f_code.co_name

def foo():
    print "foo called by %s." % mycaller()

def bar():
    foo()

bar()

**

Output:

foo called by bar.

// m



From loewis at informatik.hu-berlin.de  Wed Sep 18 03:58:56 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Sep 2002 09:58:56 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <3D881B66.1080200@something.invalid>
Message-ID: <j4wupjafun.fsf@informatik.hu-berlin.de>

Greg Ewing <see_reply_address at something.invalid> writes:

> Last I heard, Christian was moving pieces of the C
> stack while there could be pointers to it in use.

Yes, that's how stackless works, and that cannot change.

> That's definitely dangerous, as he discovered when
> it broke Tk.

What is the danger? A program crashes? It is dangerous to cross a
street at rush hour; sitting in front of a computer is rarely
dangerous.

Stackless adds a few limitations as to what C extensions can do. If
those restrictions are followed, it won't crash.

> I'd be very nervous about using Stackless until
> it stops doing that.

Why is that? Those crashes can only happen if you use the stackless
modules (and only if you use them with a stackless-unsafe extension).
If you don't import stackless, no stack is copied.

> Seems to me it could be fixed by just copying the
> C stack as a whole, instead of piecemeal.

Then it won't be stackless - one application is that the recursion
limit is only constrained by the heap size, not by the stack size.
Also, task switching becomes very expensive when copying the entire
thread. Finally, you *still* can get it to break: some extension
module might preserve the address of a local variable in global state,
but then the entire stack is copied away.

Regards,
Martin


From dag at animagicnet.no  Wed Sep  4 10:48:50 2002
From: dag at animagicnet.no (Dag)
Date: 4 Sep 2002 14:48:50 GMT
Subject: Matching a string up to a word
References: <slrnanbkbc.huj.dag@omally.animagicnet.no> <3d75ffaa$1@news.sentex.net>
Message-ID: <slrnanbti7.huj.dag@omally.animagicnet.no>

In article <3d75ffaa$1 at news.sentex.net>, Peter Hansen wrote:
> 
> Can you instead guarantee the format of the date that follows the 
> filename?  Will it always consist of the day of the week, the month,
> the date, and the year?  That would turn this into a relatively
> simple problem.

Yes I can guarentee the last four fields be in that date format.  
Actually I think that solved it.  I'll split on whitespaces into
an array, and remove the first and last four fields, and then join
whatever's left.  That should do it.  I was stuck thinking how to 
parse it left to right.  

Thanks for making look at the problem slightly differently.

Dag


From LogiplexSoftware at earthlink.net  Mon Sep 16 17:27:42 2002
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 16 Sep 2002 14:27:42 -0700
Subject: Source code
In-Reply-To: <c%qh9.102794$ub2.2120685@news1.tin.it>
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de>
	<pUdh9.9377$yt3.4604908@newssrv26.news.prodigy.com>
	<mailman.1032184821.5180.python-list@python.org> 
	<c%qh9.102794$ub2.2120685@news1.tin.it>
Message-ID: <1032211662.17056.278.camel@software1.logiplex.internal>

On Mon, 2002-09-16 at 13:17, BT wrote:
> > |If your code is that valuable, run it on your server remote from end
> > |users as a trade secret, or force them to agree to a license and protect
> > |it w/ copyright.
> >
> > I would reformulate the issue slightly:  If you BELIEVE your code is
> > that valuable, seek psychiatric help.
> >
> 
> I really don't understand posts like these.
> You think I'm fool to protect my work?

No, only a bit uninformed if you think you can get "security through
obscurity".  Obscuring the code will dissuade only the most casual or
incompetent from decompiling/disassembling your code.  Sort of like
hiding a spare key to your house under the "Welcome" mat.

> Ok, let's say I make a software in two years and I market it on internet.
> The software is good and sell some copy.
> You find the software, like it, change the name, add some feature and resell
> it.
> My time: 2 years, yours: 1 month.

And that would be illegal, hence the suggestion to *copyright* your
work.

> 
> I think you need a psychiatric help.
> Only Id shared some source code, after years...
> Noooooooo!!! IDDDDD!!! Foolish people!!! You want to sell it???!!! You want
> to live with that???!!!

They shared it because it was no longer relevant.  Sort of like opening
the source for MS-DOS 2.0 (which hasn't happened, but you should see my
point).  


> Go working in a mine, and write games for fun!!!
> I can give you the phone of my clinic, but stop it NOW, join the ideal
> world.

You mean the "real" world.  In the "ideal" world, most software would be
free, and people wouldn't pirate software that wasn't.

Incidentally, do you believe you are the first person to think of this? 
This question ("how can I obscure Python byte-code?") comes up at least
once a month on this list.  It's fairly clear that any scheme you come
up with will take longer to implement than it will to break by a
competent hacker.

Anyway, good luck!  If your software is any better than your protection
scheme I'll download it from http://www.warez.com when you release it ;)

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From h.goebel at crazy-compilers.com  Tue Sep 17 14:51:15 2002
From: h.goebel at crazy-compilers.com (Hartmut Goebel)
Date: Tue, 17 Sep 2002 20:51:15 +0200
Subject: Earing money with Python software (was: More on Protecting Source Code)
References: <mailman.1032232999.22661.python-list@python.org> <m3elbtcg8l.fsf@mira.informatik.hu-berlin.de>
Message-ID: <4jt7ma.fse.ln@lenashee.goebel-consult.de>

Martin v. Loewis wrote:

> For software, I think the economical risk is that somebody uses your
> software without giving you the money you want - not that somebody
> sits down and studies your code for days, to write a program that he
> could have written himself from scratch in the same time.

Which brings me to another (related) topic): How to earn money
distributing Python software?

I'm currently trying to make up my mind about exactly this problem
area: How can I get some reward back when writing software like
decompyle and bridgekeeper?

I would be very gald if someone could share his/her experiances on 
this topic!

- Sell it via micro-payment? 
- Declare it "Shareware" and hope somebody pays for it?
- Offer commercial support?
- Develop two different versions, a "free" one which is missing some
  features and a commercial/registered/supported one offering more
  sophisticated functionallity?

(Any other comments are welcome, too.)

Regards
+++hartmut
-- 
| Hartmut Goebel      | h.goebel at crazy-compilers.com               |
| crazy-compilers.com | compilers which you thought are impossible |


From wlfraed at ix.netcom.com  Tue Sep  3 11:15:52 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Tue, 03 Sep 2002 08:15:52 -0700
Subject: code doesn't work: question
References: <mailman.1031047683.15395.python-list@python.org>
Message-ID: <8nj2la.cr3.ln@ix.netcom.com>

Ruslan Spivak fed this fish to the penguins on Tuesday 03 September 
2002 03:03 am:


> if (tmp = var) == 5:         # this is the error line

        Unless things have changed since I last read a Python book, assignment 
is not an operator, it is a statement.

 tmp = var
 if tmp == 5:


-- 
--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From aleax at aleax.it  Wed Sep 25 05:11:14 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 09:11:14 GMT
Subject: problem with dictionaries within lists
References: <pan.2002.09.25.05.21.07.477231.167@ihatespam.shaw.ca>
Message-ID: <S4fk9.163219$pX1.5840444@news2.tin.it>

Grant Hallam wrote:
        ...
>  However, the list seems to get populated with the same dictionary
        ...
> webdatabaselist.append(dict)

Of course -- you ARE appending the same dict over ands over (AFAICT --
you're probably using tabs for indentation, making your code hard to
comprehend for those who read news with KDE's KNode [or MS's OE] as
tabs don't show there... use spaces, NOT tabs...).

Anyway, you probably need something like:

webdatabaselist.append(dict.copy())
dict = {}

either might suffice, but generally you'll need to copy the
current dict AND make dict empty for the next leg of the loop,
if I get the general drift.  If you do rebind name dict to {} for
the next leg, then you don't need the call to the copy method.


Alex



From dutoitc at hotmail.com  Thu Sep 19 13:49:04 2002
From: dutoitc at hotmail.com (Cedric Dutoit)
Date: Thu, 19 Sep 2002 19:49:04 +0200
Subject: wxPython i18n, help wanted
Message-ID: <amd2hs$i1k11@news.vtx.ch>

Hi,

I'm trying to translate an application in Chinese with i18n, but
I don't know how to display Chinese characters in wxPython.

Can anybody help me ?

(Chinese is an example language)

thks,
C.Dutoit




From Padraig at Linux.ie  Wed Sep 11 09:20:21 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Wed, 11 Sep 2002 13:20:21 GMT
Subject: regular expression for space seperated quoted string
References: <PqFf9.3627$cP3.7118@news.iol.ie> <alnbpb$ce5$1@wanadoo.fr> <%jHf9.3635$cP3.7161@news.iol.ie>
Message-ID: <pqHf9.3636$cP3.7130@news.iol.ie>

Padraig Brady wrote:
> Eric Brunel wrote:
> 
>> Padraig Brady wrote:
>>
>>
>>> Hi, I'm trying to split a string that is seperated
>>> by spaces and also contains double quoted words which
>>> can contain spaces:
>>
> 
> [snip]
> 
>> What about:
>>
>>>>> p = r'[^ \t\n\v\f"]+|"[^"]*"'
>>>>> re.findall(p, '1 2 3')
>>>>
> 
> FYI, I'm using the following:
> 
> import fileinput, re
> for line in fileinput.input():
>     #split fields
>     fields = re.findall('[^ "]+|"[^"]+"', line[:-1])
>     #remove quotes
>     fields = map(lambda field: field.replace('"', ''), listLine)
> 
> thanks again,
> P?draig.
> 

Just in case it's useful I might as well make it correct and more efficient:

import fileinput, re
reFieldSplitter = re.compile('[^ "]+|"[^"]+"')
for line in fileinput.input():
     #split fields
     fields = reFieldSplitter.findall(line[:-1])
     #remove quotes
     fields = map(lambda field: field.replace('"', ''), fields)

P?draig.



From python-list at hendel.net  Wed Sep 11 06:58:18 2002
From: python-list at hendel.net (Manuel Hendel)
Date: Wed, 11 Sep 2002 12:58:18 +0200
Subject: deleting the first and the last character of a string
Message-ID: <20020911105818.GC2054@partagas.as.de.cw.net>

I got a file with fields seperated by a "|". The problem is that there
is also a "|" at the beginning and at the end of each line although
there no fields in front or behind these "|". Now I want to delete
these two "|". Is there anything which says delete the first or last
character of a string? I checked the string module on the web, but I
couldn't find anything.

I solved it like this, but I honestly don't like it that way.

#!/usr/bin/env python
#
import sys
import string

inputfile = open(sys.argv[-1], "r")

for line in inputfile.readlines():
        fields = string.split(line, "|")
        lineN = string.join(fields[1:-1], "|")

Isn't there any better solution? ;-)

Manuel

-- 
The real purpose of books is to trap the mind into doing its own thinking. 
-Christopher Morley 



From Oschler at earthlink.net  Mon Sep 16 01:30:27 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Mon, 16 Sep 2002 05:30:27 GMT
Subject: Are generator functions thread-safe?
Message-ID: <T%dh9.6427$Lg2.1180859@news2.news.adelphia.net>

I know that a call to a generator function actually returns an independent
iterator object (mea culpa if I'm wrong on this), but are there any hidden
gotchas that would cause the use of the same generator function from
multiple threads, to result in a sharing (especially local storage) or
memory collision issue?

thx





From awinston at scn.org  Tue Sep 10 13:48:48 2002
From: awinston at scn.org (Alan Winston)
Date: Tue, 10 Sep 2002 10:48:48 -0700
Subject: PySol: "global name 'win32_gethomedir' is not defined"
References: <all3u2$qgm$1@brokaw.wa.com> <un0qp6aex.fsf@online.no>
Message-ID: <allba2$u6u$1@brokaw.wa.com>

"Syver Enstad" <syver-en+usenet at online.no> wrote in message
news:un0qp6aex.fsf at online.no...
> "Alan Winston" <awinston at scn.org> writes:
>
> > I've been trying to run the PySol 4.81 pyc files under Python 2.2.1
> > and
> >
> > WinXP and Win98, with no success.
> >
> > The Python window closes much faster than I can read the error message
> > (is
> >
> > there any way to force the window to stay open?),
>
> Can't help with the cause, but can't you just start the program from
> the command prompt instead of via explorer?

Thank you.

That makes the error messages much easier to see:

Traceback (most recent call last):
  File "pysol.py", line 19883, in ?
  File "pysol.py", line 19881, in main
  File "pysol.py", line 19838, in pysol_main
  File "pysol.py", line 18833, in __init__
  File "pysol.py", line 260, in gethomedir
NameError: global name 'win32_gethomedir' is not defined




--
Alan Winston
Seattle




From christophe.delord at free.fr  Thu Sep 19 00:45:29 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Thu, 19 Sep 2002 06:45:29 +0200
Subject: PyLog, not require SWI-Prolog right?
References: <Ey8i9.9524$Lg2.1789572@news2.news.adelphia.net>
Message-ID: <20020919064529.2a2ae37f.christophe.delord@free.fr>

On Thu, 19 Sep 2002 00:07:00 GMT
"Robert Oschler" <Oschler at earthlink.net> wrote:

> PyLog is inpendent of any Prolog interpreter right?  There's PyLog and
> PyProlog, the latter I believe is a wrapper and therefore requires
> SWI-Prolog (just like bevedere requires GNU-Prolog), but PyLog doesn't,
> is that correct?  Just looking for confirmation.

PyLog only requires Python and TPG. TPG is a parser generator and is used
to parse Prolog.
PyLog is there : http://christophe.delord.free.fr/en/pylog/
TPG is there : http://christophe.delord.free.fr/en/tpg/

Prolog is translated into Python (a predicate is an object) and then
interpreted by Python. So PyLog is independant of any Prolog interpreter
but also slower.

> 
> Also, PyLog supports backracking right?

PyLog supports backtracking but is limited by the size of the Python
stack. This may be not a problem with Stackless Python, I have to try it
some day.

> 
> thx
> 
> 
> 


-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From 2002 at weholt.org  Mon Sep 23 05:15:38 2002
From: 2002 at weholt.org (Thomas Weholt)
Date: Mon, 23 Sep 2002 09:15:38 GMT
Subject: WSDL for XML-RPC ??
Message-ID: <_YAj9.21316$0p1.377838@news2.ulv.nextra.no>

SOAP has WSDL to describe a webservice. Doeas XML-RPC have something
similar?
Is WDDX the right way to go or have I totally missed the point?

Ref :
http://www.wddx.org
http://www.openwddx.org/

Best regards,
Thomas




From wlfraed at ix.netcom.com  Tue Sep 24 21:55:02 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Tue, 24 Sep 2002 18:55:02 -0700
Subject: List/Tuple bug or feature?
References: <mailman.1032905792.2164.python-list@python.org>
Message-ID: <m15rma.pt4.ln@ix.netcom.com>

sismex01 at hebmex.com fed this fish to the penguins on Tuesday 24 
September 2002 03:12 pm:

> It's one of python's features :-)
> 
> For when you wanna write a long string, SQL for example:
> 
> Query = "SELECT what, when, where, why "
>         "FROM t1, t2, t2 "
>         "WHERE t1.id = t2.id_t1 and t3.id_t2 = t2.id and t1.id_t2 =
>         t3.id " "AND t1.answer = 'OK'"
>
        What's wrong with...

query = """select what, when, where, why
        from t1, t2, t3
        where t1.id = t2.id_t1
                and t3_id2 = t2.id ...etc """

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From tjreedy at udel.edu  Thu Sep 19 11:06:44 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 19 Sep 2002 15:06:44 GMT
Subject: surface of pygame
References: <mailman.1032440854.11721.python-list@python.org>
Message-ID: <8Kli9.36954$216.1447920@bin4.nnrp.aus1.giganews.com>

"black" <quiteblack at yahoo.com> wrote in message
news:mailman.1032440854.11721.python-list at python.org...
> I touched something about pygame today and confused with surface
type. I didnt >remember surface data type when learning python,

Because surface type is added by to PyGame and not part of standard
Python.

> and it displays strange to me, for example, I create a new surface
by loading a picture >as follows:
>
> pygame.image.load("myPic.bmp")
>
> it works fine but I coulnt see any surface appears on the screen.
> any ideas ?

Read the Pygame docs on how to display image after loading.  These are
two separate operations.  I believe there is a mailing list for PyGame
discussion.  See site.

TJR




From imbosol at vt.edu  Fri Sep  6 18:58:12 2002
From: imbosol at vt.edu (Carl Banks)
Date: Fri, 6 Sep 2002 18:58:12 -0400
Subject: Larry Wall's comment on python...
References: <mailman.1031349792.2374.python-list@python.org>
Message-ID: <allk4c$pen$1@solaris.cc.vt.edu>

mmaddox at hcsmail.com wrote:
> 
> I would argue that the "older" block-structured languages were considered
> a breakaway from the column-based languages - RPG2 and Fortran, for
> instance. Although I grant the Python structure and the column-based
> structures are not completely analogous, they BOTH burden the programmer
> with arbitrary rules to govern the physical layout of the code.

Python's forced indentation certainly doesn't burden me at all.  I
don't know about you, but I indent whether I have a block delimiters
or not.  Python doesn't force me to do anything I wouldn't do anyways.

What *is* burdensome is having to use block delimiters, AND to make
sure they are properly sychronized to the indentation.


> Although
> this layout may be considered more elegant from a readability standpoint,
> it's the enforcement of the layout that is irritating to the more
> Libertarian of us. It's a mindset difference similar to that between
> Pascal and C++.

Well, if you're complaining about your lack of freedom, then you're
complaining about your freedom to write unreadable and misleading
code.  It's a freedom I don't care for (usually).


> Having read Wall's comments on Slashdot, I think he is really just mildly
> complaining about ridiculously long page-widths in his editor, not really
> slashing Python seriously. He's just giving some props to his own baby at
> the expense of another. (He had to say something, right?) You must agree
> that Python can lead you to horizontal scrolling in your editor - the bane
> of user interface design.

No, not any more than C would.  Do you not indent in C?  If your C
code runs against the margin, is your solution to simply stop
indenting, and run all the blocks together?  (If it is, then you have
worse problems than being forced to indent.)

And, as a matter of fact, I never exceed 79 columns in any of my
Python code.  And I use eight-space tab stops.  And my editor doesn't
have horizontal scrolling; it wraps.


-- 
CARL BANKS
http://www.aerojockey.com


From tjenkins at devis.com  Fri Sep  6 10:41:56 2002
From: tjenkins at devis.com (Tom Jenkins)
Date: 06 Sep 2002 10:41:56 -0400
Subject: [DB-SIG] Re: What database should I use
In-Reply-To: <Pine.LNX.4.21.0209061629050.19270-100000@samba.crysberg.dk>
References: <Pine.LNX.4.21.0209061629050.19270-100000@samba.crysberg.dk>
Message-ID: <1031323316.6785.9.camel@asimov>

On Fri, 2002-09-06 at 10:29, Leif Jensen wrote:
> 
>     PostgreSQL, no doubt, if you use Linux, WinNT or possible WinXP !
> 
>  Leif
> 

I disagree.  Don't get me wrong, we use PostgreSQL EXTENSIVELY.  But for
2-3k records?  These "requirements" don't require anything that
powerful.

I'd look at Access, pySQLite (which intrigues me), Gadfly.

> 
> 
> On Thu, 5 Sep 2002, A wrote:
> 
> > Hello,
> > I am going to program a small application (in Python and wxPython) which consists about 
> > 2000 - 3000 records . Can you please recommend which database is the best for that. I 
> > use Windows.
> > Maybe MS Access  or FoxPro or MySQL or better?
> > I want to keep the program as small as possible.
> > Thanks
> > Ladislav
> > 
> > 
> > _______________________________________________
> > wx-users mailing list
> > wx-users at lists.wxwindows.org
> > http://lists.wxwindows.org/mailman/listinfo/wx-users
> > 
> 
> 
> _______________________________________________
> DB-SIG maillist  -  DB-SIG at python.org
> http://mail.python.org/mailman/listinfo/db-sig
-- 

Tom Jenkins
Development InfoStructure
http://www.devis.com





From emile at fenx.com  Thu Sep 19 09:56:11 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 19 Sep 2002 06:56:11 -0700
Subject: monetary applications (et al)
References: <mailman.1032390744.1503.python-list@python.org> <bsei9.130231$pX1.4674970@news2.tin.it>
Message-ID: <amcku1$9a1$1@main.gmane.org>

Alex Martelli:
> Doesn't seem to have done any damage to Python's standing in the
> field of scientific computation -- on the contrary, decoupling the

Out of curiousity, does python have any standing in the field of
business accounting?  Perhaps those with experience can chime in as to
how this has been handled effectively.  We learned long ago to keep all
amounts in whole pennies.  In python, we need to keep everything as
longs and be careful not to slip up and allow normal ints.  Not a
problem, and one that int/long unification will resolve when it gets
here.


--

Emile van Sebille
emile at fenx.com

---------






From loewis at informatik.hu-berlin.de  Mon Sep  2 12:31:26 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 02 Sep 2002 18:31:26 +0200
Subject: compiling extensions on linux
References: <3D738D4E.35B6C712@mill.co.uk>
Message-ID: <j44rd8jqu9.fsf@informatik.hu-berlin.de>

Joe Connellan <joec at mill.co.uk> writes:

> I'm trying to port an extension of mine from windows to linux and am not
> sure how to include the myExtension.def file in the compile 

You shouldn't have a .def file in the first place - I recommend that
you use distutils instead.

> eg if I use
> 
> g++ -shared -o myextension.so
> 
> it compiles fine but I get the following error when importing it into
> python
> 
> >>> import myextension
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> ImportError: dynamic module does not define init function
> (initmyextension)

Nothing to this respect is needed on Linux. Are you sure your module
does define a initmyextension function?

It might be that it defines initmyextension__Fv instead, since you
compile your extension as a C++ program.

You should compile it as a C program, unless it really is a C++
program, in which case you need to define the init function as extern
"C".

HTH,
Martin


From walterm at parque.homelinux.net  Mon Sep  9 19:19:51 2002
From: walterm at parque.homelinux.net (Walter Moreira)
Date: Mon, 9 Sep 2002 20:19:51 -0300
Subject: RPy / Win2k
In-Reply-To: <albgmp$cgv$1@newsg2.svr.pol.co.uk>
References: <albgmp$cgv$1@newsg2.svr.pol.co.uk>
Message-ID: <20020909231951.GA21801@parque.homelinux.net>

On Sat, Sep 07, 2002 at 01:19:13AM +0100, Duncan Smith wrote:
> Hello,
>          Is there anyone out there who has successfully used RPy under
> Win2k?  I get the impression (from Brian Ripley) that it should be possible
> to embed R under Win2k, but that it probably isn't straightforward.  Can
> anyone help?  Cheers.

Well, actually it is possible thanks to a patch from Rene Hogendoorn. I
have been busy but I will release a Windows version soon. The changes in
the source are not straightforward, indeed.  I'll try to do it before
end of september.

Regards:
Walter




From cliechti at gmx.net  Sun Sep  1 19:52:14 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 2 Sep 2002 01:52:14 +0200
Subject: Add methods to a class at runtime?
References: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com> <slrnan4dl3.18h.gerhard.haering@lilith.my-fqdn.de> <wmvc9.21757$FJ1.3181934@e3500-atl1.usenetserver.com> <Xns927CF1E86FFE8cliechtigmxnet@62.2.16.82> <mailman.1030921281.8891.python-list@python.org>
Message-ID: <Xns927D12FF64C84cliechtigmxnet@62.2.16.82>

holger krekel <pyth at devel.trillke.net> wrote in
news:mailman.1030921281.8891.python-list at python.org: 

> Chris Liechti wrote:
>> "Robert Oschler" <Oschler at earthlink.net> wrote in
>> news:wmvc9.21757$FJ1.3181934 at e3500-atl1.usenetserver.com: 
>> > Some interesting tidbits I
>> > found while trying out your suggestions,  as you will see in the
>> > two ".py" files that follow, I discovered that to create a method
>> > you can dynamically add to a class _instance_, you need to omit the
>> > 'self' parameter in the method declaration.  For a method you add
>> > dynamically to a class definition itself, then you must include the
>> > 'self' parameter in the method declaration.  If not, you will get
>> > an argument count mismatch complaint from the interpreter. 
>> > Fascinating stuff. 
>> 
>> that's just the differece between a "function" and a "method". a
>> method works on an object and that is passed explicit as first
>> argument. "explicit is better than implicit" (you can read this and
>> some other design ideas by typing ">>> import this" ;-)
>> 
>> > I'm running 2.1.1, anybody know if 2.2.2 is going "disable" either
>> > of these abilities? (class or class instance dynamic method
>> > addition).  I saw a note about such activities requiring a
>> > "dynamic" keyword, which is fine.  I'd really miss either of these
>> > abilities. 
>> 
>> nah. "classic classes" stay the way they are and are available until
>> 3.0 new style classes, those that inherit from "object" or "type",
>> will need the marking as "dynamic".
> 
> Can you give a pointer for further explanation?  Can't believe it :-)

robert already posted a link where its mentioned. i think i saw it too in 
that thread.
i might have choosen the wrong words... - it was it _was_ discussed, and as 
it looks, it was in the 2.2a release.

here is an older "Unifying types and classes in Python 2.2", which still 
mentiones __dynamic__, newer versions of it don't:
http://216.239.37.100/search?q=cache:Nld8Sg3osq0C:www.etsimo.uniovi.es/pyth
on/2.2/descrintro.html+__dynamic__&hl=en&ie=UTF-8

chris

-- 
Chris <cliechti at gmx.net>



From naveed_iq at hotmail.com  Wed Sep 25 13:33:21 2002
From: naveed_iq at hotmail.com (Naveed Iqbal)
Date: 25 Sep 2002 10:33:21 -0700
Subject: Whats the Difference between Windows XP and Windows 98....
Message-ID: <3fda493d.0209250933.1a99079f@posting.google.com>

Whats the Difference between Windows XP and Windows 98....when it
comes to the win32 extensions to python? I used the win32ui and
win32com.client modules to print to a printer from windows. the output
from the XP machine comes out fine, but the output from the 98 machine
looses its formating. I am using a Smart Label Printer 200. this label
printer has drivers for all windows family OS's.
does anybody have an idea as to what is going on?
thanks
Naveed


From lu_gio at hotmail.com  Tue Sep 24 02:57:16 2002
From: lu_gio at hotmail.com (Max)
Date: Tue, 24 Sep 2002 08:57:16 +0200
Subject: PyGame + PyOpenGl + ZBuffer?
References: <3D8EDCEF.70F827F6@hotmail.com> <iCGj9.186982$Jo.66084@rwcrnsc53>
Message-ID: <3D900CCC.13321011@hotmail.com>

I've inserted the set_attribute, then verify it was set.

    pygame.display.gl_set_attribute(GL_DEPTH_SIZE, 16)
    pygame.display.set_mode((640,480), OPENGL|DOUBLEBUF )
    print "ZBUFFER:",pygame.display.gl_get_attribute(GL_DEPTH_SIZE)
    print pygame.display.get_driver()

The display.get_driver() displays "directx"... DIRECTX????!!!!! WHY??? Is it
normal?

Inside the draw loop, I test if the depth test is still working...
        print "Enabled: ",glIsEnabled(GL_DEPTH_TEST)

It's a simple scene with four GL_QUADS overlapping; two with z=0 and two with
z=0.5
        glTranslate(100,100,0.5)
        glTranslate(100,100,0.0)

with a depth range from 0 to 1
    glDepthRange(0.0,1.0)

BUT... the zbuffer isn't still working... also I get a very low framerate.

HELLLP!!



Pete Shinners wrote:

> Max wrote:
> > I found my code is not working well in all computers: with a SiS 300/200
> > graphic card, the Z buffer seems disabled.
> >
> > The OpenGL initialization code is:
> >
> > pygame.init()
> > pygame.display.set_mode((640,480), OPENGL|DOUBLEBUF|HWSURFACE,32 )
> > glEnable(GL_DEPTH_TEST)
> > glDepthFunc(GL_LESS)
> > glDepthRange(0,1)
> >
> > How can I force Opengl to utilize ZBuffer?
>
> pygame has a function to control extra setting for opengl. it must be
> called before display.set_mode(). you'd probably want to pass something
> like "pygame.display.gl_set_attribute(GL_DEPTH_SIZE, 16)". perhaps try
> different sized depths?
>
> http://pygame.org/docs/ref/pygame_display.html#gl_set_attribute
>
> you might also try testing with GLUT, which usually builds with pyopengl.



From amk at amk.ca  Sat Sep  7 08:43:51 2002
From: amk at amk.ca (A.M. Kuchling)
Date: Sat, 07 Sep 2002 12:43:51 -0000
Subject: Daemonize?
References: <F7DB8D13DB61D511B6FF00B0D0F06233022B66EC@mail.hebmex.com> <mailman.1031184848.32071.python-list@python.org>
Message-ID: <unjt47pc5r8671@corp.supernews.com>

In article <mailman.1031184848.32071.python-list at python.org>, 
   Richard Jones wrote:
> Note you can also just assign an object with a write() method that does 
> nothing to sys.std(out|err) and a similar object to sys.stdin. Why the second 
> fork, btw?

No, reassigning sys.stdin and stdout isn't sufficient; you really,
truly have to close those file descriptors.  This is complicated
slightly by the fact that Python's .close() method for
stdin/stdout/stderr doesn't actually do anything.

The second fork ensures that the parent PID of your daemon process is
1.  I think it also ensures that the daemon is in a process group of
its own.

Here's the daemonize function I use; I think Greg Ward originally
gave it to me.

    def daemonize (self):
        # Fork once
        if os.fork() != 0:
            os._exit(0)
        os.setsid()                     # Create new session
        if os.fork() != 0:      
            os._exit(0)         
        os.chdir("/")         
        os.umask(0)

        os.close(sys.__stdin__.fileno())
        os.close(sys.__stdout__.fileno())
        os.close(sys.__stderr__.fileno())
        
--amk


From hwlgw at hotmail.com  Sun Sep  1 17:08:51 2002
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 1 Sep 2002 14:08:51 -0700
Subject: Python IPC options
References: <1030829572.943495@irys.nyx.net> <3D715160.2020109@itamarst.org_NOSPAM>
Message-ID: <cb035744.0209011308.7b4b3ae2@posting.google.com>

Itamar Shtull-Trauring <itamar at itamarst.org_NOSPAM> wrote:
> Twisted (http://www.twistedmatrix.com) has ...

Um, I admire the Twisted project and I hope it will mature someday. 
The Twisted people do not miss much occasions to mention Twisted. 
Which is understandable but I think they should not do that until the
Twisted documentation is better.  You really need more tutorials,
step-by-step stuff.  And also for Windows users, otherwise at least
mention it is aimed at *n*x people until then.
>From the link you provided:
% mktap ...
% twistd -f ...
They do not work and I just downloaded Twisted (.exe and .zip for
docs).  I am using Windows yes, like most people.  No idea how to do
``mktap'' using Windows so what do you think people do next?  They
remove Twisted.  And they think 'Wasted some more time with open
source stuff'.  And that is a shame and that is not what you want. 
Please stop coding and working on functionality until you have good
docs.  You have a doc explaining the difference between a framework
and a library, that is good.  More of that!  About daemons, servers,
internet protocols, tutorials, whatever.  I think your project is
great and I hope it will not die from the
open-source-amateur-be-cool-read-the-source-virus.  Open source is
great but weak without good docs.

::

It's not the mail volume that bothers me -- I can ignore 100s of
messages a day very quickly. It's the time it takes to respond to all
of them.
    -- Guido van Rossum, 20 Jan 2000


From hancock at anansispaceworks.com  Thu Sep 26 15:08:28 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Thu, 26 Sep 2002 12:08:28 -0700
Subject: How to convert unicode
References: <20020928160041.13325.57342.Mailman@mail.python.org>
Message-ID: <3D935B2C.3E2E08E1@anansispaceworks.com>

Hi Ladislav,
I think you're going to have to be a little more specific about
what exactly you want before people can help...

Ladislav:
> I have a script that downloads a webpage with  characters that are not plain ascii.From the
> web page header I can find out that there is UTF-8 coding( Unicode). I can display the page
> properly in a web browser if I send from my script
> "Content-Type: text/html; charset=UTF-8 "

Right, you're telling your browser what kind of data its receiving,
and it knows how to display it.

> But I need to save the page( the text of the page) to a file and some time later import the text
> to a database (MS Access or similar).
> How can I download the file or write to a  file so that the characters will be properly display in
> the file and in the database as well?
> Thanks for help

This is actually sort of meaningless, since databases and
files don't "display" data, they just "store" it. In general
they don't care whether that data is 7,8 or 16 bits wide,
as long as it's provided as a nice stream of bytes. (There
are exceptions -- I think some databases might trim off the
high bit, disallow "control characters", or nulls (0x00) for
"text" fields, but even those usually provide a "binary" data
type that won't mangle your data -- I can't address what 
MS Access does because I've never used it).

The question therefore is what do you want to use to *view*
the data in the file or database, and you need to specify
that.  If you want to view it with "basically everything", then
you'll need to sacrifice portability and capability by converting
to some localized encoding (the most common being ASCII, but
that may not be the most appropriate for where you are).
If that's okay, then this is a character-translation problem,
and the result will *lose data*, which you need to be sure is okay.
(This is because many applications still only support ASCII-style
8-bit encoding, despite much progress.  If nothing else, lots
of people are still using older software).

I'm guessing you want to preserve all the data though, in which
case, you're not talking about translating the data in the file
or database at all. You instead need to be asking what software
you can use (or how to configure your existing software if it
can already do it) to correctly display UTF8 encoded unicode.
You sacrifice some flexibility on the display side, but there are
a lot of unicode-aware software out there already.

If for some reason the software supports one of the other unicode
encodings, but not UTF8, then you'll need to do a unicode-to-unicode
translation, which won't lose data, but only change its
representation. UTF8 is widely supported, though, so I think this
isn't that likely.

As you can see, we'd have to start making a lot of guesses to
be more helpful than that (and I'd probably guess wrong!), so
I'll stop and give you a chance to respond with what you
want/need to use for display. :-D

Cheers,
Terry

-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
Anansi Spaceworks                 
http://www.anansispaceworks.com 
P.O. Box 60583                     
Pasadena, CA 91116-6583
------------------------------------------------------



From peter at engcorp.com  Tue Sep 24 20:18:17 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 24 Sep 2002 20:18:17 -0400
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Sep 24)
References: <up14av386chdda@corp.supernews.com>
Message-ID: <3d9100c1$1@news.sentex.net>

Dean Goodmanson wrote:
> QOTW:  "When something "is *EVERYWHERE*", how can it be "clearly VISIBLE"
> at the same time?" - Alex Martelli 
>     http://mail.python.org/pipermail/python-list/2002-September/123526.html
> 
> "... [A]necdotally speaking I find Python development goes faster more as
> a result of its very high readability than as a result of low loc/FP."
> 				 Peter Hansen     

For the record, this was from 
http://mail.python.org/pipermail/python-list/2002-February/089848.html
although I have to say I have no idea what it means... if it ever meant 
anything.  (Hansen's a real blowhard, you know.)

(I also don't know why a posting from February made it in the September
QOTW except as an accident.  But didn't someone say there's no such
thing as bad publicity? ;-)  )

-Peter



From tjreedy at udel.edu  Fri Sep 20 00:42:43 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 20 Sep 2002 04:42:43 GMT
Subject: e vs exp()?
References: <3a838998.0209191933.1f22e977@posting.google.com>
Message-ID: <7Hxi9.56777$216.2039933@bin4.nnrp.aus1.giganews.com>

"Buckshot" <c6re at sdsumus.sdstate.edu> wrote in message
news:3a838998.0209191933.1f22e977 at posting.google.com...
> Can anyone here explain why I'm getting different results from
> e**2 and exp(2)?  Which is more accurate?
>
> Python 2.2 (#1, Mar 27 2002, 14:56:58)
> [GCC 2.95.3 20010315 (release)] on sunos5
> Type "help", "copyright", "credits" or "license" for more
information.
> >>> from math import *
> >>> e**2
> 7.3890560989306522
> >>> exp(2)
> 7.3890560989306504

ActivePython 2.2.1 Build 222 (ActiveState Corp.) based on
Python 2.2.1 (#34, Apr 15 2002, 09:51:39) [MSC 32 bit (Intel)] on
win32
Type "help", "copyright", "credits" or "license" for more information.
>>> from math import *
>>> e**2
7.3890560989306495 # same as below to within roundoff, but not quite
same...
>>> exp(2)
7.3890560989306504 # matched your result exactly

The last decimal digit reported for floats is almost always flakey.
Both round to 7.3......650, so I suspect this is correct and e**2 on
your machine is not.

On the hypothesis that x**y is calculated as exp(y*log(x)), I tried
>>> exp(2*log(e))
7.3890560989306504
I expected this to end in ..495 so I am puzzled a bit also.  How is
e**x being calculated to give a different result?  Answer in the
Python source maybe, or in C library source.  In some other
situations, hypothesis seems correct
>>> pi
3.1415926535897931
>>> sq = sqrt(2)
>>> sq
1.4142135623730951
>>> pi**sq
5.0474972673709111
>>> exp(sq*log(pi))
5.0474972673709111
but not in all
>>> sq**pi
2.9706864235520198
>>> exp(pi*log(sq))
2.9706864235520194

Using 2.0 gives same result as 2:
>>> e**2.0
7.3890560989306495
>>> exp(2.0)
7.3890560989306504
>>> log(e)
1.0

Perhaps your system does not have e quite as correct:
>>> e
2.7182818284590451

Terry J. Reedy








From jdavis at empires.org  Tue Sep 24 16:37:04 2002
From: jdavis at empires.org (Jeff Davis)
Date: Tue, 24 Sep 2002 20:37:04 GMT
Subject: cgi post problem
References: <FVRj9.192186$Jo.71059@rwcrnsc53> <9aWj9.39314$V7.10439005@twister.socal.rr.com> <amq8av$2cum$1@agate.berkeley.edu> <amqgaq$2g05$1@agate.berkeley.edu>
Message-ID: <Q14k9.62350$U7.21809935@twister.socal.rr.com>


These two files worked for me (see below). Let me know if you have any more 
problems.

Regards,
        Jeff

-------------py2.cgi--------------------------
#!/usr/bin/python2.2                                                            
import cgi

def main():
    try:
        form = cgi.FieldStorage()
        input = form.keys()

        print "Content-type: text/html"
        print
        print("""<?xml version='1.0' encoding='UTF-8'?>                                 
<!DOCTYPE html PUBLIC '-//W3C//DTD XHTML 1.1//EN'                               
'http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd'>                                 
<html xmlns='http://www.w3.org/1999/xhtml' xml:lang='en'>                           
<head>                                                                              
<title>test</title>                                                         
</head>                                                                         
<body>""")
        print "<p>",input,"</p>"
        print "</body></html>"
    except:
        pass

if __name__ == "__main__":
    main()

---------------py2.html------------------------------------------
<html>
<body>

<form action="py2.cgi" method="post">
     <input type="text" name="Name" size="20" />
     <input type="text" name="PartNumber" size="20" />
     <input type="text" name="Inventory" size="20" />
     <input type="text" name="Cost" size="20" />
     <input type="text" name="WholesalePrice" size="20" />
    <input type="submit" value="Update Records" />
</form>

</body>
</html>



jano wrote:

> There was an error in my prior post.  See ^^^^
> 
> 
> "jano" <notvalid at hotmail.com> wrote in message
> news:amq8av$2cum$1 at agate.berkeley.edu...
>>
>>
>> "Jeff Davis" <jdavis at empires.org> wrote in message
>> news:9aWj9.39314$V7.10439005 at twister.socal.rr.com...
>> > Hi,
>> >
>> > It looks as though you have a lot of code. It might be helpful to try
>> > to narrow the problem further by removing a lot of the code unrelated
>> > to
> your
>> > problem, which appears to be that "keys" does not contain any items.
> After
>> > you work that out, then put the other code back in.
>> >
>> > My first impression is that you have some "weird" form element names
>> > because they contain the "/" character. That might be perfectly legal,
> but
>> > perhaps the cgi.FieldStorage class does not properly handle such
>> > element names, or perhaps it considers such names erroneous.
>> >
>> > Next, it seems that the line "keys = form.keys()" might be out of
>> alignment
>> > with the rest of the code in it's block. Perhaps it's just the way it
>> > looks after you copied your code into the post.
>> >
>> > Regards,
>> >         Jeff
>> >
>> >
>>
>>
>>
>> Yes, the problem is that keys does not contain any items.  The 'weird'
> names
>> are not the problem, because i tried it also with regular 1-word names
>> (no
>> slashes), with the same results.  Also, the keys=form.keys() line is
>> properly indented in the code -- it was just a copy/paste problem.  Here
> is
>> the code again, with the extraneous stuff stripped out:
>>
>> #!/usr/local/bin/python
>>
>> import cgitb; cgitb.enable()
>> import sys, traceback, cgi
>>
>> def main():
>>     try:
>>         form = cgi.FieldStorage()
>>         input = form.keys()
>>
>>         print "Content-type: text/html\n\n"
>>         print("""<?xml version='1.0' encoding='UTF-8'?>
>>         <!DOCTYPE html PUBLIC '-//W3C//DTD XHTML 1.1//EN'
>> 'http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd'>
>>         <html xmlns='http://www.w3.org/1999/xhtml' xml:lang='en'>
>>             <head>
>>                 <title>test</title>
>>             </head>
>>             <body>""")
>>         print "<p>",keys,"</p>"
> 
> #                              ^^^^ should be 'input' not 'keys', as it
> #                              is
> in my script.
> 
>>         print "</body></html>"
>>    except:
>>         pass
>>
>> if __name__ == "__main__":
>>      main()
>>
>> And the form is:
>>
>> <form action="http://someurl.com/cgi-bin/updatexml.py" method="post">
>>      <input type="text" name="Name" size="20" />
>>      <input type="text" name="PartNumber" size="20" />
>>      <input type="text" name="Inventory" size="20" />
>>      <input type="text" name="Cost" size="20" />
>>      <input type="text" name="WholesalePrice" size="20" />
>>     <input type="submit" value="Update Records" />
>> </form>
>>
>>
>> The script, again, is currently just supposed to return the keys, which
>> I
> am
>> expecting to be Name, PartNumber, etc., but it returns only an empty
>> list. Any ideas.
>>
>> thanks,
>>
>> jano
>>
>>
>>



From dance_code at hotmail.com  Mon Sep 30 11:12:38 2002
From: dance_code at hotmail.com (lee lion)
Date: Mon, 30 Sep 2002 15:12:38 +0000
Subject: What's wrong with it?
Message-ID: <F78PoHZgzDgtZKpjE5a000055b1@hotmail.com>

Hi, What's wrong with it?
class Borg(object):
  _shared_state = {}
  def __new__(cls):
    return cls._shared_state

class Singleton(Borg):
  def __new__(cls, arg):
    self = Borg.__new__(cls)
    self['val'] = arg
    return self
  
def main():
  x = Singleton('sausage')
[snip]

When running:
>>> 
Traceback (most recent call last):
  File "D:\Programming\Python22\my_Python_programs\MyBorgSingleton.py", 
line 27, in ?
    main()
  File "D:\Programming\Python22\my_Python_programs\MyBorgSingleton.py", 
line 16, in main
    x = Singleton('sausage')
ValueError: dictionary update sequence element #0 has length 1; 2 is 
required

Could anyone tell me why it broke with that strange runtime error?

Thank in advance,
Lion

"The best way of living is to earn a little more than you cost and to cost 
just as you need to."
                                      Li Hang

_________________________________________________________________
?????????????? MSN Messenger: 
http://messenger.microsoft.com/cn/




From aglyportREMOvethispart at nospam.yahoo.com  Tue Sep  3 23:11:45 2002
From: aglyportREMOvethispart at nospam.yahoo.com (Anton Graph)
Date: Tue, 03 Sep 2002 20:11:45 -0700
Subject: Coredumps (Python 1.4)
Message-ID: <3D7579F1.1090500@nospam.yahoo.com>

PyObject *
getfeed(char *feedfilename)
{
	int p[2];
	// create a pipe dolayout would generate output into
	if( pipe(p) < 0 ) {
		raise("piping failed");
	}
	char cmd[256];
	sprintf(cmd, "%s %s %d >/tmp/spt.dolayout 2>&1",
		sptbindir("dolayout"), feedfilename, p[1]);
	// dolayout is killed on SIGPIPE delivery caused by sptdumptags
	// termination
	FILE *fp = popen(cmd, "w");
	close(p[1]);
	if(!fp) {
		raise("can't run dolayout");
	}
	PyObject *layout[2];
	layout[1] = PyFile_FromFile(fp, feedfilename, "w", fclose);
	// line buffering
	if(setvbuf(fp, 0, _IOLBF, 0)) {
		raise("can't make out_feed linebuffered");
	}
	errno = 0;
	fprintf(fp, "lblheight\n");
	fp = fdopen( p[0], "r" );

	char s[128]="";
	if(!fgets(s, sizeof(s), fp)) {
		// l_fatal?
		l_warn("there is nothing in the pipe");
	}else{
		if(verbose)
			fprintf(stderr, "lblheight is %s", s);
	}
	layout[0] = PyFile_FromFile(fp, feedfilename, "r", fclose);
	PyObject *tuple = PyTuple_New(2);
	PyTuple_SET_ITEM(tuple, 0, layout[0]);
	PyTuple_SET_ITEM(tuple, 1, layout[1]);
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ I coredump on this 
line ever since I linked with Electric Fence memory debugging lib. Any 
ideas why? This is with Python 1.4 on SCO. Any know issues with memory 
allocation there? I'm building 2.2 now, but it will take a while.
This code seem to work fine on Linux.

	return tuple;
}



From martin at v.loewis.de  Thu Sep  5 19:39:49 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 06 Sep 2002 01:39:49 +0200
Subject: Python Source Code Documentation
References: <al69st$n1f$1@driftwood.ccs.carleton.ca> <G5Bd9.427560$m91.16660188@bin5.nnrp.aus1.giganews.com> <m3bs7cubp3.fsf@mira.informatik.hu-berlin.de> <al8k02$186$1@driftwood.ccs.carleton.ca> <9FQd9.324581$Aw4.13464627@bin2.nnrp.aus1.giganews.com>
Message-ID: <m3ofbcngze.fsf@mira.informatik.hu-berlin.de>

"Terry Reedy" <tjreedy at udel.edu> writes:

> When I looked a few years ago, I mentally divided the source files
> into four groups:
>   compilation phase (a few files)
>   interpretation phase (1 file, the main loop -- ceval.c?)
>   builtin datatypes (one for each -- int, long, tuple, etc -- each
> with similar structure)
>   other

There is the obvious division Parser, Python, Modules. Python can be
further subdivided as you indicate.

Regards,
Martin



From syver-en+usenet at online.no  Mon Sep  9 17:35:02 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Mon, 09 Sep 2002 21:35:02 GMT
Subject: Undocumented file behaviour (mode='r+')
Message-ID: <ulm6argnk.fsf@online.no>

It would be nice for those of us who are not familiar with all the in's
and out's of file processing to include in the documentation for file
objects the behaviour of the write and read methods when opening a
file for updating ('r+' or 'rb+'). 

I finally understood what was happening when I looked up the MSDN C-runtime
library documentation for fopen. Before that I was pretty amazed that
my calls to write failed to have an impact on the file that it was
called on. Pretty that both write and read fails without notice
methinks, but when that is the behaviour of the underlying library I
understand it might be difficult to fix.

Here's the behaviour I am talking about:

class FileBugTestForWin2k(unittest.TestCase):
    """ Demonstrates that you have to do a
    flush after writing, if you are gonna read
    and a seek after reading if you are gonna write.

    It seems that the MS documentation is faulty, the unix man page
    documents the same behaviour as we're seeing here
    """
    def setUp(self):
        f1 = open('aTestingFile.bin', 'wb')
        f1.write('lots of shit')
        f1.close()

    def tearDown(self):
        import os
        os.unlink('aTestingFile.bin')

    def testFailureWithoutSeek(self):
        f2 = open('aTestingFile.bin', 'rb+')
        assert f2.read(2) == 'lo'
        f2.write('s') # here is the bug, nothing is written to the file
        f2.seek(0)
        # should have been 'loss of shit' but write fails silently
        self.assertEquals('lots of shit', f2.read())

    def testFailureWithFlush(self):
        """ This should be successful according to the docs
        """
        f2 = open('aTestingFile.bin', 'rb+')
        assert f2.read(2) == 'lo'
        f2.flush() # MS C-runtime docs say we have to flush or
                   # seek when switching from read to write
        self.assertRaises(IOError, f2.write, 's') # write fails here with
                                        # errno 0
    def testSuccessfulWithSeek(self):
        f2 = open('aTestingFile.bin', 'rb+')
        f2.read(2)
        f2.seek(2) # ATTENTION, I didn't know that you had to seek when
                   # switching between read and write operations on the file.
        f2.write('s')
        f2.seek(0)
        self.assertEquals('loss of shit', f2.read())

    def testWriteToReadFailure(self):
        f2 = open('aTestingFile.bin', 'rb+')
        f2.write('s')
        assert len(f2.read()) > 300 # reads a load of crap

    def testWriteToReadFailure(self):
        f2 = open('aTestingFile.bin', 'rb+')
        f2.write('s') # advancing file pointer 1 step into the file
        f2.flush()
        self.assertEquals('ots of shit', f2.read()) # reading from seek(1)


-- 

Vennlig hilsen 

Syver Enstad


From jubafre at brturbo.com  Mon Sep 23 20:52:07 2002
From: jubafre at brturbo.com (jubafre at brturbo.com)
Date: Mon, 23 Sep 2002 21:52:07 -0300 (GMT-03:00)
Subject: file = open(path,'wb')?????
Message-ID: <611064730.1032828727023.JavaMail.nobody@webmail2.brturbo.com>

i open the a file for writing in a binary mode, but doesnt work

final3=['0020 0000x9', '0030 0000xa', '00B4 0000x8', '0030 0000xb', '00F0', '0001', '0003', '0005']
destino=raw_input("Digite o nome do arquivo .mem: ")
f = open(destino,'wb')
f.write('\x03AHM ')
for i in final3:
    f.write(i+' ')
f.close()


 i open the file in the notepad and my file appears like a text file and not as a binary file, i dont know why??

>>file.mem
AHM 0020 0000x7 0030 0000x8 0010 0000x9 00F0 0003 0002 0000 

>>i want the file like this, how??
AHM  &#8364; 0 ?  &#402; ? 



Juliano Freitas
www.gebrasil.hpg.com.br

From aleax at aleax.it  Wed Sep 25 06:26:53 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 10:26:53 GMT
Subject: comparing all values of a list to regex
References: <mailman.1032947731.3297.python-list@python.org>
Message-ID: <Nbgk9.140934$ub2.3077596@news1.tin.it>

Manuel Hendel wrote:

> I have to compare all values of a list to a regex. If the regex
> matches all list items the list should be added to a new list. If the
> regex matches just one list item, the list should be added to another
> new list. And last but not least, if the regex doesn't mutch at all,
> the list should be added again to another new list.
> 
> Can someone help me with this? This is driving me crazy.

E.g.:

num_matches = [ there.match(item) is not None for item in alist ].count(1)

if num_matches == 0:
    againanothernewlist.append(alist)
elif num_matches == 1:
    anothetnewlist.append(alist)
elif num_matches == len(alist):
    anewlist.append(alist)


Note that there's some ambiguity on what to do if alist is empty --
literally this is BOTH "matches all list items" AND "doesn't match
at all".  Here I've chosen the second interpretation, but depending
on your specs you can choose to test num_matches in different ways.


Alex
      


From skip at pobox.com  Mon Sep  9 21:26:33 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 9 Sep 2002 20:26:33 -0500
Subject: how to build wxPython w/ Sun's C++ compiler?
In-Reply-To: <m3k7luojpq.fsf@mira.informatik.hu-berlin.de>
References: <mailman.1031607011.28525.python-list@python.org>
        <m3k7luojpq.fsf@mira.informatik.hu-berlin.de>
Message-ID: <15741.19017.938561.815463@12-248-11-90.client.attbi.com>

    >> How do I convince it to use CC for files with .cpp extensions.

    Martin> Currently, you cannot: distutils does not support usage of
    Martin> separate tools for C and C++; feel free to contribute patches.

I received a little frontend script due to Robin Dunn which did the trick
for me.  It examines the args on the command line.  If any end in ".c" it
assumes C, otherwise C++.  I realize that's a hack.

As for patches, it seems to me that distutils needs a generic way to
override such settings (CC, CFLAGS, LDFLAGS, etc) via command line
arguments, environment variables or ConfigParser files.  It's not clear
which way would be most Pythonic.  I'd lean toward environment variables,
but that's just because the setup.py command line is cluttered enough as it
is and I'm a Unix-centered geek.

Skip



From campbells4 at ozemail.com.au  Fri Sep 20 12:31:08 2002
From: campbells4 at ozemail.com.au (Alistair Campbell)
Date: Sat, 21 Sep 2002 02:31:08 +1000
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com>
Message-ID: <b0Ii9.7044$Sr6.243147@ozemail.com.au>

Hi,

I am a bit in the same position ben. I am a psychologist by profession but I
have always been interested in programming. But, there is no way I would do
it as a job. I think doing it for other people would be too full on for me.

I get my inspiration for things to program from my other interests. So, I am
in the middle of producing a computerised assessment of executive brain
function. Next I want to write a program that enables me to develop and
administer questionnaires, surveys, and semi-structured interviews on a PC
(even TCP/IP!) and link it back to a central clinical information database.
Then I would like to write a small statitistics program that enables users
to compute clinical significance of change pre and post treatment. I would
also like to expand this to more of a suite of obscure statitics tools in
the social sciences. I also have a project in mind to write a program that
does a thing called repertory grids.

All in all these are areas of psychology that I have an interest in anyway
and the idea of programming aspects of them simply comes because I have the
programming as a tool. If I were you I'd stick to programming in your other
domain of knowledge first so that you're not having to learn the language
and the knowledge at the same time. If you have the knowledge it is then
technical thing about how to automate the process as efficiently and as
elegantly as possible.

All the best.

Alistair




From maxm at mxm.dk  Thu Sep 26 14:05:16 2002
From: maxm at mxm.dk (Max M)
Date: Thu, 26 Sep 2002 20:05:16 +0200
Subject: How do you pronounce "__" (double underscore)?
References: <525eaba.0209260956.72bf6786@posting.google.com>
Message-ID: <3D934C5C.4000603@mxm.dk>

Pat Notz wrote:
> @ = "at"
> ! = "bang"
> . = "dot"
> * = "star"
> 
> Is there a short (one or two syllable) pronounciation for "__" (double
> underscore).  "Underscore underscore" and "double underscore" are kind
> of long.  Somthing like "splat"?  Anyone?

_ = 'under'

__ = 'under-under'



regards Max M

The reason I don't reach any higher is that I stand on the shoulders of 
midgets.



From imbosol at vt.edu  Mon Sep 23 23:38:44 2002
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 23 Sep 2002 23:38:44 -0400
Subject: Three dumb questions (ordered by dumbness descending)
References: <amnvit$7kqdt$1@ID-77524.news.dfncis.de> <pan.2002.09.23.20.43.59.461504.1216@io.com>
Message-ID: <amomou$knu$3@solaris.cc.vt.edu>

Steven Feil wrote:
> On Mon, 23 Sep 2002 16:03:21 -0500, Thorsten Kampe wrote:
> 
>> Okay, here they are:
> 
> Ok, I'll take a stab at the first two questions. I do think that a
> little more clarity on your part could make it a bit easer.
> 
> The first thing to remember is that Python is a dynamically typed
> language. It is typed and each object has a type, It would be a little
> easer if you would disclose the type for each object you are referring
> to.
> 
> 
>> 1. Why is 'zip(zip(x)) != x' (should be the same because it's the
>> transposed)
> 
[snip]
> 
> I will need to know more about where you got zip from.

It's a builtin, since Python 2.0.  Very handy for when you want to
loop over two sequences side-by-side.  For example:

    for x,i in zip(seq,range(len(seq)):
        print "Item %d of seq is %s" % (i,x)



>> 2. Peter Norvig mentions in "Python for Lisp Programmers" some
>> "don'ts": "[x] + y" and "x[1:]". Are there more things to avoid
> 
> humm, "[x] + y", again it comes down to type.
[snip]


In this case, the types of x and y are obvious to someone familiar
with functional programming.  Yet I wonder what the point of these
"don'ts" is.

I imagine he means that adding and slicing lists is Python is not as
efficient as it is in LISP?  I wonder, then, what the recommended way
do functional stuff like that.


-- 
CARL BANKS
http://www.aerojockey.com


From tdelaney at avaya.com  Thu Sep 12 19:09:14 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 13 Sep 2002 09:09:14 +1000
Subject: "str.contains(part)" or alternatives?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A594@natasha.auslabs.avaya.com>

> From: Stefan Schwarzer [mailto:sschwarzer at sschwarzer.net]
> 
> Thank you (and all others) for their answers. I didn't expect 
> to get so many.
> Btw, Greg, the .count approach is much nicer than ".find != -1".

The .count approach can also be considerably slower - it must scan the
entire string, rather than being able to stop when it first finds a match.

Tim Delaney



From maxm at mxm.dk  Tue Sep  3 03:58:54 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 03 Sep 2002 09:58:54 +0200
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <mailman.1030907703.26266.python-list@python.org> <akuifb$h0e$2@solaris.cc.vt.edu> <1030962064.628232@ampungk.ozonline.com.au> <al0djh$a33$1@solaris.cc.vt.edu>
Message-ID: <3D746BBE.4030008@mxm.dk>

Carl Banks wrote:

>>What would the equivalent of r"(.)(.)(.)\3\2\1"
>>This means a "palindrome of 6 characters"
>>But it is unlikely that the human readable processor would understand 
>>that (isn't it??)



Why not just write the parts of the regex as named strings?

like::

     name = '[a-zA-Z0-9_.]+'
     at = '@'
     dot = '\.'
     topDomain = 'com|org|dk'
     email = name + at + name + dot + topDomain

instead of::

     email = '[a-zA-Z0-9_.]+@[a-zA-Z0-9_.]+\.com|org|dk'

Well my syntax is most likely wrong as I suck at regex, but the meaning 
should be clear enough.



From TuxTrax at fortress.tuxnet.net  Sun Sep 22 04:13:49 2002
From: TuxTrax at fortress.tuxnet.net (TuxTrax)
Date: Sun, 22 Sep 2002 08:13:49 -0000
Subject: cursor positioning and color
Message-ID: <slrnaoqv2q.hn4.TuxTrax@fortress.tuxnet>

Being new to python, I would like to know how one goes about such 
tasks as precise cursor positioning and output to screen in color.

For instance, If I want to print a character or line, at line 10, 
column 18, in bright white on red, how would I accomplish this?

Is there a python library specifically for setting up a text screen 
to look and function the way you want it to, without having to use a GUI?

The SLRN screen is a good example. Obviously, text positioning and coloring
commands are being utilized here, as well as such things as clearing
the screen and controlled scrolling.

How would these things be done in python, and what reference material is
available on line? I couldn't find anything in the Python manual, so any
URL's would be appreciated

Cheers,

Mathew

-- 
TuxTrax   (n.) An invincible, all terrain, Linux driven armored assault
vehicle that can take as much fire as it gives ;-)

Yes, I am a Penguin cult high priest. Flipper readings upon request.

ROT13 this email address to mail me:
bar jbeq abg guerr, uvtu qrfreg zna, ng lnubb qbg pbz



From creyes at ea.com  Thu Sep 12 21:33:53 2002
From: creyes at ea.com (Christian Reyes)
Date: Thu, 12 Sep 2002 18:33:53 -0700
Subject: Problem with HTTPS request 
Message-ID: <uo2fveepf88175@corp.supernews.com>

I got the following response when trying to make an HTTPS request.

<h1>SSL required to access this page</H1>

Then I read in the documentation the following

Note: HTTPS support is only available if the socket module was compiled with
SSL support.

Does anyone know where i can get my hands on a socket module that was
compiles with SSL support?

TIA,
Christian




From cliechti at gmx.net  Mon Sep 23 22:48:32 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 24 Sep 2002 04:48:32 +0200
Subject: Python and cgi and <a> problem!
References: <ammn6e$lj7$1@lust.ihug.co.nz> <Xns9292EA042C259cliechtigmxnet@62.2.16.82> <slrnaov5q3.nrp.jon+usenet@snowy.squish.net>
Message-ID: <Xns929330F5A61B0cliechtigmxnet@62.2.16.82>

Jon Ribbens <jon+usenet at unequivocal.co.uk> wrote in
news:slrnaov5q3.nrp.jon+usenet at snowy.squish.net: 

> In article <Xns9292EA042C259cliechtigmxnet at 62.2.16.82>, Chris Liechti
> wrote: 
>> print '<A href="%s">%s</A>' % (filename, filename)
> 
> I hope none of your filenames have '"', ">" or "&" in.

yeah i avoid characters that are special to the shell in my filenames - i 
even don't use " " very often. <wink>

sure there are improvements, but its a start.

chris

-- 
Chris <cliechti at gmx.net>



From mcherm at destiny.com  Wed Sep 18 12:38:09 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Wed, 18 Sep 2002 12:38:09 -0400
Subject: Python-style license?
References: <jikm5-p5l.ln1@junker.stroeder.com>
Message-ID: <amaa4q$i5t$1@main.gmane.org>

Michael Str?der wrote:
> HI!
> 
> In some Python modules there's only a note about a "Python-style 
> license". Now I wonder what that means. Any clues?
> 
> Ciao, Michael.
> 

http://www.python.org/2.2/license.html






From manuel.klimek at web.de  Fri Sep 20 14:27:11 2002
From: manuel.klimek at web.de (Manuel Klimek)
Date: Fri, 20 Sep 2002 20:27:11 +0200
Subject: Embedding Python in C++ and get rid of it again
Message-ID: <amhhcq$ur4$05$1@news.t-online.com>

Hi folks,

I've got a problem:
I'm coding on xpermud, a scriptable mud client.
This one features multiple scripting languages
(perl and python at the very moment). The
big problem is, that we want the user to be
able to choose his scripting language at 
runtime (saved, for example, in a bookmark).
It was quite hard to get perl to do this part,
but unloading the library and reloading it
made this possible. At first this seamed to
work with python, too. In fact, it does, as
long as you don't start any modules that
depend on .so files. They just don't find
the symbols python exports. So I did lot's of
research and found out that setting the
LTDL_GLOBAL flag helps with this one. But then
the python .so has it's symbols globally and
can't be unloaded any more. Even worse, perl.so
can't be started afterwards (probably) because
of a symbol name clash. So why does python want
to have "hard"-exported symbols in .so libraries?
Only so that python module developers can
be more "dirty" in their programming ways? I
think from a software design point of view, that
this can't be a good solution (and this was
the _only_ thing where perl seemed to be designed
better, big compliment to the python developer
crew :-)
Now, our current solution is to disable loading
of other interpreters once python is loaded.
I'm not really glad with this one. Does
anybody know enough of python internals to
hint me some other solution? We tried so far:
- unloading the .so (doesn't work if there
  are global symbols)
- just letting the .so stay in (makes perl
  crash when called)
- forbidding the user to switch languages
  after going to python (which he probably'll
  do anyways ;-), but this is still annoying...)
- not using LTDL_GLOBAL, but this disables
  the use of .so modules (for example time)

Special question: Is there a different way to
get the python modules find the symbols they
need, some magic internal undocumented python
function we could call on the C++ side...?

thanks in advance for any hint,
  Manuel



From awinston at scn.org  Tue Sep 10 14:10:56 2002
From: awinston at scn.org (Alan Winston)
Date: Tue, 10 Sep 2002 11:10:56 -0700
Subject: Pysol Solitaire for Windows wanted
References: <algahc$e8q$1@news7.svr.pol.co.uk> <2MPe9.47925$Jo.5395@rwcrnsc53> <alh591$sqa$1@brokaw.wa.com> <alh9r7$vni$1@brokaw.wa.com> <fzwirhzl.fsf@morpheus.demon.co.uk>
Message-ID: <allcjg$sr9$1@brokaw.wa.com>

> There are bugs in the Windows code for the latest version
> (4.80). IIRC, it also depended on a version of Tk which isn't the
> standard. None of the problems were impossible to solve, but they were
> a mild pain, and the result is *not* the same as the original, as the
> supplied .pyc file is somehow built as a single file from multiple
> source files.

Ah.

> The author is also effectively uncontactable, which makes getting
> fixes included in the source pretty much impossible.
>
> It's a real pain, as PySol is a really good program. If anyone could
> package it up again, I'm sure the result would be of value to a lot of
> people...

The update list included with the file encourages us to wait for version 5.

I think I will go on waiting, using the available ( "Try going to
http://www.gamehippo.com and search for pysol." ) PySol 4.60 in the
meantime.



Thanks,


Alan




From quiteblack at yahoo.com  Mon Sep 23 04:00:57 2002
From: quiteblack at yahoo.com (black)
Date: Mon, 23 Sep 2002 01:00:57 -0700 (PDT)
Subject: os module and sys module~
Message-ID: <20020923080057.955.qmail@web21310.mail.yahoo.com>


Hi, all~

    I couldnt tell the difference between them immediately, and some same method confused me much, for example they both have "path" attribute but what's the difference ?



---------------------------------
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020923/259004a6/attachment.html>

From max at alcyone.com  Fri Sep 27 15:30:42 2002
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 27 Sep 2002 12:30:42 -0700
Subject: file open verification
References: <9eabe547.0209270710.17ca112d@posting.google.com>
Message-ID: <3D94B1E2.77541C55@alcyone.com>

John Howard wrote:

> Situation - I have several files on an apache server that are being
> accessed by
> python programs that are receiving data from html forms.
> 
> q1: How can the status of a file be checked?  If two forms have been
> submitted
> and the data is being added to a common file, how do I know that data
> from both forms are being written correctly to the common file?
> 
> q2: Related to above - really more of a cgi question - does the apache
> server keep all these file accesses separate some way?

No, it's your job to keep them all separate.  Not keeping them separate
will cause disaster.  Use lockfiles.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ An ounce of hypocrisy is worth a pound of ambition.
\__/ Michael Korda
    Crank Dot Net / http://www.crank.net/
 Cranks, crackpots, kooks, & loons on the Net.


From jdavis at empires.org  Tue Sep 24 05:23:17 2002
From: jdavis at empires.org (Jeff Davis)
Date: Tue, 24 Sep 2002 09:23:17 GMT
Subject: cgi post problem
References: <FVRj9.192186$Jo.71059@rwcrnsc53>
Message-ID: <9aWj9.39314$V7.10439005@twister.socal.rr.com>

Hi,

It looks as though you have a lot of code. It might be helpful to try to 
narrow the problem further by removing a lot of the code unrelated to your 
problem, which appears to be that "keys" does not contain any items. After 
you work that out, then put the other code back in.

My first impression is that you have some "weird" form element names 
because they contain the "/" character. That might be perfectly legal, but 
perhaps the cgi.FieldStorage class does not properly handle such element 
names, or perhaps it considers such names erroneous.

Next, it seems that the line "keys = form.keys()" might be out of alignment 
with the rest of the code in it's block. Perhaps it's just the way it 
looks after you copied your code into the post.

Regards,
        Jeff

jano wrote:

> Hi,
> 
> I am trying to post a form to a cgi script.  The form is as follows:
> 
> <form action="http://someurl.com/cgi-bin/updatexml.py" method="POST">
>     <input type="text" name="/Product/Name" size="20" />
>     <input type="text" name="/Product/PartNumber" size="20" />
>     <input type="text" name="/Product/Inventory" size="20" />
>     <input type="text" name="/Product/Cost" size="20" />
>     <input type="text" name="/Product/WholesalePrice" size="20" />
>     <input type="submit" value="Update Records" />
> </form>
> 
> At the other end, the updatexml.py file currently just echoes back the
> form data, as follows (with some wrapper code):
> 
> #!/usr/local/bin/python
> 
> import sys, traceback, cgi
> 
> excText = ""
> 
> 
> def main():
>     global excText
>     try:
>         form = cgi.FieldStorage()
>  keys = form.keys()
> 
>         print "Content-type: text/html\n\n"
>         print("""<?xml version='1.0' encoding='UTF-8'?>
>         <!DOCTYPE html PUBLIC '-//W3C//DTD XHTML 1.1//EN'
> 'http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd'>
>         <html xmlns='http://www.w3.org/1999/xhtml' xml:lang='en'>
>             <head>
>                 <title>CGI exception</title>
>                 <link type='text/css' rel='stylesheet' href='style.css'
>                 /> <script language=\"javascript\" type=\"text/js\"
> href=\"scripts/Netscape4Fix.js\"></script>
>             </head>
>             <body>
> 
>         """)
>         print "<p>** ",keys," **</p>"
>         print "</body></html>"
> 
> 
>     except:
>         print "Content-type: text/html\n\n"
>         import time
>         errtime = '--- '+ time.ctime(time.time()) +' ---\n'
>         errlog = open('cgi_errlog', 'a')
>         errlog.write(errtime)
>         traceback.print_exc(None, errlog)
>         errlog.close()
>         print("""<?xml version='1.0' encoding='UTF-8'?>
>         <!DOCTYPE html PUBLIC '-//W3C//DTD XHTML 1.1//EN'
> 'http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd'>
>         <html xmlns='http://www.w3.org/1999/xhtml' xml:lang='en'>
>             <head>
>                 <title>CGI exception</title>
>                 <link type='text/css' rel='stylesheet' href='style.css'
>                 /> <script language=\"javascript\" type=\"text/js\"
> href=\"scripts/Netscape4.js\"></script>
>             </head>
>             <body>
>             <div id=\"main\">
>         """)
>         print "<p class='error'>Sorry, a problem was encountered running
>         the
> cgi script: </p><br />\n"
>  if excText != "":
>   print "<p class='citation'>", excText, "</p>"
>         print "</div></body></html>"
> 
> 
> if __name__ == "__main__":
>     main()
> 
> ==================
> 
> The output of the script is always an empty [].  Does anybody know what
> the
> problem might be?  In a previous thread, the problem was that the user
> had <input id=""...> instead of using name ="", but that is not the case
> here.
> 
> Thanks for any help.  I have checked usenet and the python faq and used
> google, to no avail.
> 
> jano





From propadovic.nenad at debitel.net  Wed Sep 18 19:54:13 2002
From: propadovic.nenad at debitel.net (Nenad Propadovic)
Date: Thu, 19 Sep 2002 01:54:13 +0200
Subject: Need advice on choosing the technology for intranet based application
References: <3d88fd1e@news.ivm.net> <20020918.191423.706043324.1365@x1-6-00-d0-09-7e-c7-3e.wido3.on.cogeco.ca>
Message-ID: <3d89108d@news.ivm.net>

Such a fast answer. Amazing!
Thank you, Mark. It's my first news-posting, you see :)
>
> Zope would be a good choice. Personally, though, I would stick to a
> combination of Apache and Python/PHP/Perl/etc for a project like this.
> Your project will be fairly small (although the GUI and chart generation
> may be somewhat complex). It can't be too huge if you plan to get it done
> by November in your spare time! :)
I'll have to make it simple :)
> I think Python (with Apache) would be
> a good choice.
Is Apache ...simple... to work with? I know it's standard :)
> One thing, you didn't mention was what languages you have used (you
> mentioned what you wrote, but not what you wrote it in). If you mentioned
> what languages you have worked with in the past could, that could help
> people give you better suggestions.
Ahm, yes that's a point. I've been doing a lot of C, a bit of C++ (don't
know it well), VBA, Ingress/Open Road... But I realy want to do the stuff in
Python.
> It really depends. You may want to generate them as images.  Or maybee
> PDF's? Would PDF be a viable option for you? ReportLAB
> (www.reportlab.com) has a very good PDF generating library for Python.
> I've
> worked with the library and it is amazing.
That sounds great!
Thanks a lot.
Nenad





From zopestoller at thomas-guettler.de  Tue Sep 10 05:04:17 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Tue, 10 Sep 2002 11:04:17 +0200
Subject: run-mailcap in python?
Message-ID: <alkcfb$9ql$07$1@news.t-online.com>

Hi!

On Unix there is a script called run-mailcap, which
opens the corresponding programm with the given file:

run-mailcap foo.html
--> Default browser starts

On windows you can use a double-click with the
explorer or use cygstart from cygwin.

Is there a portable function in python which does this?

E.g.

run("foo.html")
  --> Browser starts

run("foo.doc")
  --> Word starts

thomas



From vvenka1 at hotmail.com  Tue Sep  3 03:26:13 2002
From: vvenka1 at hotmail.com (Venkat Venkataraju)
Date: 3 Sep 2002 00:26:13 -0700
Subject: getch in python
Message-ID: <75ca4e69.0209022326.4cde72e8@posting.google.com>

Hi All

Is there any function that behaves like getch() function in C? 

Venkat


From thoa0025 at mail.usyd.edu.au  Sun Sep 15 21:50:19 2002
From: thoa0025 at mail.usyd.edu.au (:B nerdy)
Date: Mon, 16 Sep 2002 01:50:19 GMT
Subject: [xml] convert funny chars to char entites?
Message-ID: <vNah9.10101$Ee4.23987@news-server.bigpond.net.au>

ive got a string "Il Est N?" and i want to put it into a xml document

is there a quick and easy way to conver the funny characters to char
entities so i can store them?
or is there existing libraries?

cheers




From kmp at NO2SPAMatrium.fsnet.co.uk  Sun Sep 22 08:13:48 2002
From: kmp at NO2SPAMatrium.fsnet.co.uk (Ken Payne)
Date: Sun, 22 Sep 2002 13:13:48 +0100
Subject: Tk Getting a directory name
Message-ID: <amkbso$hre$1@newsg2.svr.pol.co.uk>

HI everyone -

I'm writing a Tk interface to an app and in it I want the user to choose a 
directory. There's an easy way to ask the user to pick a file - 

fileWanted=askopenfilename()

but is there a similar way to ask for a directory?

- Ken




From praveen.patil at silver-software.com  Thu Sep 19 09:43:11 2002
From: praveen.patil at silver-software.com (Praveen Patil)
Date: Thu, 19 Sep 2002 14:43:11 +0100
Subject: How pass array from  C  to python function
In-Reply-To: <NFBBLJFBNMKMLGNLJMFCKENPCCAA.praveen.patil@silver-software.com>
Message-ID: <NFBBLJFBNMKMLGNLJMFCKEOACCAA.praveen.patil@silver-software.com>

Hi ,

I have problem in passing array to python function.
Please help me passing array to python function.

Here is my  'C' program
-----------------------

void  RECEIVE_IL_STATE_S( int Instance , int vital_data[5])
{
    PyObject*arglist;
    PyObject* ret;
    PyObject* mylist;
    int count;
    
    mylist = PyList_New(5);
    for (count=0; count< 5; count++) {
          myint = PyInt_FromLong(vital_data[count]);
          PyList_Append(mylist,myint); 
    }
    arglist = Py_BuildValue("O", mylist);
    ret = PyEval_CallObject(my_callback , arglist);
    Py_DECREF(arglist);
    Py_DECREF(ret);
} 

Here is my Python program
-------------------------

G_Logfile          = None

def TestFunction(a):
    G_Logfile = open('Pytestfile.txt', 'w')
    G_Logfile.write("%d \n"% a[0])
    G_Logfile.write("%d \n"% a[1])
    G_Logfile.close


Cheers,

Praveen.

















[ The information contained in this e-mail is confidential and is intended for the named recipient only. If you are not the named recipient, please notify us by telephone on +44 (0)1249 442 430 immediately, destroy the message and delete it from your computer. Silver Software has taken every reasonable precaution to ensure that any attachment to this e-mail has been checked for viruses. However, we cannot accept liability for any damage sustained as a result of any such software viruses and advise you to carry out your own virus check before opening any attachment. Furthermore, we do not accept responsibility for any change made to this message after it was sent by the sender.]



From tim.one at comcast.net  Sun Sep 29 13:52:04 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 29 Sep 2002 13:52:04 -0400
Subject: ANN: Tim Peters' FixedPoint on SourceForge
In-Reply-To: <an755f$fm8$1@panix1.panix.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEILBHAB.tim.one@comcast.net>

[dougfort]
> This project makes Tim Peters' FixedPoint math class available
> to the community as a SourceForge project.
>
> Check out http://fixedpoint.sourceforge.net

[Aahz]
> Should I change the link on my web site to point at this?

According to me, yes -- I've only been able to make enough time to skim &
ignore the email I've gotten on this, but from the parts I didn't ignore I
think Doug has this very well in hand.  Heck, he might even change it to do
the kind of rounding some people think they expect but are unable to
articulate <wink>.




From theller at python.net  Thu Sep 19 03:01:21 2002
From: theller at python.net (Thomas Heller)
Date: Thu, 19 Sep 2002 09:01:21 +0200
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <mailman.1032329366.6813.python-list@python.org> <j4ptvbaapm.fsf@informatik.hu-berlin.de> <3D890FB0.4030408@something.invalid> <m3admeo4u1.fsf@mira.informatik.hu-berlin.de>
Message-ID: <ambso1$4g8vh$1@ID-59885.news.dfncis.de>

Martin v. Loewis wrote:
> Greg Ewing <see_reply_address at something.invalid> writes:
> 
> 
>>So, the only way to find out whether a particular
>>extension is safe to use in Stackless is to try
>>it. 
> 
> 
> That is not true. It is more reliable to inspect the source code.
> 
> In general, it is very easy to do so: if there are no callbacks to
> Python, the module is stackless-safe.

So win32all probably wxPython are incompatible with stackless?

-- 
Thomas



From javajohn at cox.net  Thu Sep 12 10:25:11 2002
From: javajohn at cox.net (John Waycott)
Date: Thu, 12 Sep 2002 14:25:11 GMT
Subject: Coding Style: Quotes
Message-ID: <bt1g9.17429$Pf7.590136@news1.west.cox.net>

I'm curious if others have adopted any standards for choice of single vs.
double quotation marks to delimit strings. A look through the standard
library reveals the choice of one over the other is rather arbitrary. I
suspect it really makes no difference, but the question has come up during a
code review.

-- John Waycott




From fb at frank-buss.de  Sat Sep 14 06:58:33 2002
From: fb at frank-buss.de (Frank Buss)
Date: Sat, 14 Sep 2002 10:58:33 +0000 (UTC)
Subject: question about dictionary type..
References: <alt22f$pf8$1@newsreader.mailgate.org>
Message-ID: <alv4op$m67$1@newsreader2.netcologne.de>

eugene kim <eugene1977 at hotmail.com> wrote:

> i'd like to have a dictionary which looks like
> table = { category1 : { category2 : { category3 : 'garbage value' } } }
> 
> category1 has many category2
> category2 has many category3
> 
> i expected
> table[ 'computer' ][ 'language' ][ 'python' ] = 0
> to create {computer : { language : { python :0 }}}
> table[ 'computer' ][ 'language' ][ 'java' ] = 0
> to becomes { 'computer' : { 'language' : { 'python' :0 , 'java' :0 }}}

If you don't insist on your syntax, a little function is all you need:

def add(tree, path, value):
  start=tree
  for key in path[:-1]:
    if start.has_key(key):
      start = start[key]
    else:
      start[key] = start = {}
  start[path[-1]] = value

You can use it like this, with tuples and lists:

>>> tree={}
>>> add(tree, ('computer', 'language', 'python'), 'cool!')
>>> add(tree, ['computer', 'language', 'java'], 'cumbersome')
>>> tree
{'computer': {'language': {'python': 'cool!', 'java': 'cumbersome'}}}

Interesting is the line "start[key] = start = {}", because first
"start[key] = {}" is evaluated, then "start = {}" (with "{}" being the 
same object as in the first assignment). You can check this with "dis", as 
I've read in another posting in this newsgroup:

>>> def f(): a=b=c
...
>>> import dis
>>> dis.dis(f)
          0 SET_LINENO               1

          3 SET_LINENO               1
          6 LOAD_GLOBAL              0 (c)
          9 DUP_TOP
         10 STORE_FAST               0 (a)
         13 STORE_FAST               1 (b)
         16 LOAD_CONST               0 (None)
         19 RETURN_VALUE

-- 
Frank Bu?, fb at frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de


From wolfson at midway.uchicago.edu  Sun Sep 22 16:25:44 2002
From: wolfson at midway.uchicago.edu (Ben Wolfson)
Date: Sun, 22 Sep 2002 20:25:44 GMT
Subject: sys.ps2 assignment has no effect
References: <uos1np8t53k396@corp.supernews.com> <3d8e0aae@news.sentex.net> <uos5dis0ogvc6f@corp.supernews.com>
Message-ID: <gp9sousalc7eq8olncn30m0312648ehhms@4ax.com>

On Sun, 22 Sep 2002 12:12:56 -0700, "Lance" <lbrannma at cablespeed.com>
wrote:

>print '...' works correctly, showing ...

Well?  Don't keep us in suspense, what does it show?

-- 
BTR
You're going to set me up as a kind of slovenly attached pig that
Jack Kornfeld can slice down in his violent zen compassion?
 -- Larry Block


From skip at pobox.com  Wed Sep 11 09:22:08 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 11 Sep 2002 08:22:08 -0500
Subject: Large file support and stat vs stat64 in extension modules
In-Reply-To: <pan.2002.09.10.16.03.39.756929.4320@remove.Stanford.EDU>
References: <pan.2002.09.10.16.03.39.756929.4320@remove.Stanford.EDU>
Message-ID: <15743.17280.884457.36169@12-248-11-90.client.attbi.com>

If I simply use the stat() system call, I get an EOVERFLOW error
(OSError: [Errno 75] Value too large for defined data type).  It seems
that I should be using the newer stat64() call, but:

Did you #include <sys/stat.h>?

    Ben> 1.  posixmodule.c appears just to use plain old stat().  I cannot
    Ben>     figure out why I get errors and it works fine.  Quite a
    Ben>     mystery.

In /usr/include/sys/stat.h on my Mandrake 8.1 system (very similar to your
RH 7.2 I believe), stat is #defined as stat64 if __USE_FILE_OFFSET64 is
defined, which it probably is if your system supports POSIX large files.

    Ben> 2.  I don't know what #if defined's to put in to choose between
    Ben>     stat and stat64 correctly.

Shouldn't need anything.

    Ben> Does anyone know how to correctly stat a file within an extension
    Ben> module?  Why does:

    Ben>     struct stat st;
    Ben>     res = stat("bigfile", &st)

    Ben> not work for me, but in posixmodule.c the same thing (apparently)
    Ben> is done and works fine?

Note that in posixmodule.c references are made to a HAVE_LARGEFILE_SUPPORT
macro.  When #defined, it causes the output struct to be interpreted
differently (the sizes of some fields change).

You might scan through your configure.in file.  If "long long" support is
present and sizeof(off_t) is greater than sizeof(long) and sizeof(long long)
is greater than or equal to sizeof(off_t), large file support is assumed.

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From nospam at bigfoot.com  Wed Sep 11 08:13:33 2002
From: nospam at bigfoot.com (Gillou)
Date: Wed, 11 Sep 2002 14:13:33 +0200
Subject: Save session in python interactive console
References: <a50631f6.0209091334.7d59476c@posting.google.com>
Message-ID: <alnb9q$2trg$1@norfair.nerim.net>

You ca have a look at the standalone ZODB (search from google) if you want
to play with (almost transparent) persistant objects.

--Gilles

"newgene" <newgene at bigfoot.com> a ?crit dans le message de news:
a50631f6.0209091334.7d59476c at posting.google.com...
> Hi, group,
>     After I was working at python interactive console(activepython
> v2.2 at Win2000 platform), I am wondering if possible to save the
> current session, so that I can load it again and continue the work
> next time. Just like R or Matlab, etc. Although python is a
> programming language, one of the best features I like is its
> interactive console. By writting my own module, I can make it as my
> personal working console. If saving session is possible, it will be
> much more convinient.
>     The pickle module provided by python can save any object,but how
> can I save all objects in current session?
>
> Thanks in advance
>
>
> CW




From fredrik at pythonware.com  Wed Sep 18 14:23:56 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 18 Sep 2002 18:23:56 GMT
Subject: Tkinter:select directory dialog?
References: <20020918105332.00910.00007029@mb-ch.aol.com> <mailman.1032362543.493.python-list@python.org>
Message-ID: <0x3i9.13118$HY3.2916659@newsc.telia.net>

Martin Franklin wrote:

> > A nice graphical "choose a directory"  Dialog would help.
> > Does such a thing exist?
>
>
> from tkCommonDialog import Dialog
> class Chooser(Dialog):
>     command = "tk_chooseDirectory"
>     def _fixresult(self, widget, result):
>         if result:
>             self.options["initialdir"] = result
>         self.directory = result # compatibility
>         return result
>
> def askdirectory(**options):
>     return apply(Chooser, (), options).show()

(hmm.  I think I've seen that code before...)

the askdirectory function and a Directory class (same as
the Chooser above) was added to the tkFileDialog module
in Python 2.2:

    from tkFileDialog import askdirectory

    directory = askdirectory(
        title="select directory",
        mustexist=1
    )

    if not directory:
        print "cancelled"
    else:
        print "selected", directory

</F>

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




From marklists at mceahern.com  Sat Sep 14 15:34:22 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sat, 14 Sep 2002 14:34:22 -0500
Subject: Converting a string to time - simple task but don't know how to do it!
In-Reply-To: <7A6D2E53461F2BF2.AC83A3D5E1A01468.48CC780BD6530537@lp.airnews.net>
Message-ID: <JHEOKEOOLIGLDHCMAHMOAELGDOAA.marklists@mceahern.com>

Try:

  import time
  s = "<<your time string>>"
  t = time.strptime(s)

I think strptime is only supported on Unix?

So if you get an AttributeError when you try that, your best best is to use
the mx.DateTime package (a google search should find that).

mx.DateTime works like this:

  import mx.DateTime
  d = mx.DateTime.DateTimeFrom(your_time_string)
  d2 = mx.DateTime.now()
  diff = d2 - d1
  print diff

// m

-




From aglyportREMOvethispart at nospam.yahoo.com  Thu Sep  5 21:12:39 2002
From: aglyportREMOvethispart at nospam.yahoo.com (Anton Graph)
Date: Thu, 05 Sep 2002 18:12:39 -0700
Subject: Memory fault - core dumped   while compiling python 2.2.1 on
  SCO
References: <3D757EFA.9050401@nospam.yahoo.com> <yu99sn0qqw1r.fsf@europa.research.att.com> <m3bs7etem9.fsf@mira.informatik.hu-berlin.de> <3D765A05.6030004@nospam.yahoo.com> <j48z2h61j9.fsf@informatik.hu-berlin.de> <3D768753.8030302@nospam.yahoo.com> <m3n0qx8j3w.fsf@mira.informatik.hu-berlin.de> <3D76A139.1030304@nospam.yahoo.com> <al69pn$sg$0@216.39.172.122> <ff3c0649.0209051224.3da59b27@posting.google.com>
Message-ID: <3D780107.9090907@nospam.yahoo.com>

J. L. Schilling wrote:
> bokr at oz.net (Bengt Richter) wrote in message 
> 
>>>[...]
>>
>>How about
>>    cc -v
>>?
> 
> 
> Just for the record, cc -V is the option on the SCO compiler
> to report its version.  It should show something like this
> on an OSR 5.0.5 system:
> 
> 217$ /bin/cc -V
> SCO UNIX Development System  Release 5.1.1A 27Jul98
^^^^^^^^^^^^^^^^^^ 5.1.1A is what I have also.



From robin at jessikat.fsnet.co.uk  Sat Sep 21 04:50:10 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 21 Sep 2002 09:50:10 +0100
Subject: Decimal arithmetic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org>
 <amabpb$5rm$1@news.nems.noaa.gov> <3D8B1537.4010803@thinkware.se>
 <ad052e5c.0209201128.1a53108@posting.google.com>
 <amgcgf$5e1fd$1@ID-125932.news.dfncis.de>
 <7xn0qcfafp.fsf@ruckus.brouhaha.com>
 <amgpm4$5si22$1@ID-125932.news.dfncis.de>
 <7xwupf52jx.fsf@ruckus.brouhaha.com>
Message-ID: <TXOVyjACLDj9EwHM@jessikat.fsnet.co.uk>

In article <7xwupf52jx.fsf at ruckus.brouhaha.com>, Paul Rubin <phr-
n2002b at NOSPAMnightsong.com> writes
>Christopher Browne <cbbrowne at acm.org> writes:
>> REALITY CHECK for a moment.
>> 
>> What do you think A Real Live bank uses to compute interest?
>
>If I had to guess, I'd guess they use floating point.
>
>I do know that financial modelling software I worked on that was sold
>to banks used floating point.  Also, spreadsheets use floating point.
>
>But it's possible (likely even) that some bookkeeping applications use
>BCD.  One that I worked on actually used exact, multi-precision
>rational arithmetic, but I think that's unusual.
>
>Note that the x86/x87 still supports BCD arithmetic in microcode.  In
>reality, though, programs that use BCD to keep track of money don't do
>all that much arithmetic, so software emulation is plenty fast enough.
Well I also worked in finance/research and I agree with Paul and
Christopher. Most trading/research software uses FP, most transactions
are computed using BCD.

The thing about most real financial instruments is that they use fairly
complicated sets of rules to calculate the actual interest and since the
rules evolved before modern computers are mostly computable using
fingers and counting. In particular almost all the calculations have to
be with pre-specified precision and rounding rules.

EG a bond pays fixed amounts of money at pre-specified intervals, no
calculations are involved at all. Of course when I want to compute the
yield or npv I need to do some arithmetic and FP is usually easiest and
faster as well.

Anyhow it seems to me that BCD is pretty good only for additive type
calculations. I can't see that BCD is any better than FP at representing
1/3 unless we go fully rational, but that's another story.
-- 
Robin Becker


From gerhard.haering at gmx.de  Fri Sep 20 01:50:09 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Fri, 20 Sep 2002 07:50:09 +0200
Subject: newbie seeks advise.
In-Reply-To: <77c84db2.0209192137.51167ebb@posting.google.com>
References: <77c84db2.0209192137.51167ebb@posting.google.com>
Message-ID: <20020920055009.GA2163@lilith.ghaering.test>

* akirasugiura <kodokuchef at yahoo.com> [2002-09-19 22:37 -0700]:
> Hi,
> 
> I've been learning python for about a month. I had no prior knowledge
> of programming.
> 
> My question is, I have a tuple which is, data = (69, 40, 74, 41, 56,
> 57, 38, 58, 26, 55). How I can make a function that picks up a largest
> number from this "data" tuple. If "data" was a  'List' I would be able
> to use "sort" but since sort doesn't  work with tuple. I am stuck
> here.

max(data)

Now that was too simple, wasn't it?

If you want to program a max() function yourself, you can convert the
tuple to a list using list(data), then sorting it, then finding the
largest element yourself.

-- Gerhard



From syver-en+usenet at online.no  Tue Sep 17 06:37:52 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Tue, 17 Sep 2002 10:37:52 GMT
Subject: Book for win32?
References: <am430k$9l3$03$1@news.t-online.com> <uhegq710h.fsf@online.no> <3D868FA2.4040503@nospam.free.fr>
Message-ID: <it146gwl.fsf@online.no>

laotseu <bdesth at nospam.free.fr> writes:
 
> I'd just add one advice : if you're not familiar with the w32 API, go
> and get yourself a good reference of it ! (I think you can get the
> win32 API ref on borland's site, with the free version of bcc55).
> 

Yes, that is good advice. But isn't it just as well to use
microsoft's own version at http://msdn.microsoft.com. 

-- 

Vennlig hilsen 

Syver Enstad


From gerhard.haering at gmx.de  Fri Sep 20 14:56:13 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 20 Sep 2002 18:56:13 GMT
Subject: Newbie questions: write permission, interpreter minimizing
References: <mailman.1032538785.8777.python-list@python.org>
Message-ID: <slrnaomrqd.j7.gerhard.haering@lilith.my-fqdn.de>

pcarey at lexmark.com wrote in comp.lang.python:
> I use PythonCardPrototype to make UI's.
> When launching these apps (double-clicking),
> the interpreter launches (as it should) and remains in the
> background (as it should)... but is there any way to automatically
> minimize the interpreter?

You can give your script a .pyw instead of a .py extension, which will
result in the script being launched with pythonw.exe intead of
python.exe. This way, you have a normal Windows GUI app, without any
console window. OTOH it's harder to track exceptions, but AFAIK
wxPython (and thus maybe PythonCard, too) has a way to show exceptions
in a GUI window.

-- Gerhard


From christophe.delord at free.fr  Mon Sep  9 16:56:54 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Mon, 9 Sep 2002 22:56:54 +0200
Subject: a re question
References: <pan.2002.09.09.16.10.38.482129.20940@presidency.com>
Message-ID: <20020909225654.44e16058.christophe.delord@free.fr>

match applies the regular expression at the beginning of the string. Try 'search' instead:

>>> s='001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23'
>>> p = re.compile(r'(\d{9,} {3,}\d)')
>>> print p.match(s)
None
>>> print p.search(s)
<_sre.SRE_Match object at 0x400aed20>
>>> print p.search(s).groups()
('123456789   7',)


You can also use a different expression:

>>> p = re.compile(r'.*(\d{9,} {3,}\d)')
>>> print p.match(s)
<_sre.SRE_Match object at 0x400aed20>
>>> print p.match(s).groups()
('123456789   7',)




Christophe.


On Mon, 09 Sep 2002 16:10:39 -0400
Rajarshi Guha <rajarshi at presidency.com> wrote:

> Hi,
>   I have a file with lines of the format:
> 
> 001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23
> 001 Xya FGh   143557789   7 100 09/05/2002 20:23:23
> 
> I am trying to extract the 9 digit field and the single digit field
> immediatley after that.
> 
> When I use Visual Regexp to try out the regexp 
> 
> (\d{9,} {3,}\d)
> 
> it highlights the 2 fields exactly. 
> 
> But when I use the following Python code I get None:
> 
> >> s='001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23'
> >> p = re.compile(r'(\d{9,} {3,}\d)')
> >> print p.match(s)
> >> None
> 
> Could anybody point out where I'm going wrong?
> 
> Thanks,


-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From loewis at informatik.hu-berlin.de  Wed Sep  4 14:38:30 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 04 Sep 2002 20:38:30 +0200
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com>
Message-ID: <j4d6rt61nd.fsf@informatik.hu-berlin.de>

Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:

> >Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:
> >
> >> So why hasn't this been extended when nested scopes were introduced,
> >> e.g. by reusing the global directive or some other one? 
> >
> >I think the main reason is that nobody could propose an acceptable
> >syntax. Reusing the global directive does not work.
> 
> Can you explain why it does not work?

Consider

x = 1

def foo():
  x = 2
  def bar():
    x = 3
    def baz():
      global x
      x = 4

If you invoke baz(), which of the variables is changed? How do you
denote that you want to change the others?

Regards,
Martin



From wlfraed at ix.netcom.com  Tue Sep 24 22:04:03 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Tue, 24 Sep 2002 19:04:03 -0700
Subject: Pythonwin (win32all build 148) on Python 2.2.1 on Win2K, bogus syntax e
References: <u47k9.46401$V7.11091779@twister.socal.rr.com>
Message-ID: <ji5rma.4v4.ln@ix.netcom.com>

A. Dimsdale Fetus A. Moser fed this fish to the penguins on Tuesday 24 
September 2002 05:04 pm:

> I am running Pythonwin (win32all build 148) on Windows 2000
> Professional with Python 2.2.1. When I attempt to run a Python script
> in it, it halts the running of the script on supposed syntax errors,
> which are completely bogus. Examples:
> First line of our (Aaron L. Dimsdale and Alex S. Moser) first
> semi-significant Python script, called fight:
> """fight.py 5.0 -- crappy primitive fighting game by alex s. moser &
> aaron l. dimsdale"""
>         ^^^ (note: no linebreak in that line)
> """if one of the players does not say "atk" or "rest" when asked, they
> will \ forfeit their turn"""
> 
> Pythonwin barfed on the quote marks marked by carets, returning:
> "Failed to run script - syntax error - invalid syntax".
> I removed those quote marks and the quote marks at the beginning of
> the second documentation line, and left those two lines as one string.

        Is that supposed to be "doc-string"? (since it isn't assigned to 
anything?)... Try:

"""
fight.py 5.0 -- crappy primitive fighting game by alex s. moser & ...

if one of the players does not say "atk" or "rest" when asked,
they will forfeit their turn
"""

        Triple quotes work to delimit multiple line strings.


> Then, several lines later:
> 
> p1char = str(raw_input("Player 1, choose your character class: "))

        I'd almost suspect a misplaced quote mark somewhere, but I presume 
Pythonwin is color-coding properly...

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From akakakhel at attbi.com  Mon Sep  9 20:54:52 2002
From: akakakhel at attbi.com (Ali K)
Date: Tue, 10 Sep 2002 00:54:52 GMT
Subject: IDLE
Message-ID: <wpbf9.264320$aA.45691@sccrnsc02>

Why won't my IDLE work?




From brueckd at tbye.com  Sun Sep 15 23:06:40 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Sun, 15 Sep 2002 20:06:40 -0700 (PDT)
Subject: can wxpython do this?
In-Reply-To: <am3d0b$jj$1@news01bc.so-net.ne.jp>
Message-ID: <Pine.LNX.4.44.0209152005070.19839-100000@ns0.tbye.com>

On Mon, 16 Sep 2002, Erick wrote:

> Thank you very much Lothar.
> 
> 
> > There is no problem to setup an installer easily that takes
> > Python,WXPysthon and whatever you need. It simply blowes up the
> > installer package size.
> 
> Isn't this the way to distribute wxpython apps?

Yeah, it is. He meant "blows up" as in "bloats" rather than 
"explodes/breaks".

-Dave




From nanotech at europa.com  Fri Sep 13 12:29:10 2002
From: nanotech at europa.com (czrpb)
Date: Fri, 13 Sep 2002 09:29:10 -0700 (PDT)
Subject: Help with regular expression using findall and .*?
In-Reply-To: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220D18@hendrix.empolisuk.com>
Message-ID: <Pine.BSO.4.21.0209130905130.6866-100000@thetics.europa.com>

Harvey:

Great thanks!! And thanks for sticking to my question's requirements. <wink!>

Ok, this is what we thought around here. But what I do not understand is why any backtracking data is being kept? The '?' in '.*?' means it is non-greedy right? When would backtracking ever occur using '.*?'? What am I missing?

<<q


On Fri, 13 Sep 2002, Harvey Thomas wrote:

> czrpb wrote
> > 
> > Could anyone help out with rewriting (still using regular expressions)
> > the following so that it does not cause an exception:
> > 
> > import re
> > 
> > s1=('macro\n'+'a'*200+'\norcam\n')*10
> > s2=('macro\n'+'a'*20000+'\norcam\n')*10
> > 
> > p=re.compile(r'macro.*?orcam',re.DOTALL)
> > 
> > for x in re.findall(p,s1):
> >     print x
> > 
> > for x in re.findall(p,s2):
> >     print x
> > 
> > thanks!! Quentin Crain
> > 
> 
> You need to be very careful about using .*? as the engine "only" allows 10,000 backtracks
> 
> Try this
> 
> p = re.compile('macro(?:[^o]+|o(?!rcam))*orcam')
> for x in p.findall(s2):
>     print x
> 
> HTH
> 
> Harvey
> 




From gerhard.haering at opus-gmbh.net  Wed Sep  4 10:55:05 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 4 Sep 2002 14:55:05 GMT
Subject: How to force built-in commands over imported.
References: <3D761C39.4020802@akst.tn.tudelft.nl>
Message-ID: <slrnanc7jg.24k.gerhard.haering@haering.opus-gmbh.net>

Maurice van de Rijzen wrote:
> Dear all.
> I'm a newbie on the field of python, therefore the answer to my question 
> may be simple.
> There is a built-in commans called open(fileName). There is also a 
> open() command in the os-module open which needs more than the fileName. 
>   Originally I didn't import the os-module and everything worked fine. 
> However, when I imported the os-module I received an error that open 
> needed more arguments. How can I force to take the built-in open() 
> instead of the open() in the os-module.

The technical answer is to use __builtins__["open"].

A good advice is to avoid the use of "from ... import *" in most cases.
Just "import os", then use the desired function by prefixing it with the
module name, like "os.open".

-- Gerhard


From strat_addict at yahoo.com  Sat Sep 21 21:51:24 2002
From: strat_addict at yahoo.com (strat_addict)
Date: 21 Sep 2002 18:51:24 -0700
Subject: COM, Word problem
Message-ID: <e00ba67.0209211751.1411ad7c@posting.google.com>

hello all,
I'm getting a strange error when calling Word with Dispatch. I just
noticed it, since I haven't played around with Word in quite some
time. Anyway, here is what is getting spit back:
>>> from win32com.client.dynamic import Dispatch
>>> word = Dispatch("Word.Application")
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
  File "C:\PYTHON21\win32com\client\dynamic.py", line 84, in Dispatch
    IDispatch, userName =
_GetGoodDispatchAndUserName(IDispatch,userName,clsctx)
  File "C:\PYTHON21\win32com\client\dynamic.py", line 81, in
_GetGoodDispatchAndUserName
    return (_GetGoodDispatch(IDispatch, clsctx), userName)
  File "C:\PYTHON21\win32com\client\dynamic.py", line 72, in
_GetGoodDispatch
    IDispatch = pythoncom.CoCreateInstance(IDispatch, None, clsctx,
pythoncom.IID_IDispatch)
com_error: (-2147417846, 'The message filter indicated that the
application is busy.', None, None)

I've seen some book code that doesn't use dynamic in the import
statement and tried this to no avail. All other applications work
(Excel, Access, even Vim)
I've looked at some bug reports on the board that are similar, but I
can't even create an instance. Has anyone had a similar problem? Just
trying to get a sense of what I should try next.

All help is appreciated,
strat addict


From aleax at aleax.it  Fri Sep 27 03:34:40 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 27 Sep 2002 07:34:40 GMT
Subject: Silly question about ungzipping a file
References: <7396d2b2.0209261311.1fdd2d3a@posting.google.com>
Message-ID: <kSTk9.149109$ub2.3289741@news1.tin.it>

Lemniscate wrote:
        ...
>>>> import gzip
>>>> myFile = gzip.open('LL_tmpl.gz')
>>>> file('output.txt', 'w').write(myFile.read())
>>>> 
> 
> Now, is there any way to do this so that less memory is used?  I mean,

import shutil
shutil.copyfileobj(myFile, open('output.txt', 'w'))

> 129MB.  I'm sure I'm just missing something simple, but why is that?

I think the reason you're missing this simple something is that you
are not entirely familiar with Python's rich standard library -- a
defect common to us all (I don't recall _every_ useful module either:-).

A partial help is /F's "Python Standard Library" (O'Reilly) -- it
won't magically infuse you with knowledge of every such module, but
by giving at least one example script for each module in the
library (all in a remarkably concise 280 pages!) it makes it quite
a bit more likely that more modules will stay in your subconscious
memory.  Python's online reference, Beazley's New Riders book, my
Nutshell when it comes (the delay ain't my fault THIS time - the
complete, technically-revised draft is currently in the editor's
hands!-), each give a different perspective on Python's standard
library, and enhance the likelihood that you'll know where to find
excellent code ripe for reuse.  Probably the best single piece of
advice is the one in the Python online docs -- "sleep with the
Library Reference under your pillow" (and use a thin pillow, so that
the knowledge has an easier time filtering up into your brain:-).


Alex



From vtail at yandex.ru  Mon Sep 30 12:56:23 2002
From: vtail at yandex.ru (vtail)
Date: 30 Sep 2002 09:56:23 -0700
Subject: McMillian installer + win32com.client.constants - how to use them together?
Message-ID: <9ff6b66.0209300856.5fcdf961@posting.google.com>

Greetings,

can anybody please explain me the Right Way to use McMillian installer
with application that uses constants from MSWord library?

I've generate them with win32com/client/makepy.py and use them in my
code via
====>
import win32com.client
...
a = win32com.client.constants.wdBlack  #  application crashed here
...
<===

When I build .exe from my applications it is crashed on marked line
with AttributeError exception.

I've tried to study this subject myself, but now I'm totally confused
with all this gencache.EnsureModule(), prog_id, CLSID etc. I can't
find any clear documentation on the subject.

I'll greatly appreciated any help on this, even RTFM links ;).

Victor.


From debl2nonosabsolutelynospam at bellatlantic.net  Mon Sep  9 21:48:51 2002
From: debl2nonosabsolutelynospam at bellatlantic.net (David Lees)
Date: Tue, 10 Sep 2002 01:48:51 GMT
Subject: IDLE
References: <wpbf9.264320$aA.45691@sccrnsc02>
Message-ID: <3D7D4F84.318CDB0F@bellatlantic.net>

This is a really great group, with lots of helpful folks, but few of
them are mind readers :)  Why don't you describe your hardware, OS,
Python distribution, a description of how you installed and exactly what
you mean when you say IDLE does not work.

david lees


Ali K wrote:
> 
> Why won't my IDLE work?


From tim.one at comcast.net  Thu Sep 19 00:00:58 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 19 Sep 2002 00:00:58 -0400
Subject: Remarks to Python 2.3
In-Reply-To: <20020918222421.328088eb.use-net@schabi.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCCECIBFAB.tim.one@comcast.net>

[Markus Schaber, on 2.3 features]
> ...
> - Mutable sets support some in-place operations - should those, just as
> the in place operations on lists, like sort(), return None instead of
> the new list to make clear that they are in place operations and for the
> sake of consistency?

Yes, and that's already been done.  The docs may be lagging here:  be a
prince and submit a doc patch!




From jhndnn at yahoo.com  Thu Sep 26 19:46:14 2002
From: jhndnn at yahoo.com (John Dunn)
Date: 26 Sep 2002 16:46:14 -0700
Subject: Changing global variables in embedding Python
Message-ID: <2cc96dc7.0209261546.22f71021@posting.google.com>

Hello-

I have embedding Python in a C++ application. All is well, except I
cannot figure out how to modify globals via Python script. After I
load in my script files, I do this-

module = PyImport_AddModule("<embed>");
dict =  PyModule_GetDict(module); 
PyDict_SetItemString( dict, "__builtins__", PyEval_GetBuiltins());
PyObject* args = PyTuple_New(1);

... fill out arguments to function here....

PyObject* func = PyDict_GetItemString( dict, func_name );
static PyObject* test = PyString_FromString( "test string" );
PyDict_SetItemString( dict, "test", test );      
PyObject* result = check( PyEval_CallObject( func, args), 0 ); 

The Python function can access test fine, but it can't modify test. In
the following code, the first line works properly, but the second will
fail with the error 'local variable 'test' referenced before
assignment'. The message object is an argument to the function.

message.print( test )
test = "test2"

Any idea what I might be doing wrong?

Thanks for any help-

John
--
John Dunn
Peak Audio, a division of Cirrus Logic Inc.
http://www.peakaudio.com


From manuel at hendel.net  Wed Sep 25 04:10:11 2002
From: manuel at hendel.net (Manuel Hendel)
Date: Wed, 25 Sep 2002 10:10:11 +0200
Subject: problems defining an empty list
In-Reply-To: <200209250100.31255.shalehperry@attbi.com>
References: <20020925070654.GB57911@partagas.as.de.cw.net> <200209250040.12209.shalehperry@attbi.com> <20020925075416.GC57911@partagas.as.de.cw.net> <200209250100.31255.shalehperry@attbi.com>
Message-ID: <20020925081011.GD57911@partagas.as.de.cw.net>

Thanks a lot, that worked fine.

Manuel

On Wed, Sep 25, 2002 at 01:00:31AM -0700, Sean 'Shaleh' Perry wrote:
> On Wednesday 25 September 2002 00:54, you wrote:
> > On Wed, Sep 25, 2002 at 12:40:12AM -0700, Sean 'Shaleh' Perry wrote:
> > > it sounds like you have a hidden piece of whitespace.  Perhaps the code
> > > is indented when it should not be?  If the code is not in a loop, class,
> > > if structure, etc then it should not be indented.
> >
> > Actually it is in a ... structure.
> > [....]
> > # empty list for lines per all domains
> > linesperdomains = []
> >
> > for domain in domains:
> >     # empty list for lines
> >     domainlines = []
> >     for neededline in neededlines:
> >         if domain == neededline[0]:
> >             domainlines.append(neededline)
> >     print len(domainlines)
> >
> >     # empty list for lines per domain
> >     linesperdomain = []
> >     linesperdomain = [domain, domainlines]
> >     print linesperdomain
> > [....]
> >
> > This ist just a part of the script.
> >
> 
> One of python's annoyances is it treats a tab and a space as unique indent 
> levels.  So if you happen to use a space everywhere but one line or vice 
> versa that can mess things up.
> 
> I meant to comment on something else too.
> 
> linesperdomain = [] is not needed since you initialize it to a new list on the 
> next line.
> 
> 

-- 
Please don't lie to me, unless you're absolutely sure I'll never find out the 
truth. 
-Ashleigh Brilliant 
(contributed by Chris Johnston) 



From aleax at aleax.it  Fri Sep 20 10:32:41 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 14:32:41 GMT
Subject: implementing reduce protocol in C
References: <3D8A7DA1.7090506@hotmail.com> <62Ai9.136323$pX1.4852975@news2.tin.it> <3D8AF51D.5040305@hotmail.com>
Message-ID: <dkGi9.138788$pX1.4922161@news2.tin.it>

Tom Widrick wrote:
        ...
> static PyObject *
> proto_reduce(self)
> HyProtoObject *self;

I recommend you use "ANSI C" style as a habit:

static PyObject*
proto_reduce(HyProtoObject* self)

it may not matter here, but one day you'll be glad you got into
this habit.  And since Python does now require an ANSI/ISO
compatible C compiler, there's no downside.

> {
> PyObject *args;
> PyObject *result;
> 
> args = PyTuple_New(5);
> if(args == NULL)
> return NULL;
> 
> PyTuple_SET_ITEM(args, 0, self->po_name);
> Py_INCREF(self->po_name);
> PyTuple_SET_ITEM(args, 1, self->po_parents);
> Py_INCREF(self->po_parents);
> PyTuple_SET_ITEM(args, 2, self->po_children);
> Py_INCREF(self->po_children);
> PyTuple_SET_ITEM(args, 3, self->po_doc);
> Py_INCREF(self->po_doc);
> PyTuple_SET_ITEM(args, 4, self->po_dict);
> Py_INCREF(self->po_dict);
> 
> result = Py_BuildValue("(OOO)", self->ob_type, Py_None, args);
> Py_DECREF(args);
> return result;
> };
> 
> Any better?

Much better, I think.  I believe you can save the last
Py_DECREF by using "(OON)" as the first arg to Py_BuildValue,
telling it to steal a reference to the third argument, but
that's not crucial.


Back to your main problem however -- things are NOT rosy.
Check out:
http://mail.python.org/pipermail/python-bugs-list/2002-March/010707.html
If I read this correctly, the case where __reduce__ returns a
second argument of None is being deprecated and is not well
supported anyway.  Thus, you probably want to avoid it.  I think
this makes the rest of the discussion moot.

Rather, you probably want to expose a factory function to be
passed as the first item in the tuple returned by __reduce__
(you may then want to return a 2-items tuple rather than a
3-items one from __reduce__, perhaps).  Just make sure said
first item exposes an attribute named __safe_for_unpickling__ 
with a true value, or register it as a "safe constructor" with
module copy_reg.

(As an aside -- this is the kind of tasks that I usually handle
with a PyRun_String call in my C-coded-module's init function:
they're SO much easier to express in Python, and they're one-off,
thus performance-insensitive tasks anyway; however, I've been
criticized by important and knowledgeable Pythonistas for this
habit -- didn't really understand the reason for the criticism,
but, anyway, use this hint at your own risk:-).


Alex



From tchur at optushome.com.au  Thu Sep 19 20:32:46 2002
From: tchur at optushome.com.au (Tim Churches)
Date: Fri, 20 Sep 2002 10:32:46 +1000
Subject: Need advice on choosing the technology for intranet based 
 application
References: <3d88fd1e@news.ivm.net> <uokpn32h8dhqfa@corp.supernews.com>
Message-ID: <3D8A6CAE.1C925928@optushome.com.au>

"A.M. Kuchling" wrote:
> 
> In article <3d88fd1e at news.ivm.net>, Nenad Propadovic wrote:
> > -I assume the simplest way to present a graph/chart kind of graphic is to
> > create a jpg/giff/tiff on the server side, and just show it on the client?
> 
> Yes.  PIL, the Python Imaging Library, would be one possibility; so is
> the gd module.  For example, the pricing charts on the pages at
> http://www.mems-exchange.org/catalog/ are generated from Quixote using
> gd.  PDF likely isn't helpful, because browsers won't display PDF
> images inlined in a web page, though it's certainly well-suited for
> multi-page reports that need to be printed.

Yes, but the ReportLab tools can produce very nice raster images as well
as PDFs.

The R statistical package, embedded in Python via RPy (see Walter
Moreira's recent message on yhis list
regarding this) is also an excellent way to produce very sophisticated
graphs which can then be
streamed back to the user's browser.

Tim C
> 
> --amk
> --
> http://mail.python.org/mailman/listinfo/python-list



From fperez528 at yahoo.com  Tue Sep 17 13:07:40 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 17 Sep 2002 11:07:40 -0600
Subject: Python equivalent to perldoc?
References: <Olvh9.7898$Lg2.1389903@news2.news.adelphia.net> <mailman.1032226040.2391.python-list@python.org> <9ywh9.7955$Lg2.1406050@news2.news.adelphia.net>
Message-ID: <am7ngs$rk9$1@peabody.colorado.edu>

Bob X wrote:

> Skip Montanaro wrote:
>>     Bob> In Perl you can run "perldoc warnings" and get the definition of
>>     Bob> what "warnings" does. Is there a equivalent in Python?
>> 
>> Try "pydoc warnings".
>> 
> What a funny guy!  :-)

$ pydoc warnings
NAME
    warnings - Python part of the warnings subsystem.

FILE
    /usr/lib/python2.2/warnings.py

CLASSES
    exceptions.Exception
        _OptionError

    class _OptionError(exceptions.Exception)
     |  Exception used by option processing helpers.
     |
... snipped rest.

What's funny about it? Looks like exactly what you asked for.

Cheers,

f.


From olivier.ravard at novagrid.com  Thu Sep 26 17:58:17 2002
From: olivier.ravard at novagrid.com (olivier Ravard)
Date: Thu, 26 Sep 2002 23:58:17 +0200
Subject: Doc for Tix / Python ?
Message-ID: <3D9382F9.7040407@novagrid.com>

Hi,

Is someone know where we can find some documentation
or tutorial about using Tix with Python ?

I try to use PanedWindow without succes...

O.R.



From see_reply_address at something.invalid  Thu Sep 12 21:20:38 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Fri, 13 Sep 2002 13:20:38 +1200
Subject: Destructor never called ???
References: <76710511.0209111450.40999dd8@posting.google.com> <tp70ou0r0j11as6enhpsaqul4oqa3mlu6t@4ax.com>
Message-ID: <3D813D66.3080109@something.invalid>

Tim Roberts wrote:

> However, Python is a garbage-collected system.  Unlike in COM, an object is
> NOT destroyed as soon as its reference count goes to 0.


Actually, in CPython, it is. If your object isn't being
destroyed, it must be because there is still a reference
to it somewhere, possibly due to a cyclic data structure.

If it's a cyclic data structure, it will eventually be
found by the cyclic garbage collector. But the __del__
method *still* won't be executed, because the garbage
collector doesn't know what order the destructors of
a cyclic structure should be run in, and it doesn't
try to guess. Instead, it puts any objects it finds
which have __del__ methods into an "uncollectable
garbage" list.

So the bottom line is, don't rely on __del__ methods,
unless you're sure the object can't possibly be part
of a cyclic structure. Even then, it's not really a
good idea -- if something really must be done, do it
explicitly when it needs to be done.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From see_reply_address at something.invalid  Mon Sep  2 01:55:05 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Mon, 02 Sep 2002 17:55:05 +1200
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de>	<874rd9phtd.fsf@pokey.henrik-motakef.de> <mailman.1030916106.31312.python-list@python.org>
Message-ID: <3D72FD39.10902@something.invalid>

Fran?ois Pinard wrote:

> [Henrik Motakef]
> 
> PLEX has a Python way to describe regular expressions.  It is likely available
> stand-alone.

http://www.cosc.canterbury.ac.nz/~greg/python/Plex/

You get documentation as well if you get it from there. :-)

Note that Plex's RE implementation is very special
purpose -- you couldn't use it as a direct replacement
for the re module. But a wrapper for the RE module
which uses the same syntax could easily be made.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From propadovic.nenad at debitel.net  Wed Sep 18 18:30:01 2002
From: propadovic.nenad at debitel.net (Nenad Propadovic)
Date: Thu, 19 Sep 2002 00:30:01 +0200
Subject: Need advice on choosing the technology for intranet based application 
Message-ID: <3d88fd1e@news.ivm.net>

Hello pythonistas,
I'm about to write an application related to injury prevention in big
industrial companies. It's supposed to be intranet-based (because the admins
are reluctant to install it on x (where x>>1) clients). So: the client is a
web-browser. I still don't know whether I will get a server on my own, or
I'll have to fit into an unknown web server environment, both options being
realistic.
By now it seems the logic of the application will be simple. It's going to
be a relational database with several tables, a modest number of users with
write priviledges, some more with read priviledges only. Performance is not
a big issue.
The GUI has a part I consider complex: I have to present some statistics in
form of charts or graphs.
Now I've never done any web based application in my life, have a very
rudimentary knowledge of HTML, have been doing som database hacking (O.K.,
engineering :-) ) with MS Access and Ingress/ Open Road, have once written a
socket-based combined telnet/ftp client.
So I kindly ask for knowledgeable opinions on the following question:
-which web/application server? I'm considering Zope (seams rather BIG when
you download & unpack it, and it seams all possible web applications have
been written for it) and twisted (smaller, seams lower lewel and somewhat
easier to learn), and I'm wondering if there is something else out there,
which is both powerfull and flexible and easy to learn and complete. Oh yes,
I have both a time constraint (mid november, having only my spare time for
the project), and don't want end up rewritting the stuff under another
system, because the chosen one isn't flexible enough. Database included
would be nice.
-I assume the simplest way to present a graph/chart kind of graphic is to
create a jpg/giff/tiff on the server side, and just show it on the client?
Thanks to everybody taking the time to read the message, and even more to
those who answer it :-)
Cheers,
Nenad Propadovic
P.S. I'm just introducing Python to my group at BMW (a completely different
project, though), and I'll let you know if there are some bigger success
stories.




From whisper at oz.net  Tue Sep 17 15:00:03 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 17 Sep 2002 12:00:03 -0700
Subject: CGI question
In-Reply-To: <slrnaoespa.2i8.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <GCEDKONBLEFPPADDJCOEEEHMFAAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Gerhard H?ring
> Sent: Tuesday, September 17, 2002 11:29
> To: python-list at python.org
> Subject: Re: CGI question
>
>
> In article <slrnaoerk7.mgo.jboy at otaku.freeshell.org>, John D. Boy wrote:
> > Hello there!
> >
> > I am working on a simple CGI script to use as a framework for a website
> > of mine. Basically, I want it to work as follows: When somebody accesses
> > the site, a CGI (called e.g. index.py) is run. It takes an html-template
> > and fills it with content from another file, called e.g. index.dat, and
> > returns the completed html-document. When I want to present another
> > page, I want the url to look like this: http://my.domain.name/?page. In
> > that case, index.py would fill the same template with the content of a
> > file called e.g. page.dat.
> >
> > Now, my problem is to find out what follows behind the question mark in
> > my CGI script. Unless the key has a value
> > (http://my.domain.name/?key=value), it is not in the cgi.FieldStorage
> > dictionary. Is it simply impossible, or am I not using the right
> > function calls?
>
> You need to configure your webserver to rewrite URLs from your
> desired form to
> the "/scriptname?key=value form" . I believe the relevant Apache module is
> called mod_rewrite.
>
> -- Gerhard

>From what I've heard, the difficulties of writing mod_rewrite rules is
enough to make even seasoned apache warriors shudder.

You might consider using a URL like http://my.domain.name/?doc=page to save
yourself the trouble.

HTH...

Dave LeBlanc
Seattle, WA USA




From jmj8237 at cs.rit.edu  Sat Sep  7 22:19:29 2002
From: jmj8237 at cs.rit.edu (jason jurkowski)
Date: 7 Sep 2002 19:19:29 -0700
Subject: regarding serve_forever() and non-blocking connection
Message-ID: <52aa2822.0209071819.7560c8c0@posting.google.com>

yeah i know this is an odd way to reply to a post but its the only way
i could do it.  here is the response:
really the proxy stresses the cpu?  if http_proxy.handle_request()
compiles to more than a handfull of bytecodes and the requests are
concurrent the interpretor is probably doing too much context
switching.  by default the interpretor performs a context switch
between threads after executing 10 bytecodes.  you can change this w/
sys.setcheckinterval().


From alf at logilab.fr  Wed Sep 18 07:23:14 2002
From: alf at logilab.fr (Alexandre Fayolle)
Date: Wed, 18 Sep 2002 11:23:14 +0000 (UTC)
Subject: Logical Programming with Python?
References: <3D8771FB.4000600@web.de>
Message-ID: <slrnaogond.g6a.alf@virgo.logilab.fr>

Dans l'article <3D8771FB.4000600 at web.de>, Ingo Linkweiler a ?crit :
> Hello,
> 
> i am looking for examples for logical programming (Prolog like) with 
> Python. I have found some projects, but no examples how to use them.
> Most of them seem "unuseable" or only can solve equations:

Well, Logilab's constraint package sure is not dead or unuseable. 
We are using it for project scheduling (project Pygantt on 
http://www.logilab.org/) and it works very well, especially with the
latest not yet released version of the constraint module. 

http://www.logilab.org/python-logic/documentation.html gives an example
of how the module can be used. 

Now it is not a prolog engine, but a CSP-solver. 

PyLog seems closer to what you're looking for, but your description is
not very precise, so I can't tell for sure. 

PyProlog (http://sf.net/projects/pyprolog) is a wrapper around
SWI-prolog. I think there's another project wrapping GNU-prolog
somewhere, though I can't find it right now. 

Some more resource are available on http://www.logilab.org/python-logic/
and in the mailing list archives. 


-- 
Alexandre Fayolle                              LOGILAB, Paris (France).
http://www.logilab.com   http://www.logilab.fr  http://www.logilab.org
Advanced computing - Python - Customized trainings - Consulting - XML
Informatique avanc?e - Python - Formations sur-mesure - Conseil - XML


From hat at se-46.wpa.wtb.tue.nl  Fri Sep 20 08:00:47 2002
From: hat at se-46.wpa.wtb.tue.nl (Albert Hofkamp)
Date: Fri, 20 Sep 2002 12:00:47 +0000 (UTC)
Subject: general thoughts [off-topic]
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com>
Message-ID: <slrnaom3rn.ht6.hat@se-46.wpa.wtb.tue.nl>

On Fri, 20 Sep 2002 07:28:52 GMT, Benjamin <phncontact.libero.it@> wrote:

> i have an unusual problem with programming python,
> nothing tecnical, something which goes "deeper".

After reading your post, I am not even sure that it is Python related,
it just happened while learning Python.

> you may call it inspiration, see me as an artist
> who lacks of inspiration (i know there are many
> people who think programming is an art).

I do think that, but w.r.t solving the problem only.

> some weeks ago i began learning python as my first
> (i hope some others will follow) programming language,
> i've been going through some tutorials, but they begin
> to bore me. the only interesting thing (from a beginners

A programming language is just a tool, just like a hammer. You can read
about hammers as much as you like, but after a few descriptions you
understand what a hammer is, and tend to get bored (at least, I do).

> point of view, of course) i programmed is a small
> database script which receives user input and stores
> things in a txt file. that was real fun, writing code,

Applying the hammer to solve a problem is a very different matter however.
I agree that using a tool to solve something is very satisfying.

> the problem, in one word: coding. but now i don't know
> what to programm. sounds weird, doesn't it? :) what
> should i program?? at the moment i have windowsXP

One the problems of a computer is its flexibility :-)
I wouldn't recommend fixing that, but it does imply that there are little
limitations w.r.t. the type of problems you can solve.

The best answer I can think of, is 'whatever you consider fun to program'.
Preferably a stupid time-consuming task (since computers are good at that).
That varies (for me) from computing the minimal amount of money that needed
to be given or paid to balance a cost-sheet in the student-house I lived
in to solving logical puzzles.
(i.e. a small problem that is easy to understand, yet often surprisingly
complicated to program (especially when you generalize the problem, such as
balancing for n students instead of 3).

> installed, but i hope to switch soon to some linux
> distribution or freeBSD. maybe my problem gets better

Probably not. A unix system is much more flexible, and can be modified at
much more points (i.e. there is lots to explore!), but it is still a tool.
If you don't have a problem, there is no need to apply the tool.

Ask yourself the question with everything that you do "can a computer do
this better?". If you think the answer is yes, you have a problem to solve.
Another trick can be to see what simple repetitive job you don't like (and
washing the dishes is not simple :-) ).

> i've been learning python not for becoming a programmer
> (i study modern languages at the university), but for having
> fun. my personal motto is "i want to know how it works".

Good motto, I use it too.
With Python, it seems that you now know how it works, so it seems to be
time to move on. That does not mean you will never use Python any more. You
know it, and when you stumble upon a problem that you want to be solved,
you go back to the tool that can help you.

I do this quite often. I look at software packages a lot, and keep in my
mind that software for solving such and such problems exists. Then, when I
find a problem that fits, I go back to the package, and try it.

> to write the ultimative windows security patch that fills
> all security holes in m$'s os & co. and get rich ^_^). it

There are easier ways to get rich IMHO :-)

> certainly wouldn't be boring. i don't think i will ever write

I tend to disagree here, finding and patching holes also gets boring when
you do it long enough.

> the opposite. i think if it's a job, it's something i MUST
> do - and therefore not interesting anymore -, and not something
> i CAN do if i want.

Correct, so I do the boring stuff while I work, and the fun stuff during my
free time (if I feel like it).

> well, i hope i didn't bore you :) i would really like if you
> share your opinions about those things with me/us.

Ok, but it's a whee bit off-topic here.



Albert
-- 
Unlike popular belief, the .doc format is not an open publically available format.


From jkraska at san.rr.com  Fri Sep 20 02:13:52 2002
From: jkraska at san.rr.com (Courageous)
Date: Fri, 20 Sep 2002 06:13:52 GMT
Subject: strong/weak - dynamic/static [Was: Getting started]
References: <mailman.1032444745.27364.python-list@python.org>
Message-ID: <h0floucibce0b30c56pbsum246rnjk257j@4ax.com>

On Thu, 19 Sep 2002 09:09:19 -0500, sismex01 at hebmex.com wrote:

>So, you see, in C (a strong, static typed language) I obtained
>a char from a floating-point number; not the string representation,
>just the first byte.

C is a statically typed language, but to call it "strongly" typed
is a bit of a stretch. C is at best, middle of the road in type
strength.

C//



From Oschler at earthlink.net  Thu Sep 12 05:14:05 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Thu, 12 Sep 2002 09:14:05 GMT
Subject: Any C++/Python sig members here?
References: <l1Kf9.1470$Lg2.157618@news2.news.adelphia.net> <alp5rc$luo$1@bob.news.rcn.net>
Message-ID: <xVYf9.1972$Lg2.309842@news2.news.adelphia.net>

"David Abrahams" <david.abrahams at rcn.com> wrote in message
news:alp5rc$luo$1 at bob.news.rcn.net...
>
>
> To answer your question from
> http://mail.python.org/pipermail/c++-sig/2002-September/002055.html:
> Yes, Boost.Python works on Linux. See a list of supported platforms and
> compilers for Boost.Python v1 at www.boost.org/libs/python/doc/index.html.
> Boost.Python v2 is compatible with even more platforms and compilers.
>
> -Dave

Dave,

Yes, unfortunately I can't find any info on getting it to work with
Borland's Kylix 3/C++.  I guess I'll have to use gcc for now, which is a
shame since Kylix 3/C++ is such a terrific IDE.

thx





From observer at NOSPAM.space.pl  Thu Sep  5 06:49:39 2002
From: observer at NOSPAM.space.pl (Johann)
Date: Thu, 05 Sep 2002 12:49:39 +0200
Subject: urgent: binary driver for MySQL/Windows
References: <ivbenucnekgan8tp6vm2351h5pq9s5u488@4ax.com>
Message-ID: <aldenu8jspt7irgh2n4fg961s9hl6gt3o5@4ax.com>

Forget it. I solved it using MySQLdb.

--
Johann


From sismex01 at hebmex.com  Thu Sep 26 17:47:52 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Thu, 26 Sep 2002 16:47:52 -0500
Subject: Who knows somefunction?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062330260501D@mail.hebmex.com>

> Erik Max Francis <max at alcyone.com> writes:
> > > Is there a function <somefunction> in Python
> > > for which
> > > 
> > > something is <somefunction>(id(something))
> > > 
> > > returns invariably true?
> > 
> > Not offhand, but there's an easier test.  x is y really 
> just means id(x)
> > == id(y), so you could change this test to:
> > 
> > 	id(something) == id(somethingElse)
> > 
> > and you probably wouldn't have the need for your 
> hypothetical function.
> 
> No I think he's asking for a way to dereference an id of an object.
> And the answer is there's no way to do that.
> 

Actually, it would be possible, in CPython at least;
but dangerous, and set a bad precedent.

Just imagine....

>>> x = deref(0x98237128)
>>> print x

--> and who knows what would happen here, because, what object
    would exist at that address?

Nope, it's a bad idea.

-gus







Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From uwe.schmitt at procoders.net  Fri Sep 27 07:53:35 2002
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 27 Sep 2002 11:53:35 GMT
Subject: mod_python
Message-ID: <an1grv$moqid$1@hades.rz.uni-sb.de>

Hi,

I tried mod_python these days and found it quite usefull.
Are there any modules for handling sessions, cookies, and
persistent data ? 

Greetings, Uwe.

-- 
Uwe Schmitt                    Computer science is no more about Computers,
dr.schmitt at procoders.net       than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           


From erikprice at mac.com  Wed Sep  4 10:37:32 2002
From: erikprice at mac.com (Erik Price)
Date: Wed, 4 Sep 2002 10:37:32 -0400
Subject: Python critical mass /turning point
In-Reply-To: <899f842.0209040617.711664fe@posting.google.com>
Message-ID: <D8840A3F-C013-11D6-BB50-00039351FE6A@mac.com>

On Wednesday, September 4, 2002, at 10:17  AM, Anthony_Barker wrote:

> 1. Interpreter Available on every desktop/server
> Perl has this on the server
> Java had this in the browser space
>
> Solution(s)
> Parrot/Perl6 integration - the if python is included
> with the perl interpreter it would be everywhere
> Lobby Sun & IBM to include Python on their Unix CDs
> Lobby Microsoft to get added to their resource kit
> Other options are already available such as freeze....

At least it's shipping by default on Mac OS X now (as of 10.2).

Unfortunately readline wasn't compiled in.



Erik





--
Erik Price

email: erikprice at mac.com
jabber: erikprice at jabber.org




From brucer at theramsays.com  Mon Sep  2 20:35:19 2002
From: brucer at theramsays.com (Bruce Ramsay)
Date: 2 Sep 2002 17:35:19 -0700
Subject: Why it is so dramatical?
References: <mailman.1030957743.4021.python-list@python.org> <Xns927D6DBDE418Cduncanrcpcouk@127.0.0.1> <mailman.1030962662.4274.python-list@python.org>
Message-ID: <88a4765d.0209021635.54ddbf5d@posting.google.com>

"Bo M. Maryniuck" <b.maryniuk at forbis.lt> wrote in message news:<mailman.1030962662.4274.python-list at python.org>...
> On Monday 02 September 2002 11:57, Duncan Booth wrote:
> > None of this is specific to Python, the same issues arise in almost any
> > programming language (although not always the same solutions).
> Almost... Keep in mind, that here is the range 0-0xffffff, what is quite
> bigger that 0-0xfff:
> ----------------------8<---------------------------
> #!/usr/bin/perl
> 
> $b = '';
> $data = 'AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
> AAAAAAAAAAAAAAAAAAA';
> for $a (0..0xffffff) {
>     $b += $data;
> }
> ----------------------8<---------------------------
> 
> Since time() always show me zero, I've used UNIX time command:
> ----------------------8<---------------------------
> *bo at oberon:(~/work/spike/benchmarks) time perl str vs list.pl 
> 
> real    0m8.734s
> user    0m8.710s
> sys     0m0.010s
> ----------------------8<---------------------------
> 
> As you see, adding string to the string isn't a huge problem for Perl as 
> for Python.
> What's "wrong" with Perl?

The Perl += operator does not concatenate strings.

Bugs aside, I suspect that Perl has an optimisation for this case, or
a special string data structure.  Perl run times seem to grow linearly
for loop counts from 1e5 to 1e7.  Python run times go from 0.053s to
18.2s when changing the loop count from 1e3 to 1e4.  This is on Python
2.1 on an Athlon 1600.


From hancock at anansispaceworks.com  Fri Sep  6 19:06:10 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Fri, 06 Sep 2002 16:06:10 -0700
Subject: Python class ( Prolog ) ???
References: <20020905234600.11264.60328.Mailman@mail.python.org> <3D7914FD.5D713A48@anansispaceworks.com>
Message-ID: <3D7934E2.3C431FF9@anansispaceworks.com>

Well, I haven't seen any posts go by on the python-logic
SIG, so I thought I'd carry over this topic from the
python list as a conversation starter:

While we're on the subject -- I've taken a couple of
tries at Prolog in the past, but didn't really get
anywhere with it.  I understand the concept of
expressing logic in code, but got rather bogged down
when it came to any practical examples.

Python, with it's object/procedural focus seems like
it would complement Prolog well (or vice-versa). Can
anyone fluent in both compare what sort of applications
you might apply a prolog-python solution to?  Where
would you draw the line between the two? What would
using prolog help me with? What's a good problem to
solve as a learning project?  Any more book or website
recommendations for a learner?

Conversely is there anyone who thinks Prolog is
a waste of time for a Python programmer, or that
it would be better to use something else in such
a role?

Terry Hancock wrote:
> From: Christophe Delord <christophe.delord at free.fr>
> > On Thu, 05 Sep 2002 21:57:20 GMT
> > "Joseph A. Knapka" <jknapka at earthlink.net> wrote:
> > > > Luis wrote:
> > > > Does anyone have a class that provides the functionality of prolog ???
> >
> > > There's a SourceForge project that enables you to call
> > > GNUProlog from Python:
> > > <URL: http://bedevere.sourceforge.net/>
> >
> > I have written a module in Python to mix Prolog and Python. It was just a funny
> > exercise. It is not complete and rather slow but if you are curious you can download
> > it: http://christophe.delord.free.fr/en/pylog/
> 
> You may also be interested in the "unofficial"
> python-logic SIG:
> 
> http://www.logilab.org/python-logic/
> 
> Which was recently mentioned on the edu-sig at python.org
> list.

I suppose follow-ups should go to the python-logic
list?

Cheers,
Terry

-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
Anansi Spaceworks                 
http://www.anansispaceworks.com 
P.O. Box 60583                     
Pasadena, CA 91116-6583
------------------------------------------------------



From peter at engcorp.com  Thu Sep 19 08:51:54 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 19 Sep 2002 08:51:54 -0400
Subject: StructuredText -> HTML module??
References: <wbii9.19869$0p1.359879@news2.ulv.nextra.no> <3d89b69d$1@news.sentex.net> <imji9.23500$sR2.405010@news4.ulv.nextra.no>
Message-ID: <3d89c85f@news.sentex.net>

Thomas Weholt wrote:
 >"Peter Hansen" <peter at engcorp.com> wrote:
 >>Maybe you could help us by telling us *what* you expected.
 >>Otherwise we're likely to waste our time making suggestions
 >>that are just as inadequate from your point of view.

> I need to take a piece of structured text as a string into a method and get
> html out in the other end. As described on the page on Zope.org under
> StructuredText. My hacking with PyWiki gave me a list where each element had
> a '<pre>' in front of it etc. I might be using it the wrong way, but ....
> 
> I was under the impression that StructuredText as described on the site
> mentioned earlier was sort of a standard, that the parser etc. would produce
> the same output.
> 
> All I'm actually after is to be able to write stuff like :
[snip example]

Unfortunately, StructuredText is *not* standardized, so you may not
get the results you want with some, or any, of the implementations.
StructuredText is more of a concept than a standard, although there
are certain rules that are defacto standards (well, barely even that).

The presence of lots of <pre> elements in your experiment suggests
strongly to me that you unexpectedly triggered some rule in
the particular implementation you tested (as opposed to suggesting
that it just doesn't work right).

Maybe you could show an example with a few lines of code that
gives that behaviour.  StructuredText can be tricky; it might have
been just the presence of a single bit of punctuation which changed
the output from what you expected into that mess with <pre>.

Zope has one that seems to work okay, although I bet PyWiki does too:

 >>> import StructuredText
 >>> p = 'This is a test.\012\012- This is a test\012\012- as is 
this.\012\012What is up?\012\012 testing a paragraph\012 with indentation.'
 >>> print p
This is a test.

- This is a test

- as is this.

What is up?

  testing a paragraph
  with indentation.
 >>> dir(StructuredText)
['Basic', 'ClassicDocumentClass', 'Document', 'DocumentClass', 'HTML', 
'HTMLClass', 'HTMLNG', 'StructuredText', '__builtins__', '__doc__', 
'__file__', '__name__', 'html_quote', 'html_with_references', 'letters', 
're', 'string', 'sys']
 >>> html = StructuredText.HTML(p)
 >>> html
'<html>\012<body>\012<p>This is a test.</p>\012\012<ul>\012<li><p>This 
is a test</p></li>\012<li><p>as is 
this.</p></li>\012\012</ul>\012<h1>What is up?</h1>\012<p> testing a 
paragraph\012 with indentation.</p>\012</body>\012</html>\012'
 >>> print html
<html>
<body>
<p>This is a test.</p>

<ul>
<li><p>This is a test</p></li>
<li><p>as is this.</p></li>

</ul>
<h1>What is up?</h1>
<p> testing a paragraph
  with indentation.</p>
</body>
</html>


-Peter



From stephen.boulet at motorola.com  Fri Sep  6 17:57:19 2002
From: stephen.boulet at motorola.com (Stephen Boulet)
Date: Fri, 06 Sep 2002 16:57:19 -0500
Subject: how can i make a executable program with a python2's script
References: <dd083388.0209060003.1188000c@posting.google.com>
Message-ID: <3D7924BF.805@motorola.com>

You can put

    #/usr/bin/env python

at the top of the file and do a chmod +x file. Is this the answer to the 
right question?

-- Stephen

jgm2163 wrote:
> os     :RedHat 7.2
> Python :Python2.2.1
> 
> now,i had writen a python's script (extend with .py)
> how can i make a executable program on RedHat 7.2?




From loewis at informatik.hu-berlin.de  Mon Sep  2 09:42:09 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 02 Sep 2002 15:42:09 +0200
Subject: Must have Python modules?
References: <_THc9.42153$vY2.1039887@e3500-atl2.usenetserver.com> <slrnan6k4s.eg.gerhard.haering@haering.opus-gmbh.net> <FCIc9.30407$FJ1.5585669@e3500-atl1.usenetserver.com>
Message-ID: <j4y9akjyoe.fsf@informatik.hu-berlin.de>

"Robert Oschler" <Oschler at earthlink.net> writes:

> In C++, which has thousands of libraries, snippets, etc. there were only a
> few that I personally used religiously.  

[...]

> In python I already have two that I'm very find of, Secret Labs
> RegExp module and the xml.dom module.  (And the wonderful WingIDE
> environment for Python development!)

Python is different here; it comes with batteries included. Both SRE
and xml.dom.minidom are part of the standard Python distribution since
2.0.

So if your question includes standard library modules as well: my
absolutely must-have module is "os", followed by "sys".

Regards,
Martin


From tdelaney at avaya.com  Wed Sep  4 19:00:39 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 5 Sep 2002 09:00:39 +1000 
Subject: Why Python?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A573@natasha.auslabs.avaya.com>

> From: Max M [mailto:maxm at mxm.dk]
> 
> Gerhard H?ring wrote:
> > Markus Jais wrote:
> 
> >>so maybe you should learn both PHP and Python (and Ruby, of 
> course :-)
> > 
> > I'd skip PHP. But then, I'm biased. And proud of it :-P
> 
> Well PHP will not teach you a lot as a programmer, but it can 
> have good commercial, practical and pragmatic value.
> 
> It is often used in the wild and it is nice to have a general 
> feel for it.

And I'm struggling mightily with it at the moment and have decided that I
would never recommend it for a project. In my very short time with it
(approx 4 weeks), I've come across several very annoying things, and just
the other day a probable show-stopper.

And searching for help (newsgroups, web sites, etc) hasn't turned up
anything useful for what I've been trying to do.

Finding that much wrong in such a short time is an immediate flag that there
is something fundamentally wrong with a system.

Tim Delaney



From cliechti at gmx.net  Sat Sep  7 18:05:57 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 8 Sep 2002 00:05:57 +0200
Subject: Killing/Terminating a running thread
References: <mailman.1031419029.11871.python-list@python.org> <Xns9282E6FBFE1C2cliechtigmxnet@62.2.16.82> <mailman.1031434090.15525.python-list@python.org>
Message-ID: <Xns9283FF1199cliechtigmxnet@62.2.16.82>

Heiko Wundram <heikowu at ceosg.de> wrote in
news:mailman.1031434090.15525.python-list at python.org: 

> Am Sam, 2002-09-07 um 22.42 schrieb Chris Liechti:
>> sound like the news thread "Interrupting Python" could be
>> interresting fo 
> r 
>> you...
> 
> I've looked at that, but the main problem is that this thread deals
> with interrupting fullblown processes on the fly (from within the
> code), not with interrupting a process by an exterior signal.

ah, sorry. i dont follow that thread.

>> the clean solution is to test for a flag and terminate if the flag is
>> set 
> .
>> 
>> e.g. you could provide a function:
>> 
>> def interrupted():
>>          return global_flag_if_user_wants_to_abort
>> 
>> and in you function in the queue:
>> 
>> def worker():
>>          while not interrupted():
>>                   do_stuff()
> 
> This is what I do to shutdown the worker cleanly. The problem is:
> do_stuff() doesn't return until it's complete, and I have no control
> whatsoever about the internal workings of do_stuff().

then you have a problem... if you could enforce the use of the intterupted 
function or flag it would work flawlessly on all platforms.

> What I actually wish to do is while the worker still is in do_stuff(),
> to kill the worker as a whole, to discard all its data, when a timeout
> on returning a value from do_stuff() occurs. It's not about clean
> termination of a thread, but about killing the thread in case it takes
> too long (in case do_stuff() takes to long).

it's maybe possible to kill threads on linux, but i think its a risky thing 
to do anyway. it could have an impact on stability of the other/main 
threads.

you cold start worker _processes_ you can kill those cleanly (maybe not on 
Win9x but on every real OS ;-). that requires inter-process communication. 
there are diffrent sorlutions to achieve that, e.g. "pyro" or a corba 
implementation.
if you realy need to kill a worker and have no controll over its 
implementation, then i think that you can't avoid using processes.

> Thanks for the advice anyway! :) I guess I at least did the general
> interrupt on the worker right then... :)

chris
-- 
Chris <cliechti at gmx.net>



From exarkun at meson.dyndns.org  Tue Sep 17 20:46:16 2002
From: exarkun at meson.dyndns.org (Jp Calderone)
Date: Tue, 17 Sep 2002 20:46:16 -0400
Subject: hexacodes in python?
In-Reply-To: <-1995099244.1032292636635.JavaMail.nobody@webmail1>
References: <-1995099244.1032292636635.JavaMail.nobody@webmail1>
Message-ID: <20020918004616.GA19175@meson.dyndns.org>

On Tue, Sep 17, 2002 at 04:57:16PM -0300, jubafre at brturbo.com wrote:
> what is the hexacode for '?' in python?
> where can i take the hexacodes?	
> 	
> Juliano Freitas
> www.gebrasil.hpg.com.br 

  Assuming you're refering to the ASCII value of a character, the function
you want is 'ord'.

>>> ord('?')
63

  If you want this in hexadecimal, 'hex':

>>> hex(63)
'0x3f'


    Jp

-- 
        "I quite agree with you," said the Duchess; "and the moral of
that is -- Be what you would seem to be' -- or, if you'd like it put
more simply -- Never imagine yourself not to be otherwise than what it
might appear to others that what you were or might have been was not 
otherwise than what you had been would have appeared to them to be
otherwise.'"       -- Lewis Carrol, "Alice in Wonderland"
--
 8:00pm up 119 days, 20:53, 7 users, load average: 0.00, 0.03, 0.06
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 248 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020917/ea328c07/attachment.sig>

From solonas73 at hotmail.com  Mon Sep 16 07:04:07 2002
From: solonas73 at hotmail.com (Solon Kazantzidis)
Date: 16 Sep 2002 04:04:07 -0700
Subject: upload/download to a palm os device with python
Message-ID: <24165ed4.0209160304.7a3c08af@posting.google.com>

Hello,

I want to write a python (for windows) program which can establish a
connection to a palm device so that I am able to upload (from the
desktop pc to the palm device) and download (palm-->desktop) data.

I am totally new to this. I have basic experience with python. I have
basic (theoretical from reading articles and writing Java2ME programs
from the tutorials) experience with Palm devices.

Some of my Questions (maybe you can help me to answer them):

1. Is there a python library which I can use in order to access a palm
device and work on the databases (create new, delete existing, browse
existing) stored on it?

CODE LIKE:
from <lib> import *

con = lib.OpenSerialConnection(...)
con.GetDatabase(...)
con.CreateDatabase(...)
con.DeleteDatabase(...)
con.PutDatabase(...)

2. If there is no python library. Can this be solved by writing C
source and making a module for Python? (I have limited time as this is
work for my thesis)

3. Are the thoughts I had in 1. and 2. well suited to solve the
following problem

Have a standalone windows application which can upload/download data
from/to a palm device which allows for further work (storing in
database, transmitting to ...) with the data from the palm device on
the desktop pc?


I thank you for your help.

p.s.: please excuse my newbie mistakes in writing this posting and
point on them so that i can learn for the future.


From jetman516 at hotmail.com  Thu Sep 12 18:06:18 2002
From: jetman516 at hotmail.com (The Jetman)
Date: 12 Sep 2002 15:06:18 -0700
Subject: Starting external functions in Python applications
References: <1d3ffc6b.0209120345.64cf64a1@posting.google.com>
Message-ID: <a38d43b0.0209121406.fb4aa98@posting.google.com>

stencelp at web.de (Paulus) wrote in message news:<1d3ffc6b.0209120345.64cf64a1 at posting.google.com>...
> Hi.
> How can I link an external dll in a Python programs
> and (when possible) start any function from this DLL ? 
> Thanks in advance
> paulus

http://www.oreilly.com/catalog/python2/chapter/ch15.html

Depending on the DLL, you may be able to access it via COM.  This 
sample chapter from the 2nd Ed of Lutz has an example of scripting 
COM objects from Python (ex 15-13) and lots of other stuf.  HTH...Jet


From duncan at NOSPAMrcp.co.uk  Tue Sep 10 10:00:34 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 10 Sep 2002 14:00:34 +0000 (UTC)
Subject: default variable value and cmd module
References: <hmmf9.201$GH5.711@news.hananet.net>
Message-ID: <Xns9285975DAA56duncanrcpcouk@127.0.0.1>

"konghyeogjun" <toto2001 at hananet.net> wrote in 
news:hmmf9.201$GH5.711 at news.hananet.net:

> This sample using 'cmd' module do not prints default variable value, 5.
> 
> import cmd
> class x(cmd.Cmd):
>     def do_y(self, value = 5):
>         print value
> z = x()
> z.cmdloop()
> 
> but, the following sample prints '5'. why don't the sample above prints?

>From the documentation for cmd.Cmd:
    3. A command `foo' is dispatched to a method 'do_foo()'; the do_ method
       is passed a single argument consisting of the remainder of the line.

So the value passed to do_y as 'value' is whatever you type on the rest of 
the command line after your 'y' command. It doesn't print 5 because the 
default value is always overridden, even if you don't type anything after 
the command name.


-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From jdavis at empires.org  Thu Sep 19 05:08:19 2002
From: jdavis at empires.org (Jeff Davis)
Date: Thu, 19 Sep 2002 09:08:19 GMT
Subject: shared mem & semaphores module
References: <YWbi9.17569$V7.4455639@twister.socal.rr.com> <pan.2002.09.19.08.02.16.550254.723@thomas-guettler.de> <i6ei9.112316$ub2.2394246@news1.tin.it> <7xr8fqquee.fsf@ruckus.brouhaha.com>
Message-ID: <7ugi9.18276$V7.4769331@twister.socal.rr.com>

> 
> mmap has no provision at all for semaphores.  Try:
> 
>   http://gigue.peabody.jhu.edu/~mdboom/omi/source/shm_source/shm.html
> 
> I wish the standard library would include this.

The threading module, as someone else pointed out, includes a Semaphore 
class. It seems like it might be wrapped around locking functions in the 
thread module, however, which seems strange. The one you point to seems 
like what I'd expect in the standard library.

It's good to hear that I'm not crazy for thinking that shared memory 
functions should be in the standard library.

Which brings up an interesting question: why are python libraries not more 
categorized? It would make sense to me that instead of doing "import pg" 
you should do something more like "import db.pg" so as not to clutter the 
module namespace with 50 different database drivers (mysql, postgres, 
etc).

With that model they could make us do our imports like "import unix.shm" or 
whatever and it wouldn't get in anyone's way.

Regards,
        Jeff







From clpy at snakefarm.org  Mon Sep  2 11:33:31 2002
From: clpy at snakefarm.org (Carsten Gaebler)
Date: Mon, 02 Sep 2002 17:33:31 +0200
Subject: Newbie's unsurmountable pb with IO stream trick
References: <aba30b75.0209020633.26ba4c70@posting.google.com>
Message-ID: <al00cb$o0a$1@news.schlund.de>

Spendius wrote:

> For those who know SQL*Plus y'all know you get a prompt:
> "SQL> " 
> (remove the "'s to get the string) and then this frontend
> Oracle software waits for your input. This is somewhat tricky 
> to deal with,

You'd be probably better off using a Python->Oracle interface like 
cx_Oracle (http://www.computronix.com/utilities/) to access your 
database directly than trying to mess with an interactive program.

cg.



From peter at engcorp.com  Mon Sep 23 15:04:04 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 23 Sep 2002 15:04:04 -0400
Subject: Non-interpreted Python?
References: <S5Jj9.360$UC2.34013@read2.inet.fi>
Message-ID: <3D8F65A4.38124945@engcorp.com>

Tapio Kelloniemi wrote:
> 
> I'm new to Python and would like to know, wheter it is possible to
> compile Python code to binary executables (and on which platforms)? 

Yes, at least on Linux and Win32, if you are not expecting large increases
in performance as a result.  This is done by "freezing" the bytecode files
and bundling them into a single executable, usually with the libraries
(.DLL or .so as appropriate) bundled but separate.

There are other approaches, but there's no point describing them until
you explain *why* you want to do this.  What are your requirements?

> Or better, if it is possible to compile code into object format and then
> link with C/C++/whatever code.

Python has excellent facilities for embedding in applications written
in C/C++.  Check the documentation for the relevant sections.

-Peter


From pinard at iro.umontreal.ca  Tue Sep 10 07:27:14 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Tue, 10 Sep 2002 07:27:14 -0400
Subject: Python 2.3
In-Reply-To: <3d7dc8cd_1@news.newsgroups.com> (JB's message of "Tue, 10 Sep
 2002 12:34:56 +0200")
References: <3d7cfc19_5@news.newsgroups.com>
	<5t0qnu0dfbebm1uvk1v93msk4tespgp6bn@4ax.com>
	<3d7dc8cd_1@news.newsgroups.com>
Message-ID: <oqk7luozjh.fsf@titan.progiciels-bpi.ca>

[JB]

> But there is probably no way to write a function f Rat x Rat 
> --> Rat and then at the input prompt write f(2/3,-3/2).

You need to build some special notation for creating your rational constants.
But Python makes it quite easy to do.  You could aim writing something like:

   f(Rat(2, 3), Rat(-3, 2))

or if you like ultra-concision, which is excusable interactively, do:

   r = Rat
   f(r(2,3), r(-3,2))

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



From narnett at mccmedia.com  Wed Sep  4 19:27:38 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Wed, 4 Sep 2002 16:27:38 -0700
Subject: Bind threads to addresses -- Windows & urllib?
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F06233022E50FC@mail.hebmex.com>
Message-ID: <PHEDLIMPMIAGAEPEFBGCEEFICHAA.narnett@mccmedia.com>


--
Nick Arnett
Phone/fax: (408) 904-7198
narnett at mccmedia.com

> That's not gonna help you at all, because all threads will
> be capped by the interface's top throughput, or by your
> machine's processing power.

You're making an incorrect assumption about the purpose of using multiple
addresses.  It has nothing to do with my end of the connection; it is to
cope with servers that regard a reasonably well-behaved spider (in my
opinion, at least) as a denial-of-service attack.  If the server operators
would reveal what they regard as well-behaved, I wouldn't have to resort to
this, but as one might expect, nobody wants to disclose the parameters of
their DOS defenses.  I can't get the relevant sites to even respond to
inquries... and they don't even have a robots.txt file.

On the other hand, the more I think about this, the less interested I am in
bothering, since it would surely be easy for a server to block a range of
addresses.

I'm also working the other obvious solution, heuristics for the spider to
set its own speed so that it won't trigger  defenses -- but it appears
there's something more than simple rules at the other end.  The robot wars I
expected years ago have arrived...

No lectures on what consists good robot behavior, please -- I've been
operating *the* list on that subject for years
(http://www.mccmedia.com/mailman/listinfo/robots)

Nick




From glingl at aon.at  Thu Sep 26 17:24:24 2002
From: glingl at aon.at (Gregor Lingl)
Date: Thu, 26 Sep 2002 23:24:24 +0200
Subject: Who knows somefunction?
Message-ID: <3D937B08.4090704@aon.at>

Hi Pythonistas!

Is there a function <somefunction> in Python
for which

something is <somefunction>(id(something))

returns invariably true?

Gregor



From bdesth at nospam.free.fr  Thu Sep 12 00:08:45 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Thu, 12 Sep 2002 00:08:45 -0400
Subject: function pointer
References: <ffcedbc9.0209110215.644c1788@posting.google.com> <mailman.1031740215.16074.python-list@python.org>
Message-ID: <3D80134D.5050802@nospam.free.fr>

Martin Franklin wrote:
> On Wednesday 11 Sep 2002 10:15 am, Christina wrote:
> 
>>Hi!
>>
>>Does Python have function pointers? If not, is there a possibility to
>>emulate them?
> 
> 
> 
> Excuse my ignorance but what is a function pointer....  
> could this be the answer:-
> 
> 
> def aFunction():
>     print "aFunction was called"
> 
> ## call it
> aFunction()
> 
> ## create a reference to it
> af=aFunction
> 
> ## call the reference
> af()
> 
> ## at the python command line:-
> 
> [dev]$ python
> Python 2.2.1 (#3, May 29 2002, 20:32:44)
> [GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> 
>>>>def aFunction():
>>>
> ....     print "aFunction was called"
> ....
> 
>>>>aFunction()
>>>
> aFunction was called
> 
>>>>af=aFunction
>>>>af()
>>>
> aFunction was called
> 
> 
> 
> Regards
> Martin
> 
> 

Well, it's not exactly a function pointer since there is no pointers in 
Python, but yes, the idea is the same.

Function pointers are often used in C to allow 'generic' behavior (like 
with the qsort() function) or as 'callback' functions to handle user 
events in GUI programming (see the Gtk+ toolkit).

Laotseu



From spyagency at hotmail.com  Sat Sep  7 02:56:08 2002
From: spyagency at hotmail.com (spyagency)
Date: Sat, 7 Sep 2002 18:56:08 +1200
Subject: Hello,python-list,questionnaire
Message-ID: <20020907065532.BOCG15160.mta5-rme.xtra.co.nz@Vsuyaj>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020907/d641c0a1/attachment.html>
-------------- next part --------------

------------------  Virus Warning Message  ------------------

charset.bat is removed from here because it contains a virus.

-------------------------------------------------------------
-------------- next part --------------
A non-text attachment was scrubbed...
Name: apix.BAK
Type: application/octet-stream
Size: 30274 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020907/d641c0a1/attachment.obj>
-------------- next part --------------
------------------  Virus Warning Message  ------------------

Found virus WORM_KLEZ.H in file charset.bat
The uncleanable file is deleted.

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

From tenthousand.NOSPAM at yahoo.com  Mon Sep 30 18:09:23 2002
From: tenthousand.NOSPAM at yahoo.com (Bart Van Hemelen)
Date: Tue, 01 Oct 2002 00:09:23 +0200
Subject: idea for a much needed application
References: <576c1752.0209291326.56dcdf31@posting.google.com> <pasepuc0r73mh5vlcsu1l7erl4e2b17qvi@4ax.com> <an831g$bfrqn$1@ID-157863.news.dfncis.de>
Message-ID: <ujihpu0gvloetcsadfeg706rcs64o51mjj@4ax.com>

In the dim and distant past on Mon, 30 Sep 2002 00:40:24 +0100, it was
rumoured that "mike henley" <rsrchstr at msn.com> spake thus on the
subject of "Re: idea for a much needed application" in comp.lang.perl:

>"other than the ones already present on a web-server", exactly! i want to be
>ably to use whatever language, modules or applications i choose, and over
>the year use others, freely. if i choose perl, php, or even python, ruby or
>tcl, then i want to be able to do that.

Would be sys admin hell. 

Quite frankly, you'd probably be cursing about your hosting service
being crap because this would cause so much hassle to the people who
have to keep the servers running (patches and upgrades to software,
plus scripts by inexperienced users going haywire). 

I know that if I'd was working at a firm offering such things, they'd
have to pay me a lot since it woudl mean I'd have to be extremely good
at a whole lot of things, which in turn would mean expensive
subscriptions for users.

--
Bart Van Hemelen
http://pr1nc3.com/BVH/
--------------------------------------
Life would be a lot easier if you could take a
look at the source code every once in a while.


From imbosol at vt.edu  Tue Sep 10 17:35:10 2002
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 10 Sep 2002 17:35:10 -0400
Subject: "str"(expr)"str" interpolation syntax
References: <Pine.GSO.4.33.0209102221070.10002-100000@techst02.technion.ac.il>
Message-ID: <allrrs$6hm$1@solaris.cc.vt.edu>

Beni Cherniavksy wrote:
> So I propose that a parenthesized expresion near a string would
> imply concatenation.  In other words,
>        "foo"(x * 3)"bar"
> would be equivallent to
>        "foo" + str(x * 3) + "bar".

Bleagh.


-- 
CARL BANKS
http://www.aerojockey.com


From marklists at mceahern.com  Fri Sep 20 23:40:22 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 20 Sep 2002 22:40:22 -0500
Subject: average?
In-Reply-To: <amggcu$195$1@news.hccnet.nl>
Message-ID: <JHEOKEOOLIGLDHCMAHMOAEFJEAAA.marklists@mceahern.com>

[Anton Vredegoor]
> Just a few more posters and we will get this function right, I am sure
>  >>> def average(seq):
>          return reduce(operator.add,seq)/float(len(seq))
                                           ^^^^^

Aw, why go through that trouble when you can just do this:

$ python
Python 2.2.1 (#1, Jun 25 2002, 10:55:46)
[GCC 2.95.3-5 (cygwin special)] on cygwin
Type "help", "copyright", "credits" or "license" for more information.
>>> from __future__ import division
>>> import operator
>>> def average(seq):
...     return reduce(operator.add, seq) / len(seq)
...
>>> l = range(1, 5)
>>> a = average(l)
>>> print a
2.5
>>> print l
[1, 2, 3, 4]
>>>

<running and ducking for cover>

// m



From marklists at mceahern.com  Sun Sep 15 09:13:34 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sun, 15 Sep 2002 08:13:34 -0500
Subject: AspectP?
In-Reply-To: <jOQg9.6057$Lg2.882928@news2.news.adelphia.net>
Message-ID: <JHEOKEOOLIGLDHCMAHMOMEMIDOAA.marklists@mceahern.com>

[Robert Oschler]
> Anybody seen a robust (fairly ready for primetime) implementation
> of Aspect Oriented Programming for Python?

Ignoring the criteria of robust...

Search the archives (http://groups.google.com/ -- use advanced search,
filter on comp.lang.python).  You'll find an amateurish metaclass
implementation that I posted.  You'll find some ideas and implementation
posted by Pedro Rodriguez.  In Pedro's posts, you'll also find references to
the TransWarp stuff of Philip Eby.  I'm too lazy to search the archives
right now and give you the exact URLs but there should be enough clues in
this post to help you.

Cheers,

// mark

-




From astavale at yahoo.co.uk  Wed Sep 18 13:57:59 2002
From: astavale at yahoo.co.uk (Alistair Thomas)
Date: Wed, 18 Sep 2002 17:57:59 GMT
Subject: Signals for GtkEntry widgets
References: <mailman.1031866395.5190.python-list@python.org> <6qqg9.509135$Aw4.21217664@bin2.nnrp.aus1.giganews.com> <da4423a8.0209171024.38ee828a@posting.google.com>
Message-ID: <H83i9.119235$AR1.4976137@bin2.nnrp.aus1.giganews.com>

In article <da4423a8.0209171024.38ee828a at posting.google.com>, "D.
Shifflett" <shifflett at nps.navy.mil> wrote:

> I have found that the 'activate' event/signal occurs when the Enter key
> is pressed in a GtkEntry widget. 

Thanks, much simpler solution than I've been using!

A resource that I've just found and seems to have some useful tips is the
PyGtk FAQ:

http://www.async.com.br/faq/pygtk/index.py?req=index

Although it doesn't mention the Enter key in GtkEntry

--
       Al


From fperez528 at yahoo.com  Wed Sep 18 19:46:37 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 18 Sep 2002 17:46:37 -0600
Subject: linux python ideas
References: <3D88E578.6010001@diespammerdieuselesspython.com>
Message-ID: <amb38t$7qs$1@peabody.colorado.edu>

Rob Andrews wrote:

> The idea is not "how to program in Python", which would take more than
> one quick presentation, but a one-off demonstration of ways in which
> Python is particularly useful for such a group.
> 
> Any suggestions would be appreciated, since my notes are pretty generic
> so far.

Well, here's another one. It's a short script to fetch bibliographic 
information from the SPIRES database in BibTex format, from preprints listed 
in the XXX Los Alamos server. I use it constantly to update my references 
database with zero effort:

$ getbibtex.py NNNNN >> references.bib

and I'm done. Coupled to Pybliographic (also written in python, incidentally) 
I have fabulous reference management with perfect coupling to LyX.

It's specific to my needs, but shows how a simple code snippet can fetch very 
useful info off the internet, and could easily be modified for other 
purposes.

Cheers,

f.

# code follows:

#!/usr/bin/env python
"""Get BibTex references from preprint archives supported by SPIRES.

Usage:

  getbibtex preprint1 [preprint2 ....]

The list of preprints can also be fed in from standard input.

Preprint references can be given as:

 - name/number, as in hep-lat/0104015.

 This sets the archive name to 'name', and subsequent references can be given
 only as numbers. That is:

   hep-lat/n1 n2 n3 hep-ph/n4 n5
   
 will fetch n1,n2,n3 from hep-lat and n4,n5 from hep-ph.

 - numbers only. In this case, a preset default is used (you can change this
 value by editing getbibtex.py). The current default is: `%(def_archive)s`.

All archive names supported by SPIRES will work.
 
Output is given as a list of bibtex references, so you can simply do:

  getbibtex .... > mybiblio.bib

and mybiblio.bib will be a valid BibTex file.
"""

__author__ = "Fernando Perez <fperez at pizero.colorado.edu>"

#---------------------------------------------------------------------------
# Global defaults
def_archive = 'hep-lat'

import re,urllib,sys

#---------------------------------------------------------------------------
# Function definitions
def make_url(archive,number):
    """Make a SPIRES search url given an archive name and preprint number.

    make_url(archive,number) -> url string."""

    return ('http://www.slac.stanford.edu/spires/find/hep/www?'
            'eprint=%s&eprint=%s&format=wwwbriefbibtex'
            % (archive,number) )

def build_preprints_list():
    """Build the list of (archive,number) pairs from input parameters."""
    if len(sys.argv)>1:
        input_list = sys.argv[1:]
    else:
        input_list = sys.stdin.read().splitlines()
    pp_list = []
    archive = def_archive
    for item in input_list:
        if '/' in item:
            archive,number = item.split('/')
            pp_list.append([archive,number])
        else:
            pp_list.append([archive,item])
    return pp_list

#----------------------------------------------------------------------------
# Main code

# give user help if requested with -h... or --h...
try:
    if sys.argv[1].startswith('-h') or sys.argv[1].startswith('--h'):
        print __doc__ % globals()
        sys.exit()
except IndexError:
    pass

bibtex_re = re.compile( r'(@Article{.*})',re.DOTALL)

for archive,number in build_preprints_list():
    search_page = urllib.urlopen(make_url(archive,number)).read()
    try:
        print bibtex_re.search(search_page).group(1)
    except:
        print >> sys.stderr, 'ERROR: Failed to get results 
for:',archive,number



From nas at python.ca  Sat Sep  7 13:49:57 2002
From: nas at python.ca (Neil Schemenauer)
Date: Sat, 7 Sep 2002 10:49:57 -0700
Subject: Daemonize?
In-Reply-To: <20020907161037.GA70503@hishome.net>
References: <F7DB8D13DB61D511B6FF00B0D0F06233022B66EC@mail.hebmex.com> <mailman.1031184848.32071.python-list@python.org> <unjt47pc5r8671@corp.supernews.com> <20020907161037.GA70503@hishome.net>
Message-ID: <20020907174957.GA25058@glacier.arctrix.com>

Oren Tirosh wrote:
>     devnull = os.open('/dev/null', 0)
>     os.dup2(devnull, 0)
>     os.dup2(devnull, 1)
>     os.dup2(devnull, 2)
> 
> This is safer because it prevents these file descriptors from being reused.

Better:

    os.close(sys.__stdin__.fileno())
    os.close(sys.__stdout__.fileno())
    os.close(sys.__stderr__.fileno())
    os.open('/dev/null', 0)
    os.dup(0)
    os.dup(0)

 pedanticly-y'rs Neil



From marklists at mceahern.com  Mon Sep 30 11:52:08 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 30 Sep 2002 10:52:08 -0500
Subject: Question about __getattribute__/__getattr__
In-Reply-To: <Pine.LNX.4.44.0209301035390.1837-100000@mercury.sabren.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOIEJDECAA.marklists@mceahern.com>

>         if cached_value == None:

This is an aside, but comparison to None should be via identity rather than
equality; e.g.,

  if x is None:

or:

  if x is not None:

Why?  Because you may be comparing an object that barfs when compared to
None via its own __eq__ method.  It's a subtle point and I don't think I'm
explaining it very well.  Suffice it to say, if you really are comparing to
None, use is/is not.  If you want to test whether an object is/is not zero,
then use:

  if x:

or:

  if not x:

// m




From mak at trisoft.com.pl  Mon Sep 16 07:50:26 2002
From: mak at trisoft.com.pl (Grzegorz Makarewicz)
Date: Mon, 16 Sep 2002 13:50:26 +0200
Subject: upload/download to a palm os device with python
References: <24165ed4.0209160304.7a3c08af@posting.google.com>
Message-ID: <3D85C582.50101@trisoft.com.pl>

Solon Kazantzidis wrote:
> 
> 1. Is there a python library which I can use in order to access a palm
> device and work on the databases (create new, delete existing, browse
> existing) stored on it?
> 

sf.net/projects/pilot-link - for Palm palms with http://www.pyrite.org/ 
for offline database manipulations.

wincerapi for Windows CE devices

mak




From roy at panix.com  Mon Sep 16 09:56:34 2002
From: roy at panix.com (Roy Smith)
Date: Mon, 16 Sep 2002 09:56:34 -0400
Subject: Converting Python app to C++ completely
References: <aklm6c$r1r$1@news.swman.net.uk> <3D85D9EB.9020102@thinkware.se>
Message-ID: <roy-2A7271.09563416092002@reader1.panix.com>

Magnus Lycka <magnus at thinkware.se> wrote:
> A common approach is to factor out the performance critical
> parts (once you have actually measured) and rewrite them in
> C or C++. The profiler is a good tool for this measuring.

I'm in the process of doing exactly that.  I've got an app I wrote in 
Python which needed speeding up.  Profiling showed that 99% of the CPU 
time was spent in one class, which did a lot of low-level character 
processing.  Mostly what it did was thrash around creating string 
objects and tearing them down again in a rather un-pythonic way.

Seemed like a perfect opportunity to learn C++, so I re-wrote the class 
in that language.  Somewhat surprisingly, I only got about a 2x speedup.  
I think the reason is because now instead of spending all my time 
allocating and deallocating Python strings, I'm doing the same with C++ 
strings :-)

I'm going to take a shot at re-writing it again using C-style arrays of 
characters.  It'll be interesting to see how much speedup I get.  I 
figure one of two things will happen; either the C version will be much 
faster than the C++ version, or it won't.  Either way, I figure I will 
have learned something about C++, and maybe about Python too :-)


From jeremy at alum.mit.edu  Thu Sep  5 15:30:13 2002
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 5 Sep 2002 12:30:13 -0700
Subject: newbie-question: Nested functions and variable scope
References: <7a7157fd.0209041503.7b9a0f44@posting.google.com>
Message-ID: <b0f083db.0209051130.7323be95@posting.google.com>

leoboiko at linuxbr.com.br (Leonardo Boiko) wrote in message news:<7a7157fd.0209041503.7b9a0f44 at posting.google.com>...
> Hi, I'm a Python newbie. I'm writing a program that uses expat. Some
> code looks like this:
> 
> def generate_main(filename="main.xml", (...)):
>     (...)
>     outfile = open (filename, "w")
>     in_reply = 0 # true if we are inside "reply" element
> 
>     def start_element(name, attrs):
>         if name == "reply":
>             in_reply = 1
>         elif not in_reply:
>             outfile.write("<" + name.encode(encoding) + ">")
   [...]
> If I put "in_reply" outside everything and use "global in_reply"
> inside each nested function, it works, but that's evil! What would be
> a better way of doing it? Do I need to use a class?

You'd be better off using a class.  You can't rebind a name in a
nested scope.  So there is no way to make in_reply a local variable in
generate_main() and rebind it in start_element().

Jeremy


From aicolburn at yahoo.com  Thu Sep 12 12:56:50 2002
From: aicolburn at yahoo.com (Alan C.)
Date: 12 Sep 2002 09:56:50 -0700
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <ac677656.0209111225.4a141bc5@posting.google.com>
Message-ID: <1a3a9ffa.0209120856.1de06f57@posting.google.com>

Your comments are well taken, Tom, but I think a little care is needed
when using IDLE for learning Tkinter, just because the
"root.mainloop()" command will hang IDLE in some circumstances.

I enjoyed this tutorial. I had only two other thoughts/suggestions for
you Steve.

First, perhaps include a brief explanation early on about why you're
creating the program within a class structure. As a Tkinter beginner
myself, I'm thinking all the widgets could be added within the root
frame, i.e., see the Tkinter notes at
http://bembry.org/tech/python/index.shtml. [The reasons become a
little clearer toward the end of the tutorial; I'm just thinking the
overall document might be clearer to readers if some explanation was
also included early on.]

The other addition that I think newbies would appreciate is in the
section of the tutorial where you're talking about events. I think it
would be educational to distinguish between, say,
"self.button1Click()" and "self.button1Click" I don't know for sure,
but I bet there are folks who get confused distinguishing between
them.

Again, thanks for putting this together. I read it carefully :-)

Al


From mikemccand at bigfoot.com  Wed Sep 25 10:28:39 2002
From: mikemccand at bigfoot.com (Mike McCandless)
Date: 25 Sep 2002 07:28:39 -0700
Subject: how to check if a function signature "matches" without calling the function?
References: <a76c1561.0209241305.454f7ae6@posting.google.com> <wCfk9.140788$ub2.3074390@news1.tin.it>
Message-ID: <a76c1561.0209250628.5e13765f@posting.google.com>

Fantastic -- thank you!  Python never ceases to amaze me ...

Mike

Alex Martelli <aleax at aleax.it> wrote in message news:<wCfk9.140788$ub2.3074390 at news1.tin.it>...
> Mike McCandless wrote:
> 
> > Does Python 2.2.1 make it possible for me to see some details about
> > the expected arguments that a function takes, without actually calling
> > the function?
> 
> Yes.  inspect.getargspec lets you learn all you need:
> 
> import inspect
> argnames, has_varargs, has_kdws, default_values = inspect.getargspec(func)
> 
> 
> > Specifically, if I have a function or method object, and I have a
> > tuple + dict of arguments I'd like to apply it on, is there a function
> > I can call that would check if the tuple + dict can be properly
> > "coerced" to the formal parameters?
> 
> You still have some work to do for your checking, but it doesn't
> seem to be much.  You need to check you supply each mandatory
> argument exactly once, and each optional argument 0 or 1 time;
> you can have other positional args only if has_varargs, and
> other named args only if has_kwds.  E.g. for clarity:
> 
> 
> class NoWay(Exception): pass
> 
> def times_supplied(argname, argposition, atuple, adict):
>     return (argname in adict) + (argposition < len(atuple))
> 
> def check_args(argnames, default_values, atuple, adict):
> 
>     number_mandatory = len(argnames) - len(default_values)
> 
>     for pos in range(len(argnames)):
>         name = argnames[pos]
>         n = times_supplied(name, pos, atuple, adict)
>         if n<1 and pos<number_mandatory:
>             raise NoWay, "Mandatory argument %s (%s) missing" % (name, pos)
>         elif n>1:
>             raise NoWay, "Argument %s (%s) given twice" % (name, pos)
> 
> def check_funcargs(func, atuple, adict):
>     import inspect
>     names, has_varargs, has_kdws, defaults = inspect.getargspec(func)
>     check_args(names, defaults, atuple, adict)
>     if not has_varargs and len(atuple)>len(names):
>         raise NoWay, "Supplied %d positional args, max is %d" % (
>             len(atuple), len(names) )
>     if not has_kwds:
>         for name in adict:
>             if name not in names:
>                 raise NoWay, "Unknown argument %s" % name
> 
> 
> You can express this more concisely &c, of course, or use other
> means than raising a dedicated exception to diagnose problems
> (I only diagnose the first problem if any, that can be tweaked
> too, of course).
> 
> 
> Alex


From gumuz at looze.net  Tue Sep  3 04:14:02 2002
From: gumuz at looze.net (Gumuz)
Date: Tue, 3 Sep 2002 10:14:02 +0200
Subject: what about things like __*** ?
References: <7153651c.0209021927.16a3efaf@posting.google.com> <al1p8n$12d$1@wanadoo.fr>
Message-ID: <3d746df6$0$222$4d4ebb8e@news.nl.uu.net>

ah great, I found asking this myself last night, too.

isn't this the same for:   __methodName(self, parm) which makes it a private
method?



"Eric Brunel" <eric.brunel at pragmadev.com> wrote in message
news:al1p8n$12d$1 at wanadoo.fr...
> black wrote:
> > Hi, all~
> >
> > when reading the tutorial I met some trouble.(Maybe trifle to you all
> > but really trouble to me)
> >
> > That is I couldnt figure out what __*** is. Here is the description of
> > that article but even confused me:
> >
> > There is limited support for class-private identifiers. Any identifier
> > of the form __spam (at least two leading underscores, at most one
> > trailing underscore) is now textually replaced with _classname__spam,
> > where classname is the current class name with leading underscore(s)
> > stripped.
> >
> > What's it trying to say please ???
>
> As a good practice in object-oriented programming, the structure of a
> class, i.e. its actual attributes, are often hidden from the outside
world.
> It's called "encapsulation". It helps to separate the interface of the
> class from the actual implementation of the methods, that shouldn't be
> known outside the class itself.
>
> Languages like C++ or Java have a native support for encapsulation, by
> allowing to declare attributes "private" or "protected". A value of a
> private or protected attribute can't be read or set by another class than
> the owner class (or one of its sub-classes for protected attributes).
>
> IIRC, before Python 2.0, there were no support at all for encapsulation in
> Python. So you could always do:
>
> class C:
>   def __init__(self):
>     self.a = 12
> o = C()
> o.a = 13
> print o.a
>
> The attribute "a" of instances of class C can be read or modified outside
> the class, and there's no (simple...) means to prevent that.
>
> From Python 2.0, there's a (limited) support for encapsulation: if the
name
> of an attribute begins with a double-underscore, it's just as if it had
> been declared private:
>
> class C:
>   def __init__(self):
>     self.__a = 12
>   def get_a(self):
>     return self.__a
> o = C()
> o.__a = 13       # Does not work
> print o.__a      # Again, does not work
> print o.get_a()  # Works
>
> class CC(C):
>   def __init__(self):
>     C.__init__(self)
>     print self.__a    # Does *not* work
>
> The attribute __a is private: it can be seen only in the class C. It's
> impossible to read or set it from outside the class, even in one of its
> sub-classes. This sounds like a big restriction, but is actually very
often
> used in OO programming.
>
> The actual mechanism involves a "name-mangling", and that's what the
> comment tried to explain. But it's not that important: the final result is
> just that any attribute beginning with a double-underscore is private.
>
> HTH
> --
> - Eric Brunel <eric.brunel at pragmadev.com> -
> PragmaDev : Real Time Software Development Tools -
http://www.pragmadev.com




From moc.q-dnan-p at p-nand-q.com  Mon Sep  2 01:02:23 2002
From: moc.q-dnan-p at p-nand-q.com (Gerson Kurz)
Date: Mon, 02 Sep 2002 05:02:23 GMT
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <un50ccotaak86b@corp.supernews.com>
Message-ID: <3d72f007.1990343@News.CIS.DFN.DE>

On Sun, 01 Sep 2002 21:07:24 -0000, "A.M. Kuchling" <amk at amk.ca>
wrote:

>The problem with a new syntax is that no one else would be using it, so
>you'd still need to learn the existing syntax for use with grep, vi, Perl,
>&c.  (It wouldn't surprise me if Perl 6's revised regexes run into this very
>difficulty and don't gain much adoption.)

A tounge-in-cheek-answer: I don't use grep, I use python. I don't use
vi, I use python. I don't use perl, I use python :) 

Seriously, my thinking was, the re.compile function is there to
compile an expression to a binary representation for optimized
searching. So maybe, a "clean syntax" -> "ugly re syntax" compiler
would be good? 

Besides, you could use the output ("ugly re syntax") in grep, vi, perl
if you really so intended. 




From DennisR at dair.com  Tue Sep 17 17:26:21 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Tue, 17 Sep 2002 21:26:21 GMT
Subject: Python comment stripper
Message-ID: <16Nh9.248$Wc6.17087796@newssvr13.news.prodigy.com>

Does anyone have a Python source comment stripper or can anyone recommend a
parser that could be adapted for same?

Why?  ... well there is a better than 2:1 byte size reduction in both raw
and compressed .py files if comments are stripped.  Results of analyzing
dospath.py in :

                               raw Kb    compressed Kb
                             ----------    --------------------
original                    10.44               2.98
4 spaces -> 1 space    8.85               2.86
+ strip comments       4.65               1.34

--

Dennis Reinhardt

http://www.dair.com




From loewis at informatik.hu-berlin.de  Mon Sep  9 05:17:13 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 09 Sep 2002 11:17:13 +0200
Subject: Computing directory sizes problem
References: <d32d294c.0209061035.3e6fa400@posting.google.com>
Message-ID: <j4fzwjbjza.fsf@informatik.hu-berlin.de>

nmichael at yahoo.com (Neophytos Michael) writes:

> Dear Python experts,
> I wrote the following function to compute the size of a directory
> (being new to python there maybe better ways to do this - if so please
> advice).  It works quite well except when the directory in question
> contains files whose names that are not made up of ascii (english)
> characters (this typically happens on my machine for my non-english
> mp3 files).

What Python version are you using? What localized Windows version?
What are the file names in question? 

If the file names are non-ASCII, it should still work - provided the
file names are supported in the system locale (i.e. ANSI code page).

It might be easiest to adjust your system locale.

Regards,
Martin


From cmay4 at yahoo.com  Fri Sep 20 16:30:18 2002
From: cmay4 at yahoo.com (Chuck May)
Date: Fri, 20 Sep 2002 20:30:18 GMT
Subject: Python Sybase slow compared to Perl?
Message-ID: <Xns928FA7E6BA09Dcmay4yahoocom@165.112.130.8>

I have always written my database scripts in Perl.  I am very interested 
in Python, and would like to start using it for all my scripting.  I am 
not trying to start a flame war, I'd just like to understand why a 
particular script I wrote runs so much faster under Perl.    The script in 
question takes roughly 3 times as long to run under Python.  It is very 
possible that there is a faster way to do this under Python.

I cut out most of the script, leaving only the part of the scrit that does 
all the work.  I am basically selecting all the fields from a table, 
looping over the results, and outputting the results in a fixed-width text 
file.  Seems simple enough.  

Here is the Perl code snippet:

   # output the actual data
   $sth = $dbh->prepare( "SELECT $select FROM $table->[0]" );
   $sth->execute();
   $num_recs = 0;
   $tempfile = "$dumpfile.tmp.gz";
   unlink $tempfile;
   open( DUMP_FILE, "| gzip -c > $tempfile" )
   while ( @data = $sth->fetchrow_array() ) {
       $line = "";
       for ( $i = 0; $i < scalar @lengths; $i++ ) {
           $line .= pack( "A$lengths[$i]", $data[$i] );
       }
       print DUMP_FILE "$line\n";
       $num_recs++;
   }
   close DUMP_FILE;
   $sth->finish();

and here is the Python code snippet:

   # output the actual data
   c = db.cursor()
   c.execute( "SELECT %s FROM %s" % ( select, table ) )
   num_recs = 0
   if os.path.isfile( "%s.gz" % dumpfile ):
       os.remove( "%s.gz" % dumpfile )
   file = os.popen( "gzip -c > %s.gz" % dumpfile, "w" )
   while 1:
       row = c.fetchone()
       if not row : break
    
       line = []
       for i in range( len( row ) ):
           line.append( "%-*s" % ( lengths[i], row[i] ) )

       print >> file, "".join( line )
       num_recs += 1

   file.close()
   c.close()

No matter what table how many records there are, the Python script always 
takes longer to run (usually by a facter of 3 or more).  

I am still a Python newbie, so I may be doing something wrong.  Has anyone 
else experienced similar results?  Is this due to a poorly optimized 
Sybase module?  I really want to use Python, and I don't mind giving up a 
little speed, but this is more than I can accept.

I ran the Python script using the profiler and the output is below.  Any 
help would be much appreciated.  Thanks.

Chuck May
IMS, Inc.

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

profile output:

 Ordered by: standard name

 ncalls  tottime  percall  cumtime  percall filename:lineno(function)
      1    0.000    0.000 1230.540 1230.540 <string>:1(?)
      4    0.000    0.000    0.000    0.000 Sybase.py:114(_get_diag)
      2    0.000    0.000    0.000    0.000 Sybase.py:141(_ct_errors)
      3    0.000    0.000    0.000    0.000 Sybase.py:148(_row_bind)
 900345   96.240    0.000   96.240    0.000 Sybase.py:166(_extract_row)
 900348  177.670    0.000  273.910    0.000 Sybase.py:180(_fetch_rows)
      3    0.000    0.000    0.010    0.003 Sybase.py:229(__init__)
      3    0.000    0.000    0.000    0.000 Sybase.py:246(__del__)
 900361   83.230    0.000  267.090    0.000 Sybase.py:258(_lock)
 900361   86.720    0.000  269.000    0.000 Sybase.py:262(_unlock)
      3    0.000    0.000    0.010    0.003 Sybase.py:312(close)
      3    0.000    0.000    0.000    0.000 Sybase.py:328(execute)
 900348  178.280    0.000  988.260    0.001 Sybase.py:365(fetchone)
      1    0.000    0.000    0.000    0.000 Sybase.py:408(fetchall)
      3    0.000    0.000    0.000    0.000 Sybase.py:451(_fetch_rowcount)
      3    0.000    0.000    0.000    0.000 Sybase.py:466(_start_results)
      2    0.000    0.000    0.010    0.005 Sybase.py:494(__init__)
      2    0.000    0.000    0.000    0.000 Sybase.py:522(_raise_error)
      2    0.010    0.005    0.010    0.005 Sybase.py:527(connect)
      2    0.000    0.000    0.000    0.000 Sybase.py:564(__del__)
      4    0.000    0.000    0.000    0.000 Sybase.py:570(close)
      3    0.000    0.000    0.010    0.003 Sybase.py:613(cursor)
      2    0.000    0.000    0.000    0.000 Sybase.py:618(execute)
      2    0.000    0.000    0.000    0.000 Sybase.py:640(_fetch_results)
      2    0.000    0.000    0.010    0.005 Sybase.py:667(connect)
      1    0.000    0.000 1230.540 1230.540 bsi_dump.py:176(main)
      3    0.000    0.000    0.000    0.000 bsi_dump.py:238(outputMessage)
      1  242.210  242.210 1230.520 1230.520 bsi_dump.py:39(dumptable)
      1    0.030    0.030    0.030    0.030 commands.py:50
(getstatusoutput)
      1    0.000    0.000    0.000    0.000 getopt.py:121(do_shorts)
      1    0.000    0.000    0.000    0.000 getopt.py:136(short_has_arg)
      1    0.000    0.000    0.000    0.000 getopt.py:35(getopt)
      1    0.010    0.010    0.010    0.010 getpass.py:18(unix_getpass)
      1    0.000    0.000    0.000    0.000 getpass.py:71(_raw_input)
      2    0.000    0.000    0.000    0.000 posixpath.py:194(isfile)
      1    0.080    0.080 1230.620 1230.620 profile:0(main())
      0    0.000             0.000          profile:0(profiler)
      2    0.000    0.000    0.000    0.000 stat.py:29(S_IFMT)
      2    0.000    0.000    0.000    0.000 stat.py:54(S_ISREG)
      2    0.000    0.000    0.000    0.000 string.py:121(join)
      2    0.000    0.000    0.000    0.000 threading.py:39(__init__)
1800738   60.630    0.000   60.630    0.000 threading.py:44(_note)
1800738   60.440    0.000   60.440    0.000 threading.py:581
(currentThread)
      2    0.000    0.000    0.000    0.000 threading.py:64(RLock)
      2    0.000    0.000    0.000    0.000 threading.py:69(__init__)
 900369  123.010    0.000  183.860    0.000 threading.py:81(acquire)
 900369  122.060    0.000  182.280    0.000 threading.py:99(release)



From max at alcyone.com  Fri Sep 13 15:12:08 2002
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 13 Sep 2002 12:12:08 -0700
Subject: transform list of int in a list of string??
References: <mailman.1031940917.9935.python-list@python.org>
Message-ID: <3D823888.AF9C6B6F@alcyone.com>

jubafre at brturbo.com wrote:

> How can i transform a list of integer numbers in a list of string
> numbers, like that:
> 
> x=[1,2,3,4] to y=['1', '2'...]????????

	map(str, L)

Or if you prefer list comprehensions,

	[str(x) for x in L]

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ To be a man means to be a fellow man.
\__/ Leo Baeck
    PyUID / http://www.alcyone.com/pyos/uid/
 A module for generating "unique" IDs in Python.


From mikael.lexen at spam.no.volvo.com  Thu Sep 26 10:08:50 2002
From: mikael.lexen at spam.no.volvo.com (Mikael =?iso-8859-1?Q?Lex=E9n?=)
Date: Thu, 26 Sep 2002 16:08:50 +0200
Subject: unicode woes
References: <amuhrk$4cn$05$1@news.t-online.com> <3d92d120$0$76313$e4fe514c@dreader4.news.xs4all.nl> <amuldd$bur$01$1@news.t-online.com> <s2eb6-433.ln1@atlas15.dnp.fmph.uniba.sk> <amv3o4$p1s$02$1@news.t-online.com>
Message-ID: <3D9314F2.1C30196A@spam.no.volvo.com>

You can find information about Python 2.3 in this document

http://www.python.org/dev/doc/devel/whatsnew/section-encodings.html

/Mikael

Ulli Stein wrote:

> Radovan Garabik wrote:
>
> > Ulli Stein <mennosimons at gmx.net> wrote:
> >  : But nevertheless you would have to use everywhere the Python unicode()
> >  : function.
> >
> >  : Or how would you do this (nonsense string):
> >  : str = str + "bla?" + str[:5] + "??
> >
> >  : Would you write then:
> >  : str = str + unicode("bla?") + str[:5] + unicode("??)?
> >
> > I ran into the same problem and find out that python2.3
> > handles this.
> > Just add the line
> > # -*- coding: UTF-8 -*-
> > as the second line of your python script and then
> > you can use all your strings in UTF-8 encoding
> > (or any other, whatever fits you)
> >
>
> From where do you have this information? Is there some official changelog
> for python2.3?
>
> Thanks.



From edgebruce at yahoo.com  Mon Sep 30 18:51:40 2002
From: edgebruce at yahoo.com (Bruce Edge)
Date: Mon, 30 Sep 2002 15:51:40 -0700
Subject: Id, or stinky primo?
Message-ID: <7A4m9.25416$V64.5149538@newsfeed.slurp.net>

Is the Id it really worth the $$$$?
I've been looking at a kona stinky primo, and it seems like a comparable 
bike for less dough.
What's the general opinion out there? 
Why is the Id better?

-Bruce.



From wolfson at midway.uchicago.edu  Thu Sep  5 02:56:24 2002
From: wolfson at midway.uchicago.edu (Ben Wolfson)
Date: Thu, 05 Sep 2002 06:56:24 GMT
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <un50ccotaak86b@corp.supernews.com> <al0e96$oms$0@216.39.172.122> <4pid9.5883$e5.1033514@newsb.telia.net> <al5m4u$agv$0@216.39.172.122> <e0Dd9.6122$e5.1077418@newsb.telia.net>
Message-ID: <8vvdnucccqp2och24gigcmmc91njbjkhhk@4ax.com>

On Thu, 05 Sep 2002 06:41:14 GMT, "Fredrik Lundh" <fredrik at pythonware.com>
wrote:

>Bengt Richter wrote:
>
>> >you can ask SRE to dump the internal parse tree
>> >to stdout:
>> >
>> >>>> sre.compile("[a-z]\d*", sre.DEBUG)
>> >in
>> >  range (97, 122)
>> >max_repeat 0 65535
>> >  in
>> >    category category_digit
>> >
>> >turning this into 'English' is left as an exercise etc.
>>
>> Interesting, thanks. Does the above mean that sre can't fully match
>>  'a'+'9'*65537
>> ?
>
>in this context, 65535 represents any number:

Doesn't that cause problems for something like this?

>>> m=re.compile(r'\d{0,65535}a').match(('9'*1000000)+'a')
>>> len(m.group(0))
1000001

-- 
BTR
You're going to set me up as a kind of slovenly attached pig that
Jack Kornfeld can slice down in his violent zen compassion?
 -- Larry Block


From martin at v.loewis.de  Sat Sep 14 16:50:04 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 14 Sep 2002 22:50:04 +0200
Subject: Python11
References: <xwMg9.1433$Dy4.464@nwrddc03.gnilink.net>
Message-ID: <m3admk47pf.fsf@mira.informatik.hu-berlin.de>

"Jeff Sasmor" <jeff at sasmor.com> writes:

> Is there any definitive information on the next Python conference,
> Python 11?  I've read the conferences-discuss archive on
> Python.org (http://mail.python.org/pipermail/conferences-discuss/)
> and the discussion seems to be tapering off.  IIRC, by last year
> at this time there was already a call for papers.

Python11 will be colocated with the O'Reilly Open Source Convention
(OSCON), which will take place July 7-11, 2003 in Portland, USA.

I believe O'Reilly will organize the specific schedule, including
setting dates for the call for papers, submission deadlines, etc.
Given that this is still ten months away, I don't see any urgency.

HTH,
Martin


From GavinT at ResMed.com.au  Tue Sep 17 23:48:58 2002
From: GavinT at ResMed.com.au (Gavin Tomlins)
Date: Wed, 18 Sep 2002 13:48:58 +1000
Subject: PythonScript
Message-ID: <362133AB12B2CE4EA3C01E12CD6CCC6E0254E08B@osprey.corp.resmed.org>

Greetings, 

Firstly apologies if this is slightly out of the scope of the mailing list
but after searching the web, archives I've been unable to find references of
PythonScript with its use with the activescript engine. I own March Hammond
& Andy Robinson's Python Win32 Programming and have gone over there chapter
of active scripting though there examples focus more on python and
activescripting via the command shell. 

The problem I seem to be hitting is utilising python in applications that
embed an ActiveScript Engine. To date I am unable to get one python script
to execute. Other languages vbscript, jscript,perl etc, work in the same
environment so I'm left wondering if its a configuration problem or
something that I might be overlooking. If anyone could give me pointers,
url's, snippets etc. it would be interested. The product I specifically
having problems is FinalBuilder (http://www.atozedsoftware.com) though have
seen this problem in other applications as well.

I've tried both ActiveState and Python distribution 2.2.1 with Win32
1.48all.

Regards

Gavin


Warning:  Copyright ResMed.  Where the contents of this email and/or attachment includes materials prepared by ResMed, the use of those materials is subject exclusively to the conditions of engagement between ResMed and the intended recipient.

This communication is confidential and may contain legally privileged information. By the use of email over the Internet or other communication systems, ResMed is not waiving either confidentiality of, or legal privilege in,the content of the email and of any attachments.
If the recipient of this message is not the intended addressee, please call ResMed immediately on  +61 2 9886 5000 Sydney, Australia.




From tjreedy at udel.edu  Wed Sep 25 10:56:15 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 25 Sep 2002 14:56:15 GMT
Subject: What function called this function ??
References: <98jk9.22340$0p1.391334@news2.ulv.nextra.no>
Message-ID: <j8kk9.271749$z91.11732486@bin3.nnrp.aus1.giganews.com>

"Thomas Weholt" <2002 at weholt.org> wrote in message
news:98jk9.22340$0p1.391334 at news2.ulv.nextra.no...
> Is it possible, by hacking the traceback/stack/etc, to get
information about
> the function, that functions class and in what module that function
and
> class are defined, from inside a function located elsewhere?
Something like
> a way of answering the question of  "who called me?" without sending
the
> answer along as a parameter.

Yes.  Traceback hack has been posted previously.  If not in the FAQ,
you might find posts thru groups.google.com.

TJR




From andrea.baglioni at redturtle.net  Fri Sep  6 04:25:14 2002
From: andrea.baglioni at redturtle.net (andrea.baglioni at redturtle.net)
Date: Fri, 6 Sep 2002 10:25:14 +0200
Subject: .py to .pyc
Message-ID: <OF7755EF7E.E950A855-ONC1256C2C.002DD99D@redturtle.it>

Hi everybody,

I'm a python newbie and i've a problem.

I want to compile a python file called File.py and i want to produce a file
called File.pyc, so i do

python File.py

I always find the following error:

Traceback (most recent call last):
  File "File.py", line 18, in ?
    import Globals, StructuredText, string, utils, re
ImportError: No module named Globals

Can anybody help me to find the problem? I think it's a problem of
$PYTHONPATH and $PYTHONHOME but i'm not sure.

Please...help!

Thanx in advance

Andrew





From martin at v.loewis.de  Mon Sep  9 01:39:28 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 09 Sep 2002 07:39:28 +0200
Subject: Unicode -> String problem
References: <20010708204337.MJGP11827.femail4.rdc1.on.home.com@jparlar> <3B48D1DB.C3F726C9@ActiveState.com> <mailman.1031520251.6883.python-list@python.org>
Message-ID: <m3lm6bvi0f.fsf@mira.informatik.hu-berlin.de>

pinard at iro.umontreal.ca (Fran?ois Pinard) writes:

> For now, I would presume that `codecs.ascii_encode(UNICODE_STRING)' does the
> same as `UNICODE_STRING.encode("ascii")' and consequently, we do not need to
> know about `ascii_encode'.  Is that right?

In the context of the OP's question, you'd use
UNICODE_STRING.encode("ascii", "replace"), or, perhaps,
codecs.getencoder("ascii")(UNICODE_STRING, "replace").

Regards,
Martin


From rjones at ekit-inc.com  Tue Sep 24 18:01:17 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Wed, 25 Sep 2002 08:01:17 +1000
Subject: Difference between MOD_PYTHON and the CGI module
In-Reply-To: <3d904321$0$234$4d4ebb8e@news.nl.uu.net>
References: <3d904321$0$234$4d4ebb8e@news.nl.uu.net>
Message-ID: <200209250801.17174.rjones@ekit-inc.com>

On Tue, 24 Sep 2002 8:55 pm, Gumuz wrote:
> Suppose I'd like to use python as a 'replacement' for PHP, what do I use?

Visit http://www.python.org/cgi-bin/moinmoin/WebProgramming and browse the 
offerings, then choose whatever solution feels best to you.


    Richard




From david at fielden.com.au  Fri Sep 13 00:05:36 2002
From: david at fielden.com.au (DG)
Date: Fri, 13 Sep 2002 14:05:36 +1000
Subject: problems with getpass.getpass() under Linux
In-Reply-To: <40197710-C6C9-11D6-B141-00039351FE6A@mac.com>
Message-ID: <008601c25ada$d1a5f580$3101a8c0@borg.fielden.com.au>

Greetings,

I have Slackware Linux 8.0 with the (as distributed) Python 2.0.1 installed.
When I import getpass and try to getpass.getpass() I get the following error
and the password is echoed:

"Warning: Problem with getpass. Passwords may be echoed."

Looking into getpass.py it does this:

try:
        import termios, TERMIOS

If that fails, it goes on to try to import Windows and then Mac-specific
modules, with a fallback to just calling raw_input (effectively).

I have TERMIOS.py in /usr/lib/python2.0/plat-linux2 but I do not have
termios.py anywhere.

I have the Python 2.2.1 tarball (not (yet) installed), it contains
TERMIOS.py and but not termios.py (altho it does have a termios.c).  I
notice that the 2.2.1 getpass.py only tries to import termios.

What do I need to do to have getpass.getpass() work correctly?

tia

Rowdy




From jacob at boris.cd.chalmers.se.cd.chalmers.se  Wed Sep 25 10:00:43 2002
From: jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen)
Date: 25 Sep 2002 14:00:43 GMT
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <3D7958DB.50903@nospam.free.fr> <3D87402E.6020700@thinkware.se>
Message-ID: <amsfib$cq1$1@nyheter.chalmers.se>

In article <3D87402E.6020700 at thinkware.se>,
Magnus Lyck?  <magnus at thinkware.se> wrote:
...
>My personal experience from many years of coding C++ etc
>with curlies and Python without, is that I tend to get
>lost a lot more in my C++ code than in my Python code.
>Not to mention the code of others... I've certainly had
>a number of debugging sessions where I had to help some
>mediocre C++ coder who couldn't match his own braces...
>The problem is usually that the blocks get to big... The
>fact that you can get away with incorrect indentation in
>the brace-languages makes it worse.
>
>Most modern editors have tools like indentation guides and
>folding that can make it easier to see block structures.
>
>Was it one of Beethoven's symphonies that ended so abruptly
>that the audience didn't understand when to applaud? It was
>considered a big failure at it's time, and it's considered
>one of the best works of a genious today...

I think there is a good explanation for why code with explicit delimiters
is harder to read and follow than normal Python code.

There is an absolutely outstanding series of books by Edward Tufte
about how to create charts, tables and other information presentation
graphics. In his books he introduces a measurable unit called "Chart junk".
This is the percentage of the ink used in the chart that does not convey
information. Ink does not convey information if it can be erased, or
if by changing the position of other ink in the diagram, you can leave the
ink out.

Tufte then goes on to practically show that the unit is a relevant one to
ease of understanding. The books are beautiful and should be required reading
for anyone who ever designs a user interface.

In any case, I think that there exists something we can call source-junk, which
is the programming language equivalent of chart-junk. It is all the "ink"
in the source code that does not add to your understanding of what the code
does. Block starts and ends can be represented by white areas. Redundant
parenthesis can be thrown away. Declarations are seldom needed by the
programmer in order to understand the program.

The most important parts of your program, for understanding how it works
is the names of your objects (variables, classes, methods, functions, modules).
Second to that is the operations you do on your objects - how you combine
them. That's about it. Please note that I didn't say anything about block
delimiters or parentheses in this. I think they were put into programming
languages to make it easy to write fast compilers. I also think that
one of the factors making C so popular was that the BEGIN/END of Pascal and
ADA were even more intrusive than {}, so people considered C to be much
more readable than the competitors.

The lack of source junk is certainly the biggest thing that attracts me to
Python. The rest of the benefits are fairly pedestrian. This one is the
really big win.

Jacob Hall?n

A


-- 


From gumuz at looze.net  Tue Sep  3 11:15:42 2002
From: gumuz at looze.net (Gumuz)
Date: Tue, 3 Sep 2002 17:15:42 +0200
Subject: Design question.... parent/child class
References: <3d74b021$0$234$4d4ebb8e@news.nl.uu.net> <3d74b283$1@news.sentex.net>
Message-ID: <3d74d0c9$0$225$4d4ebb8e@news.nl.uu.net>

I'm sorry, I don't think I understand it correctly.

Do you mean i should create a third Interface class just to communicate to
the other class? Isn't that then just the same but with an extra class?

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3d74b283$1 at news.sentex.net...
> Gumuz wrote:
> > hi all,
> >
> > i have a SessionManager class, which contains a list of Session
> > objects(instantiated from a Session class). I have a design problem.
> >
> > Each session object has a small 'queue' of messages for the specific
> > session. How can a session send a message to another session without
having
> > some kind of parent-reference to the SessionManager object?
> >
> > i am a bit puzzled, or is this parent-reference not such a bad idea
after
> > all?
>
> Increasing coupling is always bad, but you could do this in a way that
> doesn't really increase coupling (at least in Python).  The Session
> objects don't need to know about SessionManagers specifically, just
> about some thing that has a particular interface.  In this case, maybe
> it's a "post" method.  Pass in a reference to the SessionManager,
> or even the post method itself (a "bound" method, bound to a specific
> SessionManager instance), but don't do an "import sessionmanager"
> or anything like that in the session.py module itself.
>
> Note that passing messages between Sessions could be considered a
> function not appropriate for something called merely "SessionManager".
> Maybe you want to do this with another mechanism in parallel to this,
> or rename that class.  Personally, I wouldn't do that unless the
> whole design was starting to get complex.
>
> -Peter
>




From tdelaney at avaya.com  Wed Sep 11 19:37:17 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 12 Sep 2002 09:37:17 +1000
Subject: "str.contains(part)" or alternatives?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A590@natasha.auslabs.avaya.com>

> From: Stefan Schwarzer [mailto:sschwarzer at sschwarzer.net]
> 
> "if part in s" disturbs the pattern "if item in sequence" 
> somewhat but that has
> already been weakened by "if key in dict" (that I like, nonetheless).

In fact, this was the deciding factor in Guido changing his position on 'if
s1 in s2'. A string has sequence-like behaviour (e.g. it is iterable,
individual characters can be indexed), but also has its own behaviour (e.g.
substring containment, all elements are single-character strings).

Tim Delaney



From wlfraed at ix.netcom.com  Sun Sep  1 23:10:44 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Sun, 01 Sep 2002 20:10:44 -0700
Subject: curses: AttributeError: initscr
References: <un5dhs4f7qlte8@corp.supernews.com>
Message-ID: <krkuka.va3.ln@ix.netcom.com>

Cesar fed this fish to the penguins on Sunday 01 September 2002 05:52 
pm:

> 
> 
> I'm running python 1.5/2.2.1 under redhat linux 7.2.
> 
        So which is it when you get the error?

        I could be wrong (I've not kept track of when certain changes took 
effect) but does 1.5 understand library directories? On this system 
(Mandrake) curses is a directory with an __init__.py file.

        What does (from python) dir(curses) return?

-- 
--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From yaipa at aol.com  Wed Sep 18 17:59:24 2002
From: yaipa at aol.com (yaipa)
Date: 18 Sep 2002 14:59:24 -0700
Subject: When we say 'One obvious way...',  do we mean 'generally obvious' or does that matter?
Message-ID: <8d148763.0209181359.68cd789f@posting.google.com>

Code from mail thread:

http://groups.google.com/groups?dq=&hl=en&lr=&ie=UTF-8&frame=right&th=ef69520e5c
dd3170&seekm=mailman.1031971577.27090.python-list%40python.org#link1


I LIKE the below code snippet. While it might be Pythonic, it will
only be 'obvious' to a Python programmer with some experience under
his/her belt.

*** Comments begin again after code snippet ****
-------------------------------------------------------------------------

#!/usr/bin/env python

import inspect

def make_undefined(cls, attr):
    def undefined(self, *args, **kwargs):
        msg = "'%s' object has no attribute '%s'" % (cls.__name__, attr)
        raise AttributeError(msg)
    return undefined

def is_magic(attr):
    prefix = suffix = '__'
    return attr.startswith(prefix) and attr.endswith(suffix)

class Restricted(type):

    common = ['__class__', '__defined__', '__dict__', '__doc__',
              '__getattribute__', '__init__', '__new__']

    def __new__(cls, classname, bases, classdict):
        defined_key = 'defined'
        defined = classdict.get(defined_key, []) + Restricted.common

        # If a non-magic attribute from any base class is not in common or
        # defined, hide it with a method that raises a descriptive
        # AttributeError (mimicking Python).
        for base in bases:
            for k, v in inspect.getmembers(base):
                if not is_magic(k) and k not in defined:
                    classdict[k] = make_undefined(cls, k)
        return type.__new__(cls, classname, bases, classdict)

class RestrictedDict(dict):

    __metaclass__ = Restricted

    defined = ['put', 'get', 'keys', 'items', 'values']

f = RestrictedDict()
print f.keys
print f.keys()
f['a'] = 'b'
print f.keys()
try:
    f.clear()
except AttributeError:
    pass
print f.keys()

-------------------------------------------------------------------------
*** More Comments ***
-------------------------------------------------------------------------

The flip side of this would be to build up accessors with a root 
class that can only 'get' a key or keys and finish with a child
class that can get/set/add and destroy key(s) in the dictionary.  Then
the user of such a library can inherit from the accessor class that which 
fits his/her needs.  While this is practiced in other languages, I
don't see it much used in Python.  Does the 'one obverse' rule mean the
one way that is more 'Generic' or more 'Pythonic'? 

As a note, here are my accessors and a couple of data dictionary  
classes 
-------------------------------------------------------------------------
class StaticFinalDictionaryMethods:
   #1. class only allows the getting of dictionary names/values.

   # Private Class level Vars.
   def __init__ (self):
      pass

   def getKeys(self):
      return self._dictionary.keys()

   def getNamePairs(self):
      return self._dictionary.items()

   def getByKey(self, _key, _index = 'None'):
      # test to make sure key is valid
      if self._dictionary.has_key(_key):
         if _index == 'None' or (_index >= self.min and _index < \
                                    len(self._dictionary[_key])):
            if _index == 'None':
               return self._dictionary[_key]
            else:
               return self._dictionary[_key][_index]
         else:
            return 'None'
      else:
         return 'no_such_key'

class BasicDictionaryMethods(StaticFinalDictionaryMethods):
   #1. class allows the getting of dictionary names/values.
   #2. class allows the setting of dictionary values by hash name.

   # Private Class level Vars.
   def __init__(self):
      # Python requires us to explicity call the super class
      StaticFinalDictionaryMethods.__init__(self)

   #Class private Setter for Task startup initilization
   # Does not support updating single element of a tuple.
   def setByKey(self, _key, _value, _index = 'None'):
      # test to make sure key is valid
      if self._dictionary.has_key(_key):
         if _index == 'None':
            # set tuple here
            self._dictionary[_key] = _value
      else:
         print _key, " - ERROR, TaskCore.setByKey().error - no_such_key"

class DynamicDictionary(BasicDictionaryMethods):
   def __init__(self):
      BasicDictionaryMethods.__init__(self)   

   def addkey(self, _key, _value):
      self._dictionary.setdefault(_key,[]).append(_value)
      
# end accessors      
#---------------------------------------------------------------
# 
class ExitTerms(TaskDataAccess.StaticFinalDictionaryMethods):

   # Private Class level Vars.
   def __init__ (self):
      # explicit call to the super class
      TaskDataAccess.StaticFinalDictionaryMethods.__init__ (self)
      
      # private dictionary, only 'self.' class methods can access.
      self._dictionary = {
        'ttf'                       :   None,
        'exitKey'                   :  "unset_exitKey",
        'symptom'                   :  "unset_symptom",
        'message'                   :  "unset_message",
      }
      
class DynamicKeys(TaskDataAccess.DynamicDictionary):

   # Private Class level Vars.
   def __init__ (self):
      # explicit call to the super class
      TaskDataAccess.DynamicDictionary.__init__ (self)
      self.min = 0

      #  Build DataStructure and __init__ each '_key' in dictionary.
      # -------------------------------------------------------------

      # private dictionary, only 'self.' class methods can access.

      self._dictionary = {}


From jhorneman at pobox.com  Sun Sep 29 16:09:11 2002
From: jhorneman at pobox.com (Jurie Horneman)
Date: Sun, 29 Sep 2002 22:09:11 +0200
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com> <3d9739c7$0$19818$91cee783@newsreader01.highway.telekom.at>
Message-ID: <3d975d86$0$24792$91cee783@newsreader01.highway.telekom.at>

"Jurie Horneman" <jhorneman at pobox.com> wrote in message
news:3d9739c7$0$19818$91cee783 at newsreader01.highway.telekom.at...
> "Jurie Horneman" <jhorneman at pobox.com> wrote in message
> news:9ac02e81.0209290839.4f4d6475 at posting.google.com...
>
> However, I do stand by my criticisms of the Python docs.

No - I've changed my mind. Slightly. Although a few functions are
undocumented, the documentation of the imp module did contain the
information I wanted. Plus there's a rather good example which I should have
examined more closely... alright, some egg on my face there.

Jurie.





From ken at hotmail.com  Thu Sep 26 02:19:08 2002
From: ken at hotmail.com (Ken)
Date: Thu, 26 Sep 2002 16:19:08 +1000
Subject: How to send Email using Python?
Message-ID: <amu8qh$93n60$1@ID-49758.news.dfncis.de>

Hi all, I like to know how to send a HTML featured Email from Python? I know
how to do a text based version using SMTPLib, but not sure how to do it with
a HTML featured Email. Can someone help me? Can someone provide me with
sample code?

Thanks




From GavinT at ResMed.com.au  Wed Sep 18 19:26:25 2002
From: GavinT at ResMed.com.au (Gavin Tomlins)
Date: Thu, 19 Sep 2002 09:26:25 +1000
Subject: PythonScript
Message-ID: <362133AB12B2CE4EA3C01E12CD6CCC6E0254E097@osprey.corp.resmed.org>

Hi Mark,

thanks for the replay and apologies for mispelling your name. The
application that I'm trying to get working is FinalBuilder
(http://www.atozedsoftware.com/ a build/release tool. Would it be possible
to suggest a small thirdparty product support PythonScript and
ActiveScripting that would be good for a test harness so as to do
comparative testing ? Is there any fundamental difference between the Win32
1.48all and the ActiveState distribution ? I'm assuming that you are right
in terms of the application support as the internet explorer test pages
appear to work after registering the python activescript script.

Regards

Gavin

You located in Melb/Syd ?

-----Original Message-----
From: Mark Hammond [mailto:mhammond at skippinet.com.au]
Sent: Wednesday, 18 September 2002 10:58 PM
To: python-list at python.org
Subject: Re: PythonScript


Gavin Tomlins wrote:
> Greetings, 
> 
> Firstly apologies if this is slightly out of the scope of the mailing list
> but after searching the web, archives I've been unable to find references
of
> PythonScript with its use with the activescript engine. I own March
Hammond
> & Andy Robinson's Python Win32 Programming and have gone over there
chapter
> of active scripting though there examples focus more on python and
> activescripting via the command shell. 
> 
> The problem I seem to be hitting is utilising python in applications that
> embed an ActiveScript Engine. To date I am unable to get one python script
> to execute. Other languages vbscript, jscript,perl etc, work in the same
> environment so I'm left wondering if its a configuration problem or
> something that I might be overlooking. If anyone could give me pointers,
> url's, snippets etc. it would be interested. The product I specifically
> having problems is FinalBuilder (http://www.atozedsoftware.com) though
have
> seen this problem in other applications as well.
> 
It is likely to be some strange interaction between PythonScript and the 
host.  Python works fine in all MS hosts, and quite a number of 3rd 
party ones.  What application is it?

Mark.

-- 
http://mail.python.org/mailman/listinfo/python-list


Warning:  Copyright ResMed.  Where the contents of this email and/or attachment includes materials prepared by ResMed, the use of those materials is subject exclusively to the conditions of engagement between ResMed and the intended recipient.

This communication is confidential and may contain legally privileged information. By the use of email over the Internet or other communication systems, ResMed is not waiving either confidentiality of, or legal privilege in,the content of the email and of any attachments.
If the recipient of this message is not the intended addressee, please call ResMed immediately on  +61 2 9886 5000 Sydney, Australia.




From aleax at aleax.it  Tue Sep 24 08:12:50 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 24 Sep 2002 12:12:50 GMT
Subject: how to build up a guest book with python ?
References: <mailman.1032854050.26863.python-list@python.org>
Message-ID: <6FYj9.157588$pX1.5677010@news2.tin.it>

black wrote:

>     I'm learning python for almost 1 month and decide to build up a guest
>     book with it, but I really know nothing about web programming, any
>     tutorials for me please ?

Steve Holden's "Python Web Programming" (New Riders) should be
perfect for your needs -- it touches on all the technologies you
need to be acquainted with for the purpose, and gives a good
tutorial grounding in each of them.  (I'm biased: Steve is a
friend and I was a tech editor for the book -- but I have some
counter-bias to compensate, as I could be recommending my own
books instead of his, so it probably balances out:-).


Alex



From wyojustin at hotmail.com  Tue Sep 24 23:07:15 2002
From: wyojustin at hotmail.com (Justin Shaw)
Date: Wed, 25 Sep 2002 03:07:15 GMT
Subject: Removing all occurences of a character from a string.
References: <amr8nh$rd8$1@driftwood.ccs.carleton.ca>
Message-ID: <DL9k9.201596$216.7001230@bin4.nnrp.aus1.giganews.com>

> How can I remove the single quotes so that "'Hello World'" becomes "Hello
World"?

Try this one:

>>> s = "'Hello World'"
>>> print s.replace("'", "")
Hello World
>>>

Justin

> I have a string, say "'Hello World'", where Hello World is in single
> quotes(i.e., 'Hello World'). How can I remove the single quotes so that
> "'Hello World'" becomes "Hello World"?
> Thanks
> Sean Ross
>
>




From leader730 at hotmail.com  Sat Sep  7 16:03:57 2002
From: leader730 at hotmail.com (Leader)
Date: Sat, 7 Sep 2002 16:03:57 -0400
Subject: getting clients ip address
Message-ID: <r0te9.25654$6W3.867995@weber.videotron.net>

Hello, I am using Python for my website, more specifically Spyce+mod_python,
what I want to do is print out the clients ip address on the site so it can
say something like "your ip is:xxx.xxx.xxx.xxx" basicly the python version
of <?php echo getenv("REMOTE_ADDR"); ?>
I'm sure I saw something a while ago while browsing through the module docs
but I can't find it anymore.

thanks






From andremoy1 at terra.com.pe  Mon Sep  9 22:36:57 2002
From: andremoy1 at terra.com.pe (andremoy1)
Date: Mon,  9 Sep 2002 21:36:57 -0500 (PET)
Subject: Cellpadding
Message-ID: <20020910023657.3725612DCA6@smtp1.terra.com.pe>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020909/a7be9713/attachment.html>
-------------- next part --------------

-- Antivirus de Terra ha detectado un grave virus en "este mensaje". Con la finalidad de evitar su contagio, hemos procedido a borrar el virus --

border.scr is removed from here because it contains a virus.

Si tiene alguna consulta o comentario escr?banos a antivirus at terra.com.pe - Terra Networks Peru S.A. -
-------------- next part --------------
A non-text attachment was scrubbed...
Name: content14136-0.html
Type: application/octet-stream
Size: 654 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020909/a7be9713/attachment.obj>

From ahmad at NOSPAMbitbuilder.com  Mon Sep  9 16:11:46 2002
From: ahmad at NOSPAMbitbuilder.com (Ahmad Baitalmal)
Date: Mon, 09 Sep 2002 20:11:46 GMT
Subject: xml.minidom is stripping out my CRLF's in attrib values!!
References: <3D7C80C4.9070204@NOSPAMbitbuilder.com> <j4n0qr9yeo.fsf@informatik.hu-berlin.de> <3D7C8C76.8060700@NOSPAMbitbuilder.com> <Xns92849D1FAE9D8duncanrcpcouk@127.0.0.1>
Message-ID: <3D7D00AE.4080900@NOSPAMbitbuilder.com>

That is the solution I went with.
Just replaced all "\n" with "&#010;" and all "\r" with "&$013;".
Works fine now, but, I think I will try the xml:space solution also.

Thanks all!

Duncan Booth wrote:
> Ahmad Baitalmal <ahmad at NOSPAMbitbuilder.com> wrote in
> news:3D7C8C76.8060700 at NOSPAMbitbuilder.com: 
> 
> 
>>That's not what my problem is, that I know about ( crlf's between
>>nodes),, 
>>
>>Here is the deal:
>><cow>
>>    <hide value="spotted
>>with black and white"></hide>
>></cow>
>>
>>After the word "spotted" there is a crlf, -inside- the attribute
>>value. Sholdn't it be treated as part of the value?
>>
>>The value now comes stripped of that crlf.
> 
> 
> The XML specification, para 3.3.3 specifies that attribute values must
> be normalised. The normalisation will convert your newline to a space and 
> may also remove duplicate spaces:
> 
> 
>>3.3.3 Attribute-Value Normalization
>>Before the value of an attribute is passed to the application or
>>checked for validity, the XML processor must normalize the attribute
>>value by applying the algorithm below, or by using some other method
>>such that the value passed to the application is the same as that
>>produced by the algorithm. 
>>
>>All line breaks must have been normalized on input to #xA as described
>>in 2.11 End-of-Line Handling, so the rest of this algorithm operates
>>on text normalized in this way. 
>>
>>Begin with a normalized value consisting of the empty string.
>>
>>For each character, entity reference, or character reference in the
>>unnormalized attribute value, beginning with the first and continuing
>>to the last, do the following: 
>>
>>For a character reference, append the referenced character to the
>>normalized value. 
>>
>>For an entity reference, recursively apply step 3 of this algorithm to
>>the replacement text of the entity. 
>>
>>For a white space character (#x20, #xD, #xA, #x9), append a space
>>character (#x20) to the normalized value. 
>>
>>For another character, append the character to the normalized value.
>>
>>If the attribute type is not CDATA, then the XML processor must
>>further process the normalized attribute value by discarding any
>>leading and trailing space (#x20) characters, and by replacing
>>sequences of space (#x20) characters by a single space (#x20)
>>character. 
>>
>>Note that if the unnormalized attribute value contains a character
>>reference to a white space character other than space (#x20), the
>>normalized value contains the referenced character itself (#xD, #xA or
>>#x9). This contrasts with the case where the unnormalized value
>>contains a white space character (not a reference), which is replaced
>>with a space character (#x20) in the normalized value and also
>>contrasts with the case where the unnormalized value contains an
>>entity reference whose replacement text contains a white space
>>character; being recursively processed, the white space character is
>>replaced with a space character (#x20) in the normalized value. 
>>
>>All attributes for which no declaration has been read should be
>>treated by a non-validating processor as if declared CDATA. 
>>
> 
> 
> So the only way to get a newline into an attribute is to escape it in using 
> an entity reference.
> 


-- 
-  - -- ----  ----------------------------------------- --- -- -   -
*Ahmad Baitalmal***
*BitBuilder***
web: http://www.bitbuilder.com
-  - -- ----
-------------------------------------------------------------- --- -- -   -



From robin at alldunn.NOSPAM.com  Thu Sep 19 23:41:10 2002
From: robin at alldunn.NOSPAM.com (Robin Dunn)
Date: Fri, 20 Sep 2002 03:41:10 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <Xns928CED692541cliechtigmxnet@62.2.16.82> <6_Ph9.479711$q53.16315681@twister.austin.rr.com> <am9jj9$c2m$1@wanadoo.fr> <3d8868b1$0$28683@echo-01.iinet.net.au> <1Y0i9.508708$UU1.84704@sccrnsc03>
Message-ID: <pNwi9.117$FY2.85489@news.randori.com>

bap wrote:
> Does wx work with ActiveState Python? I tried to use and as soon as
> wxWindows imports, Python (2.2)
> bombs and your back out at Windows with no error message. On another group I
> was
> told that this is a known issue and wxWindows doesn't work with Active State
> Python.

It has trouble running within the PythonWin GUI, but works just fine 
with the Python interpreter itself.

-- 
Robin Dunn
Software Craftsman
http://wxPython.org  Java give you jitters?  Relax with wxPython!



From aleax at aleax.it  Thu Sep 26 14:43:16 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 26 Sep 2002 18:43:16 GMT
Subject: design question
References: <1ha6pukqdqell8f31q1k7rtor2rv5ogjs0@4ax.com>
Message-ID: <8zIk9.147208$ub2.3246077@news1.tin.it>

Gon?alo Rodrigues wrote:
        ...
>   Class1
>      |
>      |
>   Class2
> 
> Where the dashed line means a parent-child relationship. Now what I want
> is something like
> 
>   Wrapper1
>      |
>      |
>   Wrapper2
> 
> where Wrapper1 (Wrapper2) wraps Class1(Class2) and the inheritance
> reltionships between the several wrappers mirror those of the classes
> they wrap -- but as I said, the protocol that the wrappers have can be
> substantially different from the classes they wrap.
> 
> Leaving aside the question of the usefulness of doing all this, I would
> like to ask how would you proceed? Making Wrapper derive from Class is
> stupid since their protocols can differ widely and I would end up with
> lots of diamonds in the inheritance graph. On the other hand I could
> just do:
> 
> class Wrapper1(object)
>     def __init__(self, <more args>):
>         self.__wrap = Class1()
>         <more initialization code>
> 
> But then if I did the same with Wrapper2, it would be at least quirky
> since it inherits from Wrapper1 but it wraps Class1 <scratches head in
> puzzled bewilderment>

Don't use __wrap as it interferes with communication between Wrapper1
and Wrapper2; use _wrap, with a single underscore, so that such
communication within your own hierarchy of wrappers is unimpeded,
whild (by convention) you're still communicating to the outside
world that this is an internal implementation detail.

Declare all your wrapper's __init__ method with a similarly named
optional argument, a tuple that determines what factory callable
to call to get the object to be wrapped:

class Wrapper1(object):

    def __init__(self, <whatever args>, _wrap=(Class1,)):
        self._wrap = _wrap[0](*_wrap[1:])
        <more init code>


class Wrapper2(Wrapper1):

    def __init__(self, <whatever args>, _wrap=(Class2,)):
        super(Wrapper2, self).__init__(<whatever>, _wrap=_wrap)
        <more init code>


Alex



From jimb417 at NO*  Fri Sep  6 22:40:39 2002
From: jimb417 at NO* (Jim)
Date: Fri, 06 Sep 2002 20:40:39 -0600
Subject: PyQT picture display
Message-ID: <albp27030dq@enews4.newsguy.com>

Hello to all,

I've a widget with three buttons and a label. I want to display a JPEG 
picture when either of the first two buttons is clicked. When I run the 
code below, the widget displays on screen but when either button is clicked,
the picture does not appear( the text in the label disappears as it should 
). Can anyone tell me what I'm doing wrong here?

Thanks,
Jim

from qt import *
import sys

class Picts(QWidget):
        def __init__( self, parent = None, name = None ):
                QWidget.__init__( self, parent, name )

                global Picture1
                global Picture2

                Picture1 = QPixmap(  "!Pict1.jpg" )
                Picture2 = QPixmap(  "!Pict2.jpg" )

                self.label = QLabel( "Click to see a picture!", self )
                self.label.setGeometry( 50, 50, 300, 300 )

                button1 = QPushButton( "Picture 1",self )
                button1.setGeometry( 10, 10, 80, 30 )

                button2 = QPushButton( "Picture 2",self )
                button2.setGeometry( 100, 10, 80, 30 )

                button3 = QPushButton( "Quit", self )
                button3.setGeometry( 190, 10, 80, 30 )

                self.connect( button1, SIGNAL( "clicked()" ),  self.setPicture1 )
                self.connect( button2, SIGNAL( "clicked()" ),  self.setPicture2 )
                self.connect( button3, SIGNAL( "clicked()" ), qApp, SLOT( "quit()" ) )

        def setPicture1( self ):
                self.label.setPixmap( Picture1 )

        def setPicture2( self ):
                self.label.setPixmap( Picture2 )

app = QApplication( sys.argv )
w = Picts()
w.setGeometry( 100, 100, 500, 500 )
app.setMainWidget( w )
w.show()
app.exec_loop()

-- 
Registered Linux User #269187
http://counter.li.org



From tim.one at comcast.net  Fri Sep 20 15:33:25 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 20 Sep 2002 15:33:25 -0400
Subject: e vs exp()?
In-Reply-To: <7Hxi9.56777$216.2039933@bin4.nnrp.aus1.giganews.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEIMBFAB.tim.one@comcast.net>

[Terry Reedy]
> ...
> On the hypothesis that x**y is calculated as exp(y*log(x)),

Bad hypothesis.  The implementation of a decent x**y function is one of the
most difficult tasks in a platform's math library.  The "naive" expression
you give *can* be used if the library has an efficient way to do the
internal computations using extended precision.  Else it's a disaster:  if
all computations are done in the result's precision, log(x) introduces an
error of its own, then that error is effectively multipled by y, and then
the magnified error in that is used as an exponent!  As a rule of thumb, the
number of trailing bits in the result that are complete nonsense can be as
bad as the number of bits in the internal float exponent field (this is
specific to using that specific expression, of course).  However, a decent
modern libm supplies an x**y with error strictly less than 1 unit in the
last bit.  I've written such a beast without benefit of an extended hardware
precision; I wouldn't want to do it again <wink>.

WRT exp(y) vs math.e**y, math.e isn't the mathematical e, of course, and
whatever representation error math.e suffers for cramming e into 53 bits is
magnified by the "**y" part.  exp(y) is going to be more accurate, unless
the platform libm plain sucks (Ping Tak Peter Tang published a way to
compute exp(y) that's very efficient, uses only native precision, and has
worst-case error less than 0.53 units in the last place; I had the extreme
pleasure of collaborating with him on a math library in a previous life).




From steven_shaw at iprimus.com.au  Tue Sep 10 12:01:07 2002
From: steven_shaw at iprimus.com.au (Steven Shaw)
Date: Wed, 11 Sep 2002 02:01:07 +1000
Subject: dynamism
References: <66715c02.0209091923.265fe77e@posting.google.com> <Xns92855D48384F9duncanrcpcouk@127.0.0.1>
Message-ID: <3d7e209a_1@news.iprimus.com.au>

"Duncan Booth" <duncan at NOSPAMrcp.co.uk> wrote in message
news:Xns92855D48384F9duncanrcpcouk at 127.0.0.1...
> steven_shaw at adc.com (Steven Shaw) wrote in
> news:66715c02.0209091923.265fe77e at posting.google.com:
>
> >
> > I thought you could just extend an object's attributes at run-time
> > like this:
> >
> >>>> p = object()
> >>>> p.x = 1
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in ?
> > AttributeError: 'object' object has no attribute 'x'
>
> New style classes (which are those with 'object' or other builtin classes
> as a base class) do not all allow you to add arbitrary attributes:
>
> If you have an object which is a builtin type (object, int, dict, ...)
then
> you cannot add attributes. If you have an object which is a subclass of a
> builtin type, then you can add attributes unless the subclass defined a
> member __slots__.
>
> The __slots__ attribute in a class, if defined, lists all of the attribute
> names settable in instances of that class. If no __slots__ attribute is
> directly defined in a class then you may add arbitrary attributes (this
> applies even if __slots__ is set in a base class).
>
> An object of a builtin type, or with __slots__ defined in the class and
> all base classes does not have a __dict__ attribute.

Does this mean that the attributes are stored in a more efficient way?

Is there a way to implement "record-based" object layout rather than
"associative-array based" object layout?


Cheers, Steve.




From sholden at holdenweb.com  Mon Sep 30 21:41:08 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 30 Sep 2002 21:41:08 -0400
Subject: PROBLEM with form in cgi
References: <anas2k$i2b$1@lust.ihug.co.nz>
Message-ID: <dZ6m9.36062$F21.1411@fe02>

"CheapSkate" <gua81 at XXXyahoo.com> wrote in message
news:anas2k$i2b$1 at lust.ihug.co.nz...
> HI,
> I'm having problem trying to figure out what is wrong with my cgi script.
>
> here's the code that's creating the problem in the apple.cgi
>
The name of your script isn't very relevant. While it's good you tried to
remove extraneous code, unfortunately you may have gone too far. You often
find that this kind of message is actually due to an earlier error in the
source.

> print """<form action="delete.html" name="myform" method="post">"""
>                              ^
> it keeps on saying invalid syntax in the action(as indicated)

The caret is pointing at the dot in "delete.html"?

> and if I try to change it to this
> print """<form action="delete.html" name="myform" method="post">"""
>                              ^
Isn't this exactly the same as the lines above?

> it gives the error at action.
>
I would suggest looking for an unclosed triple-quote (""") above this
statement. That's typical of the kind of error that generates this anodyne
and not-very-helpful message. Given the causes, it would be difficult to
know what error message you *could* use to give the programmer more idea
what went wrong.

Good luck.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From andymac at bullseye.apana.org.au  Fri Sep 13 20:25:31 2002
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Sat, 14 Sep 2002 11:25:31 +1100 (edt)
Subject: Dynamic loading of c modules,question.
In-Reply-To: <1810446.1031909856@dbforums.com>
Message-ID: <Pine.OS2.4.32.0209141122170.1817-100000@tenring.andymac.org>

[posted & mailed]

On Fri, 13 Sep 2002, hpyhpy wrote:

{...}

> /****************************/
> example.i
>
> /* File : example.i */
> %module example
>
> extern int    gcd(int x, int y);
> extern double Foo;
> /****************************/
> And I do :
> #gcc -c example.c
> #swig -python example.i
> #gcc -c example_wrap.c -I/usr/local/include/python2.2/
> #gcc -G -Wl example.o example_wrap.o -o _examplecmodule.so

                                          ^
why the underscore here?

{...}

> So I wonder where is the "example.py"???
> So I can't "import example"
>
> Traceback (most recent call last):
> File "", line 1, in ?
> ImportError: No module named example

There would only be an example.py if you wrote it yourself.  SWIG
generates the extension module, which you can directly use without a
wrapper in Python.

if you rename _examplemodule.so to examplemodule.so, it should work.

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andymac at bullseye.apana.org.au  | Snail: PO Box 370
        andymac at pcug.org.au            |        Belconnen  ACT  2616
Web:    http://www.andymac.org/        |        Australia




From martin at v.loewis.de  Thu Sep 19 15:25:21 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 19 Sep 2002 21:25:21 +0200
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <mailman.1032329366.6813.python-list@python.org> <j4ptvbaapm.fsf@informatik.hu-berlin.de> <3D890FB0.4030408@something.invalid> <m3admeo4u1.fsf@mira.informatik.hu-berlin.de> <ambso1$4g8vh$1@ID-59885.news.dfncis.de> <j4fzw6tofh.fsf@informatik.hu-berlin.de> <mailman.1032441028.15013.python-list@python.org> <j4elbqrrex.fsf@informatik.hu-berlin.de> <mailman.1032458669.32270.python-list@python.org>
Message-ID: <m37khh93z2.fsf@mira.informatik.hu-berlin.de>

Christian Tismer <tismer at tismer.com> writes:

> > Academically, I don't think the results are unpredictable, as
> > micro-thread switch occur at predictable points in processing.
> 
> Yes, right now, but this will change pretty soon.

You mean, tasklet switches will occur at unpredictable times?

This sounds like a bad idea.

> I am not saying that the random module is not thread safe (although
> I'm not sure if it is). 

To quote from the docstring:

  Multi-threading note: the random number generator used here is not
  thread- safe; it is possible that two calls return the same random
  value.

> The point is: Microthread safety is more than thread safety, since
> microthreads don't care about thread locks, since they are still *in
> the same thread*.

The issues are the same: the values of data may change at points where
you don't expect them to change. Thread-safety means to be aware of
such changes, and protect against them, by means of synchronization.

> I have to do rigid locking of every unknown piece of code, until
> somebody explicitly claims some code to be switchable.

This, again, sounds terrible. I doubt it will help, and users will
complain.

Regards,
Martin


From gumuz at looze.net  Wed Sep  4 07:14:15 2002
From: gumuz at looze.net (Gumuz)
Date: Wed, 4 Sep 2002 13:14:15 +0200
Subject: encryption (passwords)
References: <3d75d115$0$221$4d4ebb8e@news.nl.uu.net> <7xd6ru2gfi.fsf@ruckus.brouhaha.com>
Message-ID: <3d75f0f0$0$230$4d4ebb8e@news.nl.uu.net>

Actually, I am trying to create a sort of simple instant messenger
server+client.

although this is just a play-learn-project and security is not really an
issue, i'd like to include encryption in it as well to get experience in
this, too.



"Paul Rubin" <phr-n2002b at NOSPAMnightsong.com> wrote in message
news:7xd6ru2gfi.fsf at ruckus.brouhaha.com...
> "Gumuz" <gumuz at looze.net> writes:
> > My limited understanding of encryption tells me that I need a 'key' to
> > decrypt stuff. So I figure that this key will bey written(hard-coded?)
> > somewhere in my Python-script and it needs to be known at both sides in
> > order to encrypt and decrypt. Somehow, I can't believe this is true. I
can't
> > get over the idea that if my application becomes open-source anyone
could
> > see the key and decrypt my messages.
> >
> > I am fairly certain that it probably doesn't work this way, but I have
> > really no idea. Cann someone clarify this for me?
>
> OK, first of all, that type of question is probably better asked on
> sci.crypt rather than the Python group.
>
> Second, a word of advice: don't try to implement encryption protocols
> with your present state of knowledge, and especially don't try to
> design protocols yourself.  It's extremely easy to make mistakes that
> mess up security.  Even experts make those mistakes, all the time.
>
> I'll sketch out answers to your questions but basically I think your
> best approach is to use an unencrypted protocol between your Python
> scripts, and tunnel it through an encrypting socket layer that someone
> else has already debugged.  The two most popular choices of encryption
> layers are SSL and SSH.  SSH has a "port forwarding" feature that lets
> you set up local ports on both machines, that communicate through an
> encrypted tunnel.  Your Python program would just open the local port
> and send unencrypted data to it, and the same data would appear at the
> other end transparently decrypted.  SSL is a fancier protocol that
> doesn't incorporate forwarding, but there's a program called stunnel
> that implements forwarding over SSL.
>
> Between the two, I'd say SSH is a easier to set up if you're just
> trying to secure a single socket pair between two machines.  If you're
> trying to have a bunch of machines that talk to each other, and/or
> they're not all controlled by the same people (e.g. you're a vendor
> that needs to exchange secure traffic with various different
> customers), you probably want to use SSL.  That type of network is
> called an extranet and various books have been written about how to
> build and manage them.  If you say more specifically what you're
> trying to do, I or someone else here might be able to advise you more
> usefully.
>
> Finally, if you're building any kind of secure system that's intended
> to protect valuable data, I recommend you read "Security Engineering"
> by Ross Anderson all the way through before deploying your system.  It
> will give you an idea of what you're up against and what kinds of
> measures you can take.
>
> Re your basic questions:
>
> Yes you're correct that you need a secret decryption key.  The key
> should be unique for your specific installation, NOT part of the
> program that gets distributed.  It's like the password that you log
> into your computer with.  The login program is distributed freely as
> source code, but your password is not part of it.  You install the
> program, THEN set your password for your specific account.
>
> Then there's the problem of how to store the key, so your program can
> use it.  There are several approaches, each with good and bad points:
>
>     1) Include the key, in cleartext form, as part of the program
>     configuration that gets loaded at program startup.  Good points:
>     convenient.  Bad point: anyone with access to the config file
>     (possibly gotten by breaking into your system) gets the key.
>     This method is not recommended if the data being protected is
valuable.
>
>     2) Use a passphrase as a key that you enter when you start the
>     program, so it stays in memory and doesn't get stored on disk.
>     Good point: better security (key not stored).  Bad points: you
>     have to enter the key every time you start the program.  If you
>     want someone else to start the program for you, you have to tell
>     them the passphrase.  If the program has to start without human
>     attention at reboot if the system crashes, you're hosed.
>
>     3) Encapsulate the key in a special hardware device that does the
>     cryptography.  This can range from a $5 smart card or USB dongle, to a
>     multi-kilobuck sealed module on your PCI or SCSI bus.  This is the
>     preferred method for high security applications (banking, etc.).  Good
>     points: best security of all, key can't be stolen without physical
>     access to the machine, can implement its own access controls, etc.
>     Bad points: special hardware required; cheaper devices are slow.




From david at no.westcontrol.spam.com  Wed Sep 11 11:17:07 2002
From: david at no.westcontrol.spam.com (David Brown)
Date: Wed, 11 Sep 2002 17:17:07 +0200
Subject: Multi-threading python
References: <almvvt$34p$1@news.netpower.no> <mailman.1031751975.3372.python-list@python.org>
Message-ID: <alnmcs$gdt$1@news.netpower.no>

"Michael Chermside" <mcherm at destiny.com> wrote in message
news:mailman.1031751975.3372.python-list at python.org...
> David Brown wrote:
> > I've just been reading a little about the GIL and threading in python,
and I
> > have a couple of questions that I hope someone can answer.
>
> I'm not an expert, but I'll try.
>
> > Am I right in thinking that threading in Python does not use the
underlying
> > OS's threads, but effectively implements its own scheduler?  Or does it
use
> > the OS's threads, but uses the GIL to ensure that only one is runable at
a
> > time?
>
> It uses the OS's threads, and then uses the GIL to ensure that only one
> is executing Python bytecodes at a time. However, other threads may be
> runable... they could be blocked waiting on I/O, or executing some C
> code from a module that releases the GIL before performing a
> long-running computation.
>

If C code (or other extensions) that take significant time, or block for
other reasons, release the GIL then there I suppose there is not going to be
much of a problem.  The point is to maximize the throughput, and avoid
threads blocking each other unnecessarily.

> > On a dual-processor system, does this mean that only one thread of a
Python
> > program can run at a time, even when both CPUs are free, so that two
> > seperate python interpreter processes are needed to take advantage of
the
> > two CPUs?
>
> Yes.

It sounds like that is only true for code that does a lot of work in pure
python (rather than with time-consuming, GIL-released C code).

>
> > What about when a Python thread is blocked for some other reason, such
as
> > while waiting for a file read - will that block other threads, or will
the
> > waiting thread definitely release the GIL and allow other threads to
run?
>
> Depends on what it's blocked for. If it's blocked for I/O, then the GIL
> is released and the other threads will run. I'm not sure what else it
> would block for that might be a problem. But you could certainly CREATE
> a problem if you tried... simplest way would be to write C code which
> ran for a long time WITHOUT releasing the GIL.
>
> It would be much more difficult to block things through badly written
> Python code, since after each bytecode the interpretor can switch to a
> different thread.
>
> > If the GIL does block all threads in this manner, are there any plans to
> > introduce more fine-grained locking to improve scalability, analagous to
> > locking in the Linux kernel?
>
> No. The general consensus is that eliminating the GIL would be quite
> difficult and that there's not really all that much to gain. (Most
> casual users don't have multi-processor systems; serious users of this
> sort probably use Numeric and/or put the tight loops into C anyhow.)
> However, if you know someone who's good with this stuff and wants to
> volunteer, I'm sure they wouldn't turn down working code (unless it
> introduced other problems).
>
> -- Michael Chermside
>

Thanks for the information - it was definitely helpful.  It looks like the
system used by Python will work fine in all but a few situations (SMP system
in which a lot of work is being done in Python itself, rather than in
extension modules - or in badly written extensions which don't release the
GIL before doing heavy work).  Such situations can be dealt with by running
two python processes - if a program really needs to do hard SMP work in
python, then that is a small price to pay since I image the GIL simplifies a
lot of things.  I had been worried that the Python threading scheme would
have caused more inefficiences (such as all threads stopping while a single
thread was blocked for I/O).




From fredrik at pythonware.com  Thu Sep  5 02:41:12 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 05 Sep 2002 06:41:12 GMT
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com> <eppstein-56D20B.14244204092002@news.service.uci.edu>
Message-ID: <c0Dd9.6121$e5.1077322@newsb.telia.net>

David Eppstein wrote:

> I'm sure there were good reasons for this decision, but it leads to ugly
> workarounds like making each nonlocal variable into a list and assigning
> to x[0] instead of x...

when you find yourself doing that, it's a sure sign that you're
no longer using python to write python code...

</F>




From kkuroda at crl.ucsd.edu  Sat Sep  7 01:51:30 2002
From: kkuroda at crl.ucsd.edu (Kow Kuroda)
Date: Fri, 6 Sep 2002 22:51:30 -0700
Subject: Larry Wall's comment on python...
In-Reply-To: <wkr8g7rn4c.fsf@yahoo.com>
Message-ID: <DB7BFD3A-C225-11D6-9F6D-00050287C0A6@crl.ucsd.edu>

On Friday, September 6, 2002, at 11:25 AM, Rod Stephenson wrote:

snipped
> "Python is cool to look at small bits of, but I think the "outline"
> syntax breaks down with larger chunks of code. I'm with Aristotle on
> the structure of discourse--a story should have a beginning, and
> middle, and an end. So should blocks"
snipped
> Any comments?

Well, Python code may look awkward --- if not awk-like ---
if you don't know how to customize the tab width in your
editor of choice. For example, vi and emacs indent by 8
characters by default, which I suspect is too wide for Python.
Some code with deep nesting wouldn't look very nice with
this default.

Kow




From fperez528 at yahoo.com  Sat Sep 28 21:15:38 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Sat, 28 Sep 2002 19:15:38 -0600
Subject: Lambdaizing Python Code
References: <mailman.1033238258.18496.python-list@python.org>
Message-ID: <an5k7r$6ds$1@peabody.colorado.edu>

Terry Hancock wrote:

> 
> Not that I want to sign up for the style police or
> anything, but ...
> 
> Would respectfully like to point out that this could
> as well be represented:
> 
> def plur(word, number):
>     if number > 1:
>         word = word + 's'
>     return '%d %s' % (number, word)

Well, not quite. You need 'if number !=1'. Otherwise this happens:
In [36]: plur('cat',0)
Out[36]: '0 cat'

when you want:

In [37]: plur0('cat',0)
Out[37]: '0 cats'

Anyway, I know that the above is equivalent to my lambda. I just like and 
often use lambda for quick, simple one-liners. The page originally referred 
to by the OP is a nightmare, and a funny example of abusing the lambda idea 
to death. But I don't mind them in simple cases, nor do I  find them 
particularly unreadable.

My original:

plur=lambda s,n:'%s %s' % (n,s+('','s')[n!=1])

is fine for me. I see exactly what it does in one pass of my eyes. That's why 
_I_ like lambdas, and where I use them. But this is absolutely a matter of 
_personal_ preference.

I'm a big defender of code clarity and readability over conciseness/cleverness 
(I'm an ex-perler, for good reason). It's just that in cases like the above, 
_I_ find the lambda form quicker to visualize mentally and overall simpler to 
my taste. Feel free to differ.

Cheers,

f.


From james.kew at btinternet.com  Tue Sep  3 17:31:33 2002
From: james.kew at btinternet.com (James Kew)
Date: Tue, 3 Sep 2002 22:31:33 +0100
Subject: Another stylistic question: passing on arguments
Message-ID: <al39nn$1mtarg$1@ID-71831.news.dfncis.de>

I have a class hierarchy in which the __init__ method must always have the
same signature: the classes are all validators for various sorts of file,
and by having them all expose the same initialisation and checking methods I
can happily use them polymorphically. So far so good.

However, sometimes in derived classes I need to initialise additional
instance data at __init__ time. Up to now, I've happily overriden __init__
in derived classes, done my derived-class-specific stuff, and then delegated
down to the base class __init__.

Today, though, I wanted to add an extra argument to the __init__ method. The
extra argument is only meaningful to the base class. So, I merrily added it
to the base class __init__, blissfully forgot to modify the derived classes
__init__s, ran it up, and of course it blew up at runtime as now some
classes take three parameters on __init__ and some take two.

Alex posted this a while ago, which I tucked away in my Inbox and which
burrowed itself into my subconscious:
http://groups.google.co.uk/groups?hl=en&lr=&ie=UTF-8&oe=UTF-8&selm=kNw%258.1
31213%24vm5.4318544%40news2.tin.it

particularly his advice that:
> In fact: when you wrap a library class, please CONSIDER accepting
> arbitrary keyword parameters and passing them up to the superclass's
> method that you're overriding-and-delegating-to.  Just in case this
> sort of thing happens tomorrow.  Pretty please.

My C/C++ background doesn't get me beyond varargs and default values as far
as variable argument lists go, but I'm starting to see now the power of the
positional and keyword argument * and ** syntax. I'm thinking that, had I
written my derived class __init__s to catch and pass on additional arguments
to the base class, my base-class-only change would have been transparent to
derived classes.

But is that just using a nifty Python tool because it's there? I suspect my
"same __init__ signature everywhere" requirement would be better met by
defining __init__ only in the base class and having it delegate out to an
empty do_init method which derived classes can override.


Um: I think I've just argued myself into the second solution. But I'll be
playing with the *args, **kwargs syntax to add it to my toolbox; and I
wanted to thank Alex for helping this newbie along the Python path.

--
James Kew
james.kew at btinternet.com




From sholden at holdenweb.com  Sat Sep  7 11:54:34 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 7 Sep 2002 11:54:34 -0400
Subject: Newbie's unsurmountable pb with IO stream trick
References: <aba30b75.0209020633.26ba4c70@posting.google.com> <3D738EE2.3A8756D@engcorp.com> <aba30b75.0209030009.3ccb0207@posting.google.com>
Message-ID: <xkpe9.15245$l_4.9955@atlpnn01.usenetserver.com>

"Spendius" <spendius at muchomail.com> wrote ...
> Peter Hansen <peter at engcorp.com> wrote in message...
>
> OK I forgot to say I'm trying to develop my stuff under Windoze.
>
> > at once here.  For one thing, why are you trying to use
> > global variables at all when the original code was in
> > Java, which doesn't have such things?
> Do you mean Java doesn't have global variables or I'm mistaken ?!?
>
Java maintains a much stricter control over named access to non-local
variables than Python does. If you get an "UnboundLocalError" this
invaribaly means you are trying to use a variable which has not previously
been bound to a value. At the very least this makes your program logic
questionable.

> > And what does not declaring variables have to do with
> > the "global keyword issue" (whatever that is)?
> I kept getting the error message
> # UnboundLocalError: local variable 'cpt' referenced before assignment
> where I really didn't expect to get such an error, and noticed in this
> forum a lot of guys who know Python complain and talk themselves about
> a "mess" (my 'cpt' var. is a counter incremented every time a method
> gets called and I must sometimes re-initialize it to 0 inside an 'if'
> statement: merely IMPOSSIBLE).
>
Not at all. The problem is that, if you are using a global variable, you
must either initialize it before first use outside the functions that refer
to it, or you must ensure that your first call to any such function will
initialize it. This is simply sound programming practice.

> > If you have an SQL database to access, why not use one of the
> > standard DBAPI modules to get to it?  You might be doing a lot
> > of extra work when you don't have to.
> And why not ? Anyway SQL*Plus contains plenty of built-in commands
> that allow you to format your output how you like: I'd like to
> keep that functionality. And let's say that I want to do that for
> the sole purpose of learning the handling of IO streams in Python !
>
Well, that's a bit like saying you want to learn how to handle irregular
verbs in Esperanto, since the Java streams concept doesn't transliterate
directly into Python. You are simply bringing too much of your mental Java
baggage into the Python domain, where it will get in your way rather than
helping you to solve problems.

As far as inefficiency goes, sure, you may wish to solve your problem by
interacting with an external SQL/Plus process even if there are better ways
to go (and there are, as Peter mentioned). However, you should not expect
Java-style solutions to be appropriate in the Python world, where we have
popen() and friends to talk to external processes.

> > If you're just learning Python, you've taken on a fairly
> > complicated and large task right at the beginning.  Why not
> I admit it. But I did *exactly* the same with Java (writing
> the app I'm trying to port) but never got stuck into problems
> that blocking...
>
Don't try to port the Java, try to re-implement it in Python :-)

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                        pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From bokr at oz.net  Mon Sep  9 10:50:53 2002
From: bokr at oz.net (Bengt Richter)
Date: 9 Sep 2002 14:50:53 GMT
Subject: python program to watch a file - example.
References: <mailman.1031486051.16547.python-list@python.org>
Message-ID: <alicgd$t85$0@216.39.172.122>

On Sun, 8 Sep 2002 07:52:56 -0400, "Wayne R" <wayne at tbnets.com> wrote:

>Sorry for the html,  I forgot to turn it off (as I usually turn it off
>as soon as I install a new system, forgot this time).
>
>The platform is Gentoo Linux,  I am writing a piece of code for the
>Twiggi project to patch in adduser support.
>
>
>|-> > >I am trying to write a python program that will run as a deamon
>and
>|-> > >watch a file for data to arrive and then wake up and perform
>duties
>|-> on
>|-> > >that data.   I found ways to make a python program run as a
>deamon
>|-> but I
>|-> > >am at a loss to find anything related to watching a file for data
>so
>|-> > >that it can then process the data.  I didn't really want to do a
>|-> sleep
>|-> > >loop or something like, I am sure there is a better way.
>|-> > >
>|-> > >Any ideas or links to more info?
>|-> > >
>|-> > Don't annoy people you want help from, for starters.
>|-> > What system are you targeting? (Outlook in the headers says you're
>on
>|-> windows
>|-> > generating this abominable mix of actual message and useless
>cruft).
>|-> >
>[Wayne R] 
>
Sorry 'bout that. I don't very often express myself so testily, I hope.

Anyway, if you are going to have to read new data periodically, you could just
poll by doing it, I suppose. Unless you can control the writer or have access to
something like Isaac describes, you might almost as well poll, ugly as that is.

E.g., this seems to work, and I made it back off if there's errors due to
permissions etc. The question would be if open/seek/tell/seek/read/close
could cause an unacceptable error for some writer trying to get exclusive
access or such. But it should be more robust than that, probably.

If you control the source, it's a different ballgame, of course.
Anyway, try this on Linux (mine is slackware 2.2.16 on an old p90 with Python 1.5.2 ;-)

--< monf.py >--
#!/usr/bin/python
def mon(fname, firstpos, period):
    import os, time, string
    last_len = firstpos
    next_errp = 1
    ioerr_ctr = 0
    while 1:
        f = None; data = ''
        try:
            f = open(fname)
            f.seek(0,2) # to EOF
            told = f.tell()
            if told < last_len:
                print 'File shortened from %s to %s. Starting from 0.'%(
                    last_len, told
                )
                last_len = 0
            f.seek(last_len)
            data = f.read()
            ioerr_ctr = 0    # fresh sequence
            next_errp = 1
        except IOError, e:
            ioerr_ctr = ioerr_ctr + 1
            if ioerr_ctr >= next_errp:
                next_errp = next_errp*2
                print '%s IOErrors: %s' % (ioerr_ctr, e)
        if f is not None: f.close()
        if len(data):
            print '----< %s bytes of new data >----\n%s' % (
                    len(data), `data`
            )
            last_len = last_len + len(data)  # use actually read length
            print '----< end, now totaling %s bytes >---' % last_len
        time.sleep(period)

if __name__ == '__main__':
    import sys, string
    try:
        fname, firstpos, period = sys.argv[1:4]
        firstpos = int(firstpos)
        period = float(period)
        mon(fname, firstpos, period)
    except Exception, e:
        print '%s: %s' % (string.split(str(e.__class__),'.')[-1], e)
        print 'Usage monf.py filename firstpos(byte offset) period(seconds)'
---------------

Running it locally via telnet from windows NT, and messing with
the file directly on the Linux local terminal on the desk at my left:

bokr at springbok:~/junk$ monf.py junk.txt 0 5
1 IOErrors: [Errno 2] No such file or directory: 'junk.txt'
2 IOErrors: [Errno 2] No such file or directory: 'junk.txt'
4 IOErrors: [Errno 2] No such file or directory: 'junk.txt'
----< 16 bytes of new data >----
'A line of data.\012'
----< end, now totaling 16 bytes >---
----< 22 bytes of new data >----
'Another line of data.\012'
----< end, now totaling 38 bytes >---
File shortened from 38 to 18. Starting from 0.
----< 18 bytes of new data >----
'Replaced totally.\012'
----< end, now totaling 18 bytes >---
KeyboardInterrupt:
Usage monf.py filename firstpos(byte offset) period(seconds)
bokr at springbok:~/junk$

Here I'll start monitoring at position 10, and then chmod it -+r
from the other computer.

bokr at springbok:~/junk$ monf.py junk.txt 10 5
----< 8 bytes of new data >----
'otally.\012'
----< end, now totaling 18 bytes >---
----< 35 bytes of new data >----
'Will now chmod it -r then +r again\012'
----< end, now totaling 53 bytes >---
1 IOErrors: [Errno 13] Permission denied: 'junk.txt'
2 IOErrors: [Errno 13] Permission denied: 'junk.txt'
4 IOErrors: [Errno 13] Permission denied: 'junk.txt'
----< 38 bytes of new data >----
'Ok now add some data to make it read.\012'
----< end, now totaling 91 bytes >---
KeyboardInterrupt:
Usage monf.py filename firstpos(byte offset) period(seconds)
bokr at springbok:~/junk$

(I hit Ctrl-C via telnet to kill it).

This is just simple a hack, not tested beyond what you see. YMMV with
your environment. You can mod to suit your task, hopefully. It
would have been better to make a directory keyed by error strings
so unique info wouldn't get supressed in the total. Plus a time-tagged
log file would be better than a scrolling screen. Etc., etc., but
it was an experiment. You could also wath the file size with

    int(string.split(os.popen('ls -ls' % fname).read())[4]) # or something like that ;-)

but I imagine that's a fair bit more overhead than open/seek/tell.
Of course it would also give you a chance to monitor other aspects
of the file, or a directory. I guess if you had the privileges and
the time, you could possibly create a special file system that you
could symlink the file into and that would provide side effect hooks
to block on and read event-driven special data. Maybe that's what sgi did?

Regards,
Bengt Richter


From bdesth at nospam.free.fr  Mon Sep 23 00:38:16 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Mon, 23 Sep 2002 00:38:16 -0400
Subject: Overriding all methods in a class
References: <mailman.1032532107.31766.python-list@python.org>
Message-ID: <3D8E9AB8.9090000@nospam.free.fr>

Joao Prado Maia wrote:
> Hi,
> 
> I have been looking for an answer for this question for a while, but 
> couldn't find anything that would work for what I want to do.
> 
> I know that you can use __getattr__ to overload methods that your class 
> don't have, but what about overloading all methods, so I could in a way 
> 'intercept' calls to my methods and do something different if I wanted, is 
> that possible ?
> 
> What I want to do basically is create a caching system for a specific 
> class, and have __getattr__ (or any other magic method name) handle the 
> decision to sue the cached return value or to actually call the method.
> 
> Maybe it would be just easier to rename my current class to _Classname and 
> create a new Classname with just __getattr__.
> 
> Any suggestions ?
> 
> Cheers,
> Joao
> 
> 

from stupidQ import whyBlackMagic

Well, if it's only about caching, why do you need to go into 'black 
magic' to implement it ? Couldn't you just handle the cache etc in the 
method itself ?


laotseu



From Oschler at earthlink.net  Wed Sep 11 12:18:25 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Wed, 11 Sep 2002 16:18:25 GMT
Subject: Any C++/Python sig members here?
Message-ID: <l1Kf9.1470$Lg2.157618@news2.news.adelphia.net>

I joined the C++ sig for Python, to talk with others about BoostPython.  I
joined more than a day ago and got my subscription confirmation request from
the sig server.  I have not seen any messages yet, is it a very low traffic
sig or something?

thx





From aglyportREMOvethispart at nospam.yahoo.com  Wed Sep  4 15:09:57 2002
From: aglyportREMOvethispart at nospam.yahoo.com (Anton Graph)
Date: Wed, 04 Sep 2002 12:09:57 -0700
Subject: Memory fault - core dumped   while compiling 2.2.1 on SCO
References: <3D757EFA.9050401@nospam.yahoo.com> <yu99sn0qqw1r.fsf@europa.research.att.com>
Message-ID: <3D765A85.9080904@nospam.yahoo.com>

Andrew Koenig wrote:
> Anton> Any ideas?
> 
> Looks like the same symptom I had.
> 
> If you are using binutils 2.13, try going back to 2.12.1 and
> see if that fixes the problem.
> 

I don't think I use binutils:


Wed Sep  4 11:37:53 PDT 2002 bongo:/u5/src/Python-2.2.1 $ as --version
Assembler: (null)
  aline 1        : Warning: Illegal flag (-) - ignored
  aline 1        : Warning: Illegal flag (v) - ignored
  aline 1        : Warning: Illegal flag (e) - ignored
  aline 1        : Warning: Illegal flag (r) - ignored
  aline 1        : Warning: Illegal flag (s) - ignored
  aline 1        : Warning: Illegal flag (i) - ignored
Nonexistent file



Wed Sep  4 11:37:48 PDT 2002 bongo:/u5/src/Python-2.2.1 $ ar --version
ar: ERROR: Illegal option -- -
Usage: ar [-V?] -key[arg] [posname] afile [name ...]
   where key[arg] is one of the following: r[uabi], m[abi], d, q, t, p, x


Can't think of other utils in binutils.



From markus.vonehr at ipm.fhg.de  Tue Sep 10 06:05:49 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Tue, 10 Sep 2002 12:05:49 +0200
Subject: compile error on SWIG file
References: <3D7C55EE.D9278B1C@ipm.fhg.de> <3D7D97A6.A63610AF@ipm.fhg.de>
Message-ID: <3D7DC3FD.815BF6E0@ipm.fhg.de>

I put the file in the DLL directory.

In a pyhon shell I typed:

import mvelib
Traceback (most recent call last):
  File "<pyshell#2>", line 1, in ?
    import mvelib
SystemError: _PyImport_FixupExtension: module mvelib not loaded


So python doesn't find the module, I guess?
Any ideas?

Markus

Markus von Ehr schrieb:
> 
> Hi Lyle,
> 
> thanks for your answer and your efforts!
> I changed everything to .cpp
> The first error has gone
> Now I have an error (translated from german):
> 
> mvelib.obj: error LNK2001: unresolved external symbol initmvelib
> build\temp.win32-2.1\Release\mvelib.lib : fatal error LNK1120: 1
> unresolved reference
> LINK : fatal error LNK1141: Creation error of export file
> error: command 'c:\Programme\VC98\BIN\link.exe' failed with exit status
> 1141
> 
> Markus
> 
> Markus von Ehr schrieb:
> >
> > Hi,
> >
> > compiling my c-file generated from SWIG
> > (swig -c -python -o mvelib.c mvelib.i)
> > with a setup-file, I receive the error:
> >
> > c:\programme\vc98\include\eh.h(32) : fatal error C1189: #Error: "eh.h
> > is only for C++!"
> >
> > Anybody knows what I have to change???
> >
> > Thanks Markus
> >
> > ****************************
> > setup.py
> > ****************************
> > from distutils.core import setup, Extension
> >
> > setup (name = "mvelib",
> >         version = "1.0",
> >         maintainer = "Markus von Ehr",
> >         maintainer_email = "markus.vonehr at ipm.fhg.de",
> >         description = "mvelib Python module",
> >
> >         ext_modules = [Extension('mvelib',sources=['mvelib.c'])]
> >  )


From loewis at informatik.hu-berlin.de  Mon Sep 23 07:29:25 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 23 Sep 2002 13:29:25 +0200
Subject: UpToDate SOAP-module ??
References: <ATAj9.21315$0p1.376966@news2.ulv.nextra.no> <j4fzw1ghpv.fsf@informatik.hu-berlin.de> <vLCj9.24765$sR2.439699@news4.ulv.nextra.no>
Message-ID: <j4vg4xeyga.fsf@informatik.hu-berlin.de>

"Thomas Weholt" <2002 at weholt.org> writes:

> Anyway, if you want to provide more information about why I safely can
> ignore SOAP feel free. It would make my project a bit less complicated so
> I'd be glad to skip it. :-)

It all depends on what your users want to use; and I'm talking about
*real* users here, not people that just have wishlists, but don't
seriously plan to use your service.

I think that clients will prefer whatever makes development of their
applications simplest. Which technology is simplest depends on various
aspects, such as the complexity of the interface, the target language
for developing clients, and any libraries that client developers may
have.

If the interface is a simple one, clients might be happy with XML-RPC,
except for the result-parsing part (if they use Python or Perl, they
might be happy with XML-RPC even for results).

If the interface is complex and out-of-the-box interoperability is
important for unknown clients, I'd prefer CORBA over SOAP.

If you are not only developing the server, but also the client, I'd
prefer CORBA over SOAP also.

About the only scenario where offering SOAP might be needed is that
people want to develop clients with C#, in .NET, since there are no
.NET ORBs available, yet.

Regards,
Martin



From timh at zute.net  Wed Sep 25 09:30:11 2002
From: timh at zute.net (Tim Hoffman)
Date: Wed, 25 Sep 2002 21:30:11 +0800
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amabpb$5rm$1@news.nems.noaa.gov> <3D8B1537.4010803@thinkware.se> <ad052e5c.0209201128.1a53108@posting.google.com> <mailman.1032664289.615.python-list@python.org> <amo0aq$5nf$1@nntp2-cm.news.eni.net>
Message-ID: <3D91BA63.7090203@zute.net>

> 
> There ought to be severe penalties for idiots that use floating point
> dollars for financial applications.  If forced to use floating point
> (e.g. because customer demands BASIC), then keep money amounts in
> whole pennies (or whatever the smallest currency unit for the country
> is), and divide by 100 (or whatever) for printing only (or just add the
> decimal point yourself).
> 

In many circumstances this is exactly what you can't do.
The minute you need to calculate percentages you will get in to
the things like .0001254 of a cent, no matter what is you
smallest unit of currency.

I worked for a insurance company (using Business Basic) many
years ago on MAI and Prime gear, and we did everything as floats
with 14 places after the decimal point,and only rounded when
a human needed to see a number, and then used standard accounting
practices for rounding,

Rgds

Tim



From merman at snafu.de  Sat Sep 14 16:13:06 2002
From: merman at snafu.de (T. Kaufmann)
Date: Sat, 14 Sep 2002 22:13:06 +0200
Subject: newbie-question: abstract classes
Message-ID: <3D839852.2060405@snafu.de>

Hi everybody,

I have a problem:

I want to create an abstract class but it doesn't work.
What' s wrong here?

class Abstract:

     def __init__(self): return
     def method(self):   raise NotImplementedError

class Concrete:

     def __init__(self):
         Abstract.__init__(self)
         return
     def method(self):
         print "\n ... in Concrete!\n"

if __name__ == '__main__':

     #a = Abstract()
     #a.method()
     c = Concrete()
     c.method()

thanx for help

o-o

thomas



From jfaris at sbcglobal.net  Thu Sep 12 10:21:45 2002
From: jfaris at sbcglobal.net (Jordan Faris)
Date: Thu, 12 Sep 2002 07:21:45 -0700
Subject: Best tutorial
Message-ID: <00c901c25a67$b9dcecb0$e780cf3f@leader1>

Of the several fine tutorials on the Web, which one would be the most direct
in getting a novice up to snuff on Python?
Thanks.......




From jacob at boris.cd.chalmers.se.cd.chalmers.se  Wed Sep 25 10:03:32 2002
From: jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen)
Date: 25 Sep 2002 14:03:32 GMT
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <mailman.1031377930.4225.python-list@python.org> <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com> <3D8E6353.7030303@something.invalid>
Message-ID: <amsfnk$cq1$2@nyheter.chalmers.se>

In article <3D8E6353.7030303 at something.invalid>,
Greg Ewing  <greg at cosc.canterbury.ac.nz> wrote:
>Courageous wrote:
>
>> One should use spaces
>> and not tabs in Python code. Otherwise, any python file where
>> tabs and spaces are mixed is also a choice of which indentation
>> level other viewers of the code MUST use in their editors.
>
>
>If you use spaces only, you're also choosing how other
>people will see it in their editors (unless they're
>willing to use a tool to re-indent it).
>
>Using *tabs* only is the only way to get truly
>preference-independent indentation.

Exactly the point! My code is my work of art. Why should anybody be allowed
to change it? We don't allow people to go painting moustaches on the
Mona-Lisa, do we?

Jacob Hall?n

-- 


From LogiplexSoftware at earthlink.net  Wed Sep 11 17:31:52 2002
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 11 Sep 2002 14:31:52 -0700
Subject: deleting the first and the last character of a string
In-Reply-To: <1031777408.17116.65.camel@software1.logiplex.internal>
References: <20020911105818.GC2054@partagas.as.de.cw.net>
	<mailman.1031756234.22942.python-list@python.org>
	<alo5df$hoa$03$1@news.t-online.com> 
	<200209111246.03080.shalehperry@attbi.com> 
	<1031777408.17116.65.camel@software1.logiplex.internal>
Message-ID: <1031779912.17116.77.camel@software1.logiplex.internal>

On Wed, 2002-09-11 at 13:50, Cliff Wells wrote:

> >>> def chopd(s, d):
> ...     s = s.strip()
> ...     return s[s[0] == d : {0:len(s), 1:-1}[s[-1] == d]]

Er, slightly better:

          return s[s[0] == d : (len(s), -1)[s[-1] == d]]

(Note to self: dictionaries aren't answer to everything...)

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From pereira at cis.upenn.edu  Sat Sep 21 22:21:12 2002
From: pereira at cis.upenn.edu (Fernando Pereira)
Date: Sat, 21 Sep 2002 22:21:12 -0400
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com> <azFi9.163782$5r1.6708151@bin5.nnrp.aus1.giganews.com>
Message-ID: <B9B2A158.136B7%pereira@cis.upenn.edu>

On 9/20/02 9:40 AM, in article
azFi9.163782$5r1.6708151 at bin5.nnrp.aus1.giganews.com, "Terry Reedy"
<tjreedy at udel.edu> wrote:

> 
> "Benjamin" <phncontact.libero.it@> wrote in message
> news:ifjloug0dkla5o3lvli464mmou5mbh47d5 at 4ax.com...
>> i've been learning python not for becoming a programmer
>> (i study modern languages at the university), but for having
>> fun. my personal motto is "i want to know how it works".
> ['but I'm getting bored']
> 
> Perhaps you should look into some area of computer-aided language
> analysis and processing.

In that vein, some suggestions for further reading:

http://nltk.sf.net
http://www.cis.upenn.edu/~cis530/
http://www.norvig.com/python/python.html

-- F



From duduca007 at yahoo.com  Tue Sep 24 14:00:22 2002
From: duduca007 at yahoo.com (Carlos Moreira)
Date: Tue, 24 Sep 2002 11:00:22 -0700 (PDT)
Subject: Pyro Error.
In-Reply-To: <20020924174641.40449.qmail@web11403.mail.yahoo.com>
Message-ID: <20020924180022.33444.qmail@web11408.mail.yahoo.com>

Oh right, oh right....I found the problem (a silly
problem :) ) and fix it.
I need more coffe...

	[]'s

Cadu Moreira
KaduSoft President
------------------------
[...Please god, put down the dollar ( U$ 1 == R$ 3.57
)...]



--- Carlos Moreira <duduca007 at yahoo.com> wrote:
> Testing the simplest example of docs/8-example.html,
> on client side, generate an error message:
> 
> Traceback (most recent call last):
>   File "./client.py", line 10, in ?
>     print jokes.joke("Irmen")
>   File
>
"/home/pgs/carlos/python/Python-2.2/lib/python2.2/site-packages/Pyro/core.py",
> line 315, in __invokePYRO__
>     self.adapter.bindToURI(self.URI)
>   File
>
"/home/pgs/carlos/python/Python-2.2/lib/python2.2/site-packages/Pyro/protocol.py",
> line 224, in bindToURI
>     raise ProtocolError('connection failed')
> Pyro.errors.ProtocolError: connection failed
> 
> Where I must seek for inconsistences? Pyro.conf?
> 
> 	[]'s
> 
> Cadu Moreira
> KaduSoft President
> --------------------------
> [......]
> 
> __________________________________________________
> Do you Yahoo!?
> New DSL Internet Access from SBC & Yahoo!
> http://sbc.yahoo.com
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list


=====
"Liberdade, Igualdade, Fraternidade e Foda-se"
-- Millor Fernandes

__________________________________________________
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
http://sbc.yahoo.com



From sismex01 at hebmex.com  Mon Sep  9 10:26:47 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Mon, 9 Sep 2002 09:26:47 -0500 
Subject: xml.minidom is stripping out my CRLF's in attrib values!!
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062330237C4A7@mail.hebmex.com>

Probably not sripped, but normalized:

for example, your original string:
>>> s = "spotted\x0Awith black and white"

becomes the normalized string:
>>> s = "spotted with black and white"

Is something like this happening?

If it is, then it's the correct behaviour, according
to the XML standard: all text should be normalized,
except if it's in a CDATA block.

HTH

-gus

Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From gmuller at worldonline.nl  Tue Sep 17 15:14:36 2002
From: gmuller at worldonline.nl (GerritM)
Date: Tue, 17 Sep 2002 21:14:36 +0200
Subject: Will "Python in a Nutshell" be too short?
References: <mailman.1032288261.15607.python-list@python.org>
Message-ID: <am7v2u$n32$1@reader1.tiscali.nl>

"Hamish Lawson" <hbl at st-andrews.ac.uk> schreef in bericht
news:mailman.1032288261.15607.python-list at python.org...
> O'Reilly doesn't yet list the forthcoming "Python in a Nutshell" by Alex
> Martelli, but according to Amazon.com's page for the book, it will have
400
> pages. Is that indeed the plan?
<...snip...>
> By comparison, O'Reilly's "Perl in a Nutshell" is 800 pages and has the
> following chapters
<...snip...>
>
> Hamish Lawson

By comparison "Java in a nutshell" is 600 pages and on top of that "Java
examples in a nutshell" is another 400 pages. By the way I liked the set up
of this second book. It is close to the cookbook idea. (The Python Cookbook
is GREAT!).

regards Gerrit

--
www.extra.research.philips.com/natlab/sysarch/





From pcarey at lexmark.com  Fri Sep 20 12:17:54 2002
From: pcarey at lexmark.com (pcarey at lexmark.com)
Date: Fri, 20 Sep 2002 12:17:54 -0400
Subject: Newbie questions: write permission, interpreter minimizing
Message-ID: <OFC78165B0.5B28155A-ON85256C3A.0058C33A@lexmark.com>

Win2K, Python2.2
Hello list.

I'm relatively new to Python.

When I want to test for write permissions (on Win2K),
I use try: except: clauses, something like this:

try:

      os.mkdir("X:\\anydir")

except OSError (e,s):

      print s

os.rmdir("X:\\anydir")



Is there an easier way?

I tried os.access("X:\\anydir", W_OK), but it always returns 1,
so I guess I am misunderstanding that.



Also, (and forgive me if I betray woeful ignorance)...

I use PythonCardPrototype to make UI's.
When launching these apps (double-clicking),
the interpreter launches (as it should) and remains in the background
(as it should)... but is there any way to automatically minimize the
interpreter?



Thank you,

Pete






From padraig at linux.ie  Sun Sep 15 15:50:04 2002
From: padraig at linux.ie (Padraig Brady)
Date: Sun, 15 Sep 2002 19:50:04 GMT
Subject: regular expression for space seperated quoted string
References: <PqFf9.3627$cP3.7118@news.iol.ie> <alnbpb$ce5$1@wanadoo.fr> <%jHf9.3635$cP3.7161@news.iol.ie> <pqHf9.3636$cP3.7130@news.iol.ie>
Message-ID: <3D84E4B4.4040305@linux.ie>

Padraig Brady wrote:
> Padraig Brady wrote:
> 
>> Eric Brunel wrote:
>>
>>> Padraig Brady wrote:
>>>
>>>
>>>> Hi, I'm trying to split a string that is seperated
>>>> by spaces and also contains double quoted words which
>>>> can contain spaces:
>>>
>>>
>>
>> [snip]
>>
>>> What about:
>>>
>>>>>> p = r'[^ \t\n\v\f"]+|"[^"]*"'
>>>>>> re.findall(p, '1 2 3')
>>>>>
>>>>>
>>
>> FYI, I'm using the following:
>>
>> import fileinput, re
>> for line in fileinput.input():
>>     #split fields
>>     fields = re.findall('[^ "]+|"[^"]+"', line[:-1])
>>     #remove quotes
>>     fields = map(lambda field: field.replace('"', ''), listLine)
>>
>> thanks again,
>> P?draig.
>>
> 
> Just in case it's useful I might as well make it correct and more 
> efficient:
> 
> import fileinput, re
> reFieldSplitter = re.compile('[^ "]+|"[^"]+"')
> for line in fileinput.input():
>     #split fields
>     fields = reFieldSplitter.findall(line[:-1])
>     #remove quotes
>     fields = map(lambda field: field.replace('"', ''), fields)

Just to be more complete the compilation of the regular expression
above gave a 1.3% speedup. However changing the map(lambda...) to
a list comprehension gives another 4% speedup.

import fileinput, re
reFieldSplitter = re.compile('[^ "]+|"[^"]+"')
for line in fileinput.input():
     #split fields
     fields = reFieldSplitter.findall(line[:-1])
     #remove quotes
     fields = [field.replace('"', '') for field in fields)]

P?draig.



From syver-en+usenet at online.no  Mon Sep 16 05:12:57 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Mon, 16 Sep 2002 09:12:57 GMT
Subject: basic question
References: <ad3fe3.0209152217.1d7d4468@posting.google.com>
Message-ID: <ud6re70xo.fsf@online.no>

s3v3n3v3s73 at yahoo.com (7) writes:

> I am using Pythonwin. How can I always make sure when I make a change
> to a script and save it, that when I go to run it, that I run the
> changed version. I know there is a very basic answer, this has been
> getting to me.

reload(modulename) 

should do the trick

-- 

Vennlig hilsen 

Syver Enstad


From jingguoming at yahoo.com.cn  Wed Sep 11 22:22:55 2002
From: jingguoming at yahoo.com.cn (jgm2163)
Date: 11 Sep 2002 19:22:55 -0700
Subject: how can i make a binary executable program with a python2's script
Message-ID: <dd083388.0209111822.7598320d@posting.google.com>

os:redhat7.2
python:2.1
how can i make a binary executable program with a python2's script


From neil.padgen at mon.bbc.co.uk  Fri Sep  6 05:06:50 2002
From: neil.padgen at mon.bbc.co.uk (Neil Padgen)
Date: Fri, 06 Sep 2002 09:06:50 +0000
Subject: How to force built-in commands over imported.
References: <mailman.1031181848.23769.python-list@python.org>
Message-ID: <al9r1m$sie$1@nntp0.reith.bbc.co.uk>

Delaney, Timothy wrote:
> [ unit testing framework snipped]
>
> which allows me to perform each set of tests independently, and run them
> all through a single command. Even so, it should probably specify the
> actual names of the test cases in case there is a conflict somewhere, but
> that would fail to pick up any new test cases.

There's a fabulous regression testing script in Dive Into Python
which takes care of this for you, by running all tests in the
current (or given) directory which are in modules called *test.py.

http://www.diveintopython.org/regression_divein.html

I've modified this to work in the current directory and below,
but haven't been able to test it thoroughly in anger yet.  (I'm not
yet confident enough at writing tests...)

-- Neil



From sismex01 at hebmex.com  Wed Sep 18 11:48:50 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Wed, 18 Sep 2002 10:48:50 -0500
Subject: isFloat: Without Exception-Handling
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233024CBBD6@mail.hebmex.com>

> 
> Is there a way to write the following
> method without using exceptions?
> 
> def isFloat(string):
>      is_float=1
>      try:
>          float(string)
>      except:
>          is_float=0
>      return is_float
> 
> print isFloat("asdf") # --> 0
> print isFloat("0.1")  # --> 1
> 
> Somehow I don't like to catch exceptions. Maybe
> that's because my first language was C.
> 
>   thomas
> 

Catching exceptions in Python is quicker than in other
languages, because of internal implementation issues.
So, you could rewrite the following like this:

def isFloat(S):
  R = 0
  try:
    float(S)
    R = 1
  finally:
    return R

A bit shorter, and more concise, since using "except:" is
bad form (you're not specifying which exception you're
interested in catching), and "finally:" says what you
really meant: "In the end, return R, no matter what
happens".

HTH

-gus




Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From bfarwick at juno.com  Sun Sep 22 06:42:06 2002
From: bfarwick at juno.com (brrrent)
Date: 22 Sep 2002 03:42:06 -0700
Subject: module name errors (trying to import curses functions from modules as needed)
Message-ID: <1b6dca50.0209220242.d9afc12@posting.google.com>

I use some very simple curses functions with Python 1.5.2.  I don't
have a problem with curses per se.

I use the old top level setup code shown in Andrew Kuchling's early
curses HOWTO.  It works well.

But ...
I want to keep all my various curses input and output functions in
separate modules, calling them when needed, instead of defining them
all at the top of my main program.

If I include top level curses setup code in each module, they import
with no problem.  I call them with the desired parameters and they do
their job.

The real problem is that when a function imported with one module
finishes, so does curses (albeit briefly).

Then of course the screen flashes annoyingly (very briefly) before the
next module once again reactivates curses.

If I try to omit the curses top level setup in the modules and put it
in my main program, of course Python complains about things such as
"stdscr" name errors.  This makes sense, but I don't know how to work
around it.

I can of course finish my program, putting up with the brief screen
flashes, and when finished, pull all my function definitions into my
main program.

If there's a better way, I would be grateful for advice on same.

Sincerely,

b


From at_news at pontos.net  Sat Sep 21 07:46:17 2002
From: at_news at pontos.net (ilias)
Date: Sat, 21 Sep 2002 14:46:17 +0300
Subject: LISP - [SPECS ERR] - Backquote - please confirm. - [#V0.3]
References: <amc48r$hc0$1@usenet.otenet.gr>
Message-ID: <amhlmk$456$1@usenet.otenet.gr>

I've found a simple bug in the ANSI Common Lisp specification.

At least i think so.

I've published this in comp.lang.lisp, and asked for validation.

The bug is very simple.

But i got no responses that leads to a confirmation.

I had this again. When i was claiming that:

----------------------------------------------------------------
() can be changed to [] with 2 lines of ANSI Common Lisp Code

An ANSI Conforming Common Lisp implementation *must* execute:

(set-syntax-from-char #\] #\) )  ; => T
(set-syntax-from-char #\[ #\( )  ; => T
[+ 3 2]                          ; => 5

http://groups.google.com/groups?selm=alot2c$pkv$1 at usenet.otenet.gr

(LispWorks *does*, Franz Allegro *does not*, Corman Lisp *does not*, 
other versions: don't know, i got no feedback !
----------------------------------------------------------------

But this case is much simpler.

Interested in "The 'Savages' of C.L.L." ?

Interested in LISP?

Interested in Software Engeneering?

Here's a showcase.

The Medium 'Public Newsgroups'.

Used for quick 'Language Evolution'.

==============================================================

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

NESTED BACKQUOTE

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

The [online-version of the] standard-document states:

http://www.lispworks.com/reference/HyperSpec/Body/02_df.htm

"[...] If the backquote syntax is nested, the innermost backquoted form 
should be expanded first. This means that if several commas occur in a 
row, the leftmost one belongs to the innermost backquote.[...]"

=> "the *innermost* backquoted form *should be* expanded first"

This definition is wrong.

------------------------------------------------------------------------
Testing with implementations:

Allegro:
(setq d 1  c 'd  b 'c  a 'b )
`(,a `(,b ,,c))               ; => (B (EXCL::BQ-LIST B D))
              ;;; which is same as: (B `(,B ,D))

LispWorks:

`(,a `(,b ,,c))               ; => (B (SYSTEM::BQ-LIST B D))
              ;;; which is same as: (B `(,B ,D))


We see in the processing of both implementations:

- The *outermost* backquoted form is expanded first
=> direct contradiction to the specification

- The *innermost* backquoted form is processed *partially* by the 
outermost backquote to evaluate the ,,c => ,D.

This confirms: The definition of nested backquote is wrong.

------------------------------------------------------------------------
one could say now:

The specification is correct. The implementors are wrong.

--------------------------------------------------------------------
The 'natural' flow of things will help out.

Naturally (as given by reading in the stream) the outermost backquote is 
expanded first.

The Lisp-Reader cannot preview more than *one* char. Thus it is 
impossible to detect the existence of the innermost quote to avoid 
evaluation of the outermost backquote.

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

(setq d 1  c 'd  b 'c  a 'b )

`(,a `(,b ,,c))               ; => (b `(c ,d))
      |_|__||
|_|________|

the only  , prior to a will be processed from the outermost `
the only  , prior to b will be processed from the innermost `
the left  , prior to c will be processed from the innermost `
the right , prior to c will be processed from the outermost `

--------------------------------------------------------------------
The author meant probably this:

If the backquote syntax is nested, the innermost backquoted form should 
be *served*[not expanded] first [with a comma]. This means that if 
several commas occur in a row, the leftmost one belongs to the innermost 
backquote [and so on].

in any case:

http://www.lispworks.com/reference/HyperSpec/Body/02_df.htm

"[...] If the backquote syntax is nested, the innermost backquoted form 
should be expanded first.[...]"

is wrong.

--------------------------------------------------------------------
'Looking around'

i follow an suggested link in the specs:

http://www.lispworks.com/reference/HyperSpec/Body/02_dfa.htm

"[...] Implementors who have no particular reason to make one choice or 
another might wish to refer to IEEE Standard for the Scheme Programming 
Language,[...]"

http://www.cs.indiana.edu/scheme-repository/doc.standards.html

nothing found there,

but i reached finally this point:

http://www.schemers.org/Documents/Standards/R5RS/HTML/r5rs-Z-H-7.html#%_sec_4.2.6
"[...]
Quasiquote forms may be nested. Substitutions are made only for unquoted 
components appearing at the same nesting level as the outermost 
backquote. The nesting level increases by one inside each successive 
quasiquotation, and decreases by one inside each unquotation.

`(a `(b ,(+ 1 2) ,(foo ,(+ 1 3) d) e) f)
                 ===>  (a `(b ,(+ 1 2) ,(foo 4 d) e) f)
(let ((name1 'x)
       (name2 'y))
   `(a `(b ,,name1 ,',name2 d) e))
                 ===>  (a `(b ,x ,'y d) e)
[...]
"

Which confirms what already is obvious:

The ANSI Common Lisp specification about nested Backquote is wrong.

My main interest now is:

I've lost *many* time due to this wrong specification, which ruins my 
understanding process for days.

I've invested *many* time to find out whats happened and to write 
everything down, so other newcomers to lisp (i'm a lisp-novice) don't 
run into the same problem.

Now look at the behaviour of comp.lang.lisp.

And explain me please:

- Is this a common behaviour of Lisp people?

- Or is this only the behaviour of the few people online in cll?

- are other languages-lovers like this people? i mean, declaring the 
specs as a religion which has no errors? Believe - don't doubt.

Are this people not able to say:
- Yes, it's maybe an bug.
- Yes, it's an bug.
- Ok, we've a bug in the specs, so what?
- Ok, we've a bug in the specs, lets look what to do.

Simply to raise the image and the quality of comp.lang.lisp. And of 
Common Lisp.

I've a strange feeling, that the nature of Common Lisp is like the 
nature oth the people in c.l.l.

Talking, talking, talking.

And silence, when it is time to talk.

-

I'm a lisp-novice.

Be happy.



From peter at engcorp.com  Tue Sep 10 08:35:43 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 10 Sep 2002 08:35:43 -0400
Subject: Windows/DOS: double clicking a .py file
References: <mailman.1031613313.24086.python-list@python.org> <3d7d6f26@news.sentex.net> <just-D942FB.09231610092002@news1.xs4all.nl>
Message-ID: <3d7de70a$1@news.sentex.net>

Just wrote:
> Not only that, but I want to only keep the window open when an exception 
> occurred.
> 
> But thanks for all the input!

Just, maybe your users would be happy to enable an option
that would just *mail* you the exception traceback when
a problem occurs, rather than having to read it in an
ugly DOS window and perhaps make mistakes in transcribing
it when they tell you about it.

There was at least one recent thread which showed examples
of how to do this.  If you can't dig them up (might be
in the Cookbook too) ask and someone can find a link or
repost.

If these are internal users, that is certainly a better
option than just printing the traceback.

-Peter



From tjreedy at udel.edu  Fri Sep 27 23:25:21 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 28 Sep 2002 03:25:21 GMT
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <amvcev$cf6$1@s1.read.news.oleane.net> <7nHk9.172208$pX1.6117207@news2.tin.it> <ofLk9.304267$z91.13145351@bin3.nnrp.aus1.giganews.com> <dITk9.175519$pX1.6209886@news2.tin.it> <t_3l9.334924$AR1.14880141@bin2.nnrp.aus1.giganews.com> <mailman.1033165663.5015.python-list@python.org> <an325r$hir$1@peabody.colorado.edu>
Message-ID: <Ai9l9.339471$AR1.15314440@bin2.nnrp.aus1.giganews.com>

"Fernando P?rez" <fperez528 at yahoo.com> wrote in message
news:an325r$hir$1 at peabody.colorado.edu...
> Mike Rovner wrote:
>
> >
> > "Terry Reedy" <tjreedy at udel.edu> wrote in message
> > news:t_3l9.334924$AR1.14880141 at bin2.nnrp.aus1.giganews.com...
> >> na=1; np=3; nl=1; ng=0
> >> print "Found: %d apple%s, %d plum%s, %d lemon%s, and %d grape%s."
%\
> >> (na, na!=1 and 's' or '', np, np!=1 and 's' or '', nl, nl!=1 and
's'
> >> or '', ng, ng!=1 and 's' or '')
> >>
> >> Found: 1 apple, 3 plums, 1 lemon, and 0 grapes.

> > 4. Try:
> > use_s=('','s')
> > print "Found: %d apple%s, %d plum%s, %d lemon%s, and %d grape%s."
%\
> > (na, use_s[na!=1], np, use_s[np!=1], nl, use_s[nl!=1], ng,
use_s[ng!=1])

Of course.

> Why not just a simple function?
>
> In [23]: plur=lambda s,n:'%s %s' % (n,s+('','s')[n!=1])
> In [27]: na=1; np=3; nl=1; ng=0
> In [28]: print "Found %s, %s, %s, %s" %
(plur('apple',na),plur('plum',np),
>    ....: plur('lemon',nl),plur('grape',ng))
> Found 1 apple, 3 plums, 1 lemon, 0 grapes

Nice.  I like both of these better than my version.  Thanks.

Terry J. Reedy




From syver-en+usenet at online.no  Tue Sep 17 07:01:57 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Tue, 17 Sep 2002 11:01:57 GMT
Subject: don't understand error message
References: <mailman.1032251060.4906.python-list@python.org> <Xns928C66E3994D7duncanrcpcouk@127.0.0.1> <mailman.1032255920.1725.python-list@python.org>
Message-ID: <elbs6fsg.fsf@online.no>

Manuel Hendel <manuel at hendel.net> writes:

> But I already got a new one.
> 
> [....]
> if domainfield in domain:
>     domainlines.append(line)
> [....]
> 
> This should work if the domainfield and domain got the following
> values, shouldn't it?
> 
> domainfield = 1234.ab.cdef.de
> damain = 1234.ab.cdef.de

No I don't think so. It will only work when the conditions are as
described in the error message. I think it would be wise to assume
that your code is at fault first, instead of assuming that the error
is elsewhere. Assume that you've just made a simple mistake first, and
only after having ruled that out, progress to more and more unlikely
error sources, usually ending with blaming the compiler or
interpreter, which should be the last bug source that you consider.

In my understanding *in* works as follows. It takes an
element as the left argument and a collection as the right element. I
is therefore quite natural that you have to have a single character on
the left and a string on the right.


> 
> But I get this error message:
> 
> TypeError: 'in <string>' requires character as left operand

Yes, read the error message. It tells you that what you did is
invalid, and it also tells you what you have to do if you are going to
use this construct. I think what you want is closer to this:

if domain.find(domainfield) != -1:
    # do stuff

>From the python documentation:

find(sub[, start[, end]]) 
Return the lowest index in the string where substring sub is found, such that sub is contained in the range [start, end). Optional arguments start and end are interpreted as in slice notation. Return -1 if sub is not found. 





-- 

Vennlig hilsen 

Syver Enstad


From tjreedy at udel.edu  Thu Sep  5 00:48:41 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 05 Sep 2002 04:48:41 GMT
Subject: How do i print to a printer using python?
References: <6afb9d2c.0209041044.43c5f5f1@posting.google.com>
Message-ID: <JmBd9.427851$m91.16671782@bin5.nnrp.aus1.giganews.com>

"Naveed Iqbal" <iqbal at colorado.edu> wrote in message
news:6afb9d2c.0209041044.43c5f5f1 at posting.google.com...
> I have to print to a standard printer. I take the data from the
> keyboard and print it using the printer. Please help me....my job
> depends on this!!!!

On *nix or dos/windows, you open the printer port as if it were a
file, just like other ports and devices.  On *nix, this maybe be
something like '/dev/lpt' or a the numbered serial or parallel port
that your printer is attached to.  On Dos and my win98 machine, this
does the trick:

>>> p=file('LPT1:','w')
>>> p.write('Test Line\n\f')
>>> p.flush() # p.close() also flushes

and out come a sheet with 'Test Line' printed.  Note that \f is
formfeed char; without it, nothing is printed on page printers (*jets,
etc.).

Terry J. Reedy






From sadams123 at optushome.com.au  Wed Sep  4 15:03:43 2002
From: sadams123 at optushome.com.au (Steven)
Date: Thu, 5 Sep 2002 05:03:43 +1000
Subject: How do i print to a printer using python?
References: <6afb9d2c.0209041044.43c5f5f1@posting.google.com>
Message-ID: <3d765914$0$26216$afc38c87@news.optusnet.com.au>

"Naveed Iqbal" <iqbal at colorado.edu> wrote in message
news:6afb9d2c.0209041044.43c5f5f1 at posting.google.com...
> I have to print to a standard printer. I take the data from the
> keyboard and print it using the printer. Please help me....my job
> depends on this!!!!

What operating system are you using? you'll need o post a few more details.

Steven




From josef.buergi at bluewin.ch  Sun Sep 29 05:21:36 2002
From: josef.buergi at bluewin.ch (Josef Bürgi)
Date: Sun, 29 Sep 2002 10:21:36 +0100
Subject: Python and GUI question
Message-ID: <3d96b808_3@news.bluewin.ch>

Dear NG

I designed a CMS using mySQL and PHP. Now I like to write an administrativ
GUI Client Tool using Python. So I bought a few books about it, saying that
there is more then one GUI interface (Tk, Qt, xvWindow). So can anyone tell
me which one you use and why ? And how do you Design GUI interfaces (IDE) ?

Best wishes fromSwitzerland

Joe




From aleax at aleax.it  Fri Sep 20 10:10:35 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 14:10:35 GMT
Subject: Python sequences reference - problem?
References: <mailman.1032443760.5578.python-list@python.org> <YZli9.114624$ub2.2439313@news1.tin.it> <3D8B0F86.9000600@one.net>
Message-ID: <v%Fi9.138621$pX1.4918425@news2.tin.it>

Michael Schneider wrote:

> I also like python.
> 
> When doing pure math  I have been using J for prototyping algorithms.
> J is basically APL without the special symbols .

Brings me back a LONG time -- APL and APL2 were all the rage in 
IBM Research back in the '80s, which is where I first worked.  I
also used APL for the prototype of my master thesis in the '70s
(later completed it in Vax/VMS Fortran and Assembler -- and, for
the specific job that program was doing [compiling AHPL code into
a gate-level circuit design], Fortran with a bit of Assembler
turned out to be much more productive than APL, too:).

However, I just can't get into J: what used to be "at first weird,
soon learned and readable" symbols have mutated into "totally
obscure forever" symbols (in my opinion).  I still *THINK* in APL
for some problems: I want an `expand` primitive, don't really
care about its syntax or how it's spelled, a `reduce` one too,
etc, etc... but J's cconcrete syntax is just too weird for me.


Alex



From robin at jessikat.fsnet.co.uk  Tue Sep  3 18:17:43 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 3 Sep 2002 23:17:43 +0100
Subject: [Q] win32com/Excel
References: <mailman.1031083266.9747.python-list@python.org>
Message-ID: <GXRQCIAHUTd9EwVZ@jessikat.fsnet.co.uk>

In article <mailman.1031083266.9747.python-list at python.org>, Mike
Brenner <mikeb at mitre.org> writes
>Hi Robin, 
>
>That was great, but most importantly, it was FAST!
>
>Would it be possible to tell how you did that so quickly? Is there a reference 
>that would tell me that x1LastCell is in constants? or that you put the 
>activeWindow in front of ActiveCell?

OK let me tell you how I did this. I'm getting old so my memory is
fairly poor these days so I rely on the source of the interface file
generated by win32com\client\makepy.py. I scrolled down and built the
interface for my excel which appeared in the list as 'Microsoft Excel
9.0 Object Library (1.3)'.

The I edited the generated file in win32com\gen_py
it was called 00020813-0000-0000-C000-000000000046x0x1x3.py derived from
the GUID I guess.

Scanning I see the xlLastCell constant in a class called constants. I
don't normally use these constants so I checked what happens to
constants; right at the bottom I see

win32com.client.constants.__dicts__.append(constants.__dict__)

ie the constants are added into the win32com.client.constants thing's
dict so should appear as attributes of win32com.client.constants.

Then I looked for something to hang this constant on, seems like a range
has this method. The vb seemed to like using ActiveCell as the start
point and scanning for that method reveals a method  in a class called
Window.

As for the book, sheet and window things I seem to remember them from
some previous era when I did a lot of excel (horror show).


As for the nitty gritty details of how to turn off various bits of
interactive code I can only recommend detailed reading of any available
documentation.

My boss is Andy Robinson co-author with Marc Hammond of the win32 python
book so from the python side I know that's a very good start. On the
other side I recommend asking an expert or consulting the appropriate
newsgroup. Writing an application in excel obviously requires answers to
your questions. A quick check reveals that there is a way to stop
checking for viruses in macros (at least in my excel) that might be the
way to go.

Googling with 'microsoft project option' reveals a lot of hits so
perhaps that might be a good source.

>
>Do you know of any references as to how to do it in Microsoft Project whose 
>object model seems much less documented than Excel's is?
>
>I know how to use Project Explorer in the Visual Basic Editor inside the 
>Microsoft OLE applications like Microsoft Project, and I know how to get the 
>Python list of objects using makepy. However, that does not seem to be enough to 
>figure out many questions in Microsoft Project (and even some questions in 
>Microsoft Excel).
>
>What would be really nice is a document that explains what each object in the 
>list is used for. For example, when opening a Microsoft Project file, it asks a 
>bunch of questions like: are macros okay? should I open related files? should I 
>save the changes I think you made to the previous file? 
>
>I would like to learn how to cancel out those messages when I open a Microsoft 
>Project file from Python. What book or application do you recommend I purchase? 
>Or what file can I download to learn how to do these things?
>
>Thanks,
>Mike
>
>
>
>>I would like to emulate this Visual Basic statement from Python, using
>>Com:
>>
>>        Range(Range("A1"), ActiveCell.SpecialCells(xlLastCell)).Select
>>
>>this statement selects the active worksheet (which is, incidentally,
>>not similar to UsedRange).  
>>
>>any clues?  
>>
>>thank you,
>from win32com.client import Dispatch
>from win32com.client import constants
>xl = Dispatch("Excel.Application")
>AB = xl.ActiveWorkbook
>AS = AB.ActiveSheet
>AW = xl.ActiveWindow
>R = AS.Range('A1',AW.ActiveCell.SpecialCells(constants.xlLastCell))
>print AB.Name, AS.Name, R

-- 
Robin Becker


From henk.derudder at nospam.sciatl.com  Thu Sep 26 07:37:55 2002
From: henk.derudder at nospam.sciatl.com (henk.derudder at nospam.sciatl.com)
Date: Thu, 26 Sep 2002 11:37:55 GMT
Subject: getting system time 
Message-ID: <3d92ec67.596514453@news.skynet.be>

Hi,

I wondered whether it is possible (and if possible, how) to monitor
changes in the system time on a Windows machine. Suppose one resets
the clock: where is this info stored (registry,file). Which format?
Can I get this using Python?

thanks,

Henk


From roffe at aqualene.uio.no  Sat Sep 28 09:45:42 2002
From: roffe at aqualene.uio.no (Rolf Marvin Bøe Lindgren)
Date: Sat, 28 Sep 2002 15:45:42 +0200
Subject: [Mac] Editor on Macintosh?
References: <1fj7za2.1erbqps1uvzpj8N%news@agapow.net>
Message-ID: <lbzbs6ifcs9.fsf@aqualene.uio.no>

[Paul-Michael Agapow]

| <http://www.hekkelman.com>.) I found it the best editor for Python on
| the Macintosh, having never been really happy with the alternatives. 

GNU Emacs is available on the Mac.  there's a carbonized version
available, no need for X11.

-- 
Rolf Lindgren                                            http://www.roffe.com/
roffe at tag.uio.no


From robin at alldunn.NOSPAM.com  Thu Sep 19 22:24:39 2002
From: robin at alldunn.NOSPAM.com (Robin Dunn)
Date: Fri, 20 Sep 2002 02:24:39 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <Xns928CED692541cliechtigmxnet@62.2.16.82> <6_Ph9.479711$q53.16315681@twister.austin.rr.com>
Message-ID: <GFvi9.114$FY2.82713@news.randori.com>

Lee Gray wrote:
> 
> For a newbie, what's easier, Tk or wxWindows?
> 

I've found that it depends on how your brain is wired.  There are some 
people who love Tkinter because it just fits the way they think but just 
can't understand wxPython no matter how much they try.  There are other 
folks where the opposite is true, I am one of them and that is the main 
reason wxPython was created.

-- 
Robin Dunn
Software Craftsman
http://wxPython.org  Java give you jitters?  Relax with wxPython!



From jmiller at stsci.edu  Mon Sep  9 14:01:08 2002
From: jmiller at stsci.edu (Todd Miller)
Date: Mon, 09 Sep 2002 14:01:08 -0400
Subject: Can't exponentiate zero in NumPy
References: <al5gdq$18g$1@naig.caltech.edu> <alifc4$gsi$1@naig.caltech.edu>
Message-ID: <3D7CE1E4.3000600@stsci.edu>

Rick Muller wrote:
> I found out more about this problem, and I'm still hoping that someone 
> can suggest a good fix.
> 
> The problem comes when I take a power of a NumPy array that has very 
> small values in it:
> 
>  >>> from Numeric import *
>  >>> a = zeros(3,Float)
>  >>> a**2
> array([0., 0., 0.])
>  >>> a[1] = 1.e-310
>  >>> a**2
> Traceback (most recent call last):
>    File "<stdin>", line 1, in ?
> OverflowError: math range error
> 
> This crashes on Linux, but not on Mac OS X.
> 
> Is this a bug or a feature?
> 
> If it's a feature (which it might be), can someone offer a one-line 
> workaround? What I would like to do is set everything below a small 
> value (say 1.e-16) to zero. Obviously,
>  >>> for i in range(len(a)): if a[i] < 1.e-16: a[i] = 0
> or something like it should work, but it will also be slow, and I was 
> hoping that one of the Numeric UFuncs might do the same thing more 
> effectively.
> 
> Thanks in advance for any help anyone can offer.
> 
> Rick
> 

I think what you're looking for is "where".  You use it like:

 >>> b = where( abs(a) < 1.e-16, 0, a)

Which means "where the absolute value of a is less than 1.e-16, use 0, 
else use a".  The "<" operator only works like this for Python-2.1 and 
up;  prior to that you have to substitute a "lessthan" function call.


Todd



From andrew.thompson at ashecastle.com  Mon Sep 23 13:15:37 2002
From: andrew.thompson at ashecastle.com (Andrew Thompson)
Date: Mon, 23 Sep 2002 18:15:37 +0100
Subject: e vs exp()? / financial applications
In-Reply-To: <LNBBLJKPBEHFEDALKOLCCEKKBFAB.tim.one@comcast.net>
Message-ID: <011a01c26324$d6300940$4802a8c0@blair>

One must be aware of the representational issues and accuracy of
arithmetic within a datatype when using numerical types.  This is not so
much a Python issue as one of representing the real number line with
float types.

For example, the following on Python 2.2.1 #34 running on Windows XP
Pro, Celeron Processor (probably different elsewhere)

W=0.89 / 4

Now evaluate

1+w+w+w+w

and 

w+w+w+w+1

which are both *different* according to the == operator, but for the
purposes of numerical accuracy are identical.

This behaviour is quite mild, but it shows the point that rounding
errors can easily creep in, and using == to test for equality is not
always sensible.

Caveat Emptor!








From aleax at aleax.it  Mon Sep 30 05:00:45 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 30 Sep 2002 09:00:45 GMT
Subject: Best way to create a class dynamically on the fly?
References: <tFml9.13043$Lg2.3865279@news2.news.adelphia.net> <DGml9.13046$Lg2.3865365@news2.news.adelphia.net>
Message-ID: <1pUl9.160835$ub2.3602794@news1.tin.it>

Robert Oschler wrote:
        ...
> I forgot to mention, I would like the classes I create dynamically to be
> fully "pickle-able".  If there are any caveats or limitations in this area
> please let me know.

Classes are pickled by name (assuming that using modules pickle or cPickle 
as they stand is what you mean by "fully pickle-able").  So, in general, 
classes you pickle (or, classes whose instance you pickle) must be 
top-level names exported by some module.  Dynamically created classes
obviously don't qualify -- saving all the data needed to recreate such
classes, and recreating them just as dynamically when unpickled, is a
harder problem.  You can register your own arbitrary factories for 
unpickling, but I don't know if that satisfies your criteria for "fully 
pickleable".


Alex



From james.kew at btinternet.com  Wed Sep  4 17:29:33 2002
From: james.kew at btinternet.com (James Kew)
Date: Wed, 4 Sep 2002 22:29:33 +0100
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Sep  4)
References: <448A0F008E0AFE11.19B9EF86BEE2B191.107D5020B234F743@lp.airnews.net>
Message-ID: <al5tvu$1mtjjr$1@ID-71831.news.dfncis.de>

"Ian Bicking" <ianb at colorstudy.com> wrote in message
news:448A0F008E0AFE11.19B9EF86BEE2B191.107D5020B234F743 at lp.airnews.net...
> QOTW:  The canonical, "Python is a great first language", elicited,
> "Python is a great last language!" -- Noah Spurrier
> <c9d82136.0208251844.63a6d187 at posting.google.com>

>     A thoughtful note on one person's decision to use Python in a
>     scientific environment:
>         <3D60A1E2.7FD246E0 at anansispaceworks.com>

*cough* My newsreader turns this into an email address, not a news article.

I can't persuade Google to find this message-id; any chance of a link? I
don't remember that thread but it sounds interesting.

--
James Kew
james.kew at btinternet.com






From DennisR at dair.com  Tue Sep 17 10:59:33 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Tue, 17 Sep 2002 14:59:33 GMT
Subject: Windows AutoRunning Python Hello -- feedback requested
References: <Hmuh9.70$Ha6.5066815@newssvr14.news.prodigy.com> <8a6ba1da.0209170148.23bca1e8@posting.google.com>
Message-ID: <prHh9.169$UJ1.17889587@newssvr14.news.prodigy.com>

> "Netscape" or "Internet Explorer", but given a system like yours plus
> a framework for building an HTML user interface (based on one of the
> many web application frameworks for Python) I think this system is
> just as valid as Tk, wxWindows, Qt or whichever other 'real' toolkit
> you care to name.
I think you got it.  My system allows one to prototype an app on your web
page.  Using server logs, you can see how people use it and what the usage
problems are.  Then a version of the web page can be turned into an
executable.  Also, client-server apps can be implemented with the client and
server displaying in different frames (example at
http://www.autorunning.com/present.cgi click on "here" for client part).

> ... is it simply a toolkit for distributing
> and running browser-based code?
Yes

> Zope-based Windows applications, anyone?
Gasp.

I am now in my third day of testing Python support.  I have previously
implemented Perl but have not distributed it publicly (i.e. it is an
undocumented capability).

Zope would pose a number of challenges.  But, it is an interesting idea.
I'll put it into the hopper.  Thanks for your reply.
--

Dennis Reinhardt

http://www.dair.com





From rgkindya at yahoo.co.in  Sun Sep  8 22:13:43 2002
From: rgkindya at yahoo.co.in (rgkindya)
Date: 8 Sep 2002 19:13:43 -0700
Subject: getting input in a embeded GUI app ...
References: <b2df190a.0209070836.9713b08@posting.google.com> <Xns9282C063D9443cliechtigmxnet@62.2.16.82>
Message-ID: <b2df190a.0209081813.73a72d65@posting.google.com>

Chris Liechti <cliechti at gmx.net> wrote in message news:<Xns9282C063D9443cliechtigmxnet at 62.2.16.82>...

> tha pythonwin way: you can write your own input function in c/c++ that gets 
> the input from a message box. it should return Python string. (i assume 
> you've found out how to write python funcs in c/c++, i.e. like those for in 
> a module)
> then you can replace the builtin function raw_input with your function 
> (assign it to __builtin__.raw_input). take a look at the python win sources 
> on how to do this in python.
>  
> chris

thank chris,

as a newbie , i request you to provide that code ,that will help me to
save my work nights and headache. please Chris i'm a newbie ...


From richie at entrian.com  Thu Sep 12 10:58:40 2002
From: richie at entrian.com (Richie Hindle)
Date: Thu, 12 Sep 2002 15:58:40 +0100
Subject: Coding Style: Quotes
References: <bt1g9.17429$Pf7.590136@news1.west.cox.net>
Message-ID: <hfa1ouodkntb3t6hbq60d9ckf2olcsah79@4ax.com>

> I'm curious if others have adopted any standards for choice of single vs.
> double quotation marks to delimit strings.

I tend to use double-quotes for user-visible strings, and single quotes
for internal ones.  For instance:

    print "Hello, %s." % userRecord[ 'name' ]

In a library module, docstrings count as user-visible because the
"users" include the programmer using the library:

    def greetUser( userRecord ):
        """Says hello to the user."""
        print "Hello, %s." % userRecord[ 'name' ]

Don't ask me why I starting using this standard, or where the idea came
from, 'cos I have no idea.  It just seems to make sense.

-- 
Richie Hindle
richie at entrian.com



From roy at panix.com  Mon Sep 16 22:22:03 2002
From: roy at panix.com (Roy Smith)
Date: Mon, 16 Sep 2002 22:22:03 -0400
Subject: Converting Python app to C++ completely
References: <aklm6c$r1r$1@news.swman.net.uk> <3D85D9EB.9020102@thinkware.se> <roy-2A7271.09563416092002@reader1.panix.com> <3D86209F.4080501@earthlink.nospam> <am5am0$fi6$1@panix2.panix.com> <3d868a77$1@news.sentex.net>
Message-ID: <roy-32ADCF.22220316092002@reader1.panix.com>

Peter Hansen <peter at engcorp.com> wrote:
> Hmmm... if instead you choose to wait until you actually *need* C++
> to solve a problem that Python can't handle, you just might find that
> you don't ever need to learn C++.  Then the second one is actually
> not a problem at all! :-)
> 
> This thought courtesy of XP's philosophy of avoiding unnecessary
> work, which is acronymized as YAGNI (You Aren't Going to Need It).

I'm reasonably convinced that I'll never need C++ to solve real 
problems.  I'm also reasonably convinced that I do need C++ to get 
invited to job interviews.  This thought courtesy of today's economy.

Could be worse.  I could be trying to learn Cobol, for which I see a lot 
more job postings than I do for Python :-(


From b.maryniuk at forbis.lt  Mon Sep  2 06:31:58 2002
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Mon, 2 Sep 2002 12:31:58 +0200
Subject: Why it is so dramatical?
In-Reply-To: <200209021228.58453.b.maryniuk@forbis.lt>
References: <mailman.1030957743.4021.python-list@python.org> <Xns927D6DBDE418Cduncanrcpcouk@127.0.0.1> <200209021228.58453.b.maryniuk@forbis.lt>
Message-ID: <200209021231.58678.b.maryniuk@forbis.lt>

On Monday 02 September 2002 12:28, Bo M. Maryniuck wrote:
> Since time() always show me zero, I've used UNIX time command:
Well, not exactly... ;-)

-- 
Regards, Bogdan

Microsoft is not the answer.
Microsoft is the question.
"No" is the answer.




From rune.hansen at viventus.no  Tue Sep 17 07:28:37 2002
From: rune.hansen at viventus.no (Rune Hansen)
Date: Tue, 17 Sep 2002 13:28:37 +0200
Subject: tkinter asyncore.loop
Message-ID: <am73d3$o4v$1@oslo-nntp.eunet.no>

Hello, Some months back I sendt a question to the list where I asked
if anyone knew how to catch errors in an asyncore loop.
I got an answer to override the "handle_error" in the client. I
still can't get it to trap any errors though. The error I need to trap is
"Connection lost". If I remove the network connection and try to
update the stream or post to the socket nothing happens. When I
reconect the network my application just continues to serve like
nothing has happened. This is of course nice, but not the
functionallity I'm looking for. 
Here's the code I'm working with:

class StreamApp(StreamAppTK):
    def __init__(self):
        ...
    def streamOn(self):
        ...
        self.sc = client(host[0],int(port[0]),self,self.readPasswordFile())
        self.poll()
    def poll(self):
        try:
            asyncore.poll(0.05)
            if self.sc.connected:
                self.master.after(100, self.poll)
        except:
            pass   

class client(asyncore.dispatcher):
    def __init__(self,host,port,StreamApp,userPwd):
        asyncore.dispatcher.__init__(self)
        try:
            self.host = host
            self.port = port
            self.StreamApp = StreamApp
            self.fp = ""                
            username, password = userPwd
            self.buffer = "login "+username+" 
"+md5.new(password).digest()+"\n"
            self.create_socket(socket.AF_INET, socket.SOCK_STREAM)
            self.connect( (self.host, self.port) )
        except socket.error,errMsg:
            self.StreamApp.messageBar.message('state','Error %s, %s' % 
(errMsg))
    def handle_connect(self):
        pass
    def handle_read(self):
        try:
            self.fp = self.fp + self.recv(65565)
            if "\n" in self.fp:
                tmp = self.fp.split("\n")[:]
                for i in tmp[:-1]:
                    self.StreamApp.insertItem(i)
                self.fp = tmp[len(tmp)-1]
        except socket.error,errMsg:
            self.StreamApp.messageBar.message('state','Error %s, %s' % 
(errMsg))
    def writeable (self):
        return (len(self.buffer) > 0)
    def handle_write(self,buffer=""):
        try:
            if buffer:
                self.buffer = buffer
            sent = self.send(self.buffer)
            self.buffer = self.buffer[sent:]
        except socket.error,errMsg:
            self.StreamApp.messageBar.message('state','Error %s, %s' % 
(errMsg))
    def handle_error(self, etype, evalue, etb):
        print 'handling error: %s, %s, %s' % (etype, evalue, etb)
    def handle_close(self):
        self.connected=0
        self.close()
        

As always I'll apreciate any help...

regards

/rune



From aleax at aleax.it  Wed Sep 25 05:33:18 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 09:33:18 GMT
Subject: Shadow password handling
References: <3d90e36d$0$123$7b0f0fd3@reader.news.newnet.co.uk>
Message-ID: <ypfk9.163335$pX1.5843300@news2.tin.it>

Phil Edwards <"phil (at) linux2000.com"> wrote:

> Hi:
> 
> Does anybody know of a Python module which will allow username/password
> credentials to be validated ona Linux box which is configured to use md5

Perhaps PyPAM, http://www.pangalactic.org/PyPAM/ , but I haven't tried
it.  PAM is supposed to let you access any installed authentication
module on your Linux box, isn't it?


Alex



From torkil at fast.no  Wed Sep 18 04:21:42 2002
From: torkil at fast.no (Torkil Grindstein)
Date: Wed, 18 Sep 2002 10:21:42 +0200
Subject: Regular expression help needed
References: <mailman.1032262640.11685.python-list@python.org>
Message-ID: <3D883796.AF573144@fast.no>

Thank you all for your help!

I appreciate it a lot.

Torkil


From amdescombes at qualicontrol.com  Mon Sep  9 03:43:52 2002
From: amdescombes at qualicontrol.com (Andre Michel Descombes)
Date: Mon, 9 Sep 2002 09:43:52 +0200
Subject: Python and Jython inconsistencies when encoding strings
References: <ala00d$dij$1@wanadoo.fr> <j4ofbbclfp.fsf@informatik.hu-berlin.de> <3d78ce33$1_3@news.bluewin.ch>
Message-ID: <alhjck$chv$1@wanadoo.fr>

Thanks,

> sticking to ascii there can avoid some troubles :).

unfortunately it is not always possible :(

Cheers,
Andre

"Samuele Pedroni" <pedronis at bluewin.ch> wrote in message
news:3d78ce33$1_3 at news.bluewin.ch...
>
> Martin v. L?wis <loewis at informatik.hu-berlin.de> wrote in message
> j4ofbbclfp.fsf at informatik.hu-berlin.de...
> > >>> s
> > u"\u0153"
> >
> > Now, U+0153 is LATIN SMALL LIGATURE OE. It so happens that \x9c (what
> > the terminal sends) is U+0153 in CP 1252 (which is the ANSI code page
> > on your Windows installation). This might be a bug in Java, which
> > assumes that bytes sent by the terminal are in the ANSI code page,
> > when they are really in the OEM code page.
>
> no it's more the Jython parser that does that, things can be fixed running
> Jython as
>
> jython -Dpython.console.encoding=cp850
>
> on the other hand output seems buggy for:
>
> print s.encode("cp850")
>
> [I have reported that on our SF bug tracker]
>
>
>  > > Does anybody know what is causing this inconsistency? Is there any
way
> to
> > > avoid it?
> >
> > Yes. Don't use the console.
>
> sticking to ascii there can avoid some troubles :).
>
> regards
>
>




From gerhard.haering at gmx.de  Tue Sep  3 16:23:55 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 3 Sep 2002 20:23:55 GMT
Subject: https in Python?
References: <ac259413.0209031134.3778c6ac@posting.google.com> <m37ki2su6r.fsf@mira.informatik.hu-berlin.de>
Message-ID: <slrnana6iq.t8.gerhard.haering@lilith.my-fqdn.de>

Martin v. Loewis wrote in comp.lang.python:
> s_gherman at yahoo.com (Sorin Gherman) writes:
> 
>> Is it possible to handle https using Python's libraries? 
> 
> Sure. Just use httplib.HTTPS. This requires SSL support in the
> socket module, which is only available on Unix, and only if you have
> OpenSSL.

See my page http://www.cs.fhm.edu/~ifw00065/pyssl/index.html for
Windows.

I wonder why ActiveState doesn't provide SSL support with their Python
distribution. I thought they want to distinguish themselves by adding
value ;-)

-- Gerhard


From erikprice at mac.com  Mon Sep  9 23:52:55 2002
From: erikprice at mac.com (Erik Price)
Date: Mon, 9 Sep 2002 23:52:55 -0400
Subject: IDLE
In-Reply-To: <wpbf9.264320$aA.45691@sccrnsc02>
Message-ID: <C9E2977C-C470-11D6-8068-00039351FE6A@mac.com>

On Monday, September 9, 2002, at 08:54  PM, Ali K wrote:

> Why won't my IDLE work?

If your engine chokes within a second or so of turning over, your idle 
could be set too low.  If it feels like it's running at a higher RPM, 
then it's probably set too high.  I'm not sure about your exact model, 
but if it's older you may also wish to inspect the carburetor.  In a 
newer one I wouldn't even suggest doing the repairs yourself, since 
most of the time proprietary tools are required for maintenance 
(including something as simple as accessing the plugs in a 
mid-late-90's Honda).  It's deplorable what they're doing these days.





Erik





--
Erik Price                                                  (zombies 
roam)

email: erikprice at mac.com
jabber: erikprice at jabber.org




From aleax at aleax.it  Thu Sep 19 11:36:33 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 19 Sep 2002 15:36:33 GMT
Subject: strong/weak - dynamic/static [Was: Getting started]
References: <uojinmcfapjf45@corp.supernews.com> <tyf8z1y5cp5.fsf@pcitapi22.cern.ch>
Message-ID: <5ami9.114674$ub2.2440492@news1.tin.it>

Jacek Generowicz wrote:
        ...
> Could someone provide a serious reference (i.e. something which would
> have gone through some sort of editorial/peer/technical review, rather
> than someone's random opinions on the web) explaining or defining the
> concepts of weak, strong, dynamic and static typing.

For example, Amer Diwan's lessons on Fundamentals of Programming
Languages at the University of Colorado are on the web as PDF files,
all info and links at http://www.cs.colorado.edu/~diwan/5535/ --
I wouldn't consider these "somebody's random opinions on the web"...
it IS a regular, qualified university course.

The introduction to types, in particular, is at:
http://www.cs.colorado.edu/~diwan/5535/lectures/2-types.PDF

and the already-referenced URL gives some fundamental
bibliography (I'm prejudiced in favor of Carlo Ghezzi's
text, but that, of course, is just because he's a compatriot:-).


Alex



From pan-news at asheeshenterprises.com  Fri Sep 20 17:53:38 2002
From: pan-news at asheeshenterprises.com (Asheesh Laroia)
Date: Fri, 20 Sep 2002 17:53:38 -0400
Subject: access to palm with pilot-link pisock library (python binding)
References: <24165ed4.0209200308.4d6b3b7a@posting.google.com>
Message-ID: <amg597$50j$1@news.hcf.jhu.edu>

You could try the jpilot-Python tools that integrate Python and JPilot.

While it's not a perfect solution, it's the best I can find that works for
Python right now.

You can get it from http://www.nickpiper.co.uk/jpilot-python.php3.  If
something doesn't work for you, *make sure to email the author*; he is
very responsive.

Best luck!  Feel free to reply!

-- Asheesh.

On Fri, 20 Sep 2002 04:08:43 +0000, Solon Kazantzidis wrote:

> Hello,
> 
> I want to achieve the following:
> 
> - I want to use the python language to access my palm pilot Vx - I found
> the pilot-link package which provides a python binding of the pisock
> library written in c
> - Unfortunately the documentation I need as newbie to this is listed in
> the TODO list so I am searching a documentation - I want to be able to
> connect to the Palm (I have some example for a connection from the
> pyrite-publisher .py file) - I want to be able make what the pilot-xfer -l
> command does - I want to be able to create/delete/access databases stored
> on my Palm
> - I need to have access on data which I want to gather with the Palm
> 
> What I achieved so far:
> 1. downloaded pilot-link package (version 0.11.3) 2. build it and created
> the python binding (python version 2.2.1) 3. from file
> plugin_instpisock.py (pyrite-publisher) I got some starting information to
> establish connection to the palm:
> 
> -------------------------------------------------------------------------
> 
> #!/usr/bin/python2.2
> import pisock
> 
> port = "/dev/pilot"
> 
> sd = pisock.pi_socket(pisock.PI_AF_SLP, pisock.PI_SOCK_STREAM,
> pisock.PI_PF_PADP)
> if not sd:
> 	print "failed to create socket"
> 	exit
> if (pisock.pi_bind(sd, (pisock.PI_AF_SLP, port))) == -1:
> 	print "failed to bind socket"
> 	exit
> if (pisock.pi_listen(sd, 1)) == -1:
> 	print "failed to listen on socket"
> 	exit
> ret = pisock.pi_accept(sd)
> if ret == -1:
> 	print "failed to accept connection"
> 	exit
> socket = ret[0]
> pisock.dlp_OpenConduit(socket)
> pisock.dlp_ReadDBList(sd,0,pisock.dlpDBListRAM,0)
> 
> --------------------------------------------------------------------------
> 
> As there is no documentation out there and I am not the best programmer I
> can only say that the whole procedure stops with a:
> "Speicherzugriffsfehler"-->(translated) "Memory access failure"
> 
> All works fine until the "ret = pisock.pi_accept(sd)". At this point the
> python interpreter stops and waits for me to press the hotsync button.
> After I press it the above mentioned error arises.
> 
> 
> I need urgently help with this as I cannot with the start form my thesis
> until the mechanism for accessing works. (and then the real work of
> synchronising gathered data starts....)
> 
> 
> Thanks alot for your help/comments,
> 
> Solon.



From scott at projtech.com  Fri Sep  6 04:10:34 2002
From: scott at projtech.com (Scott Finnie)
Date: Fri, 6 Sep 2002 08:10:34 +0000 (UTC)
Subject: 4Suite Server bombs on win32 - suggestions?
Message-ID: <3D786370.3E5666FC@projtech.com>

As it says: I installed 4Suite (0.12.0a2) and everything seemed to go
OK, but 4ss_manager bombs (protection violation).

Python 2.2.1 on winNT sp6 (tried 2.2 on win2K with same result).

Has anyone else seen/fixed this?  Tried the mailing list but got no
reply...

Thanks,
Scott.


From jdhunter at ace.bsd.uchicago.edu  Mon Sep 23 10:58:15 2002
From: jdhunter at ace.bsd.uchicago.edu (John Hunter)
Date: Mon, 23 Sep 2002 09:58:15 -0500
Subject: numpy array assignment
Message-ID: <m2bs6on46w.fsf@mother.paradise.lost>


I have a 2 dimensional array X which is NxM and a 1 dimensional array
v which is Nx1.  I want to assign v to the k-th column of X

import Numeric
import RandomArray

Z = Numeric.zeros( (3,10), Numeric.Float  )
v = Numeric.ones( (3,1), Numeric.Float )
Z[:,4] = v

But I get the ValueError: Object too deep for desired array.  I have
read over the numpy manual on slices and array methods and the closest
thing I saw was the 'put' method, but this seems to be designed for 1d
indexing.

What's the best way to assign an arbitrary Nx1 array to a column of X?

Thanks,
John Hunter



From jeremy at alum.mit.edu  Mon Sep  9 12:56:20 2002
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 9 Sep 2002 09:56:20 -0700
Subject: Zope Status?
References: <Xns9281221A07167bosahvnetscapenet@66.185.95.104> <cSOe9.3024$io.1100@nwrddc03.gnilink.net>
Message-ID: <b0f083db.0209090856.6fd2cd00@posting.google.com>

"Jeff Sasmor" <jeff at sasmor.com> wrote in message news:<cSOe9.3024$io.1100 at nwrddc03.gnilink.net>...
> Speaking from what I understand about this subject,  it's not
> simple for the Zope Corp to merely plug-in a new version of
> Python and make a release.  Disregarding any issues about
> changes they might want to make that would take advantage of
> new features of Python 2.2 (and up), the underlying instrumentality
> of Zope, the ZODB (Zope Object DataBase) as it exists in  Zope 2.5
> and so on, will not run under Python versions > 2.1 due to
> incompatibilities of the C extension classes with the changes made
> (in Python 2.2) regarding type-class unification.
> 
> Obviously, there's already a version of ZODB for Python 2.2, but
> I think it's not something that one cannot just 'drop in' to a current
> version of Zope.

The issue is a little more subtle, actually.  The existing version of
ZODB -- the one that ships with Zope 2.5 -- works just fine with
Python 2.3.  Python works hard to be backwards compatible with C
extensions written against older versions of the C API.  The current
version of ZODB uses ExtensionClass, and that compiles cleanly and
runs with Python 2.1 and up.

The only compatibility issue is that ExtensionClass types don't mix
cleanly with new-style types.  But if you're not depending on
new-in-2.2 features, then you've got nothing to worry about.

Python 2.2 provides an opportunity to re-write ZODB to use new-style
types instead of ExtensionClass.  The new implementation will work
cleanly with new-style types.  The rewrite, dubbed ZODB4, is working,
but experimental.  We're may rewrite the API completely tomorrow <0.8
wink>.

You can checkout ZODB4 or Zope3 from cvs.zope.org to play with the new
version.

We're also discussing some ZODB issues -- basic persistence and
transaction APIs -- on the persistence-sig. 
http://www.python.org/sigs/persistence-sig

Jeremy


From phr-n2002b at NOSPAMnightsong.com  Sat Sep 21 01:34:10 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 20 Sep 2002 22:34:10 -0700
Subject: Decimal arithmetic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amabpb$5rm$1@news.nems.noaa.gov> <3D8B1537.4010803@thinkware.se> <ad052e5c.0209201128.1a53108@posting.google.com> <amgcgf$5e1fd$1@ID-125932.news.dfncis.de> <7xn0qcfafp.fsf@ruckus.brouhaha.com> <amgpm4$5si22$1@ID-125932.news.dfncis.de>
Message-ID: <7xwupf52jx.fsf@ruckus.brouhaha.com>

Christopher Browne <cbbrowne at acm.org> writes:
> REALITY CHECK for a moment.
> 
> What do you think A Real Live bank uses to compute interest?

If I had to guess, I'd guess they use floating point.

I do know that financial modelling software I worked on that was sold
to banks used floating point.  Also, spreadsheets use floating point.

But it's possible (likely even) that some bookkeeping applications use
BCD.  One that I worked on actually used exact, multi-precision
rational arithmetic, but I think that's unusual.

Note that the x86/x87 still supports BCD arithmetic in microcode.  In
reality, though, programs that use BCD to keep track of money don't do
all that much arithmetic, so software emulation is plenty fast enough.


From Lutz.Schroeer at kybernetik-manufaktur.de  Thu Sep  5 17:46:57 2002
From: Lutz.Schroeer at kybernetik-manufaktur.de (Lutz Schroeer)
Date: 5 Sep 2002 21:46:57 GMT
Subject: What database should I use
References: <mailman.1031257921.29400.python-list@python.org>
Message-ID: <Xns9280F2F7F8D12Latzikatz@139.174.2.24>

"A" <printers at sendme.cz> wrote in
news:mailman.1031257921.29400.python-list at python.org: 

> I am going to program a small application (in Python and wxPython)
> which consists about 2000 - 3000 records . Can you please recommend
> which database is the best for that. I use Windows.
> Maybe MS Access  or FoxPro or MySQL or better?
> I want to keep the program as small as possible.

>From your question I derive that you want to run the program on an MS 
Windows system.

Access seems to be the best solution. You can access the database over the 
ODBC driver without having a "real" server running in the background. You 
only need the *.mdb files.

But keep in mind, that you're stuck to Billy's system. I don't think that 
there are any modules providing access to *.mdb files on other systems. If 
there is, please enlighten me....


Latz


From fredrik at pythonware.com  Wed Sep  4 03:14:11 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 04 Sep 2002 07:14:11 GMT
Subject: Tkinter: text widget predefined key bindings
References: <2259b0e2.0208270502.5eb0a330@posting.google.com> <200208271417.21285.mfranklin1@gatwick.westerngeco.slb.com> <mailman.1030533583.4094.python-list@python.org> <akihne$kf$1@wanadoo.fr>
Message-ID: <7pid9.5887$e5.1033373@newsb.telia.net>

Eric Brunel wrote:

> Yep, but it does not work in all cases: if you type accented characters,
> they are ignored by your undo; same if you do a XWindows-style paste
> (selecting, then Button2 anywhere in your Text widget). I suppose any text
> change done via any default binding will probably also be ignored...
>
> Believe me: it's really harder than it seems... We tried to do it once and
> it never worked correctly...

the "right" way to do this is to dynamically override the "insert"
and "delete" commands, on the Tcl level.

IDLE has code to do this.

(for true enlightenment, write your own keyboard controller, and
plug it in using bind_class...)

:::

I once wrote a patch to Tkinter that added bind_insert and
bind_delete methods to the standard text widget, but I seem
to have lost the code.  If I find it, I'll post it here:

    http://effbot.org/zone/tkinter-text-bind-insert-delete.htm

</F>




From gerhard.haering at gmx.de  Sun Sep 15 22:19:03 2002
From: gerhard.haering at gmx.de (Gerhard Haering)
Date: Mon, 16 Sep 2002 04:19:03 +0200
Subject: can wxpython do this?
In-Reply-To: <am3d0b$jj$1@news01bc.so-net.ne.jp>
References: <737aefee.0209151702.3f735252@posting.google.com> <spcaou804m5oo6gah31oqaeuqsuj60h0ip@4ax.com> <am3d0b$jj$1@news01bc.so-net.ne.jp>
Message-ID: <20020916021903.GA4833@gargamel.ghaering.test>

* Erick <erick_papadakis at REMOVEFORSPAM.yahoo.com> [2002-09-16 10:42 +0900]:
> Thank you very much Lothar.
> 
> > There is no problem to setup an installer easily that takes
> > Python,WXPysthon and whatever you need. It simply blowes up the
> > installer package size.
> 
> Isn't this the way to distribute wxpython apps? Because most users will not
> have python installed on their machines, much less the wxpython libraries?
> Or is it that after compilation users do not need Python installed at all?

There is no compilation step in Python ;-) You can go both ways:

1) Put a Python installation with the necessary libraries and DLLs (wxPython +
python22.dll)  in a directory, then feed that directory to your favourite
installer maker (NSIS, whatever)

2) Use a tool like py2exe or McMillan's installer to automatically find
dependencies and make a .exe out of your app. This will still need the Python
runtime & DLLs and about the only difference is that you don't have lots of
.py{c|o} files lying around.

I personally favour option #1 and cutting down the Python runtime manually, as
I experienced that apps created with py2exe don't produce very useful
tracebacks any more. Besides, for some libraries like PIL the automatic
dependency checking of py2exe doesn't work very well, anyway.

> > >for instance, can i minimize my application to a user's desktop's taskbar
> > >and have a right-click context menu on that taskbar entry?
> 
> > I think its not comfortable but possible. Look at the API to find out.
> 
> I searched the html help for "taskbar" and there is only one function for
> Win95/NT (wxTaskBarIcon). Looks like this feature is not yet supported?

It is. Just install wxPython on Windows, start the demo app and minimize it.
Voil? a taskbar icon. I recently wrote an IMAP mail watcher for win32 myself
that doesn't have any visible window and only consists of one wxPython taskbar
icon that changes icons and can be double-clicked :-)

> > There is a really simple WxHTML Widget and at least a C++ version of a
> > wrapped IExplorer control.
> 
> The wxHTML widget supports only some very basic HTML tags, but I guess
> that'll do for now. What I really wanted to know was if there is a way for
> me to connect to the Internet, make a call to a webserver as a browser would
> do, retrieve data and display it inside my application (with or without
> HTML...does not have to be HTML at all, actually).

You don't need wxPython for that, the Python standard library has all the means
for that, including support for the most popular protocols like http, ftp,
smtp, pop3, imap and nntp. Raw sockets of course, too.

> Perhaps wxSocket is what I am talking about.

Features that Python has already (like sockets, threads, ODBC, ...) aren't
wrapped from wxWindows to wxPython. You can use standard Python libraries like
socket and threading for these.

> But I did not see anything in the "demo" that has much to do with sockets.
> Would anyone know an example of using sockets in wxpython?

There is. Check out the Slashdot headline watcher and the Hangman example. But
that's really a Python rather than a wxPython issue.

-- Gerhard



From David.Kastrup at t-online.de  Tue Sep 17 08:16:12 2002
From: David.Kastrup at t-online.de (David Kastrup)
Date: 17 Sep 2002 14:16:12 +0200
Subject: Using emacs gud/pdb python debugger under windows
References: <uvg5axb9e.fsf@online.no> <x53cse2dqw.fsf@tupik.goethe.zz> <r8fxwzdw.fsf@online.no> <x5adml216y.fsf@tupik.goethe.zz> <87elbuo46g.fsf@tleepslib.sk.tsukuba.ac.jp> <87n0qiul6j.fsf@computer.localdomain> <x5lm623vtf.fsf@tupik.goethe.zz> <ufzwam04i.fsf@online.no> <x5y9a22b95.fsf@tupik.goethe.zz> <usn09x6br.fsf@online.no> <u1y7t6zld.fsf@hotmail.com> <3cs86d38.fsf_-_@online.no>
Message-ID: <x5ofawak1f.fsf@tupik.goethe.zz>

Syver Enstad <syver-en+usenet at online.no> writes:

> Felix <aprogrammer at hotmail.com> writes:
> 
> > Syver Enstad <syver-en+usenet at online.no> writes:
> > 
> > u just need to fix a few things to make it recogonize the win32 path.
> > here's how, open gud.el and search for the variable
> > 'gud-pdb-marker-regexp',
> > 
> > then modify it to something like: 
> > (defvar gud-pdb-marker-regexp
> >   "^>
> > \\([-a-zA-Z0-9_/.:\\]*\\|<string>\\)(\\([0-9]+\\))\\([a-zA-Z0-9_]*\\|\\?\\)()\\(->[^\n]*\\)?\n")
> > 
> > this works well for me.
> 
> Ha, ha it works, it works! (Dancing around the room) I want to hug you
> Felix! I didn't think of that, obviously. So it was just pdb failing
> to recognize the output from the windows version of python.

When you are done frolicking, you should be sending this information
on to the package maintainer, so that others will not have to bear
this problem in future.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum
Email: David.Kastrup at t-online.de


From johnroth at ameritech.net  Mon Sep  2 17:49:23 2002
From: johnroth at ameritech.net (John Roth)
Date: Mon, 2 Sep 2002 17:49:23 -0400
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de>
Message-ID: <un7n06qrl1gm93@news.supernews.com>

"Gerson Kurz" <gerson.kurz at t-online.de> wrote in message
news:3d725881.345921 at news.t-online.de...
> I wrote a small wrapper module to "struct" today that will allow me to
> write C declarations and have
>
> ...
> print p.declare("""
>
> struct test1
> {
>    int a,b;
>    float this,and,that;
> };
>
> struct test2
> {
>     int count;
>     struct test1 data[80]
> };
>
> """)
> ...
> and have it create instances I can easily assign data to
> ...
> t2 = p.createInstance("test2")
> t2.data[0].a = 42
> ...
> and pack for C extensions by utilizing the struct module
> ...
> data = p.pack(t2)
> ...
>
> because of being fed up with that pack("qh34s>id",...) stuff. (I'll
> post that to my website once its in a state I can let somebody else
> see ;)
>
> Anyway, that got me thinking on why do we have to deal with regular
> expressions like r"((?:a|b)*)", when in most cases the code will look
> something like this:
>
> r = re.compile("<some cryptic re-string here>")
> ...
> r.match(this) or r.find(that)
>
> which means the real time is not spent in the compile() function, but
> in the match or find function. So basically, couldn't one come up with
> a *human readable* syntax for re, and compile that instead? Python
> prides itself on its clean syntax, and human readability, an bang -
> import re, get perl-ish code instantly!
>
> Also, I think it would already be an improvement if the syntax
> provided for clear and easy-to-understand special cases, like
>
> re.compile("anything that starts with 'abc'")
>
> and if you cannot find something in the special cases for you, you can
> always go back to
>
> re.compile("<some cryptinc re-string here>")
>
> After all, *everyone* starting with re thinks the syntax is cryptic
> and mind-boggling, and only if you get yourself into the "re mindset",
> you understand things like r"\s*\w+\s*=\s*['\"].*?['\"]" instantly. If
> we had an easier syntax, more people would be using re ;)
>
> Is the idea utterly foolish?

No, it's not utterly foolish. You might be surprised to learn that
Larry Wall agrees with you that the Perl regex syntax is much
too obtuse, and in need of a basic, ground up redesign. Even
current Perl syntax allows you a special form where you can
insert blanks for readability.

http://www.perl.com/pub/a/2002/06/04/apo5.html

http://www.perl.com/pub/a/2002/08/22/exegesis5.html

It's an interesting redesign of basic regex functionality.
Some of the things you can do with it are very, very
interesting indeed.

John Roth
>




From aahz at pythoncraft.com  Sun Sep  8 10:40:20 2002
From: aahz at pythoncraft.com (Aahz)
Date: 8 Sep 2002 10:40:20 -0400
Subject: Names and bindings (was Re: Scope of instantiated class)
References: <3D3ACCA6.7030002@lmco.com> <lku1mpl69n.fsf@pc150.maths.bris.ac.uk> <ajh9l5$3gf$1@panix1.panix.com> <lkhehvktr6.fsf@pc150.maths.bris.ac.uk>
Message-ID: <alfngk$e6o$1@panix1.panix.com>

[silent snippage ahead]

In article <lkhehvktr6.fsf at pc150.maths.bris.ac.uk>,
Michael Hudson  <mwh at python.net> wrote:
>aahz at pythoncraft.com (Aahz) writes:
>> In article <lku1mpl69n.fsf at pc150.maths.bris.ac.uk>,
>> Michael Hudson  <mwh at python.net> wrote:
>>>
>>>Common Lisp calls them "places", which is a reasonable name.  It's not
>>>that much of an issue in Python as there aren't that many of them
>>>(names, attributes, subscripts, slices).  In CL, you can define your
>>>own (tho' it's a bit tricky).
>> 
>> What exactly does "place" refer to in Common Lisp?  Is a place capable
>> of holding a value, or can places only contain references?
>
>  place n. 1. a form which is suitable for use as a generalized
>       reference. 2. the conceptual location referred to by such a
>       place[1].
>
>and 
>
>  generalized reference n. a reference to a location storing an object
>       as if to a variable. (Such a reference can be either to read or
>       write the location.)  See Section 5.1 (Generalized
>       Reference). See also place.
>
>So I guess more of the latter of your two options.

Do Common Lisp places have automatic reference/dereference semantics
like Python?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From bobx at linuxmail.org  Wed Sep 11 21:07:45 2002
From: bobx at linuxmail.org (Bob X)
Date: Thu, 12 Sep 2002 01:07:45 GMT
Subject: [FEEDBACK] Is this script efficient...is there a better way?
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <alofoq$lgu$1@peabody.colorado.edu>
Message-ID: <BNRf9.1777$Lg2.246756@news2.news.adelphia.net>

Fernando P?rez wrote:
> Bob X wrote:
> 
> 
>>I use this to parse a log file (30MB+) for keywords and write the lines
>>that are found into another file.
>>
>>Any thoughts would be appreciated...  :-)
>>
> 
> 
> How about
> 
> $ grep words infile | tee outfile
> 
> Cheers,
> 
> f.
grep is not usually on Windows...and I *want* to use Python.  :-)



From christian.schmutzer at gmx.net  Tue Sep  3 07:45:37 2002
From: christian.schmutzer at gmx.net (christian)
Date: 3 Sep 2002 04:45:37 -0700
Subject: can't access com object - why doesn't this work?
Message-ID: <edeadff.0209030345.35bc63a4@posting.google.com>

Hi

I ran into a strange problem when accesing the com interface of VC7
using python.
(I am using python and makepy to access the com interface of VC7.)

------------------------------------------
my python code looks as follows:
------------------------------------------
...
DTE = win32com.client.Dispatch('VisualStudio.DTE')
cfgs = DTE.Solution.Projects.Item(1).ConfigurationManager.Platform("Win32")
cfg = cfgs.Item(1)
print cfg
props = cfg.Properties	   
...
-----------------------------

Basically I retrieve a configuration object from the configuration
manager and then try to access a member function. But the
configuration object always seems to be broken.
When I execute the script, the print still works fine: 
output: "<win32com.gen_py.Microsoft Development Environment
7.0.Configuration>"

but accessing the python-property "Properties" causes this error:

Traceback (most recent call last):
  [...]
  File "C:\PROGRA~1\python\lib\site-packages\win32com\client\__init__.py",
line 355, in __getattr__
    return apply(self._ApplyTypes_, args)
  File "C:\PROGRA~1\python\lib\site-packages\win32com\client\__init__.py",
line 349, in _ApplyTypes_
    return self._get_good_object_(apply(self._oleobj_.InvokeTypes,
(dispid, 0, wFlags, retType, argTypes) + args), user, resultCLSID)
pywintypes.com_error: (-2147352567, 'Exception occured.', (0, None,
None, None, 0, -2147467259), None)


But I really don't understand, why this is illegal. I have researched
the com interface, its relations and internal structures in the
documentation and this should be valid.
I looked at what makepy generated, and that also says that this call
should be possible:

class ConfigurationManager(DispatchBaseClass):
[...]
	# Result is of type Configurations
	def Platform(self, Name=defaultNamedNotOptArg):
		"""Returns a collection of ConfigurationAssignment object for the
specified platform."""
		ret = self._oleobj_.InvokeTypes(0x8, LCID, 1, (9, 0), ((8,
0),),Name)
		if ret is not None: ret = win32com.client.Dispatch(ret, 'Platform',
'{B6B4C8D6-4D27-43B9-B45C-52BD16B6BA38}', UnicodeToString=0)
		return ret
...

class Configurations(DispatchBaseClass):
	"""A collection of Configuration objects."""
	CLSID = pythoncom.MakeIID('{B6B4C8D6-4D27-43B9-B45C-52BD16B6BA38}')

	# Result is of type Configuration
	def Item(self, index=defaultNamedNotOptArg):
		"""Returns an indexed member of a collection."""
		ret = self._oleobj_.InvokeTypes(0x0, LCID, 1, (9, 0), ((12,
1),),index)
		if ret is not None: ret = win32com.client.Dispatch(ret, 'Item',
'{90813589-FE21-4AA4-A2E5-053FD274E980}', UnicodeToString=0)
		return ret
...

class Configuration(DispatchBaseClass):
[...]

	_prop_map_get_ = {
		[...]
		# Method 'Properties' returns object of type 'Properties'
		"Properties": (7, 2, (9, 0), (), "Properties",
'{4CC8CCF5-A926-4646-B17F-B4940CAED472}'),
		"Type": (5, 2, (3, 0), (), "Type", None),
		
So for all I can see, this should work.
So I have also tried out a VB script to retrieve the information, and
it works just fine, so the com objects seem to be valid after all.

-------------------------------------
The VB script looks like this:
-------------------------------------
Dim cfgs As Configurations
Dim cfg As Configuration
Dim props As Properties

cfgs = DTE.Solution.Projects.Item(1).ConfigurationManager.Platform("Win32")
cfg = cfgs.Item(1)
props = cfg.Properties
--------------------------------------

This script works just fine. As you can see, apart from some minor
sytactical things, the script is identical.
I really don't get why this doesn't work in python. So what am I doing
wrong? Am I using python the wrong way, or could this be a python-com
problem?
Has anybody encountered a similar problem?

Thanks,

Christian


From ralf_ahlbrink at web.de  Sat Sep 21 12:45:02 2002
From: ralf_ahlbrink at web.de (RA)
Date: Sat, 21 Sep 2002 18:45:02 +0200
Subject: PyQt setup on Linux Mandrake 8.2
References: <3D8B4898.20808@nospam.attbi.com> <ddc19db7.0209210259.7beb52d4@posting.google.com> <3D8BC2F4.8080107@nospam.attbi.com>
Message-ID: <3d8ca20e$0$151$9b622d9e@news.freenet.de>

On Sat, 21 Sep 2002 14:33:00 +0000, dwelch wrote:

> Armin Steinhoff wrote:
>> dwelch <dwelch91 at nospam.attbi.com> wrote in message news:<3D8B4898.20808 at nospam.attbi.com>...
>> 
>>>I would like to setup PyQt 3.4 on my Mandrake 8.2 box. I am having 
>>>difficulty, and I am worried that if I do the wrong thing, KDE will go 
>>>belly-up on me since it is Qt based.
>>>
>>>First, question is: why isn't QTDIR defined by default? 
>>>
>> 
>> PyQt works with Qt2.x.x and Qt 3.x.x ... so YOU have to choose the
>> correct directory.
>> 
>> 
>>>Is QTDIR a 
>>>runtime setting or a development setting?
>>>
>> 
>> Both ..
>> 
>> 
>>>So I set QTDIR to /usr/lib/qt2 (that seems to be where the libqt2 RPMs 
>>>files are). Next question is: Since I have libqt2 already on my system, 
>>>how do I add the Qt development pieces without screwing up Qt/KDE? Is 
>>>/usr/lib/qt2 even the correct location?
>>>
>> 
>> Yes ... if you want to use only the latest version of Qt 2.x.x
>>  
>> 
>>>When I run SIP 3.4 build.py, it complains about tmake not being 
>>>available. Is tmake included with Qt2 development files?
>>>
>> 
>> It's available at the Trolltech homepage or its mirrors.
>> 
>> 
>>>SIP doesn't 
>>>seem to have any docs, so I'm not sure how to proceed.
>>>
>> 
>> Nomally you haven't to care about SIP ..
>> 
>>  
>> 
>>>Also, my RPM manager says that I have libqt2-devel (2.3.1-29mdk) 
>>>installed. No tmake listed in the files, though. And, PyQt-devel 
>>>(2.5-1mdk) is also listed as installed, but there aren't any .py files 
>>>included, only .sip files (I found this curious).
>>>
>> 
>> Corious ... there must be examples included.
>>  
>> I would suggest to dowload the latest version of Qt from the Trolltech
>> homepage, just for a simple and clean recompilation.
>> 
>> And do the same with SIP and PyQt from the Riverbank homepage ...
>> 
>> It seems to be better to forget that outdated stuff provided by the 
>> Mandrake distribution ...
>> 
>> Armin
>> 
>> 
>>>Installing software on Linux makes me feel like such an idiot!
>>>
>>>Any help or pointers would be much appreciated.
>>>
>>>Don
>>>
> 
> OK, I found and installed tmake. Now, SIP can't find some of the Qt 
> pieces it needs. So, I still don't understand the situation with Qt. Is 
> it really possible to install and run my applications that I write with 
> say Qt 3.0.5, when KDE on my system is using Qt 2.3.1? How does/would 
> KDE know which version to use? How will my apps know what version to 
> use? I am assuming there can only be one QTDIR active on a system at a 
> time...

I'm not sure. But perhaps you can also use qmake, which is located
in the libqt3-devel-3.0.5 RPM. This package contains also header files for
the compilation of  Qt3 apps.

For compilation perhaps setting QTDIR is enough. Maybe you have to set
something like --qt-dir during configure (try configure --help).
> 
> Thanks for the help,
> 
> Don

Ralf.



From max at alcyone.com  Thu Sep 26 17:34:04 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 26 Sep 2002 14:34:04 -0700
Subject: Who knows somefunction?
References: <3D937B08.4090704@aon.at>
Message-ID: <3D937D4C.2BFDF40A@alcyone.com>

Gregor Lingl wrote:

> Is there a function <somefunction> in Python
> for which
> 
> something is <somefunction>(id(something))
> 
> returns invariably true?

Not offhand, but there's an easier test.  x is y really just means id(x)
== id(y), so you could change this test to:

	id(something) == id(somethingElse)

and you probably wouldn't have the need for your hypothetical function.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Love is the wisdom of the fool and the folly of the wise.
\__/ Samuel Johnson
    Python chess module / http://www.alcyone.com/pyos/chess/
 A chess game adjudicator in Python.


From news at datatailors.com  Fri Sep 20 17:00:09 2002
From: news at datatailors.com (Peter van Kampen)
Date: Fri, 20 Sep 2002 23:00:09 +0200
Subject: saving state in Python/CGI
References: <mailman.1032544767.23458.python-list@python.org>
Message-ID: <slrnaon32o.ro.news@datatailors.com>

In article <mailman.1032544767.23458.python-list at python.org>, Frank Gibbons 
wrote:
> Hi,
> 
> I've been using Python (and Jython) for about 5 months now. So far, so 
> great. I'm building a CGI app that has about 5 different stages, and I need 
> to maintain state between them. Specifically, it's a scientific tool that 
> requires the user to upload a (possibly quite large) datafile, processes 
> it, does some computation, then hands it back (formatted) with results.
> 
> I need access to the data on each page, yet don't want to pass it as a 
> hidden field, because of its potential size. In Perl, I could just use the 
> save() method from CGI.pm to save the form on the server side, pass a 
> session key in a hidden field, then retrieve the data using the key. For 
> the life of me, I can't find an equivalent way in Python. cgiFormStorage 
> looks like a dictionary, but it's immutable, so you can't retrieve data 
> from disk and add it to the form. cgi.py appears to have no analog to 
> CGI.pm's save() method, at least not that I've been able to find. I've 
> tried pickling the form, but that doesn't work either (it saves something 
> to disk alright, but not the large data retrieved from a multi-part form 
> file upload).
> 
> Perhaps I've missed something. Maybe that's just not how you do things in 
> Python. Can anyone help?
> 
> Thanks,
> 
> -Frank Gibbons
> 

Hi Frank,

Perhaps you can use the code below. It's my first little python
project so I'm mainly posting this for some feedback. It's a simpler
version (at least to my mind) of something I found in the vaults.

It creates a dict to store sessionvars in and pickles it to the hd. You
have to setup a dir where the webserver account has write (and read)
access. It defaults to .session/ in the current dir. 

It uses cookies to pass the sessionid that's why you have be careful to
write the headers that the class generates. It uses apache's
mod_unique_id if it is enabled (and you're using apache ;-) I haven't
tested yet but it should work in any cgi-environment) which provides
nicer cookie-id's than tempfiles which is used as a fallback but I
haven't checked if tempfilenames are safe sessionids. 

You can use it like this.

#First page

import CookieSession

s = CookieSession.CookieSession()

s["whatever"] = "you want (if it can be pickled)"

s.save()

print "Content-type: text/html"
print s.headers()
print
print YourHTML

#Another page

import CookieSession

s = CookieSession.CookieSession()

restore_from_session = s["whatever"]

As I said. Comments are very welcome.

PterK


#! /usr/bin/env python

# CookieSession.py
#
# Based on stuff i found at http://wingide.com/opensource/httpsession 
# 	which in turn I located in the Vaults of Parnassus : 
#			http://py.vaults.ca/parnassus/

import os
import time
import stat
import glob
import Cookie
import cPickle
import tempfile 

#python2.2 version
#class CookieSession(dict):

class CookieSession:
	"""
	TODO: Write docstring
	
	"""

	#-----------------------------------------------------------------------
	def __init__(self, timeout = 20, \
			store = ".sessions"):
		""" Init the class.
			1. Set timeout
			2. Set where to store the pickles 
			3. Clear up old sessions
			4. Find out if a session is active
			5. Yes? Load it and 'touch' it
			6. No? Create a new one, set some attributes and save it to disk.

		"""
		
		self.__timeout = timeout * 60
		self.__session_store = store
		self.__cookies = "Cache-control: no-cache\r\n"
		self.__on_expire_is_set = 0
		
		# expire old sessions
		self.__ClearOldSessions()

		# Get the current session or create a new one
		try:
			self.__sessionid = Cookie.SimpleCookie(os.environ["HTTP_COOKIE"]) \
									["sessionid"].value
			#touch session (from an effbot-posting)
			now = time.time()
			os.utime(self.__SessionFilename(), (now, now))
			#load session
			self.__SessionDict = self.__LoadSession()
		except:
			# Create a new session
			self.__sessionid = self.__CreateSessionid()
			self.__CreateCookie()
			self.__SessionDict = {}
			self["sessionid"] = self.__sessionid
			self.save()

	#-----------------------------------------------------------------------
	def expire(self):
		"Forces the current Session to expire"
		self.__on_expire()

	#-----------------------------------------------------------------------
	def headers(self):
		"Creates all headers for the calling page"
		return self.__cookies

	#-----------------------------------------------------------------------
	def save(self):
		"Saves the sessionDict to a pickled file"
		cPickle.dump(self.__SessionDict, self.__OpenSessionFile("w"))

	#-----------------------------------------------------------------------
	def __setattr__(self, name, value):
		"Used to check whether on_expire is set. Maybe a method is better?)"
		self.__dict__[name] = value
		if name == "on_expire":
			self.__on_expire_is_set = 1
		
	#-----------------------------------------------------------------------
	def __on_expire(self, sessionid = None):
		""" Calls a function that can be assigned to the attribute on_expire
			similar to ASP's On_Session_End. 
			
			It passes the session about to expire.
		
			def OnExpire(expired_session):
				if not expired_session["TransactionComplete"]:
					#Remove cruft from database
					pass
				
			s = CookieSession()
			s.on_expire = OnExpire
			
			...
			
			s.expire()

		"""

		if sessionid is None: sessionid = self.__sessionid
		try:
			# Test if it is set to prevent unnecessary loading of pickles 
			if self.__on_expire_is_set:
				self.on_expire(self.__LoadSession(sessionid))
		finally:
			os.remove(self.__SessionFilename(sessionid))

	#-----------------------------------------------------------------------
	def __CreateSessionid(self):
		"Create a sessionid from the apache unique_id module or a tempfile."
		try:
			id = os.environ["UNIQUE_ID"]
		except:
			id = os.path.basename(tempfile.mktemp())
		return id

	#-----------------------------------------------------------------------
	def __CreateCookie(self):
		"Create a cookie to store the sessionid on the clientside"
		C = Cookie.SimpleCookie()
		C["sessionid"] = self.__sessionid
		C["sessionid"]["path"] = "/"
		self.__cookies += C.output()

	#-----------------------------------------------------------------------
	def __SessionFilename(self, sessionid = None):
		if sessionid is None: sessionid = self.__sessionid
		return self.__session_store + "/.session" + str(sessionid)

	#-----------------------------------------------------------------------
	def __OpenSessionFile(self, mode, sessionid = None):
		"Open up the on-disk session persistence file"
		if sessionid is None: sessionid = self.__sessionid
		f = open(self.__SessionFilename(sessionid), mode)
		return f

	#-----------------------------------------------------------------------
	def __LoadSession(self, sessionid = None):
		"Loads a session from a pickled file"
		if sessionid is None: sessionid = self.__sessionid
		return cPickle.load(self.__OpenSessionFile("r", sessionid))
	
	#-----------------------------------------------------------------------
	def __ClearOldSessions(self):
		"Remove any expired session files"
		files = glob.glob(os.path.join(self.__session_store, '.session*'))
		for file in files:
			filetime = os.stat(file)[stat.ST_MTIME]
			currtime = time.time()
			if filetime + self.__timeout < currtime:
				# len(".session") = 8
				sessionid = file[len(self.__session_store)+9:]
				self.__on_expire(sessionid)

	#-----------------------------------------------------------------------
	# In 2.2 this works : 
	# class CookieSession(dict):
	#
	# Python map emulation : This (below) can be dropped for python2.2
	#-----------------------------------------------------------------------
	def __len__(self):
		return len(self.__SessionDict)

	#-----------------------------------------------------------------------
	def __getitem__(self, key):
		retval = self.__SessionDict.get(key)
		if retval == None:
		  retval = ""
		return retval

	#-----------------------------------------------------------------------
	def __setitem__(self, key, value):
		self.__SessionDict[key] = value

	#-----------------------------------------------------------------------
	def __delitem__(self, key):
		del self.__SessionDict[key]

	#-----------------------------------------------------------------------
	def keys(self):
		return self.__SessionDict.keys()

	#-----------------------------------------------------------------------
	def values(self):
		return self.__SessionDict.values()

	#-----------------------------------------------------------------------
	def items(self):
		return self.__SessionDict.items()

	#-----------------------------------------------------------------------
	def has_key(self, key):
		return self.__SessionDict.has_key(key)

	#-----------------------------------------------------------------------
	def get(self, key, f=None):
		return self.__SessionDict.get(key, f)

	#-----------------------------------------------------------------------
	def clear():
		return self.__SessionDict.clear()

	#-----------------------------------------------------------------------
	def copy():
		return self.__SessionDict.copy()

	#-----------------------------------------------------------------------
	def update(b):
		return self.__SessionDict.update(b)



From e_fax_t at hotmail._ZAPME_.com  Sun Sep 22 11:16:56 2002
From: e_fax_t at hotmail._ZAPME_.com (Justin)
Date: Sun, 22 Sep 2002 17:16:56 +0200
Subject: Wanted: Tix examples
Message-ID: <3d8dde58$0$30454$afc38c87@sisyphus.news.be.easynet.net>

Hi,

I see that Activestate's builds of python 2.2.1 both
for Win32 and linux include the Tix.py module.
No example at all seems to be provided, though.
Even a visit to tix.sourceforge.net did not make it
obvious to me where to start.
The module itself mentions the tixwidgets.py demo;
where does one find it? I am primarily interested in
the tree-view widget. Any pointer appreciated,

TIA,

Justin.




From andrew.thompson at ashecastle.com  Thu Sep 26 04:25:59 2002
From: andrew.thompson at ashecastle.com (Andrew Thompson)
Date: Thu, 26 Sep 2002 09:25:59 +0100
Subject: object changing itself to another object 
In-Reply-To: <amufk0$8d1$1@wanadoo.fr>
Message-ID: <019e01c26536$58613480$4802a8c0@blair>

Smalltalk (at least the Digitalk versions) provide a means for an object
changing into another object via the #become: message.  




-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]
On Behalf Of Eric Brunel
Sent: 26 September 2002 09:22
To: python-list at python.org
Subject: Re: object changing itself to another object

Glen Murphy wrote:

> I'm creating a program that takes user input, and passes it to an
object -
> however, I want that object itself, and not the parent to change where
> that input goes. In the highly simplified example below, I would like
the
> output to be "Obj1","Obj2","Obj1" - but not suprisingly, I just get
Obj1,
> Obj1, Obj1
> 
> 
> ### code begin  ###
> 
> class Obj1:
>     def key1(self):
>         # do obj1 specific stuff
>         self = Obj2()
>         print "Obj1"
> 
> class Obj2:
>     def key1(self):
>         # do obj1 specific stuff
>         self = Obj1()
>         print "Obj2"
> 
> a = Obj1()
> 
> # simulate user keypresses
> a.key1()
> a.key1()
> a.key1()
> 
> ### code end ###
> 
> I know I could achieve the result I want by doing horribly complicated
> trees of ifs and such, but I'd like my code to be nice and expandable
(in
> this example, to easily add new Objs). I've looked through the Python
FAQ,
> Google Groups and Various O'Reilly books, but I don't really know the
> terminology for what I'm looking for, so I haven't found anything so
far,
> so does anyone have any pointers?

AFAIK, there's no way to do exactly what you request. In fact, I've
never 
seen a language allowing to dynamically change the class of an object
once 
it's been created. But there's a quite simple workaround using a factory

object. For example:

class Obj1:
  ...

class Obj2:
  ...

class ObjFactory:
  def __init__(self):
    self.__theClasses = [Obj1, Obj2]
    self.__classIndex = 0
    self.__theObject = None
  def getObject(self):
    self.nextObject()
    return self.__theObject
  def nextObject(self):
    self.__theObject = self.__theClasses[self.__classIndex]
    self.__classIndex = (self.__classIndex + 1) / len(self.__theClasses)

And now, it does what you want:

a = ObjFactory()

a.getObject().key1()     # Called on Obj1
a.getObject().key1()     # Called on Obj2
a.getObject().key1()     # Called on Obj1

I don't know if it suits your needs exactly, but you get the idea... And

since the actual objects are created by the factory, you can make them
know 
the factory and call themselves the nextObject method when needed
instead 
of calling it systematically in getObject.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools -
http://www.pragmadev.com
-- 
http://mail.python.org/mailman/listinfo/python-list




From gerhard.haering at gmx.de  Sun Sep  1 14:44:39 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 1 Sep 2002 18:44:39 GMT
Subject: Add methods to a class at runtime?
References: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com> <slrnan4dl3.18h.gerhard.haering@lilith.my-fqdn.de> <YYrc9.31010$vY2.601601@e3500-atl2.usenetserver.com>
Message-ID: <slrnan4o0m.3i4.gerhard.haering@lilith.my-fqdn.de>

Robert Oschler wrote in comp.lang.python:
> "Gerhard H?ring" <gerhard.haering at gmx.de> wrote:
>> >>> class Foo: pass
>> ...
>> >>> foo = Foo()
>> >>> foo.bar()
>> Traceback (most recent call last):
>>   File "<stdin>", line 1, in ?
>> AttributeError: Foo instance has no attribute 'bar'
>> >>> def bar(self): print "bar"
>> ...
>> >>> Foo.bar = bar
>> >>> foo.bar()
>> bar
>> >>>
> 
> Gerhard,
> 
> Thanks but I need a little bit more detail.  Suppose I want to do it from
> within code?  For example, I create a function called
> AddBarToFoo(FooObject), that given a object of type Foo (FooObject) adds the
> Bar method to it dynamically, what would be the syntax then?

To get the class of an instance, you can query its __class__
attribute.

> Would I have to put the method declaration code, which you show
> being typed in to the interpreter, in a string and then eval/exec
> it?

No, you hardly ever need to use eval or exec. Sometimes you only want
to "specialize" an existing, more generic function. For a real-life
example, here's something I experimented with yesterday:

# [snip generic_insert, generic_update]
def generic_delete(tablename, id):
    cursor = getdb().cursor()
    statement = "DELETE FROM %s WHERE id=%%s" % tablename
    cursor.execute(statement, (id,))

def make_functions(tablename):
    globals()[tablename + "_DELETE"] = lambda id: generic_delete(tablename, id)
    
# Create a specialized DELETE function for the table TEST which will
# be named TEST_DELETE (this will be a specialized generic_delete):
make_functions("TEST")

# Execute the newly created function:
TEST_DELETE(25)

I never really needed to add methods to classes on-the-fly, but the
above technique came in useful a few times. One other thing I found
useful is dynamically creating a new class (with certain class
attributes) using the new module.

As in certain other languages starting with "P", not everything you
/can/ do in Python is actually very useful, or necessarily a good
idea.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From fperez528 at yahoo.com  Wed Sep  4 17:37:08 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 04 Sep 2002 15:37:08 -0600
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Sep  4)
References: <448A0F008E0AFE11.19B9EF86BEE2B191.107D5020B234F743@lp.airnews.net> <al5tvu$1mtjjr$1@ID-71831.news.dfncis.de>
Message-ID: <al5ue4$sj2$1@peabody.colorado.edu>

James Kew wrote:

> I can't persuade Google to find this message-id; any chance of a link?

google_groups('comp.lang.python anansispaceworks.com')[0]

Cheers,

f.


From peter at engcorp.com  Fri Sep 13 21:38:55 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 13 Sep 2002 21:38:55 -0400
Subject: transform list of int in a list of string??
References: <mailman.1031945716.2586.python-list@python.org> <3d8290af$1@news.sentex.net>
Message-ID: <3d82931e@news.sentex.net>

Peter Hansen wrote:
> jubafre at brturbo.com wrote:
> 
>> my program have other modules imported, like re and sre and i put a 
>> convertion int to str but doesn?t work
>> t=str(x[i])
>> Traceback (most recent call last):
>>   File "C:\Documents and Settings\jubafre\Desktop\montador\teste.py", 
>> line 78, in ?
>>     t=str(x[i])
>> TypeError: 'list' object is not callable
> 
> 
> This error is most likely because you created a list object
> bound to the name "str".  Do *not* use "str" or other such
> reserved names as variable names if you want to preserve your
> sanity. :-)

More on this: they're not really "reserved", just built-in names
that your local definitions will hide if you aren't careful.

To see the full list:

 >>> import __builtin__
 >>> dir(__builtin__)
['ArithmeticError', 'AssertionError', .......  'xrange', 'zip' ]

Also, you can safely use them if you know what you're doing and
decide you really do want a variable with that name, but obviously
you can't use the built-in function with the same name if you do so.

-Peter



From sismex01 at hebmex.com  Thu Sep 12 13:51:09 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Thu, 12 Sep 2002 12:51:09 -0500
Subject: ANN: Thinking in Tkinter
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233023FF9D7@mail.hebmex.com>

> 
> True. It would help if the command callback sent an arguement
> that identified the widget (making it look a bit more like an
> event binding). In any event, it doesn't and command is
> unfortunately still the best way to handle button widgets.
> 

Do something like this:

# This is a kinda "ordinary" callback class.
class Callback:
  def __init__(self, function, *args, **kw):
    self._args = args
    self._kw = kw
    self._func = function
    self._button = button
  def __call__(self):
    return self._func(*self._args, *self._kw)

# And, when you define the button, do something like:
btn = Tkinter.Button(someframe, label="huuuge peeectorall muscels")
btn["command"] = Callback(some_func, btn, any-other-args)

# When "btn" is clicked on, the first argument in some_func
# will be btn.

Good luck!

-gus

--

-- please disregard offensive and arrogant legalese. :-(







> In general Tkinter has several useful callback mechanisms, 
> each of which 
> is especially easy to use for a particular type of problem 
> and may be a 
> headache for other sorts of problems (e.g. events to simulate button 
> command). Examples (aside from Button command):
> 
> Event binding examples:
> - Use mouse button events to do useful things in the Canvas 
> widget. For 
> instance you can easily drag objects around by binding to 
> <Button-1> and 
> <B1-Motion>. Or show information when the pointer is near an 
> object on a 
> Canvas using <Motion>.
> - Use keyboard events to modify the behavior of Entry widgets, for 
> example bind to the return keyto accept the data and use it 
> somehow, or 
> bind to up-arrow and down-arrow to implement a history mechanism.
> 
> Variable trace:
> - To implement input validation in Entry widgets, specify a 
> textvar and 
> put a trace on the variable. In your callback beep and 
> restore the old 
> value of the variable if the proposed new value is invalid.
> 
> There are others. My Tkinter Summary 
> <http://www.astro.washington.edu/owen/TkinterSummary.html> section 
> "Events and Callbacks" has what I believe is a complete set though 
> without examples. I'm should take another look at it and 
> expand it a bit.
> 
> -- Russell
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 

Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From anton at vredegoor.doge.nl  Fri Sep 20 14:17:04 2002
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Fri, 20 Sep 2002 20:17:04 +0200
Subject: bit operations ?
References: <amfbuq$56f28$1@ID-146704.news.dfncis.de>
Message-ID: <amfoog$15m$1@news.hccnet.nl>

On Fri, 20 Sep 2002 16:40:18 +0200, "Shagshag13" <shagshag13 at yahoo.fr>
wrote:

>i'm looking for any libs which could handle bit operations on integer.
>(i must be able to get 0 to n bits from any int., only bit j, etc.)

If you're interested in setting/getting a lot of bits of a long
integer at once, this might be faster than bitshifting. I am not
guaranteeing that it is failsafe, or even that it is fast. Just hobby
programmer code I wrote when I was chasing phantoms. It works only for
long integers >= 0L now, but this should be easy to correct.

The two functions are reciprocal, so if bits are set with the setbits
function the getbits function does the reverse.

Anton.

#bitset.py test script

from time import time
import dis

#list of bit positions that are set to one
bitget ={'0' : [], '1' : [0], '2' : [1], '3' : [0,1],
         '4' : [2], '5' : [0,2], '6' : [1,2], '7' : [0,1,2],
         '8' : [3], '9' : [0,3], 'A' : [1,3], 'B' : [0,1,3],
         'C' : [2,3], 'D' : [0,2,3], 'E' : [1,2,3], 'F' : [0,1,2,3]}

#f.e. if bit at position 1 is set, '8' changes to 'A' etc.
bitset = {  '0' : ['1','2','4','8'],  '1' : ['1','3','5','9'],
            '2' : ['3','2','6','A'],  '3' : ['3','3','7','B'],
            '4' : ['5','6','4','C'],  '5' : ['5','7','5','D'],
            '6' : ['7','6','6','E'],  '7' : ['7','7','7','F'],
            '8' : ['9','A','C','8'],  '9' : ['9','B','D','9'],
            'A' : ['B','A','E','A'],  'B' : ['B','B','F','B'],
            'C' : ['D','B','E','C'],  'D' : ['D','F','D','D'],
            'E' : ['F','E','E','E'],  'F' : ['F','F','F','F'] }
                    
def setbits(ilist):
    #set bits at the positions defined in ilist,
    #returns a long integer
    dd,m = divmod(max(ilist),4)
    if m:
        dd+=1
    res = ['0'] * (dd+1)
    for x in ilist:
        d, m = divmod(x,4)
        i = dd-d
        res[i] = bitset[res[i]][m]
    return long(''.join(res),16)

def getbits(along):
    #return a list of integers corresponding to the 
    #bitpositions that are set
    h = hex(along)[2:-1]
    res = []
    i = 0
    for j in range(len(h)-1,-1,-1):
        for pos in bitget[h[j]]:
            res.append(i+pos)
        i+=4
    return res

def test():
    n =1000
    a = 2349384569856987569824652587349873498723498L
    b = getbits(a)
    t=time()
    for i in range(n):
        getbits(a)
    print time()-t
    t=time()
    for i in range(n):
        setbits(b)
    print time()-t
    print setbits(b) == a
  #  print dis.dis(setbits)
  #  print dis.dis(getbits)

if __name__ == '__main__':
    test()





From see_reply_address at something.invalid  Tue Sep 17 21:13:53 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Wed, 18 Sep 2002 13:13:53 +1200
Subject: how to cast an instance id
References: <3D86E888.90AF569@millfilm.co.uk>
Message-ID: <3D87D351.5090508@something.invalid>

Eric Texier wrote:

> It there a way to create a new reference of  an object
> from the string representation of its id?


No. Python does not provide a way to do this, because
in general it's impossible -- the original object might
not exist any more.

You'll have to keep a reference to the object and
provide a way to get back to it. E.g. a dictionary
mapping id numbers to objects.

 > Now, I would like to give the user some way to save
 > preference

If you're trying to preserve an object from one

Python session to another, take a look at the
"pickle" module.


-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From bokr at oz.net  Wed Sep  4 15:15:42 2002
From: bokr at oz.net (Bengt Richter)
Date: 4 Sep 2002 19:15:42 GMT
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <un50ccotaak86b@corp.supernews.com> <al0e96$oms$0@216.39.172.122> <4pid9.5883$e5.1033514@newsb.telia.net>
Message-ID: <al5m4u$agv$0@216.39.172.122>

On Wed, 04 Sep 2002 07:14:08 GMT, "Fredrik Lundh" <fredrik at pythonware.com> wrote:

>Bengt Richter wrote:
>
>> I agree about new syntax, but I wouldn't mind having a re.help(regexp) function
>> for interactive use that would just explain in 'English' what the regexp expression
>> stands for.
>
>you can ask SRE to dump the internal parse tree
>to stdout:
>
>>>> sre.compile("[a-z]\d*", sre.DEBUG)
>in
>  range (97, 122)
>max_repeat 0 65535
>  in
>    category category_digit
>
>turning this into 'English' is left as an exercise etc.
>
Interesting, thanks. Does the above mean that sre can't fully match
 'a'+'9'*65537
?

Regards,
Bengt Richter


From pyth at devel.trillke.net  Sat Sep 28 16:46:45 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 28 Sep 2002 22:46:45 +0200
Subject: Best way to create a class dynamically on the fly?
In-Reply-To: <tFml9.13043$Lg2.3865279@news2.news.adelphia.net>; from Oschler@earthlink.net on Sat, Sep 28, 2002 at 06:37:13PM +0000
References: <tFml9.13043$Lg2.3865279@news2.news.adelphia.net>
Message-ID: <20020928224645.T30315@prim.han.de>

Robert Oschler wrote:
> What is the best way to create a Python class dynamically at runtime?  A
> sample or doc would be great.  I'm a veteran C++ programmer so this concept
> is very alien to me as you can imagine (dynamically created classes).  An
> example that would show the creation of a descendant (sub-class), and a
> another example showing the creation of an ancestor class would be perfect.

Chances are you don't need dynamically generated classes. python is
naturally dynamical. When i came from C++ to python i came up with
using 'exec' followed by a dynamically generated string defining the
function. 

Another way is to use the new - module, e.g.

>>> c = new.classobj('myclass', (),{})
>>> c
>>>
<class __main__.myclass at 0x82eb3b4>
>>> c()
<__main__.myclass instance at 0x82e8b2c>
>>> exec "def f(self): return 15"  # dynamically create the string
>>> setattr(c,'func',f)
>>> c().func()
15

If you describe the actual problem we might be able to give you 
a better solution than "compilation at runtime". 

regards,

    holger



From pedronis at bluewin.ch  Thu Sep  5 06:09:17 2002
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Thu, 5 Sep 2002 12:09:17 +0200
Subject: jython memory usage
References: <hVm5ydAY2hd9Eww$@jessikat.demon.co.uk><3d76366a$1_5@news.bluewin.ch> <rnLbuKAv5jd9Ewzz@jessikat.demon.co.uk><3d76494d$1_5@news.bluewin.ch> <YR13DVAUNld9EwXH@jessikat.demon.co.uk><3d765ce6_4@news.bluewin.ch> <zzKd7KAqhnd9EwxT@jessikat.fsnet.co.uk><3d767c4a$1_1@news.bluewin.ch> <YWyHyIAJKyd9EwHA@jessikat.demon.co.uk>
Message-ID: <3d772e4c_3@news.bluewin.ch>

Robin Becker <robin at jessikat.fsnet.co.uk> wrote in message
YWyHyIAJKyd9EwHA at jessikat.demon.co.uk...
> In article <3d767c4a$1_1 at news.bluewin.ch>, Samuele Pedroni
> <pedronis at bluewin.ch> writes
> >
> >Robin Becker <robin at jessikat.fsnet.co.uk> wrote in message
> >zzKd7KAqhnd9EwxT at jessikat.fsnet.co.uk...
> >> .....
> >> it, test_pdfbase_ttfonts.py doesn't look as though it's doing anything
> >> particularly odd, certainly I see no references to system/spawn in
> >> there.
> >> --
> >
> >have you tried my little program with your JVM.
> >
> >If my program work as for me, that means that your py plus jython trigger
a
> >bug of the JVM, because anyway and in any case  the JVM ought to scream
"out
> >of memory" well before eating a 1GB of memory.
> >
> >regards.
> >
> >
> OK here's the results with outofmem
>
> C:\jython-2.1\reportlab\test>javac outofmem.java
>
> C:\jython-2.1\reportlab\test>java outofmem
> 1
> 2
> 3
> 4
> 5
> 6
> 7
> 8
> 9
> 10
> 11
> Exception in thread "main" java.lang.OutOfMemoryError
>
>
> Is it possible I'm using a wrong java engine somewhere?

At this point I'm really puzzled.

To double check for the java engine you are using you can you can use
something like
this in the test code you are running (Jython code):

import sys

print sys.registry.getProperty('java.version')
print
sys.registry.getProperty('java.vm.vendor','???'),sys.registry.getProperty('j
ava.vm.version','???')












From spam at ob_scure.dk  Fri Sep  6 06:04:05 2002
From: spam at ob_scure.dk (Thomas Jensen)
Date: Fri, 06 Sep 2002 12:04:05 +0200
Subject: Using Python to call a DCOM server - "'Class not registered"
Message-ID: <3D787D95.9000907@ob_scure.dk>

Hi,

I'm sorry if this post is somewhat OT, please direct me to the correct 
group if it is.

I'm having some trouble using win32all to call a DCOM server.

win32all includes a file called testDCOM.py which contain the lines:
   clsctx = pythoncom.CLSCTX_SERVER & ~pythoncom.CLSCTX_INPROC_SERVER
   ob = win32com.client.DispatchEx("Python.Interpreter", serverName, 
clsctx=clsctx)

substituting serverName with "localhost" works fine and I get the 
extepcted "error":
   Error: The object created on server 'localhost' reported its name as 
'MJOELNER'

However when I substitute "Python.Interpreter" with my own component, I 
get a "Class not registered" error.

Code that works:
   batch = win32com.client.Dispatch("IGBatchNGBackend.IGBatch")

Code that doesn't work:
   clsctx = pythoncom.CLSCTX_SERVER & ~pythoncom.CLSCTX_INPROC_SERVER
   batch = win32com.client.DispatchEx("IGBatchNGBackend.IGBatch", 
'localhost', clsctx=clsctx)

The traceback is:

Traceback (most recent call last):
   File "test01.py", line 30, in ?
     test()
   File "test01.py", line 15, in test
     batch = win32com.client.DispatchEx("IGBatchNGBackend.IGBatch", 
'localhost',
clsctx=clsctx)
   File 
"C:\PROGRA~1\PYTHON22\lib\site-packages\win32com\client\__init__.py", lin
e 109, in DispatchEx
     dispatch = pythoncom.CoCreateInstanceEx(clsid, None, clsctx, 
serverInfo, (py
thoncom.IID_IDispatch,))[0]
pywintypes.com_error: (-2147221164, 'Class not registered', None, None)


Thanks in advance

-- 
Best Regards
Thomas Jensen
(remove underscore in email address to mail me)



From martin at v.loewis.de  Thu Sep 19 02:46:27 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 19 Sep 2002 08:46:27 +0200
Subject: Python as shared library on Unix
References: <mailman.1032384864.17580.python-list@python.org>
Message-ID: <m365x2o4sc.fsf@mira.informatik.hu-berlin.de>

"Rolf Kalbermatter" <r.kalbermatter at hccnet.nl> writes:

> There is probably little chance that this will be officially backported
> to the 2.2.1 branch of Python? 

No, there is no such chance.

Regards,
Martin


From max at alcyone.com  Sun Sep  1 17:43:11 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 01 Sep 2002 14:43:11 -0700
Subject: del(instance.attr) -- no attribute
References: <un512ii7hjgq22@corp.supernews.com>
Message-ID: <3D7289EF.CE162C2D@alcyone.com>

Lee Harr wrote:

> Is this a bug?
> 
> >>> class C:
> ...  a = 1
> ...
> >>> c = C()
> >>> c.a
> 1
> >>> del(c.a)
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> AttributeError: C instance has no attribute 'a'
> >>> hasattr(c, 'a')
> 1
> >>>
> 
> I can see that maybe del(c.a) should not be allowed to delete
> the attribute, but I think the error message is misleading.

Well, to its credit, it _does_ say that the _instance_ doesn't have the
attribute, which is precisely what the problem is.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ There is nothing so subject to the inconstancy of fortune as war.
\__/ Miguel de Cervantes
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From wayne at tbnets.com  Sun Sep  8 07:52:56 2002
From: wayne at tbnets.com (Wayne R)
Date: Sun, 8 Sep 2002 07:52:56 -0400
Subject: python program to watch a file - example.
In-Reply-To: <3d7b3437$0$12727@echo-01.iinet.net.au>
Message-ID: <004701c2572e$5738af40$02010a0a@hackncodexp>

Sorry for the html,  I forgot to turn it off (as I usually turn it off
as soon as I install a new system, forgot this time).

The platform is Gentoo Linux,  I am writing a piece of code for the
Twiggi project to patch in adduser support.


|-> > >I am trying to write a python program that will run as a deamon
and
|-> > >watch a file for data to arrive and then wake up and perform
duties
|-> on
|-> > >that data.   I found ways to make a python program run as a
deamon
|-> but I
|-> > >am at a loss to find anything related to watching a file for data
so
|-> > >that it can then process the data.  I didn't really want to do a
|-> sleep
|-> > >loop or something like, I am sure there is a better way.
|-> > >
|-> > >Any ideas or links to more info?
|-> > >
|-> > Don't annoy people you want help from, for starters.
|-> > What system are you targeting? (Outlook in the headers says you're
on
|-> windows
|-> > generating this abominable mix of actual message and useless
cruft).
|-> >
[Wayne R] 





From duncan at NOSPAMrcp.co.uk  Mon Sep 16 06:22:38 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Mon, 16 Sep 2002 10:22:38 +0000 (UTC)
Subject: representing a literal single slash
References: <dd5454fb.0209141828.1ad24260@posting.google.com> <3D84E9F9.8090101@linux.ie> <3D84FEE6.90600@linux.ie> <sh8h9.52633$AR1.2065934@bin2.nnrp.aus1.giganews.com> <xZhh9.4239$cP3.10322@news.iol.ie>
Message-ID: <Xns928B72069FBD8duncanrcpcouk@127.0.0.1>

Padraig Brady <Padraig at Linux.ie> wrote in 
news:xZhh9.4239$cP3.10322 at news.iol.ie:

>> Specifically, a raw string cannot end in a single backslash (since 
> > the backslash would escape the following quote character)
> 
> But there is no escaping when the prefix r is used?
> I guess there is a good reason for this inconsitency?

There is escaping with the prefix r: backslash followed by a quote escapes 
the quote into the string, but leaves the backslash in the string.

Whether this is a good reason depends on your viewpoint. Whether there is 
good reason for raw strings is also arguable: raw strings do not permit you 
to do anything that you could not do with non-raw strings, they violate the 
'one way to do it' rule, and they are confusing.

They do make it easier to write regular expressions, and a regular 
expression couldn't end in a single backslash anyway. If you only ever use 
raw strings for regular expressions then you are unlikely to have 
problems[1].  They are often misused to write DOS pathnames, which may end 
in a backslash.

[1] Unlikely to have problems with the raw strings that is. Problems with 
regular expressions are likely.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From wganz at texoma.net  Sat Sep  7 10:00:28 2002
From: wganz at texoma.net (Will Ganz)
Date: Sat, 7 Sep 2002 09:00:28 -0500
Subject: (slightly OT): Python and linux - very cool
In-Reply-To: <o58uia-q8q.ln@drebbelstraat20.dyndns.org>
Message-ID: <012b01c25676$ebe9e860$0202a8c0@WorkGroup>

> them first. That does serve to make badly coded Perl look *absolutely*
> awful, but it is also a great benefit.
>

And you throw into the mix the fact that the 'Perl poet' has zero experience
in design, you will have a mess that no one can figure out. Couple that with
the desire to reach Perl Wizard status by playing Perl golf to cram as much
as possible into one line of code as if any optimizing compiler wouldn't do
that anyway. Jeepers, I don't have enough digital ink in my cursor to write
all the horror stories that I could tell about that. The company that I work
for has Perl modules that are 5 years old that no one dares to touch. They
are stuck using Perl 5.04 since that is the latest version that it will run.

That is why I rebeled and use Python for my projects. No more unmaintainable
code. I still use Perl, but only when it will fit in under 100 lines of
code.

Just my view from the bottom of the fishbowl.




Will

 "We never gave up, and we never will.
In the end, if that's all they can say about us, it'll do."




From sschwarzer at sschwarzer.net  Sun Sep  1 17:32:13 2002
From: sschwarzer at sschwarzer.net (Stefan Schwarzer)
Date: Sun, 01 Sep 2002 23:32:13 +0200
Subject: Add methods to a class at runtime?
References: <9pqc9.30220$vY2.574504@e3500-atl2.usenetserver.com> <slrnan4dl3.18h.gerhard.haering@lilith.my-fqdn.de> <YYrc9.31010$vY2.601601@e3500-atl2.usenetserver.com>
Message-ID: <3D72875D.9040308@sschwarzer.net>

Hello Robert

Robert Oschler wrote:
> Thanks but I need a little bit more detail.  Suppose I want to do it from
> within code?  For example, I create a function called
> AddBarToFoo(FooObject), that given a object of type Foo (FooObject) adds the
> Bar method to it dynamically, what would be the syntax then?  Would I have
> to put the method declaration code, which you show being typed in to the
> interpreter, in a string and then eval/exec it?  Is there a more convenient
> way than that?

There was a thread on this some time ago:

http://groups.google.de/groups?hl=de&lr=&ie=UTF-8&oe=UTF-8&th=9a6dcf6380b7afe0&seekm=Xns91D662196C970duncanrcpcouk%40127.0.0.1&frame=off

Maybe that is helpful. :)

Stefan



From fredrik at pythonware.com  Thu Sep  5 02:41:14 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 05 Sep 2002 06:41:14 GMT
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <un50ccotaak86b@corp.supernews.com> <al0e96$oms$0@216.39.172.122> <4pid9.5883$e5.1033514@newsb.telia.net> <al5m4u$agv$0@216.39.172.122>
Message-ID: <e0Dd9.6122$e5.1077418@newsb.telia.net>

Bengt Richter wrote:

> >you can ask SRE to dump the internal parse tree
> >to stdout:
> >
> >>>> sre.compile("[a-z]\d*", sre.DEBUG)
> >in
> >  range (97, 122)
> >max_repeat 0 65535
> >  in
> >    category category_digit
> >
> >turning this into 'English' is left as an exercise etc.
>
> Interesting, thanks. Does the above mean that sre can't fully match
>  'a'+'9'*65537
> ?

in this context, 65535 represents any number:

>>> import re
>>> p = re.compile("[a-z]\d*")
>>> s = "a"+"9"*65537
>>> len(s)
65538
>>> m = p.match(s)
>>> len(m.group(0))
65538

</F>




From mhammond at skippinet.com.au  Tue Sep 17 23:00:03 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 18 Sep 2002 03:00:03 GMT
Subject: COM object registration
References: <afph9.9468$yt3.4669302@newssrv26.news.prodigy.com>
Message-ID: <T_Rh9.17227$Ee4.51953@news-server.bigpond.net.au>

dsavitsk wrote:
> if the module below is run from the command line, it registers and works
> fine.  if, however, it is imported and the register() method is executed it
> does not work (giving a 'localserver' error). 

What doesn't work?  Registration or subsequent use of the object?  Can 
you be more specific about the error you got, or else attach code I can 
run without modification to see the error.

Thanks,

Mark.



From creyes at ea.com  Fri Sep 13 12:44:22 2002
From: creyes at ea.com (Christian Reyes)
Date: Fri, 13 Sep 2002 09:44:22 -0700
Subject: Problem with HTTPS request
References: <uo2fveepf88175@corp.supernews.com> <b0f083db.0209130558.2aa60398@posting.google.com>
Message-ID: <uo45ara0l62cdc@corp.supernews.com>

I'm on Win 98.
i don't have a development env in which to build the python socket module.
was hopoing someone could point me to somewhere i could find a premade one.

Thanks though,
if you could point me to a site or two, that'd be great'

Jeremy Hylton <jeremy at alum.mit.edu> wrote in message
news:b0f083db.0209130558.2aa60398 at posting.google.com...
> "Christian Reyes" <creyes at ea.com> wrote in message
news:<uo2fveepf88175 at corp.supernews.com>...
> > I got the following response when trying to make an HTTPS request.
> >
> > <h1>SSL required to access this page</H1>
> >
> > Then I read in the documentation the following
> >
> > Note: HTTPS support is only available if the socket module was compiled
with
> > SSL support.
> >
> > Does anyone know where i can get my hands on a socket module that was
> > compiles with SSL support?
>
> What platform are you running?  Many people compile such a socket
> module themselves.  If that isn't an option, you'll have to look at
> binary packages for your platform.
>
> Jeremy




From tjreedy at udel.edu  Tue Sep 10 16:40:46 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 10 Sep 2002 20:40:46 GMT
Subject: Reset a generator function?
References: <%Lrf9.30107$w51.7961593@news2.news.adelphia.net>
Message-ID: <iNsf9.303942$On.12400590@bin3.nnrp.aus1.giganews.com>

"Robert Oschler" <Oschler at earthlink.net> wrote in message
news:%Lrf9.30107$w51.7961593 at news2.news.adelphia.net...
> How can I reset a generator function so that it begins yielding
values as it
> did when first called?

The generator function does not really yield values.  It returns an
iterator that yields values.  This is handled transparently by the for
loop mechanism.  The iterator cannot be reset.  The generator can be
called again, with either the same or different arguments, to return a
new iterator that will yield the same or different sequence.  It can
even be called again before you use the first iterator to get any
values.

Terry J. Reedy





From huaiyu at gauss.almadan.ibm.com  Wed Sep  4 14:16:07 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Wed, 4 Sep 2002 18:16:07 +0000 (UTC)
Subject: default tuple unpacking?
References: <slrnanaknb.1gh.huaiyu@gauss.almadan.ibm.com> <3d758180$1@news.sentex.net> <3d7584d2$1@news.sentex.net>
Message-ID: <slrnancjf7.3cl.huaiyu@gauss.almadan.ibm.com>

Peter Hansen <peter at engcorp.com> wrote:
>Peter Hansen wrote:
>> Huaiyu Zhu wrote:
>> 
>>> It is often very convenient to have default arguments such as
>>>
>>>     def func(a, b, *c): ...
>>>
>>> Is there a way to use this in statements too?  Examples:
>>>
>>>     a, (b, *c), *d = e
>
>Hmm... I think there might be an inconsistency or an unexpected
>result in this.  What if e already contains a tuple?

Bengt Richter and Bernhard Herzog have already given two different answers
to this question.  Both are consistent in themselves.  I feel that Bengt's
answer is more elegant, but it runs againt the Python syntax rule that
tuples are made from commas instead of parentheses.

Bernhard's answer can easily fit in the current syntax. I'll expand it in
more detail to make sure there is no ambiguity.  Consider

    def func0(*a):          return *a
    def func1(a, *b):       return a, *b
    def func2(a, b, *c):    return a, b, *c

For func in [func0, func1, func2], the following are equivelant

    x, y, *z = func(a, b, *c)
    x, y, *z = a, b, *c

For func in [func0, func1], the following are equivelant

    x, *y = func(a, *b)
    x, *y = a, *b

For func in [func0, func1], the following are equivelant

    x, = func(a)
    x, = a,

The following are always equivelant

    *x = func0(*a)
    *x = *a
    x = tuple(a)

Let's see if this works for all your examples:

>
>   e = (1, 2, 3)
>   *a = e        # a is ((1, 2, 3), )

This is equivalent to

    a = e  # a = 1,2,3

This equivalence may look disturbing, but I can't think of any trouble it
might cause.  Following are more equivelences of this sort

    a = *e
    *a = *e
    a = 1, 2, 3
    *a = 1, 2, 3
    a = *(1, 2, 3)
    *a = *(1, 2, 3) 

>
>That seems sensible, and it's just what happens if you pass
>"e" in to a function defined as "def func(*a)".

This is equiv to

    *a = e,   # a = (1,2,3),

>
>Now what about this:
>
>   a, *b = e     # a is 1, b is (2, 3)
>

Yes.

>That's one possible result, and presumably the one you want with
>this syntax, but it's not what you'd get with the function syntax:
>
>   def func(a, *b):
>     print a
>     print b
>
>   func(e)

That's

    a, *b = e,   # a = 1, 2, 3; b=()

>
>Here of course you get (1, 2, 3) in "a" and () in b.  To get the
>same as the proposed syntax above, you can't pass in a tuple, but
>have to pass the parameters separately:
>
>   func(1, 2, 3)  # now a is 1, b is (2, 3)

That's

    a, *b = 1, 2, 3  # a = 1; b = 2,3

>
>The problem is that in function calling, you can pass in a single
>tuple, which is one argument, or you can pass the elements separately
>as multiple arguments, which is when the special * syntax kicks in.

Always think of function arguments as a tuple. If the argument is a single
tuple, then the args is simply a single-element tuple whole element happens
to be a tuple.

Huaiyu


From hellprout at yahoo.com  Fri Sep 20 18:12:32 2002
From: hellprout at yahoo.com (hellprout)
Date: 20 Sep 2002 15:12:32 -0700
Subject: pyrhon and oracle
Message-ID: <5b8834c2.0209201412.50e77861@posting.google.com>

hi , i just want to have some detail about python and oracle 
what ca i use to query oracle with python ?

i jsut want a little exemple 

thanks


From lbrannma at cablespeed.com  Sun Sep 22 20:57:17 2002
From: lbrannma at cablespeed.com (Lance)
Date: Sun, 22 Sep 2002 17:57:17 -0700
Subject: more on weak typing..
Message-ID: <uospj56t55o858@corp.supernews.com>

I found this on the Dive Into Python site:

  statically typed language
  A language in which types are fixed at compile time. Most statically typed
languages enforce this by requiring you to declare all variables with their
datatypes before using them. Java and C are statically typed languages.
  dynamically typed language
  A language in which types are discovered at execution time; the opposite
of statically typed. VBScript and Python are dynamically typed, because they
figure out what type a variable is when you first assign it a value.
  strongly typed language
  A language in which types are always enforced. Java and Python are
strongly typed. If you have an integer, you can't treat it like a string
without explicitly converting it (more on how to do this later in this
chapter).
  weakly typed language
  A language in which types may be ignored; the opposite of strongly typed.
VBScript is weakly typed. In VBScript, you can concatenate the string '12'
and the integer 3 to get the string '123', then treat that as the integer
123, all without any explicit conversion.
So Python is both dynamically typed (because it doesn't use explicit
datatype declarations) and strongly typed (because once a variable has a
datatype, it actually matters).




From barry at python.org  Fri Sep 13 11:48:40 2002
From: barry at python.org (Barry A. Warsaw)
Date: Fri, 13 Sep 2002 11:48:40 -0400
Subject: RELEASED email package version 2.3
References: <15742.47445.811488.515719@anthem.wooz.org>
Message-ID: <15746.2264.790325.57188@anthem.wooz.org>

>>>>> "BAW" == Barry A Warsaw <barry at zope.com> writes:

    BAW> I've just released the (standalone) email package version
    BAW> 2.3.  This is identical to the version that is in the current
    BAW> cvs development tree for Python 2.3.  It is made available as
    BAW> a distutils package for installing in earlier versions of
    BAW> Python.  It is compatible with Python 2.1.3 and Python 2.2.1.

    BAW> email-2.3.tar.gz is available from the mimelib project on
    BAW> SourceForge:

    BAW>     http://sf.net/projects/mimelib/

Sorry folks, the email-2.3.tar.gz package was missing one of the test
suite files.  You'll notice this if you ran "python testall.py" and
saw failures due to the missing file msg_31.txt.

I've just released email-2.3.1.tar.gz which contains the missing
file.  There are no other substantive changes in this micro release so
if you don't care about the test suite, there's no need to update.

-Barry



From whisper at oz.net  Thu Sep 12 05:31:44 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 12 Sep 2002 02:31:44 -0700
Subject: [FEEDBACK] Is this script efficient...is there a better way?
In-Reply-To: <20020912074400.GA22578@hishome.net>
Message-ID: <GCEDKONBLEFPPADDJCOECEPFEOAA.whisper@oz.net>

<snip>
> Have you actually tested this? I used to believe that larger buffers are
> always better for performance.
>
> Wrong.
>
> I ran some test on Linux for the effect of buffer size on file reading
> speed and the results were very interesting. I started with a buffer size
> of 32 bytes, tested file I/O throughput and increased it
> logarithmically by
> about 2% for each step.  As expected, the time to read 1mb improved as the
> buffer size increased until it hit a minimum for a buffer size of around
> 4-8k (the graph is very noisy so it's hard to tell) and then rose back up
> to a value that is 10-20% worse for buffer sizes of 32-64k and remained
> more-or-less constant for anything higher.
>
> This performance curve may be a result of the CPU cache or the OS
> architecture.
>
> The chunk size used by xreadlines is 8k which is about the optimum
> value (at least for linux).
>
> 	Oren

Disk cluster size? Fragmentation?

Dave LeBlanc
Seattle, WA USA




From apederse at hotmail.com  Fri Sep 27 04:09:48 2002
From: apederse at hotmail.com (Asle)
Date: 27 Sep 2002 01:09:48 -0700
Subject: imap handler for urllib
Message-ID: <e4d8f624.0209270009.290310b0@posting.google.com>

Has anyone out there made an imap handler for urllib??
There is an IMAP url scheme described in RFC 2192.

regards,
Asle


From gregory.p.green at boeing.com  Fri Sep 20 12:57:07 2002
From: gregory.p.green at boeing.com (Greg Green)
Date: Fri, 20 Sep 2002 16:57:07 GMT
Subject: Need advice on choosing the technology for intranet based application
References: <3d88fd1e@news.ivm.net> <uokpn32h8dhqfa@corp.supernews.com>
Message-ID: <ofaslhuk.fsf@boeing.com>

"A.M. Kuchling" <amk at amk.ca> writes:

> In article <3d88fd1e at news.ivm.net>, Nenad Propadovic wrote:
> > -I assume the simplest way to present a graph/chart kind of graphic is to
> > create a jpg/giff/tiff on the server side, and just show it on the client?
> 
> Yes.  PIL, the Python Imaging Library, would be one possibility; so is
> the gd module.  For example, the pricing charts on the pages at
> http://www.mems-exchange.org/catalog/ are generated from Quixote using
> gd.  PDF likely isn't helpful, because browsers won't display PDF
> images inlined in a web page, though it's certainly well-suited for 
> multi-page reports that need to be printed.

Another possibility, which I have used, is gnuplot. This has an option
of creating image files as output instead of the dedicated
window. Works very well.

-- 

Greg Green


From jepler at unpythonic.net  Mon Sep 16 21:01:30 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Mon, 16 Sep 2002 20:01:30 -0500
Subject: pickle magically not working
In-Reply-To: <3d865dac@nntp.server.uni-frankfurt.de>
References: <3d865dac@nntp.server.uni-frankfurt.de>
Message-ID: <20020916200130.A2484@unpythonic.net>

On Tue, Sep 17, 2002 at 12:39:40AM +0200, Michael Janssen wrote:
> Hello
> 
> at least i've solved it (as a workaround), but it's pretty annoying. 
> Maybe someone has an answere :-)

It's a bug, apparently.  It's fixed in Python 2.3's CVS versions.

The tuple returned from time.localtime() is newly a 'time.struct_time'
(since version 2.2), not actually a tuple.  pickle can't handle this.
So much for the addition of named fields to tuples being "zero impact"!

If I had to guess, I'd say the "fix" was in Pickle itself, you might turn
something up by investigating Python's CVS tree.  Or somebody else might be
able to immediately indicate the relevant code...

Until then, slicing the value will work -- slicing returns a pure tuple.

Jeff



From noreply at sourceforge.net  Wed Sep  4 21:31:27 2002
From: noreply at sourceforge.net (Bruce Dodson)
Date: Wed, 4 Sep 2002 22:31:27 -0300
Subject: Jython and .NET?
Message-ID: <Qtyd9.57629$C8.155868@nnrp1.uunet.ca>

I'm thinking about Python.NET.  Then I'm thinking about Jython, and also
about the J#.Net compiler that arrived at work recently.

Can someone from the Jython camp comment whether it would be feasible to
port Jython to J#?  That is, how closely tied is Jython to the Java virtual
machine and class format?  (In interpreted mode, vs. in compiled mode.)

Thanks,
bruce




From 2002 at weholt.org  Mon Sep 23 05:09:52 2002
From: 2002 at weholt.org (Thomas Weholt)
Date: Mon, 23 Sep 2002 09:09:52 GMT
Subject: UpToDate SOAP-module ??
Message-ID: <ATAj9.21315$0p1.376966@news2.ulv.nextra.no>

I've looked around for a SOAP-module to use in a WebService-based project,
but the SOAP-implementations in python seem to be a bit ... unstable ?? Some
of them haven't been updated for a while etc. so I wondered what people
think is the most up to date and "best" module for providing SOAP-based
WebServices which includes client-side code too.

Any opinions, hints or ideas are welcome. Thoughts on ease of use, list of
features, comparisons etc. are also interesting.

Best regards,
Thomas




From aleax at aleax.it  Thu Sep 26 09:36:16 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 26 Sep 2002 13:36:16 GMT
Subject: Is there a way to find the name of the method currently being executed?
References: <mailman.1033009413.19897.python-list@python.org>
Message-ID: <k3Ek9.170802$pX1.6081475@news2.tin.it>

Craeg K Strong wrote:

> Hello:
> 
> class a:
> 
>     def func1(self):
>         print "func1 being called"
> 
>     def func2(self):
>         print "func2 being called"
> 
> In the code sample above, I would like to replace the "func1"
> and "func2" strings with constants so the same code could
> be copied verbatim into multiple functions and do the right
> thing.
> 
> Is there a way to do this in Python?  For example, maybe you could
> get your hands on the call stack and print out stack[0].__name__
> or something....

Right, that's exactly what I did in:

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66062

except that sys._getframe (gets one frame of the call stack,
not all of the stack at once) is what the recipe uses (as
usual, the version in the printed Cookbook, published by
O'Reilly, has better discussion, &c).


Alex



From brueckd at tbye.com  Mon Sep  9 22:45:38 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Mon, 9 Sep 2002 19:45:38 -0700 (PDT)
Subject: IDLE
In-Reply-To: <wpbf9.264320$aA.45691@sccrnsc02>
Message-ID: <Pine.LNX.4.44.0209091944440.12226-100000@ns0.tbye.com>

On Tue, 10 Sep 2002, Ali K wrote:

> Why won't my IDLE work?

I just tried it on your machine and it works fine.

Thanks,
Dave




From max at alcyone.com  Tue Sep 17 17:21:37 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 17 Sep 2002 14:21:37 -0700
Subject: subclassing tuple
References: <H2LMGE.1JBF@beaver.cs.washington.edu>
Message-ID: <3D879CE1.EDB15467@alcyone.com>

Douglas Zongker wrote:

> Is there any way to create a subclass of 'tuple' that has a customized
> constructor?

You want to use the .__new__ method:

	http://www.python.org/2.2/descrintro.html#__new__

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ It's like being on a diving board that you know is too high.
\__/ Sade Adu
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.


From sismex01 at hebmex.com  Fri Sep  6 10:08:57 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Fri, 6 Sep 2002 09:08:57 -0500 
Subject: Difference between static method and a function as a class at
	tribute?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F0623302318687@mail.hebmex.com>

> 
> Hi,
> 
> What are the differences, if any, between 1 and 2?
> 
> 1.
> def funtest(*args):
>     pass
> 
> class test(object):
>     method = funtest
> 
> 2.
> class test(object):
>     def funtest(*args):
>         pass
> 
>     method = staticmethod(funtest)
> 
> 
> And if there are any differences do they really matter, that is, does
> anybody has any test cases where 1 and 2 are not interchangeable?
> 
> TIA and all the best,
> Gon?alo Rodrigues

Take a look at this:

>>> class testclass:
... 	pass
... 
>>> 
>>> def functest(*args):
... 	print args
... 	
>>> 
>>> testclass.testmethod = functest
>>> 
>>> inst = testclass()
>>> dir(inst)
['__doc__', '__module__', 'testmethod']
>>> 
>>> inst.testmethod()
(<__main__.testclass instance at 0x013A5020>,)
>>> 

Any "plain" executable found as an attribute of a class
object, upon creating an instance of that class, is bound
to the instance as a method.  If not, then calling
"inst.testmethod()" should display an empty tuple.

I suspect that staticmethod(func) and classmethod(func)
wrap the function in a proxy which first extracts some
info from the actuall call (say, the instance's class)
and pass *whose* arguments to the real function.

Neat. A tribute to Python's true power, that you
can do that and not pollute the language with
wierd "special cases" and such.

-gus
--

Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From DennisR at dair.com  Thu Sep 26 18:40:14 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Thu, 26 Sep 2002 22:40:14 GMT
Subject: distributing apps w/embedded python & missing modules
References: <amvihs$252l$1@yuggoth.ucsb.edu>
Message-ID: <i1Mk9.3231$1i1.86041872@newssvr21.news.prodigy.com>

> My question: Is there an easy way
> to compile all of Python's modules into the app so that the user does not
> need to separately download the core python package?

At http://www.autorunning.com/arpyexe.cgi I have pre-compiled 24 modules
plus the core executables in a 432 K Windows package.  The user can specify
3 more modules.  Now I could have included all 167 of the lib modules but
the total library size is 1.92 Mb (pre-compression).

So, I think the answer is "yes, it is easy".  It is not clear to me where
you have gotten so far and what is roadblock you are facing.

Let me describe what I do.  Perhaps this will help.  I compress the files
described on my web page above into a self-extracting executable.  This
executable is then embedded in non-Python code and extracted into a temp
file at run time.  Control is now turned over to Python.  hth.

--

Dennis Reinhardt

http://www.dair.com




From tjreedy at udel.edu  Sat Sep  7 18:05:49 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 07 Sep 2002 22:05:49 GMT
Subject: None or 0
References: <mailman.1031418732.10826.python-list@python.org>
Message-ID: <0Lue9.503203$2p2.20193233@bin4.nnrp.aus1.giganews.com>

To summarize Tim's thorough answer:
the language reference guarantees (specifies) the short-circuit
behavior
that you discovered.  It also guarantees that
> a = a or b
is *not* equivalent to
> if a is None:   a = b
unless *you* can guarantee that 'a' will never have a null value other
than None.

I add: as far as I know, both have been true since the beginning and
will remain true thru the 2.x series.  I would be extremely surprised
if anything changed thereafter.

Terry J. Reedy





From bokr at oz.net  Mon Sep  2 00:58:46 2002
From: bokr at oz.net (Bengt Richter)
Date: 2 Sep 2002 04:58:46 GMT
Subject: stylistic question -- optional return value
References: <20020828145828.19131.h014.c001.wm@mail.dougfort.net.criticalpath.net> <mailman.1030572688.22121.python-list@python.org> <akkost$cgt$1@newsreader2.netcologne.de> <roy-C1B7F3.09024729082002@reader1.panix.com> <yu99vg5t946h.fsf@europa.research.att.com>
Message-ID: <akur66$rp6$0@216.39.172.122>

On Thu, 29 Aug 2002 13:37:26 GMT, Andrew Koenig <ark at research.att.com> wrote:

>Roy> It's also virtually always true that the real way to speed up a
>Roy> program is not code tweaking, but using better algorithms.
>
>Like most generalizations, this one is not always true.
>
>Remember, we're talking here about an interface decision, which
>means that changing it after the fact is difficult.  Moreover,
>this particular program is going to involve a large number of recursive
>function calls, every one of which will use this mechanism to pass
>information back to its caller.  In this case, therefore, the choice
>of interface will have a substantial effect on the execution speed
>of the whole program, and it will be difficult to change it later.
>
>In such circumstances, a factor of three is too much to ignore.  I am
>not saying that the faster interface is always the right one, but I am
>saying that when there's a large speed difference, and the decision
>will be difficult to change later, then there had better be a
>substantial argument in favor of the slower alternative.  Which
>I don't think there is in this case.
>
If added information is the exception (;-) rather than the rule, you might
want to consider using an exception to return the unusual bag of data.
try: is supposed to be low overhead, and the payoff is a clean main path
of excution, using the retval as-is normally, without having to test or
index into a tuple etc., and without having to pack a tuple for a normal
return value from the routine. You can define your own exception and raise it
with an arbitrary set of args, e.g.,

 >>> class MyExceptionalData(Exception):
 ...     pass
 ...
 >>> try:
 ...     raise MyExceptionalData('x value', 'optional y')
 ... except MyExceptionalData, mxd:
 ...     print mxd
 ...
 ('x value', 'optional y')
 >>> mxd
 <__main__.MyExceptionalData instance at 0x007A2460>
 >>> dir(mxd)
 ['__doc__', '__getitem__', '__init__', '__module__', '__str__', 'args']
 >>> mxd.args
 ('x value', 'optional y')
 >>> mxd[0]
 'x value'
 >>> mxd[1]
 'optional y'

You could also give your exception class an __init__ function to set
instance attributes and whatever else you'd like.

Anyway, your code might look something like

    try:                           # try overhead only
        x = foo()                  # use x knowing single return value assumption is ok
    except MyExceptionalData, e:
        x, y = e.args              # use x and y according to the exceptional situation

or alternatively

    try:                           # try overhead only
        x = foo()                  # use x knowing single return value assumption is ok
    except MyExceptionalData, e:
        x, y = e.args
        # use y here according to the exceptional situation
    # use x here either way
 
as opposed to something like

     x = foo()
     if x[1] is not sentinel:    # big index + test overhead in normal path
         x = x[0]                # indexing overhead in normal path
         # use single value
     else:
         x, y = x                # same overhead as with e.args less .args attribute lookup
         # use the unusual combo

Regards,
Bengt Richter


From bokr at oz.net  Mon Sep  9 21:40:08 2002
From: bokr at oz.net (Bengt Richter)
Date: 10 Sep 2002 01:40:08 GMT
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <un50ccotaak86b@corp.supernews.com> <al0e96$oms$0@216.39.172.122> <4pid9.5883$e5.1033514@newsb.telia.net> <al5m4u$agv$0@216.39.172.122> <e0Dd9.6122$e5.1077418@newsb.telia.net> <8vvdnucccqp2och24gigcmmc91njbjkhhk@4ax.com>
Message-ID: <aljiho$595$0@216.39.172.122>

On Thu, 05 Sep 2002 06:56:24 GMT, Ben Wolfson <wolfson at midway.uchicago.edu> wrote:

>On Thu, 05 Sep 2002 06:41:14 GMT, "Fredrik Lundh" <fredrik at pythonware.com>
>wrote:
>
>>Bengt Richter wrote:
>>
>>> >you can ask SRE to dump the internal parse tree
>>> >to stdout:
>>> >
>>> >>>> sre.compile("[a-z]\d*", sre.DEBUG)
>>> >in
>>> >  range (97, 122)
>>> >max_repeat 0 65535
>>> >  in
>>> >    category category_digit
>>> >
>>> >turning this into 'English' is left as an exercise etc.
>>>
>>> Interesting, thanks. Does the above mean that sre can't fully match
>>>  'a'+'9'*65537
>>> ?
>>
>>in this context, 65535 represents any number:
>
>Doesn't that cause problems for something like this?
>
>>>> m=re.compile(r'\d{0,65535}a').match(('9'*1000000)+'a')
>>>> len(m.group(0))
>1000001
>
Looks like a bug to me if {0,65535} acts like {0,} 

BTW, a search for \d{0,65534} seems to mean it, and compiles
so slowly that I lost patience waiting. Not very optimized, I guess.

 >>> import re
 >>> m=re.compile(r'\d{0,65535}a').search(('9'*1000000)+'a')
 >>> len(m.group(0))
 1000001

That went reasonably in time(though it's wrong), but this snoozed.
It must be brute forcing something.

 >>> m=re.compile(r'\d{0,65534}a').search(('9'*1000000)+'a')
 ^C
 [18:50] C:\pywk\junk>

Regards,
Bengt Richter


From gerhard.haering at gmx.de  Fri Sep 13 16:04:02 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 13 Sep 2002 20:04:02 GMT
Subject: Problem with HTTPS request
References: <uo2fveepf88175@corp.supernews.com> <b0f083db.0209130558.2aa60398@posting.google.com> <uo45ara0l62cdc@corp.supernews.com> <slrnao4dkn.288.gerhard.haering@lilith.my-fqdn.de> <mYqg9.8349$WV.1529@fe03>
Message-ID: <slrnao4h5c.288.gerhard.haering@lilith.my-fqdn.de>

Steve Holden wrote in comp.lang.python:
> "Gerhard H?ring" <gerhard.haering at gmx.de> wrote ...
> [complaints about crappy looking posts because of people using OE]
> Gerhard:
> 
> You're showing distinct signs of having too much time on your hands :-)

Unlikely.

> Can't you fix your brain instead so it will let you ignore other
> people's mailer problems?

Problems don't go away by ignoring them. People are so used to the
postings in crappy form (I'm only talking about form, not content)
offloaded by people like you that they need to be alerted to this from
time to time.

> an-unreconstructed-OE-user-myself-ly y'rs  - steve

I wonder why some programmers who insist on proper form of source code
don't care at all about the email/news messages they send to the
public, even though they know better.

And sorry, the Microsoft way of defining crap as standard won't work
with me.

-- Gerhard


From ianb at colorstudy.com  Tue Sep 17 15:18:34 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 17 Sep 2002 14:18:34 -0500
Subject: CGI question
In-Reply-To: <slrnaoerk7.mgo.jboy@otaku.freeshell.org>
References: <slrnaoerk7.mgo.jboy@otaku.freeshell.org>
Message-ID: <1032290315.8037.145.camel@dsl-65-184-205-5.telocity.com>

On Tue, 2002-09-17 at 13:04, John D. Boy wrote:
> Now, my problem is to find out what follows behind the question mark in
> my CGI script. Unless the key has a value
> (http://my.domain.name/?key=value), it is not in the cgi.FieldStorage
> dictionary. Is it simply impossible, or am I not using the right
> function calls? 

Look at cgi.FieldStorage.  Something like:

f = cgi.FieldStorage(keep_blank_values=1)

Will save these values.  (I don't know why it isn't set by default... by
default you shouldn't throw information away, but FieldStorage does in
this case)

  Ian





From jwpennin at bellsouth.net  Wed Sep 25 23:39:36 2002
From: jwpennin at bellsouth.net (Jonathan Pennington)
Date: Wed, 25 Sep 2002 23:39:36 -0400
Subject: Creating daily/monthly averages from datafiles
Message-ID: <20020926033935.GA2018@fiddlersnest.org>

Hello all, I've got a rather complex description/request. I'm building
a program to parse and analyze data for a watershed hydrological
database and am looking for some methods to steal so I don't re-code
what's already been done.

##################
## Description

I have some classes that iterate over text files with formats
similiar to this:

date,time,float,float,int,float,int,int # values every 15 minutes for
					# entire year, multiple years.

The data often has arbitrarily length header, and non-standard
delimiters. I've build a wxPython GUI to allow the user to delete
header lines, then a class that builds a table with a user selected
delimiter used to separate the lines into a wxGrid interface. All
pretty nice. 

Now separate from this, I have a class that will iterate over a
delimited text file and create a dictionary with keys of month number,
each of which is a dictionary with keys as day, each of which is a
list of readings from that day. Pretty basic. I then hacked up methods
(rough, not elegant at all- and not very efficient) that will create a
dictionary of values such that I can parse a datafile by the following:

>>> parser = DataParser(filename)
<...long wait as everything is done in __init__() method automagically...>
>>> parser.DailyAverages[9][16]
[43, 29, 84]
>>> parser.MonthlyAverages[4]
[35, 20, 76]

These are fields that are static to this file (ie. the class is not
very flexible as it was created to fit a specific file layout). The
actual parsing methods are also very rough, iterating over an entire
file to find all instances in a specific month, that sort of
thing. Again, not very elegant.

#################
## Request

Basically, I don't want to re-hack these average/cumtotal methods to
fit working with my wxGrid-ed data. I'm looking to see if anyone has
done any similiar work so that I don't have to reinvent the
wheel. What I need are classes/methods that can take data from a
list/grid/table and manipulate sections of that data by date/time. For
instance, take all values in column 2 and perform daily averages, take
all values in column 3 and perform monthly totals, take all values in
column 4 and perform cumulative totals by average.

Now, of course, the individual methods (ave, cumtotal, etc.) are
easy. What I'd like to find is code that someone has to easily seek
out all of the data by hour/day/month and return it to my manipulation
method, etc.

Of course, just knowing that the data is in a wxGrid is not
necessarily very helpful, the underlying data structure is what would
probably have to be manipulated. Is the wxGrid built from a
dictionary, etc. Right now, I've just dumped the values into a
subclassed wxPyGridTableBase class from the datafile, so there's no
underlying dictionary or other data structure. Ideally, I'd manipulate
the wxGrid directly, but don't know if that's
possible/economical. I'm very open to suggestions.

Does anyone know of some appropriate code, or has anyone built
similiar methods for time-constrained data mining? I've searched the
vaults and assorted other repositories, but nothing- on initial
perusal- seems too promising. Just a couple of methods would do it
really. Otherwise, I'll probably have to build a date-constrained
mining class and perhaps dump that onto the Vaults for others to use
for similiar tasks.

Thanks all,
-J



From bwucew at yahoo.com  Tue Sep 17 02:46:20 2002
From: bwucew at yahoo.com (Bwuce_Wee)
Date: 16 Sep 2002 23:46:20 -0700
Subject: Extending a C++ App
References: <996eade8.0209120912.3a653d7@posting.google.com> <n0qnhvfn.fsf@boeing.com> <996eade8.0209122352.45721938@posting.google.com> <altqiu$fjr$1@bob.news.rcn.net> <slrnao96ee.52u.Gareth.McCaughan@g.local> <am28vl$3bt$1@bob.news.rcn.net>
Message-ID: <996eade8.0209162246.7a281291@posting.google.com>

Thanks for all your help
I've been offline for a few days :)

Ive been doing some more investigations, and I thought you guys might
be interested in it.
Firstly - like you indicated - I found the shadow class code that SWIG
generates (Duh I feel stoopid)

As for the DLL thing, apparently, I just need to do this:
Py_Initialize();
SWIG_init();

... Whatever I need to do (eg. PyRun_SimpleString / File) ...

Py_Finalize();

The python code I run in there can then import from whatever I called
my module

Another thing I found was the nebula source code
  (nebula is a 3D game engine by radon labs - who released the source
code)
Nebula gave me lots to think about
(Once I had gotten over my headaches from figuring out the code)
PS: get ready for a long description :)

Nebula opens the possibility of having one of multiple scripting
languages - and easy remote scripting and serialisation.
The way nebula works with python is by having a 'root' class that
handles Cmd objects - which are simply wrappers for a function call
(with ins and outs)
Aside from this, 'root' also handles RTTI and serialisation
Each child class (inheriting root) indicates which command prototypes
it accepts - with ulong ID's for speed,etc. eg.
   AddCmd("v_MyMethod_s", 'MM  ', PointerToFunction);
The 'v_' indidates void and the '_s' indicates string meaning that
MyMethod returns nothing and accepts a string parameter
Unfortunately this *does* mean writing separate static functions for
the commands the class accepts.

The ScriptingServer class - this is the base class - runs/parses
script commands, converts (where necessary) to Cmd objects and hands
the Cmd objects to the appropriate object instances that need to
handle them.
So.. for instance, you could have a PythonScriptingServer and a
TCLScriptingServer - each capable of converting the language into Cmd
objects and handing the Cmd objects to the application.

I would envisage rather having shadow classes created automatically
based on a classes' list of Cmd prototypes. Those shadow classes could
simply be passing Cmd commands to the scripting server to hand over to
the rest of the application.
In these situations, you only really need a handful of functions
extended to python (eg. SendCmdMessage(ObjectID, Arguments) - ObjectID
would differ according to how one identifies an object instance (void*
or long ID or name,etc)

This method potentially allows for remote scripting (which nebula DOES
seem to allow) but Im still deciphering exactly how they did that (eg.
identification of object instances is still not clear to me)

Another nice side effect is serialisation
Nebula uses Cmd commands to serialise and deserialise.
When an object needs to be saved - it simply generates Cmd objects
describing the commands necessary to reconstruct the object, and
passes them to the application to save to file
To deserialise, those Cmd objects simple need to be read in and passed
to a newly created object to reconstitute it

For a while I was brainstorming how to allow my app to optionally not
use Python at all (or any scripting), but still serialise easily (eg.
Pickle). However, I could not find a reasonable way of doing this -
and nebula's method looks like a fair alternative as Cmd commmands
could still be used outside of scripting.

If you read this far - good going :)

Thanks for the help and effort


From grabiller at 3dvf.net  Thu Sep 26 01:05:05 2002
From: grabiller at 3dvf.net (Guy Rabiller)
Date: Thu, 26 Sep 2002 07:05:05 +0200
Subject: Newbie References Question
References: <3d91c72f$0$8558$79c14f64@nan-newsreader-02.noos.net> <amt4t7$f8a$1@f1node01.rhrz.uni-bonn.de>
Message-ID: <3d929578$0$5921$79c14f64@nan-newsreader-01.noos.net>

Yes you are totaly right.

I'm walking toward this way anyway, but I was just confuse about this
'reference thing' ( wich was my mystake as I forgot that my 'p' was a nested
list of coordinate ).

Thanks for the example Frank.

--
guy rabiller
3d animator / td
grabiller at 3dvf.net
http://grabiller.3dvf.net

Frank Sonnenburg wrote:
> "Guy Rabiller" <grabiller at 3dvf.net> schrieb im Newsbeitrag
> news:3d91c72f$0$8558$79c14f64 at nan-newsreader-02.noos.net...
>> Hi,
>>
>> let say I have:
>> i1 = 1
>> i2 = 2
>> and
>> p = [i1,i2]
>>
>> How can I have:
>> p=[*i1,*i2]
>> rather than
>> p=[**i1,**i2] as it is curently ?
>>
>> ( Sorry for this nasty hybrid syntax )
>>
>> Textualy, how if I want that p[0] refere to the 'i1' container, and
>> not to the refererence it contains ?
>>
>> What I want is that if now I set:
>> i1 = 4
>> that automaticaly:
>> p -> [4,2]
>> and not keeping [1,2]
>
> My proposal: THINK OO!
>
> In my opinion, this is an excellent example to use the full power of
> object oriented programming. Of course, this is a lot of overhead and
> maybe you just wanted a very short approach. But since you told, you
> are a newbie, maybe you are not quite used to this style of
> programming. Just try out and experiment a little.
>
> ===============================================================
>
> # Module point.py
>
>
> class Point:
>         def __init__(self, x=0, y=0):
>                 self._x = x
>                 self._y = y
>
>         def __repr__(self):
>                 return 'Point (%d,%d)' % (self._x, self._y)
>
>         def short_repr(self):
>                 return '(%d,%d)' % (self._x, self._y)
>
>         def set(self, x, y):
>                 self._x = x
>                 self._y = y
>
> class Figure:
>         def __init__(self, *points):
>                 self._points = []
>                 for p in points:
>                         if not isinstance(p, Point):
>                                 raise TypeError
>                         self._points.append(p)
>
>         def __repr__(self):
>                 s = 'Figure'
>                 # Derived class instance?
>                 classname = self.__class__.__name__
>                 if classname != 'Figure':
>                         s = s + ' (%s)' % classname
>                 if self._points:
>                         s = s + ', def. by'
>                         for p in self._points:
>                                 s = s + ' '+p.short_repr()
>                 else:
>                         s = s + ', no points defined'
>
>                 return s
>
>
>         #
>         # Now it's your turn ;-)
>         #
>
>         # def set, get, remove, ......
>         #       ...
>
>
> class Quad(Figure):
>         def __init__(self, p0=Point(0,0), p1=Point(0,1),
> p2=Point(1,0), p3=Point(1,1)):
>                 Figure.__init__(self, p0, p1, p2, p3)
>
>         def set_point(self, index, point):
>                 if type(index) != type(0) or index < 0 or index > 3:
>                         raise ValueError
>                 if not isinstance(point, Point):
>                         raise TypeError
>
>                 self._points[index] = point
>
> ===============================================================
>
>>>> from point import *
>>>> q = Quad()
>>>> q
> Figure (Quad), def. by (0,0) (0,1) (1,0) (1,1)
>>>> q.set_point(1,Point(555,888))
>>>> q
> Figure (Quad), def. by (0,0) (555,888) (1,0) (1,1)
>
>>>> p = Point(1,2)
>>>> p.set(4,2)
>>>> p
> Point (4,2)
>
> ===============================================================
>
>
> Frank




From itsmikeytime at hotmail.com  Wed Sep 25 15:34:29 2002
From: itsmikeytime at hotmail.com (Mikey at Work)
Date: Wed, 25 Sep 2002 15:34:29 -0400
Subject: Dynamic Event Handling in wxPython
Message-ID: <amt348$2pl$1@news-int.gatech.edu>

I'm trying to design a menu in wxPython that connects to an SQL server to
get its menu commands.  So far, I've been able to display the records from
the SQL server in the menu.  The problem, however, occurs when I try to
connect event handlers to the menu items.  I know how to connect the event
handlers (I'm either going to use EVT_MENU_RANGE or assign a EVT_MENU to
each menu item when the items refresh from the server), but the function
called by the event handler has no way of determining the id of the menu
item (at least as far as I know).

If I use...

EVT_MENU_RANGE(lowID, highID, OnMenuSelection)

... OnMenuSelection doesn't know which menu item was selected.  Is there a
way to pass the menu item's id to the function OnMenuSelection or retrieve
the id in the function?  Thanks for your help.  I'm sorry if some of my
terminology is off -- I'm kinda new to Python and programming in general.




From wlfraed at ix.netcom.com  Sun Sep 29 19:24:56 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Sun, 29 Sep 2002 16:24:56 -0700
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com> <874rc8d9c9.fsf@nospam.eml.cc> <mailman.1033319977.7828.python-list@python.org> <3d976fbc$0$23666$91cee783@newsreader02.highway.telekom.at>
Message-ID: <8428na.7vj.ln@ix.netcom.com>

Jurie Horneman fed this fish to the penguins on Sunday 29 September 
2002 02:26 pm:

> Actually, I just tried this in a DOS shell under Windows 98:
> 
> python -v script.py >output.txt
> 
> It lists the verbose output straight to the DOS shell, not to
> output.txt. The output is useless, I can't scroll back.
>
        I suspect -v output uses stderr, not stdout.

        >output.txt redirects stdout

        Don't know if MS-DOS supports redirecting stderr. On linux you'd use

        python -v script.py 2>output.txt


        However, for this situation -- ie, trying to debug your import 
statement -- don't specify the full script on the command line. Do it 
interactively with the output going to the screen.

        Oh, BTW... you NEED TWO "v"... python -vv


>>> import tmp.some     #does not exist
# trying tmp.so
# trying tmpmodule.so
# trying tmp.py
# trying tmp.pyc
# trying /usr/lib/python2.2/tmp.so
# trying /usr/lib/python2.2/tmpmodule.so
# trying /usr/lib/python2.2/tmp.py
# trying /usr/lib/python2.2/tmp.pyc
# trying /usr/lib/python2.2/plat-linux-i386/tmp.so
# trying /usr/lib/python2.2/plat-linux-i386/tmpmodule.so
# trying /usr/lib/python2.2/plat-linux-i386/tmp.py
# trying /usr/lib/python2.2/plat-linux-i386/tmp.pyc
# trying /usr/lib/python2.2/lib-tk/tmp.so
# trying /usr/lib/python2.2/lib-tk/tmpmodule.so
# trying /usr/lib/python2.2/lib-tk/tmp.py
# trying /usr/lib/python2.2/lib-tk/tmp.pyc
# trying /usr/lib/python2.2/lib-dynload/tmp.so
# trying /usr/lib/python2.2/lib-dynload/tmpmodule.so
# trying /usr/lib/python2.2/lib-dynload/tmp.py
# trying /usr/lib/python2.2/lib-dynload/tmp.pyc
# trying /usr/lib/python2.2/site-packages/tmp.so
# trying /usr/lib/python2.2/site-packages/tmpmodule.so
# trying /usr/lib/python2.2/site-packages/tmp.py
# trying /usr/lib/python2.2/site-packages/tmp.pyc
# trying /usr/lib/python2.2/site-packages/Numeric/tmp.so
# trying /usr/lib/python2.2/site-packages/Numeric/tmpmodule.so
# trying /usr/lib/python2.2/site-packages/Numeric/tmp.py
# trying /usr/lib/python2.2/site-packages/Numeric/tmp.pyc
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ImportError: No module named tmp.some
-- 
--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From KBASS1 at nc.rr.com  Mon Sep  2 06:07:21 2002
From: KBASS1 at nc.rr.com (Kevin Bass)
Date: Mon, 02 Sep 2002 10:07:21 GMT
Subject: Newby Question: os.rename
References: <AcGc9.68214$Xa.3523883@twister.southeast.rr.com>
Message-ID: <tLGc9.68215$Xa.3526150@twister.southeast.rr.com>

"Kevin Bass" <KBASS1 at nc.rr.com> wrote in message
news:AcGc9.68214$Xa.3523883 at twister.southeast.rr.com...
> I am attempting to create an archive function within my program. This
> function should create a directory with a timestamp name (such as
> /home/oracle/data/20020831) on the fly, list the contents of a current
> directory then place the contain from the previous directory into the
newly
> created directory.
>
> The newly created directory with the timestamp name is being created and
the
> list of files from the current directory has been retrieved but my problem
> comes when moving the files to the newly created directory. I am trying to
> achieve this tasks with os.rename but find it difficult to get the command
> to recognize the new directory and place the files into it.
>
> I have attempted to get os.rename to recognize the 'archloc' variable that
> contains the name of the newly created directory but that did not work.  I
> also tried to join the newly created directory with the file names then
send
> it through os.rename and that did not work either.
> What am I doing wrong? Thanks!! (My code is listed below)
>
>
> def archive_data(archive):
>
>    the_date = strftime("%Y%m%d")
>    archloc = ('/home/oracle/data/' + the_date)
>    os.mkdir(archloc)
>
>    if archive == 'yes':
>       for file in os.listdir('/home/oracle/data'):
>          file2 = os.path.join(archloc, file)
>          #print file2
>    os.rename(r'/home/oracle/data/' + file, file2)
>
> Kevin
>
>
>
>

I have figured it out. Thanks!




From jwilson at intravative.com  Fri Sep 13 23:24:22 2002
From: jwilson at intravative.com (Demi Vitalis)
Date: Fri, 13 Sep 2002 20:24:22 -0700
Subject: Part Time Help Needed
Message-ID: <uo5bbmrtuccc2d@corp.supernews.com>

Part Time Work at Home: PHP Programmers

First let me open by saying this;

I am a very seasoned programmer but have no friends or associates that
program, so I am forced to ask the general programming community online for
programming help with this project. My name is Jeremy Wilson and I reside in
Los Angeles California. This email is open to everyone no matter what
country you live in.

The Gist;
I am seeking qualified individuals to assist in the development and
programming for a new software product that will enter the marketplace in
the coming months and is being considered to be sold outright to another
software company. I am looking for dedicated individuals with the same
passion and energy necessary to launch proprietary new software in exchange
for points (percentages) of revenues and sales of the software once it is
finished and released. I am short handed and am fast approaching the project
deadline I have set upon myself, so I am offering you the chance to receive
a residual income for assisting in finishing the development of this
application. Before you start sending me responses like "is this a scam" or
whatever, please know that I am just a guy trying to get something done and
offering a piece of what I have accomplished thus far to people who want to
be involved. I will be providing my phone number and etc at the end of this
email for you to talk to me. We will draft an agreement and everything so
you will now you are not working just on spec. and on trust. And yes, I know
that there are other ways of doing this like securing funding and etc to
hire people, but I don't want to go through all of that. it takes way too
long and you have to give up way too much. Basically I am looking for some
team mates and partners to help me with this thing so I can make a lot of
money selling it. And if I am making a lot of money that means you will be
too. I am about 6 weeks away from finishing this software application and
need help to make that deadline. If I don't get help it will take another 3
months probably for me to finish it by myself. Just so you know, the
software is expected to be sold outright for around 3 million dollars. You
do the math. I have no overhead and have one partner who takes care of all
the business.

I have set aside a total of 4% of the ownership of the software to those who
are contracted to finish this project with me. This 4% will be distributed
according to the amount of contributions made and your experience level.
Basically if you are the bomb ass programmer, I have no problem giving you
the full 4%. If I have to hire multiple people to get the job done, I will
be splitting up that 4% according to the skill sets and contributions made.

Please take this as a serious offer and if you have a bunch of spare time on
your hands, you can help to finish this software and make a bunch of money.

Please feel free to contact me via an of the mediums below

Jeremy Wilson
Fax: 1-818-505-0198
Phone: 1-818-623-9850
jwilson at intravative.com
www.intravative.com
MSN Messenger: shaka0070 at hotmail.com
AOL Instant Messenger: shaka0070





From gmuller at worldonline.nl  Fri Sep 13 16:48:06 2002
From: gmuller at worldonline.nl (GerritM)
Date: Fri, 13 Sep 2002 22:48:06 +0200
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <ac677656.0209111225.4a141bc5@posting.google.com> <1a3a9ffa.0209120856.1de06f57@posting.google.com> <b16e4ef7.0209121745.44e91aa@posting.google.com> <b16e4ef7.0209130432.7fb3ee61@posting.google.com>
Message-ID: <altj22$4d5$1@reader1.tiscali.nl>

"Stephen Ferg" <steve at ferg.org> schreef in bericht
news:b16e4ef7.0209130432.7fb3ee61 at posting.google.com...
> > I can't really say WHY using a class structure is desirable.
>
> It is coming back to me now.
>
> As I was working on easygui.py (http://www.ferg.org/easygui/), I
> noticed two things. (1) I was using global variables a lot.  (2) I had
> several GUI "chunks" in one Python program, and when they were all at
> the same level, the code was really unstructured.  Very hard to
> understand.
>
> Granted, a lot of my confusion was simply newbie confusion.  But I
> thought that the next Tkinter development I did, I would see if
> structuring the GUI component into classes would help with these
> problems.  Even in the limited development I've been doing for
> "Thinking in Tkinter", it seems to me that: (1) being able to refer to
> "self." variables removes the need to make so many variables global.
> (2) Grouping the GUI material into classes really helps bring some
> structure to the program that it didn't have before.  I suspect that
> the importance of this (which may not be that obvious in small
> programs) will become overwhelmingly obvious in larger and more
> elaborate GUIs. One of my "to do" items is to test this hypothesis by
> re-writing easygui.py using a class structure.
>
> So right now, my hypothesis is that -- although Tkinter technology
> does not *require* you to structure your GUI as class -- doing so is
> probably desirable simply from the standpoint of making a large
> program more manageable.

I do recognize this evolution. In the beginning  a simple script is used and
global variables start to appear. My experience is that globals are always
suspect. So when the problem and solution is reasonably understood it is
time for refactoring, where structuring in classes frequently is the natural
pattern to use.

I would recommend to follow this logic in your stepwise introduction as
well. So start with very simple "scripts" which have globals, which are
structured in classes when the problem and solution grows and globals become
really dangerous.

Starting with classes up front is not always preferred, simply because you
don't have sufficient insight in problem and solution structure. The result
of too early classes is a lot of self induced additional
statements/structure, which only clobber up the understanding and require
more refactoring than the more pragmatic beginning with globals.

regards Gerrit
--
www.extra.research.philips.com/natlab/sysarch/





From duduca007 at yahoo.com  Tue Sep 24 17:35:28 2002
From: duduca007 at yahoo.com (Carlos Moreira)
Date: Tue, 24 Sep 2002 14:35:28 -0700 (PDT)
Subject: PyBrenda Updates and some Problems
Message-ID: <20020924213528.79691.qmail@web11403.mail.yahoo.com>

Hi,

	I did some updates in PyBrenda for my particular use:

Brenda.py
old: sock.connect(host, port)
new: sock.connect(host, port)

old: sock.bind(host, port)
new: sock.bind((host, port))

old: command = ('rsh ' +
self._nodelist[self._nodeIndex] + ' python ' +
tmpFile)
new: command = ('ssh ' +
self._nodelist[self._nodeIndex] + ' $PYTHON/python ' +
tmpFile)

Well, I need real parallelism (as you can see), so,
first of all, I performed tests/test3.py, that result
in a "error message" below:

:ts got conn <niftySocket.FileSocket instance at
0x81b3594> from ('10.16.165.38', 34043)
:tcr: command is ssh glover /home/pgs/carlos/py
/home/pgs/carlos/.PyBrenda.tmp/@13190.0
:ts: <niftySocket.FileSocket instance at 0x81b3594>
pend (in) ('workerDone',)
carlos at glover's password: 
Traceback (most recent call last):
  File "/home/pgs/carlos/.PyBrenda.tmp/@13190.0", line
14, in ?
    c = Brenda.TupleClientRsh('isere', 2012)
  File
"/home/pgs/carlos/python/Python-2.2/lib/python2.2/site-packages/Brenda.py",
line 387, in __init__
    TupleClientCds.__init__(self, host, port)
  File
"/home/pgs/carlos/python/Python-2.2/lib/python2.2/site-packages/Brenda.py",
line 297, in __init__
    self._conn = self._connectTo(host, port)
  File
"/home/pgs/carlos/python/Python-2.2/lib/python2.2/site-packages/Brenda.py",
line 313, in _connectTo
    sock.connect((host, port))
socket.gaierror: (-2, 'Name or service not known')

I need your help. Do I some mistake? What I can do for
repare this and configure PyBrenda correctly?

	[]'s

Cadu Moreira
KaduSoft President
-----------------------
[......]

__________________________________________________
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
http://sbc.yahoo.com



From tjreedy at udel.edu  Fri Sep  6 08:53:14 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 06 Sep 2002 12:53:14 GMT
Subject: .py to .pyc
References: <mailman.1031300768.16629.python-list@python.org>
Message-ID: <_y1e9.459481$m91.17838026@bin5.nnrp.aus1.giganews.com>

<andrea.baglioni at redturtle.net> wrote in message
news:mailman.1031300768.16629.python-list at python.org...
> Traceback (most recent call last):
>   File "File.py", line 18, in ?
>     import Globals, StructuredText, string, utils, re
> ImportError: No module named Globals
>
> Can anybody help me to find the problem? I think it's a problem of
> $PYTHONPATH and $PYTHONHOME but i'm not sure.

Do the following:

>>> import sys
>>> sys.path
['', 'C:\\Python22\\Lib\\site-packages\\win32',
'C:\\Python22\\Lib\\site-package
s\\win32\\lib', 'C:\\Python22\\Lib\\site-packages',
'C:\\Python22\\Lib\\site-pac
kages\\Pythonwin', 'C:\\PYTHON22\\DLLs', 'C:\\PYTHON22\\lib',
'C:\\PYTHON22\\lib
\\lib-tk', 'C:\\PYTHON22',
'C:\\PYTHON22\\lib\\site-packages\\Numeric']

to find out what is and is not being searched.

Terry J. Reedy





From theller at python.net  Thu Sep 26 10:49:37 2002
From: theller at python.net (Thomas Heller)
Date: Thu, 26 Sep 2002 16:49:37 +0200
Subject: Getting started
References: <00a801c26511$4e084750$df7e4e18@brianspiv1700> <mailman.1033021239.13549.python-list@python.org> <iaDk9.145545$ub2.3207724@news1.tin.it> <lm5olw7n.fsf@ion-tof.com> <%PEk9.146154$ub2.3217944@news1.tin.it>
Message-ID: <amv6q1$9ev20$1@ID-59885.news.dfncis.de>

Alex Martelli wrote:
> Thomas Heller wrote:
>         ...
> 
>>Hidden fields in a union by the discretion of the compiler?
>>Don't think that would be allowed by any C standard.
> 
> 
> Could you help me out by pointing to chapter and verse that would
> make it disallowed, please?
No, sorry. I just find the idea so strange (at least for C, may be
different from C++, never really used that).

Thomas



From anton at vredegoor.doge.nl  Sun Sep 15 05:17:08 2002
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Sun, 15 Sep 2002 11:17:08 +0200
Subject: How to represent sets
References: <alvd8p$1sq$1@news.hccnet.nl> <mailman.1032075137.18192.python-list@python.org>
Message-ID: <am1k1a$59d$1@news.hccnet.nl>

On Sun, 15 Sep 2002 00:31:48 -0700, "James J. Besemer"
<jb at cascade-sys.com> wrote:

>For dense sets isomorphic with integers, the long integer implementation 
>probably will be superior.
>
>For other, more general applications the dictionary approach may be better.

My script is not yet fully optimized, I just made some minor tweaks 
(the new version is at the same adress as before)
and there's some test output below.

For somewhat longer strings Set loads faster but if there are a lot of
operations on the sets, set is faster. For some other cases see below.

Could someone please check these results? It could be a programming
error since I am still tweaking the script.

from sets import Set
from universe import set
from time import time

def test():
    n = 100
    k = 10000
    r1 = [(i,j) for i in range(n) for j in range(n)]
    r2 = [(i,j) for i in range(n) for j in range(n-1,-1,-1)]
    t = time()
    S1 = Set(r1)
    S2 = Set(r2)
    for i in range(k):
        S3 = S1-S2
    print time()-t
    print "S finished"    
    t = time()
    s1 = set(r1)
    s2 = set(r2)
    for i in range(k):
        s3 = s1-s2
    print time()-t
    print "s finished"
    print 

if __name__=='__main__':
    test()

>d:\python22\pythonw -u test.py
92.5499999523
S finished
0.44000005722
s finished

from sets import Set
from universe import set
from time import time

def test():
    n = 100000
    k = 1000
    r1 = [i for i in range(n)]
    r2 = [i for i in range(0,n,2)]
    t = time()
    S1 = Set(r1)
    S2 = Set(r2)
    for i in range(k):
        S3 = S1-S2
    print time()-t
    print "S finished"    
    t = time()
    s1 = set(r1)
    s2 = set(r2)
    for i in range(k):
        s3 = s1-s2
    print time()-t
    print "s finished"
    print 

if __name__=='__main__':
    test()

>d:\python22\pythonw -u test2.py
104.090000033
S finished
2.58000004292
s finished



From zopestoller at thomas-guettler.de  Thu Sep 19 06:43:26 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Thu, 19 Sep 2002 12:43:26 +0200
Subject: ZCatalog without Zope
Message-ID: <amc9ke$e4p$02$1@news.t-online.com>

Hi!

Has anyone used ZCatalog from Zope
with a standalone ZODB (without Zope)?

Background:
I am thinking about moving away from Zope
to quixote or just simple cgi.

If you are intereseted
in the reason way I don't want to use
zope:
http://lists.zope.org/pipermail/zope/2002-August/119965.html

  thomas



From a-steinhoff at web.de  Sun Sep  1 12:36:05 2002
From: a-steinhoff at web.de (Armin Steinhoff)
Date: 1 Sep 2002 09:36:05 -0700
Subject: Python IPC options
References: <1030829572.943495@irys.nyx.net>
Message-ID: <ddc19db7.0209010836.50242309@posting.google.com>

fburton at nyx10.nyx.net (Francis Burton) wrote in message news:<1030829572.943495 at irys.nyx.net>...
[ clipp ..]
> It would be nice if Python could provide the same kind of
> communication mechanisms. Unfortunately, while Glish is
> freely available, it is heavily C++ biased and runs only on
> some Unix systems. In theory, something similar could be
> implemented on a much wider range of platforms using
> sockets and Python.
> 
> Can anyone tell me: has an interprocess communication
> mechanism specifically with a bus architecture been
> implemented in/for Python?

Corba 2.2 -> http://www.fnorb.org 
Python remote objects -> http://www.sf.net/projects/pyro
SOAP -> http://www.pythonware.com/products/soap/
SPREAD -> http://www.python.org/other/spread/
Send/Receive/Reply for QNX6 message passing -> http://www.sf.net/projects/pyqnx

 ... also PyPVM, PyMPI a.s.o.

Hope this is enough :)

Armin


From lothar at kki.net.pl  Wed Sep  4 08:52:37 2002
From: lothar at kki.net.pl (Lothar)
Date: Wed, 4 Sep 2002 14:52:37 +0200
Subject: Matching a string up to a word
References: <slrnanbkbc.huj.dag@omally.animagicnet.no>
Message-ID: <3d7601ef@news.inet.com.pl>

Best library for operating on text is re (Regular Expressions)

import re

log = open("filename.log")
for line in log:
    re.match(r"^\d+ (.*)(\s\w+){4}$",line).group(1)




> I have a string that contains an ID number a file path and a date, kind
> of like this:
>
> 23545 /this/is a/path/to/a file Sun Sep 7 2002.
>
> I want to extract the file name.  If I could be sure that the file path
> had no spaces or if it would be quited it would be easy.  Unfortunatly
> many files will have spaces and changing the format of the output I'm
> dealing with isn't possible.  I do however know that the file name starts
> after the first whitspace and that it's always followed by a three letter
> day name (like Mon, Tue etc.)  Given this is there an easy way to extract
> the whole file path from the string.
>
> Dag
>




From jepler at unpythonic.net  Sat Sep  7 11:23:00 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Sat, 7 Sep 2002 10:23:00 -0500
Subject: Larry Wall's comment on python...
In-Reply-To: <ald3g5$1p8lpp$1@ID-77524.news.dfncis.de>
References: <wkr8g7rn4c.fsf@yahoo.com> <3D7958DB.50903@nospam.free.fr> <ald3g5$1p8lpp$1@ID-77524.news.dfncis.de>
Message-ID: <20020907102252.A15633@unpythonic.net>

On Sat, Sep 07, 2002 at 04:46:07PM +0200, Thorsten Kampe wrote:
> #||<--- LOOK HERE, LARRY, THIS IS WHERE THE FUNCTION _*BEGINS*_!
> def fun(args):
>     #||<--- LOOK HERE, LARRY, THIS IS WHERE THE IF-CLAUSE _*BEGINS*_!
>     if ceci or cela:
>         do_this(with_that)
>         [snip a lot of code]
>     #||<--- LOOK HERE, LARRY, THIS IS WHERE THE IF-CLAUSE _*ENDS*_!
> #||<--- LOOK HERE, LARRY, THIS IS WHERE THE FUNCTION _*ENDS*_!

But my name's not Larry....!  Why don't you write code that's readable by
*me*?

Jeff



From whisper at oz.net  Wed Sep 18 15:14:26 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 18 Sep 2002 12:14:26 -0700
Subject: Earing money with Python software (was: More on Protecting Source Code)
In-Reply-To: <Pine.LNX.4.44.0209180912110.32336-100000@ns0.tbye.com>
Message-ID: <GCEDKONBLEFPPADDJCOECEOAFAAA.whisper@oz.net>

In my experience, the reason big corps don't buy low-priced software is that
it's... low priced!
The big boys expect to pay big prices. They also generally expect big priced
levels of support too. I think some of this stems from them knowing what it
would cost them to build rather then buy and they don't consider the lower
overhead of little companies that allows them to offer lower prices. Thus,
they question the quality of "cheap".

Another reason why big companies/the government don't buy from little
companies in general is based on concerns that 'little company' won't be
around in the future. This is especially true if 'big company' preceives
some degree of mission criticality in the software they're considering.

What dreams are made of: GE buying a 55,000 seat license at $20/seat! :)

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of brueckd at tbye.com
> Sent: Wednesday, September 18, 2002 9:32
> To: python-list at python.org
> Subject: Re: Earing money with Python software (was: More on Protecting
> Source Code)
>
>
> On Wed, 18 Sep 2002, Michael Str?der wrote:
>
> > brueckd at tbye.com wrote:
> > > free-for-personal-use-but-companies-must-pay approach when possible.
> >
> > The problem with getting some bucks from a company is the
> amount of time
> > spent during decision taking and ordering process. For you to earn 300
> > bucks for your software the company spends another 500+ bucks for
> > working time until the money gets out.
>
> Well, it really varies by company of course. Most places I've worked only
> require that much bureaucracy for larger purchases. For small
> stuff (under
> a few hundred dollars) it usually amounts to me informing a superior that
> tool X is needed and worthwhile. After that I just buy it myself and turn
> in a simple expense report.
>
> > IMHO that's why it's so hard to
> > sell small packages of rather cheap software to big companies.
>
> That's why you don't sell it for $300. Seriously: if it's a small and
> cheap package then $300 (I'm assuming US$300) is often too high for small
> businesses and many individuals anyway, and for large companies it might
> be expensive enough to trigger more purchasing overhead (at my current job
> that price could go either way). So... you sell your product with a single
> seat license for, say, $100 and then offer a site license for e.g. $500
> as well as a renewable support package for $1000 a year or whatever. You
> also give 'em discounts when you release a new version next year.
>
> -Dave
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list




From eugene1977 at hotmail.com  Wed Sep  4 08:10:21 2002
From: eugene1977 at hotmail.com (eugene kim)
Date: Wed, 04 Sep 2002 12:10:21 +0000
Subject: Makefile.pre.in ?? installing postgresql --with-python
Message-ID: <al43r1$hk8$1@newsreader.mailgate.org>

./configure --with-python gives me this error (postgresql7.1.3 config)

configure: error: The Python extension makefile was expected at 
`/usr/local/lib/python2.2/config/Makefile.pre.in\'

i searched with 
"python extension makefile"
"Makefile.pre.in"
in groups.google.com
first time i can't find a clue for the problem..

i installed python2.2.1 tgz file
thanks


From Michael.Sparks at rd.bbc.co.uk  Fri Sep 13 06:17:40 2002
From: Michael.Sparks at rd.bbc.co.uk (Michael Sparks)
Date: Fri, 13 Sep 2002 11:17:40 +0100
Subject: wrapping yield ?
References: <alqehl$9qv$1@nntp0.reith.bbc.co.uk> <Xns92885F082A21Cduncanrcpcouk@127.0.0.1>
Message-ID: <alsdpt$2nq$1@nntp0.reith.bbc.co.uk>

"Duncan Booth" <duncan at NOSPAMrcp.co.uk> wrote in message
news:Xns92885F082A21Cduncanrcpcouk at 127.0.0.1...
...
> So something like this should work:
>
> def my_generator(arg):
>    dostuff....
>    for item in waitForSomethingCheckingPeriodically():
>       if item is None:
>           yield None
>       else:
>           processit(item)

Whereas if I could wrap up yield, I simply have to do:

item = waitForSomethingCheckingPeriodically()
process(item)

Which if I have lots of this sort of thing going on, results in much
clearer (and hence maintainable) code.

If it can't (putting yield in a function, and not turning the function into
a
generator) be done, and AFAICT it can't, and this is a suggested work
around - many thanks :-)

Best Regards,


Michael.





From gerhard.haering at gmx.de  Fri Sep  6 19:44:21 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 6 Sep 2002 23:44:21 GMT
Subject: Python from C/SDL
References: <3D793B29.3198AB0E@ihug.co.nz>
Message-ID: <slrnanifek.rh.gerhard.haering@lilith.my-fqdn.de>

Gib Bogle wrote in comp.lang.python:
> Hi,
> 
> I am a complete beginner with Python.  My interest in it is focussed
> on a great piece of open-source software called pymol, a system for
> visualization of molecules that uses Python and OpenGL.  I'd like to
> get access to pymol's graphical functions from my own program, which
> is built from C, SDL and OpenGL.
> 
> I'm wondering if (with a bit of surgery) I'd be able to get access
> to the lower-level functionality of pymol, by which I mean the file
> I/O and rendering functions.

You'll have to embed Python
(http://www.python.org/doc/current/ext/embedding.html), load the pymol
modules and call functions or classes in it from C.

This means you'll have to get familiar with the Python/C API. The good
thing is that this is relatively easy. You should be able to find
documentation and examples on embedding Python. If you have any
particular questions on how to use the Python/C API, just ask here.

HTH,

Gerhard


From spam at melkor.dnp.fmph.uniba.sk  Thu Sep 26 09:18:52 2002
From: spam at melkor.dnp.fmph.uniba.sk (Radovan Garabik)
Date: Thu, 26 Sep 2002 15:18:52 +0200
Subject: unicode woes
References: <amuhrk$4cn$05$1@news.t-online.com> <3d92d120$0$76313$e4fe514c@dreader4.news.xs4all.nl> <amuldd$bur$01$1@news.t-online.com>
Message-ID: <s2eb6-433.ln1@atlas15.dnp.fmph.uniba.sk>

Ulli Stein <mennosimons at gmx.net> wrote:
 : But nevertheless you would have to use everywhere the Python unicode() 
 : function.

 : Or how would you do this (nonsense string):
 : str = str + "bla??" + str[:5] + "???"

 : Would you write then:
 : str = str + unicode("bla??") + str[:5] + unicode("???")?

I ran into the same problem and find out that python2.3
handles this.
Just add the line
# -*- coding: UTF-8 -*-
as the second line of your python script and then
you can use all your strings in UTF-8 encoding
(or any other, whatever fits you)

-- 
 -----------------------------------------------------------
| Radovan Garab??k  http://melkor.dnp.fmph.uniba.sk/~garabik |
| __..--^^^--..__         garabik @ fmph . uniba . sk       |
 -----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!


From claird at starbase.neosoft.com  Wed Sep 18 10:50:33 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 18 Sep 2002 09:50:33 -0500
Subject: minimizing Tk programs to the taskbar
References: <3D8872F1.87850856@optushome.com.au> <Vz_h9.126208$pX1.4528221@news2.tin.it> <656913E3D19D40D2.0831BD5165F06A87.FDA870F87AEF2207@lp.airnews.net> <KF%h9.126691$pX1.4542200@news2.tin.it>
Message-ID: <B7290A895D0F7BAE.4D3DC15DFC3C53E0.8688B972F296AFED@lp.airnews.net>

In article <KF%h9.126691$pX1.4542200 at news2.tin.it>,
Alex Martelli  <aleax at aleax.it> wrote:
>Cameron Laird wrote:
>
>> In article <Vz_h9.126208$pX1.4528221 at news2.tin.it>,
>>
>> Alex Martelli  <aleax at aleax.it> wrote:
>> >Richard Tardif wrote:
>> >> Is there any way to minimize programs to the taskbar? I'm talking about
>> >> windows here. By taskbar I mean the area down in the bottom right
>> >> corner, near the clock, that shows icons for what proggies you've got
>        ...
>> ?  Alex and Richard, I think you'll want to experiment with
>>   root = Tkinter.Tk()
>>     ...
>>   Tkinter.Wm.iconify(root)
>
>Nope: this (at least on Win/98 -- I have no other Win release at hand to 
>try,
>but I don't think the difference matters here) minimizes the window in the
>normal way, NOT to the system-tray, which Richard described as "the
>area down in the bottom right" and (probably mis-)named "taskbar".
>
>> Alex' intuition that a lot of the operations people say they
>> want are platform- or window-manager-specific is certainly
>> apt.
>
>Sure, but Richard did say he meant Windows (though he probably should
>have uppercased that) -- that's pretty much an acknowledgment of
>being platform-specific.  On Win in particular, crowding the system tray
>with a zillion icons was all the rage once -- so many programs did it,
>particularly device drivers and various daemons, as to make things more
>or less unusable.  Win/XP now tries to help by hiding most of the system
>tray except icons used recently, unless one specifically 'opens' the tray.
>It's still somewhat of a mess (and becomes worse, the more programmers
>learn to use Shell_NotifyIcon:-), but that's another issue.
			.
			.
			.
I applaud Alex (a non-native English speaker, moreover!)
for his careful reading.  He's accurate in all he writes
here, and, more than that, he's well informed--his are
exactly the comments that should be made.  I apologize
for confusing matters.

Many people who ask questions of the sort Mr. Tardif asked
find Tkinter's iconify() at least partially satisfying;
that's my experience, at least.  Those who want to pursue 
Tkinter management of the system tray more will probably
find <URL: http://wiki.tcl.tk/winico > interesting.
winico is a rather popular small Tk extension to manage
Windows icons.  At the very least, its source code should
be instructive.

Do properly-coded Tk extensions benefit Tkinter users 
directly?  Is there a way to load them?  I doubt it, but
I haven't thought long about it ...
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html


From tismer at tismer.com  Mon Sep 30 15:41:39 2002
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 30 Sep 2002 12:41:39 -0700
Subject: Stackless Python Howto
References: <1033385295.69247@charme.kabsi.at>
Message-ID: <3D98A8F3.6020609@tismer.com>

Johann H?chtl wrote:
> Hello!
> 
> Can someone give me a short briefing how to use stackless python? I thought
> it's all about a replacement of python.dll, and as long as you don't use
> stackless threading and continuations, at least arbitrary, lightweight
> recursion will be available.
> 
> However, the famous ackermann function with parameters ack (3, 7) and
> onwards still failes because of stack constraints.
> 
> What else do i have to specify to be go stackless?

You need to enable stack saving by sys.enable_stackless.
If you are going to use this on a regular basis, I suggest
to put this into site.py.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From mcherm at destiny.com  Wed Sep 11 09:27:46 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Wed, 11 Sep 2002 09:27:46 -0400
Subject: dynamism
References: <66715c02.0209091923.265fe77e@posting.google.com> <Xns92855D48384F9duncanrcpcouk@127.0.0.1> <3d7e209a_1@news.iprimus.com.au> <Xns92865CE1C916Aduncanrcpcouk@127.0.0.1> <Xns9286606D44C35duncanrcpcouk@127.0.0.1>
Message-ID: <alngca$vq5$1@main.gmane.org>


Duncan Booth wrote:
>  Oh, and the -O flag on
> the CVS build has no measurable effect.

There's a reason for that. Effectively, some clever tricks (and core 
loop rewrites) have been added which puts Python into (the equivilant 
of) -O mode all the time, by calculating line numbers seperately when 
needed instead of generating SET_LINENO bytecodes. At least, that's what 
I thought I understood it to be doing.

Anything that speeds up essentially all Python programs iss OK by me!

-- Michael Chermside






From gmuller at worldonline.nl  Sun Sep 22 12:03:17 2002
From: gmuller at worldonline.nl (GerritM)
Date: Sun, 22 Sep 2002 18:03:17 +0200
Subject: How to get memory usage on Windows?
References: <amilhk$bi8$1@reader1.tiscali.nl>
Message-ID: <amkpo5$kur$1@reader1.tiscali.nl>

Searching in books , documentation (a.o. win32api) and net did not yet help
I found this article on the net:
http://www.developeriq.com/Magazinestories/02jul29memory.php3
but how to call this function from Python?

The module GC is mentioned with respect to memory leaks, but as far a I
could find, no status information can be obtained.

I would expect that in the run time information the size information is
present. Can this be accessed from Python itself?

regards Gerrit

"GerritM" <gmuller at worldonline.nl> schreef in bericht
news:amilhk$bi8$1 at reader1.tiscali.nl...
> I am building an access log analyzer. For the first version I did not
worry
> about performance or memory use, the first exploration was about what is
> practical information to extract and how to present. For current result
see
> http://www.extra.research.philips.com/natlab/sysarch/GaudiStatistics.html
> and
>
http://www.extra.research.philips.com/natlab/sysarch/weblog/index2002.over.h
> tml
>
> For the second version memory use becomes an issue. The brute force
version
> keeps all information simultaneously in memory, without any size
> optimization. This results now in memory use of about 250 MByte (measured
> via very coarse windows resource tool), which scales linear with the
amount
> of (ever increasing) input. I would like to be able to measure more
specific
> how much memory is used per phase and/or per function. Is there any
function
> returning this information. I noted that on Unix the resource module gives
> access to this kind of information, but I could not find any Windows
> alternative.
>
> regards Gerrit

--
www.extra.research.philips.com/natlab/sysarch/





From oliver.smith at removethisbit.baesystems.com  Thu Sep  5 07:01:32 2002
From: oliver.smith at removethisbit.baesystems.com (Olly Smith)
Date: Thu, 5 Sep 2002 12:01:32 +0100
Subject: Returning lists from python COM servers
References: <3d7731f5$1@pull.gecm.com>
Message-ID: <3d7738f5$1@pull.gecm.com>

uhmmmmm, strangely enough, changing it to

==
return [tags,vals]
==

on the server side fixed it!

Now the question is: WHY!?

"Olly Smith" <oliver.smith at removethisbit.baesystems.com> wrote in message
news:3d7731f5$1 at pull.gecm.com...
> Using this snippet on the server side:
>
> ==
> return (tags,vals)
> ==
>
> Where tags and vals are equal length lists (actually a dict that maps
> strings to strings split up so it can be returned from the COM object)
>
> This snippet on the client side:
>
> ==
> tags,vals = self.prodcat.getPDF(id)
> ==
>
> I get this error (mildly snipped cgitb output):
>
> ==
>  547         tag,val = self.prodcat.getPDF(id)
>
> ValueError: unpack tuple of wrong size
>       __doc__ = 'Inappropriate argument value (of correct type).'
>       __getitem__ = <bound method ValueError.__getitem__ of
> <exceptions.ValueError instance at 0x00849670>>
>       __init__ = <bound method ValueError.__init__ of
<exceptions.ValueError
> instance at 0x00849670>>
>       __module__ = 'exceptions'
>       __str__ = <bound method ValueError.__str__ of <exceptions.ValueError
> instance at 0x00849670>>
>       args = ('unpack tuple of wrong size',)
> ==
>
> Any ideas on where I'm going wrong? I'm pulling my hair out trying to work
> this one out.
>
> Thanks
>
> Olly
>
>
>




From eric.brunel at pragmadev.com  Tue Sep  3 08:54:25 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Tue, 3 Sep 2002 14:54:25 +0200
Subject: hackicon, tkicon
References: <3D749F85.962D8D39@ipm.fhg.de>
Message-ID: <al2avp$7jp$1@wanadoo.fr>

Markus von Ehr wrote:

> Hi,
> 
> is there already a hackicon or tkicon module for
> python 2.2 ?
> The modules are useful for changing the window icon on win32.
> 
> Thanks for any hints,
> 
> Markus

IIRC, this is no more needed with Tcl/Tk 8.4: the standard methods 
iconbitmap and/or iconwindow should work on Windows. See:

http://www.tcl.tk/man/tcl8.4/TkCmd/wm.htm#M18

But I haven't actually tested it...

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From blarson at crary.com  Wed Sep 18 12:59:18 2002
From: blarson at crary.com (Bradley D. Larson)
Date: Wed, 18 Sep 2002 11:59:18 -0500
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de>
Message-ID: <3D88B0E6.AC08EC27@crary.com>

I might have missed it... (I read most of the posts on this subject).
Why not just use a significant info registration key. The registration
key would be sensative to date, registered name and/or some other
piece of information that would be required to get the software to work.

At least then  (if the person buying the pirated software) would have to:

    1. Set their computer date BACK to the DATE of the original reg key use.
    2. Enter the name of the original owner.
    3. Enter the other information (ie. version, address -- if they move they get a
new regkey).

This would do several things.

    1. Become pretty obvious to the purchaser that the software is pirated.
    2. The pirater and/or purchaser would have to know all items relating to
            the reg key.

You could have the reg key checked periodically during the program run time.
Hide the checking routine in some unsuspecting code (create a dummy function
like syncMufflerBelt() in a linked c module that gets called periodically.

If you do come accross an illegal copy you would be able to know the original
owner and possible trace the path to the current user (take 'em all down!).

Terje Johan Abrahamsen wrote:

> If I would like to write some code that would not be possible for others to
> read, after I distribute the program, would Python be suitable? I know that
> there are a few programs like Py2Exe, that creates .exe files. But, I also
> saw this:
>
> --------------------
>  A zip-compatible archive is built, containing all files from this
>   directory as well as your script, and appended to a custom python
>   interpreter supplied with py2exe.
> --------------------
>
> Doesn't sound like it is protected very well. Is there some alternative to
> Py2Exe that can do the job, or should I look for a compiled language
> instead?
>
> --
> http://mail.python.org/mailman/listinfo/python-list
-------------- next part --------------
A non-text attachment was scrubbed...
Name: blarson.vcf
Type: text/x-vcard
Size: 400 bytes
Desc: Card for Bradley D. Larson
URL: <http://mail.python.org/pipermail/python-list/attachments/20020918/c4296b77/attachment.vcf>

From grisha at modpython.org  Thu Sep 12 18:25:45 2002
From: grisha at modpython.org (Gregory (Grisha) Trubetskoy)
Date: Thu, 12 Sep 2002 18:25:45 -0400 (EDT)
Subject: ANNOUNCE: Mod_Python donated to ASF
Message-ID: <Pine.BSF.4.32.0209121812020.92256-100000@eden.ispol.com>

It is my pleasure to announce that Mod_Python has been donated to the
Apache Software Foundation, and is now a subproject of the httpd server
project (see http://httpd.apache.org/).

I am grateful to ASF for accepting this donation and committing resources
to further the support of Mod_Python. I believe that this action will
advance the development of Mod_Python, resulting in an ultimately better
and more popular tool for Python developers. I also believe it will serve
to better position Python as a language of choice for web development, a
need that has been expressed by many.

There are no implications to the current Mod_Python users - the license is
the same with the sole difference in that the copyright belongs to ASF
now.

As a consequense of the donation, the CVS repository is now hosted on
cvs.apache.org.  Do not use the SourceForge repository anymore, it will
soon be removed.

There will also be website and mailing changes, but the details are still
being finalized and will be announced when ready.

Regards,

Grisha Trubetskoy





From wodyk at ouray.cudenver.edu  Tue Sep 17 12:38:16 2002
From: wodyk at ouray.cudenver.edu (Wesley Dyk)
Date: Tue, 17 Sep 2002 09:38:16 -0700
Subject: Good book choice?
References: <e6c4872c.0209150635.4816ef98@posting.google.com> <05k9ou4o2m9e41puce9ni3953b5hbhn8c0@4ax.com>
Message-ID: <3D875A77.6FD89AF3@ouray.cudenver.edu>

robin at execulink.com wrote:

> eva_ag at eresmas.com (Eva) wrote:
>
> >I am new to Python but an experienced programmer in Fortran95 (not too
> >object
> >oriented). I am looking for a suited Python book. I'd like it to have
> >clear ideas about OO, as well as an straight pointing into the core
> >language.
> >
> >It seems that 'Python Essential Reference (2nd Edition)' is a good
> >choice. There is also the book 'Making use of Python' (by Rashi
> >Gupta). Is there any user that bought both? Is it good the treatment
> >of OO in the former? Is there a lack of applications point of view in
> >the former? Could anybody compare them? pros/cons?
>
> Can't help you with those, but have had luck with Learning Python
> (Mark Lutz & David Ascher, pub. O'Reilly), a great introduction to the
> language that doesn't pad out its length with stuff you can read in
> the reference documentation. I think it's the single-best Python book
> out there. So do the other developers I work with.
>
> Then you'll want to get the Python Cookbook (edited by Alex Martelli &
> David Ascher, pub. O'Reilly) for in-depth analysis of common tasks and
> thoughtful solutions. This is more advanced than your typical
> Cookbook. It's going to be by my side a long time!
>
> Start here: http://python.oreilly.com
>
> -- robin

I would have to agree with you on _Learning_Python_.  It has good
instruction on OOP with Pyhton.  I have also heard that Guido himself
likes this book the best.  I have not read the _Python_Cookbook_.  I would
start with _Learning_Python_ and move on to a reference book after that.

Aus



From tjlavoie at mb.sympatico.ca  Fri Sep 27 12:34:06 2002
From: tjlavoie at mb.sympatico.ca (Tim Lavoie)
Date: 27 Sep 2002 11:34:06 -0500
Subject: How do I set-up Multiple socket connections
References: <mailman.1033117658.2935.python-list@python.org>
Message-ID: <87n0q3qtmp.fsf@theasylum.dyndns.org>

>>>>> "Kat" == Kat Kathuw Huw <Kat> writes:

    Kat> Ok, I want to iterate through a server IP list and establish

    Kat> multiple socket connections to the various IP address' in the
    Kat> list.

Well, what would you like to do with your conversations? It can be as
simple as putting them all in a list, or each instance of some class
has its own.

If you want to create them all and then poll them sequentially, then
your job is fairly simple. It is somewhat trickier if you are doing
more arbitrary things within say, a GUI event loop. Can you elaborate
a bit?

  Tim


From thorsten at thorstenkampe.de  Sun Sep 15 03:56:25 2002
From: thorsten at thorstenkampe.de (Thorsten Kampe)
Date: Sun, 15 Sep 2002 09:56:25 +0200
Subject: IDLE
References: <wpbf9.264320$aA.45691@sccrnsc02> <3D7D4F84.318CDB0F@bellatlantic.net>
 <v4Pg9.428631$me6.49396@sccrnsc01> <bs70f7ks.fsf@online.no>
Message-ID: <am1efh$1ohti$1@ID-77524.news.dfncis.de>

* syver-en at online.no
>"Ali K" <akakakhel at attbi.com> writes:
>> Hardware: 386 enhanced, 75MHz, 32MB RAM.
>> OS: Win98
>> Copied files from other computer to mine.
>> It doesn't open when I click on ide.pyw or when I choose to edit a py
>> file
>> 
>> with it.
>
> I suggest you install the python distro from www.python.org.

You think this will work on medieval computers?

Thorsten


From alanmk at hotmail.com  Mon Sep 23 16:11:21 2002
From: alanmk at hotmail.com (Alan Kennedy)
Date: 23 Sep 2002 13:11:21 -0700
Subject: Sessionhandling
References: <c5b3549a.0209230650.34a30fca@posting.google.com>
Message-ID: <f72dac0d.0209231211.5614b398@posting.google.com>

storedmails at gmx.net wrote:
> I'm trying to program a web application with python. This tool has to
> provide diffrent languages. The user has to choose the language at the
> beginning. So i have to save the information which language the user
> chooses. I think this is a perfect example for using sessions.

The following is not related to sessions.

Are you aware that a users language preference is encoded in every
request that they make to a web server, through the "Accept-Language"
HTTP header?

Example formats for this header are (from my server logs)

Accept-Language: de
Accept-Language: en,fr-CA
Accept-Language: en-us, en;q=0.50
Accept-Language: en-us,de;q=0.7,fr;q=0.3
Accept-Language: en-us,fr-be;q=0.5
Accept-Language: en-us,zh-tw;q=0.5
Accept-Language: es-mx, en;q=0.5
Accept-Language: fr

So if you access the Accept-Language header (code below), and do some
simple string manipulation, you may be able to solve your problem
without sessions and without making the user explicitly select a
language (they've already done this in their browser).

Sample code-----------------------------------
#! /path/to/python

import os
import string

try:
    langpref = os.environ['HTTP_ACCEPT_LANGUAGE']
    langs = string.split(langpref, ',')
    firstpref = langs[0]
    secondpref = langs[1]
    thirdpref = langs[2]
    # Dealing with quality values (e.g. 'q=0.3') is left as an
exercise for the reader :-)
except:
    firstpref = 'en-ie'
-----------------------------------------------

Maybe might make your life a little simpler....

HTH,

Alan.


From loewis at informatik.hu-berlin.de  Sun Sep 15 14:08:48 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 15 Sep 2002 20:08:48 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net>
Message-ID: <j43csbruq7.fsf@informatik.hu-berlin.de>

"David Abrahams" <david.abrahams at rcn.com> writes:

> I think it still matters. You don't want people who need stackless'
> capabilities to have to use a non-standard Python forever, do you?

Unfortunately, the current implementation is processor-dependent; it
supports x86, SPARC, and PPC only (not really sure about PPC). It also
limits extensions: if they call back to Python, they must not keep
"global" variables on the stack (global in the sense that their
address is known outside their function).

These limitations require Stackless users to be more aware of the
inner workings of Python, or they risk crashes - something that would
be considered a bug in standard Python.

Regards,
Martin


From manuel at hendel.net  Wed Sep 11 08:00:34 2002
From: manuel at hendel.net (Manuel Hendel)
Date: Wed, 11 Sep 2002 14:00:34 +0200
Subject: deleting the first and the last character of a string
In-Reply-To: <20020911115118.GE2054@partagas.as.de.cw.net>
References: <mailman.1031741971.20624.python-list@python.org> <3D7F2960.20507@Linux.ie> <20020911115118.GE2054@partagas.as.de.cw.net>
Message-ID: <20020911120034.GF2054@partagas.as.de.cw.net>

On Wed, Sep 11, 2002 at 01:51:18PM +0200, Manuel Hendel wrote:
> This worked quite good, but I found another problem. There seams to be
> trailing whitespaces. I tried to strip(s) to get rid of them, but this
> doesn't work, I get the following error message:
> 
> Traceback (most recent call last):
>   File "./pop3create.py", line 9, in ?
>       print strip(line)
>       NameError: name 'strip' is not defined
> 
> What's wrong here?

Sorry for that mail, I already found the problem, I simply forgot to
put string. in front of strip. This happens when a programming
beginner tries to solve a problem for several hours.

Thanks,
    Manuel

-- 
What goes up must come down. Ask any system administrator. 



From llothar at web.de  Sun Sep 15 21:31:12 2002
From: llothar at web.de (Lothar Scholz)
Date: Mon, 16 Sep 2002 03:31:12 +0200
Subject: can wxpython do this?
References: <737aefee.0209151702.3f735252@posting.google.com>
Message-ID: <spcaou804m5oo6gah31oqaeuqsuj60h0ip@4ax.com>

On 15 Sep 2002 18:02:17 -0700, erick_papadakis at yahoo.com (Erick
Papadakis) wrote:

>hi, i am considering learning wxpython and have a few questions before
>i
>get into it to make sure it is worth it. 
>
>1. what does it give me that java or c++ dont? 
It is not java. It is a competitor to FOX and QT if you use the c++
version. It's unique when you use the Python version (QT and FOX
bindings are not so perfect at the moment).
>is it cross-platform
>too?

Yes

>2. is it a good platform to develop desktop application in, given that
>it
>is a GUI app? how will i distribute it? is there any easy installer
>that
>will allow me to install python first on a layman user's computer
>first,
>then wxpython, then my own application? is this the concept of
>wxpython
>or am i missing something? 

There is no problem to setup an installer easily that takes
Python,WXPysthon and whatever you need. It simply blowes up the
installer package size. 

>3. does it allow me access to windows (or linux x-windows) native
>APIs?
>for instance, can i minimize my application to a user's desktop's
>taskbar and have a right-click context menu on that taskbar entry?

I think its not comfortable but possible. Look at the API to find out.

>4. do i have something like a browser control built in? i want my
>desktop app to interact with my web server and display the response
>just as if it was a browser (BUT with some more client-side richness
>in features, hence the app to begin with). is this possible?

There is a really simple WxHTML Widget and at least a C++ version of a
wrapped IExplorer control. But having something like this would remove
the cross platform compatibility. 
Perhaps someone is working on a mozilla widget ?
>
>thanks very much for any ideas/resources that i can use to decide. 

Try it and then decide - don't decide on posted answers.


From mhammond at skippinet.com.au  Wed Sep 11 18:59:27 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 11 Sep 2002 22:59:27 GMT
Subject: COM, Outlook, and constants
References: <mailman.1031700793.5145.python-list@python.org>
Message-ID: <jVPf9.22829$_l1.60831@news-server.bigpond.net.au>

Ken Seehof wrote:
> I just upgraded my windows box to XP, and resumed working on a Outlook 
> COM project
> that I haven't touched for a few weeks.  For some reason it doesn't work 
> anymore.
> Anyway, the symptom I am trying to resolve now is that 
> win32com.client.constants
> doesn't seem to work anymore.  It's supposed to acquire attributes for 
> all constants
> defined in loaded COM servers through its __dicts__ attribute.
> 
> The following is typical of python COM client code.  Any hints?
> 
>  >>> import win32com.client
>  >>> ol = win32com.client.Dispatch("Outlook.Application")
>  >>> ns = ol.GetNamespace("MAPI")
>  >>> oc = win32com.client.constants
>  >>> oc.olFolderInbox
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
>   File "C:\Python22\lib\site-packages\win32com\client\__init__.py", line 
> 131, in __getattr__
>     raise AttributeError, a
> AttributeError: olFolderInbox
>  >>> oc.__dicts__
> []
>  >>>

makepy has not been run for this object library.  Your code should use 
gencache.EnsureDispatch() to ensure makepy support exists.

Mark.



From pedro_rodriguez at club-internet.fr  Fri Sep 27 05:30:22 2002
From: pedro_rodriguez at club-internet.fr (Pedro RODRIGUEZ)
Date: Fri, 27 Sep 2002 11:30:22 +0200
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <amvcev$cf6$1@s1.read.news.oleane.net> <7nHk9.172208$pX1.6117207@news2.tin.it> <amvjr0$f1v$1@s1.read.news.oleane.net> <12Uk9.175615$pX1.6212603@news2.tin.it>
Message-ID: <an186n$1h4$1@s1.read.news.oleane.net>

On Fri, 27 Sep 2002 09:47:09 +0200, Alex Martelli wrote:

> Pedro RODRIGUEZ wrote:
>         ...
>> The first one isn't possible in a Python expression AFAIK,
> 
> Of course it is -- you seem to keep ignoring the semantics of Python's
> "and" and "or" operators, which ARE short-circuiting (and return one of
> the operands, differently from C/C++'s short-circuiting && and ||, which
> return 0 or 1 whatever operands they're given).
> 

>From now on, I will know them :/ (even read ref manual on boolean
operations as a punition ;)

Always thought an expression was completely evaluated prior computation,
like in Ada - execept for and then/or else, I was wrong.

>>> def a(): print 1
...
>>> def b(): print 2
...
>>> def c(): print 3
...
>>> if a() and b() and c(): pass
...
1
>>>


Sorry for the noise

Pedro


From jdavis at empires.org  Thu Sep 19 00:42:48 2002
From: jdavis at empires.org (Jeff Davis)
Date: Thu, 19 Sep 2002 04:42:48 GMT
Subject: linux python ideas
References: <3D88E578.6010001@diespammerdieuselesspython.com>
Message-ID: <cBci9.43632$U7.15670948@twister.socal.rr.com>

I think python's strength (for me) is in it's versatility and consistency. 
That combination is very powerful. 

When I say python is consistent, I mean that there are very few special 
cases, and few special character sequences to remember. That means you 
generally only have to reference the docs for functions, and not language 
issues. 

It's also versatile, because
(a) it has a good standard library
(b) the library is easily extended with python, C, or Java (if you're into 
jython). That gives python access to the vast standard libraries, and 
extended libraries (like APIs to virtually anything in existence), of C & 
Java.

I use it for:
* simple task automation, kind of like a shell script but easier to write.
* starting to use it for web programming
* practical extraction and report, text/binary data processing or 
reformatting, whatever you want to call it.
* methematics, such as statistics, but also really simple stuff when I have 
a shell more handy than a calculator :)
* embedded scripting (because it integrates so easily with C)
* small daemons and other small applications

Regards,
        Jeff





Rob Andrews wrote:

> I'll be giving a presentation on Python to my local LUG
> (http://lugoj.org) in a few weeks, and would like to point out at least
> a few things that would be of particular interest to linux users.
> 
> The idea is not "how to program in Python", which would take more than
> one quick presentation, but a one-off demonstration of ways in which
> Python is particularly useful for such a group.
> 
> Any suggestions would be appreciated, since my notes are pretty generic
> so far.
> 
> regards,
> Rob Andrews



From owen at nospam.invalid  Thu Sep 12 13:29:48 2002
From: owen at nospam.invalid (Russell E. Owen)
Date: Thu, 12 Sep 2002 10:29:48 -0700
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid> <alln71$1s2s$1@nntp6.u.washington.edu> <b16e4ef7.0209110616.48dfe04c@posting.google.com> <3D7FFC43.10002@something.invalid> <b16e4ef7.0209120504.62d3c5bd@posting.google.com>
Message-ID: <alqiue$1ejo$1@nntp6.u.washington.edu>

In article <b16e4ef7.0209120504.62d3c5bd at posting.google.com>,
 steve at ferg.org (Stephen Ferg) wrote:

>...
>This explains why "command" binding doesn't automatically pass an
>event object the way that event binding does.  The standard behavior
>of a widget  might be defined across multiple kinds of events  (e.g.
>Button's standard behavior is defined across both a button-press and a
>button-release), so there is no *single* event that could be passed.

True. It would help if the command callback sent an arguement that 
identified the widget (making it look a bit more like an event binding). 
In any event, it doesn't and command is unfortunately still the best way 
to handle button widgets.

In general Tkinter has several useful callback mechanisms, each of which 
is especially easy to use for a particular type of problem and may be a 
headache for other sorts of problems (e.g. events to simulate button 
command). Examples (aside from Button command):

Event binding examples:
- Use mouse button events to do useful things in the Canvas widget. For 
instance you can easily drag objects around by binding to <Button-1> and 
<B1-Motion>. Or show information when the pointer is near an object on a 
Canvas using <Motion>.
- Use keyboard events to modify the behavior of Entry widgets, for 
example bind to the return keyto accept the data and use it somehow, or 
bind to up-arrow and down-arrow to implement a history mechanism.

Variable trace:
- To implement input validation in Entry widgets, specify a textvar and 
put a trace on the variable. In your callback beep and restore the old 
value of the variable if the proposed new value is invalid.

There are others. My Tkinter Summary 
<http://www.astro.washington.edu/owen/TkinterSummary.html> section 
"Events and Callbacks" has what I believe is a complete set though 
without examples. I'm should take another look at it and expand it a bit.

-- Russell


From klausman-un090902 at tuts.net  Mon Sep  9 13:08:22 2002
From: klausman-un090902 at tuts.net (Tobias Klausmann)
Date: 9 Sep 2002 17:08:22 GMT
Subject: Python plugins
References: <alf7rc$1p06l1$1@ID-41653.news.dfncis.de> <mailman.1031495592.26635.python-list@python.org> <3D7C13B3.2070606@nospam.free.fr>
Message-ID: <aliki6$1pqvck$2@ID-41653.news.dfncis.de>

Christian Tismer <tismer at tismer.com> wrote:
> If you executed a proper __import__, all modules and funcs should
> be accessible in the normal way.
> There is no need to exec.

This is what I overlooked: the lowlevel interface to __import__

Thanks a lot for your help (to laotseu, too).

Greets,
Tobias



From sa at hogia.net  Mon Sep 16 12:04:08 2002
From: sa at hogia.net (Sebastian Andersson)
Date: Mon, 16 Sep 2002 16:04:08 +0000 (UTC)
Subject: Py_CompileString and line numbers?
Message-ID: <slrnaoc065.9ki.sa@gamegirl.hogia.net>

I'm parsing a config file and from there I get embedded python code
which I send to Py_CompileString, with the config file's filename given
as filename argument to Py_CompileString. Later, when errors are found
in the python code, the wrong line numbers are reported. Is there
a way to tell Py_CompileString an offset for the first line that
I give to it?

/Sebastian


From aleax at aleax.it  Mon Sep 16 16:13:44 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 16 Sep 2002 20:13:44 GMT
Subject: Alex Martelli <aleax@aleax.it> -> search re and put in a list??
References: <mailman.1032205580.28923.python-list@python.org>
Message-ID: <YXqh9.102777$ub2.2120480@news1.tin.it>

jubafre at brturbo.com wrote:
        ...
>  Traceback (most recent call last):
>   File "<string>", line 1, in ?
>   File "C:\Documents and Settings\jubafre\Desktop\teste.py", line 27, in ?
>     final = [ big_re.sub(x, replacer)for x in S ]
> TypeError: expected string or buffer
> 
> what happened????

Exchange the arguments to big_re.sub -- it's

  final = [ big_re.sub(replacer, x) for x in S ]


Alex



From bokr at oz.net  Tue Sep  3 18:32:30 2002
From: bokr at oz.net (Bengt Richter)
Date: 3 Sep 2002 22:32:30 GMT
Subject: efficient list merging
References: <mailman.1031084711.29283.python-list@python.org>
Message-ID: <al3d9u$7cs$0@216.39.172.122>

On Tue, 3 Sep 2002 13:27:53 -0700, "Ken Seehof" <kseehof at neuralintegrator.com> wrote:

>> [Peter Saffrey]
>>
>> > I have two lists of lists of strings which I would like to merge
>> > efficiently without repeats.  [...] Or is there a better way?
>>
>> I am not sure about the relative speed of everything, but I would be
>> tempted to try:
>>
>>    dict(zip(list1 + list2, [None] * (len(list1) + len(list2)))).keys()
>>
>> It seems to work:
>>
>> >>> list1 =3D list('abcdebxyz')
>> >>> list2 =3D list('pqqr')
>> >>> list1
>> ['a', 'b', 'c', 'd', 'e', 'b', 'x', 'y', 'z']
>> >>> list2
>> ['p', 'q', 'q', 'r']
>> >>> dict(zip(list1 + list2, [None] * (len(list1) + len(list2)))).keys()
>> ['a', 'c', 'b', 'e', 'd', 'q', 'p', 'r', 'y', 'x', 'z']
>>
>> --
>> Fran=E7ois Pinard   http://www.iro.umontreal.ca/~pinard
>
>This is a bit faster, but assumes that the first list does not initially
>contain duplicates.  It eliminates duplicates that would be introduced by
>merging items from list2.  Also, this maintains the original list order
>(i.e. list1+list2 without adding duplicates).
>
>def merge(list1,list2):
>    d1 =3D dict(zip(list1,list1))
If you don't care about a throwaway list of None's (and why should you,
if you don't care about a throwaway list of tuples ;-), you could also write
instead of the above line:
     d1={}; map(d1.setdefault, list1)
>    return list1+[s for s in list2 if s not in d1]
>

I think you all missed "lists of lists" of strings in the original problem statement,
just as I did. But adding my 2 cents' worth of solution to the wrong problem, I'd want
to avoid generating a lot of temp lists in case the imput was really large. E.g.,

 >>> list1 = list('abcdebxyz')
 >>> list2 = list('pqqr')
 >>> d={}
 >>> d=reduce(lambda d,item: d.setdefault(item) or d,list1,d)
 >>> d=reduce(lambda d,item: d.setdefault(item) or d,list2,d)
 >>> d.keys()
 ['a', 'c', 'b', 'e', 'd', 'q', 'p', 'r', 'y', 'x', 'z']

Or, for a one-liner,

 >>> dict(map(None,list1+list2,[None])).keys()
 ['a', 'c', 'b', 'e', 'd', 'q', 'p', 'r', 'y', 'x', 'z']

Regards,
Bengt Richter


From joe.sixpack at whoopie.com  Wed Sep 18 16:45:58 2002
From: joe.sixpack at whoopie.com (Joe Sixpack)
Date: Wed, 18 Sep 2002 20:45:58 GMT
Subject: Is there a list comprehension to do this?
Message-ID: <H2nJ0I.36A@news.boeing.com>

Is there a list comprehension or other 'cleaner' or 'prettier' way
to do the following:

>>> xx = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10)
>>> ww = []
>>> for i in range(0, len(xx), 2):
 ww.append((xx[i+1],xx[i]))

>>> ww
[(2, 1), (4, 3), (6, 5), (8, 7), (10, 9)]
>>>

Thanks in advance...




From hellen at claggetts.net  Sat Sep 14 22:28:02 2002
From: hellen at claggetts.net (Jonathan Claggett)
Date: 14 Sep 2002 19:28:02 -0700
Subject: representing a literal single slash
Message-ID: <dd5454fb.0209141828.1ad24260@posting.google.com>

Here is an easy question (I'm hoping).

I'm trying to process a string so that all doubled up backslashes (\\)
are replaced with a single backslash (\).

The obvious approach doesn't work since a single backslash can't end a
string:

str = str.replace('\\', '\') # broken

So, how does one represent a literal single slash? unicode or hex
values perhaps?

Thanks,

Jonathan


From Oschler at earthlink.net  Wed Sep  4 08:57:01 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Wed, 4 Sep 2002 08:57:01 -0400
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com> <unbp765rbsn809@corp.supernews.com>
Message-ID: <2Knd9.87$2L.4549@e3500-atl2.usenetserver.com>

"RPM1" <rpm1deletethis at nospamfrontiernet.net> wrote in message
news:unbp765rbsn809 at corp.supernews.com...
>
>
> I think you should try coding in C.  It's used everywhere for
> everything.  It's faster than about any other language.
> Most other languages are written in C.

Including Python. :)

thx






From tjreedy at udel.edu  Thu Sep  5 09:24:03 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 05 Sep 2002 13:24:03 GMT
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com> <yu99u1l5bkra.fsf@europa.research.att.com> <al5t94$vse$0@216.39.172.122> <m3bs7da0zs.fsf@mira.informatik.hu-berlin.de> <Xns928060E75E42Fduncanrcpcouk@127.0.0.1>
Message-ID: <TVId9.437927$m91.16913283@bin5.nnrp.aus1.giganews.com>

"Duncan Booth" <duncan at NOSPAMrcp.co.uk> wrote in message
> And don't forget that when the nested function is called the frame
object
> may no longer exist: nested scope variables can continue to exist
after the
> frame in which they were created has been destroyed.

Example:
>>> def adder(x):
...   def _(y):
...     return x+y
...   return _
...
>>> a5=adder(5)
>>> a5(7)
12

Here, adder is a factory that turns out callable instances.  It could
instead be written

>>> class adder:
...   def __init__(self,x):
...     self.x = x
...   def __call__(self,y):
...     return self.x + y
...
>>> a5=adder(5)
>>> a5(7)
12

If one want writable persistent variables, one can use the class
syntax, which also allows for writable class rather than instance
variables.

Terry J. Reedy





From hancock at anansispaceworks.com  Fri Sep  6 16:50:05 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Fri, 06 Sep 2002 13:50:05 -0700
Subject: Python class ( Prolog ) ???
References: <20020905234600.11264.60328.Mailman@mail.python.org>
Message-ID: <3D7914FD.5D713A48@anansispaceworks.com>

From: Christophe Delord <christophe.delord at free.fr>
> On Thu, 05 Sep 2002 21:57:20 GMT
> "Joseph A. Knapka" <jknapka at earthlink.net> wrote:
> > > Luis wrote:
> > > Does anyone have a class that provides the functionality of prolog ???
>  
> > There's a SourceForge project that enables you to call
> > GNUProlog from Python:
> > <URL: http://bedevere.sourceforge.net/>
> 
> I have written a module in Python to mix Prolog and Python. It was just a funny
> exercise. It is not complete and rather slow but if you are curious you can download
> it: http://christophe.delord.free.fr/en/pylog/

You may also be interested in the "unofficial"
python-logic SIG:

http://www.logilab.org/python-logic/

Which was recently mentioned on the edu-sig at python.org
list.

Cheers,
Terry

-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
Anansi Spaceworks                 
http://www.anansispaceworks.com 
P.O. Box 60583                     
Pasadena, CA 91116-6583
------------------------------------------------------



From sholden at holdenweb.com  Mon Sep  9 16:07:34 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 9 Sep 2002 16:07:34 -0400
Subject: web crawler help?
References: <JARd9.6927$yt3.3340577@newssrv26.news.prodigy.com> <668b76ce.0209090652.4b074a1b@posting.google.com>
Message-ID: <Pd7f9.77370$l_4.71679@atlpnn01.usenetserver.com>

"Carl" <kingprad at mail.com> wrote in message
news:668b76ce.0209090652.4b074a1b at posting.google.com...
> I just wanted to say also that a lot of sites have a robots.txt file
> in the root directory with a list of pages the crawler shouldn't troll
> through. it's polite to honor it if you're grabbing tons of pages from
> a server. Probably fine to ignore if you're not using a lot of server
> time and only doing a few simple tasks.

Good point. I didn't mention it because I know that webchecker (with which I
am much more familiar) does honour the robots.txt file, while for some
reason (probably completeness?) websucker doesn't.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From henrik.motakef at web.de  Mon Sep 23 07:45:40 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 23 Sep 2002 13:45:40 +0200
Subject: UpToDate SOAP-module ??
References: <ATAj9.21315$0p1.376966@news2.ulv.nextra.no> <j4fzw1ghpv.fsf@informatik.hu-berlin.de> <vLCj9.24765$sR2.439699@news4.ulv.nextra.no>
Message-ID: <87elbllyjf.fsf@pokey.henrik-motakef.de>

"Thomas Weholt" <2002 at weholt.org> writes:

> Anyway, if you want to provide more information about why I safely can
> ignore SOAP feel free. It would make my project a bit less complicated so
> I'd be glad to skip it. :-)

Try a Google search including the term "REST" (or Representational
State Transfer in full length) for an alternative view on how web
services could work. xml.com also has some articles about it.

Regards
Henrik


From magnus at thinkware.se  Tue Sep 17 09:43:38 2002
From: magnus at thinkware.se (=?ISO-8859-1?Q?Magnus_Lyck=E5?=)
Date: Tue, 17 Sep 2002 15:43:38 +0200
Subject: Larry Wall's comment on python...
References: <mailman.1031339711.632.python-list@python.org>	<pc5bla.off.ln@boundary.tundraware.com>	<roy-780F0D.21015006092002@reader1.panix.com>	<vb2cla.6di.ln@boundary.tundraware.com> <20020913020740.2dae5bb6.rm@no-mail.com>
Message-ID: <3D87318A.4020405@thinkware.se>

> Well... it reminds me when I was young, and we had to program using
> ABACUS. It was really somewhat different outlook on programming.

That's nothing. We had to program in a cardboard box
in the middle of the motorway, our manager sliced us
to pieces every morning, and we had to eat rocks!

Phew, these kids these day... When you tell them how
we had it, they hardly believe us...



From DGustafson at attbi.com  Fri Sep 20 13:22:36 2002
From: DGustafson at attbi.com (Dave Gustafson)
Date: Fri, 20 Sep 2002 17:22:36 GMT
Subject: IDL help?
Message-ID: <vPIi9.376734$_91.484406@rwcrnsc51.ops.asp.att.net>

I'm still trying to get a COM server written in Python to be accessable from
.NET.

So far, I understand that I need to create a .idl file for the server and
compile that using MIDL to a .tlb type library so that the COM server can be
registered.  I *think* that the insides of this .idl file actually needs to
use the ODL syntax?  I've tried inspecting the type definitions exposed by
the ITypeInfo Viewer in the OLE/COM Object Viewer for a COM executable, but
am having a hard time translating the interface definitions into something I
can use for my Python library.

Can anyone help by pointing me to a Python based COM server that works so
that I could install it and inspect the type info as an example?  Or could
someone help with the syntax required to create the .idl for a simple
example like the one in Mark H's book...  (included below...)

# SimpleCOMServer.py - A sample COM server - almost as small as they come!
#
# We simply expose a single method in a Python COM object.
class PythonUtilities:
    _public_methods_ = [ 'SplitString' ]
    _reg_progid_ = "PythonDemos.Utilities"
    # NEVER copy the following ID
    # Use "print pythoncom.CreateGuid()" to make a new one.
    _reg_clsid_ = "{65C3A684-9077-4911-A58D-833A85B8C096}"

    def SplitString(self, val, item=None):
        import string
        if item != None: item = str(item)
        return string.split(str(val), item)

# Add code so that when this script is run by
# Python.exe, it self-registers.
if __name__=='__main__':
    print "Registering COM server..."
    import win32com.server.register
    win32com.server.register.UseCommandLine(PythonUtilities)


Thanks in advance!

Regards,
Dave






From gerhard.haering at gmx.de  Mon Sep 16 00:53:06 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Mon, 16 Sep 2002 06:53:06 +0200
Subject: Source code
In-Reply-To: <am3ms1$2e49b$1@ID-53055.news.dfncis.de>
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de>
Message-ID: <20020916045306.GA2250@lilith.ghaering.test>

* Terje Johan Abrahamsen <spoermeg at voldelig.com> [2002-09-15 23:31 -0500]:
> If I would like to write some code that would not be possible for
> others to read, after I distribute the program, would Python be
> suitable? I know that there are a few programs like Py2Exe, that
> creates .exe files. But, I also saw this:
> 
> --------------------
>  A zip-compatible archive is built, containing all files from this
>   directory as well as your script, and appended to a custom python
>   interpreter supplied with py2exe.
> --------------------

py2exe contains Python byte code. Python byte code can be transformed
into Python source code, but all comments are lost.

> Doesn't sound like it is protected very well. Is there some
> alternative to Py2Exe that can do the job, or should I look for a
> compiled language instead?

You should perhaps reconsider if and why you need to hide your source
code.

-- Gerhard



From whisper at oz.net  Mon Sep  9 19:17:08 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 9 Sep 2002 16:17:08 -0700
Subject: Windows/DOS: double clicking a .py file
In-Reply-To: <just-7B849D.00465410092002@news1.xs4all.nl>
Message-ID: <GCEDKONBLEFPPADDJCOEAECDEOAA.whisper@oz.net>

cmd.exe will accept a flag telling it to stay open after something is run.
You can create a shortcut along the lines of:

cmd /K python myscript.py <args>

You could also add this to the cmd shortcut itself if you have one, but that
generally turns out to be a major pain since all dos-window based apps will
then leave their windows open.

A nice aspect of using a custom cmd shortcut is that you can set colors and
fonts and window size just for this app which might help distinguish it if
you have a lot of dos windows open. You can set the screen buffer height to
a high number so you can scroll really far back (not a good idea to make the
screen buffer width greater then the window width though).

you can type 'help cmd' in a dos window for more info on cmd switches.

HTH,

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Just
> Sent: Monday, September 09, 2002 15:47
> To: python-list at python.org
> Subject: Re: Windows/DOS: double clicking a .py file
>
>
> [Just]
> > > impossible to actually  see  the exception... Is there a way
> > > to keep the dos prompt window in this case?
>
> [Bjorn Pettersen]
> > Try something like the following... We've had much success with
> > automatically emailing the traceback back to the developer using the
> > cgitb module -- you can never trust users to give you good error reports
> > <wink> besides the cgitb module provides much more context than the
> > standard exceptions... Let me know if you'd like the code...
> >
> > try:
> >     your code here
> > except:
> >     import msvcrt, traceback
> >     traceback.print exc()
> >     print 'Hit a key to exit'
> >     while not msvcrt.kbhit():
> >         pass
>
> Ah, that looks like a good trick, I will try that.
>
> Thanks for your code offer, but in most cases I will have to ask for the
> input data anyway, so a simple traceback is good enough for now.
>
> Just
> --
> http://mail.python.org/mailman/listinfo/python-list




From jwbaxter at spamcop.net  Sat Sep  7 18:24:18 2002
From: jwbaxter at spamcop.net (John Baxter)
Date: Sat, 07 Sep 2002 15:24:18 -0700
Subject: None or 0
References: <k7ip4-0p3.ln1@junker.stroeder.com> <jwbaxter-E9C054.09080807092002@corp.supernews.com> <3d7a2b6b.92959728@news.cybermesa.com>
Message-ID: <jwbaxter-3756A1.15241707092002@corp.supernews.com>

In article <3d7a2b6b.92959728 at news.cybermesa.com>,
 joconnor at cybermesa.com (Jay O'Connor) wrote:

> >Go into (English-speaking) classrooms, and you're going to hear chants 
> >of "5 and 1 is 6" not "5 and 1 is 1".
> 
> Because in that contexts "and" means "plus" and in this context "and"
> means "logical and"

Agreed and understood.  But not helpful to the programming for everyone 
cause.

  --John


From bokr at oz.net  Sat Sep 28 14:03:57 2002
From: bokr at oz.net (Bengt Richter)
Date: 28 Sep 2002 18:03:57 GMT
Subject: Python threading?
References: <mailman.1033054839.28121.python-list@python.org>
Message-ID: <an4qud$rmm$0@216.39.172.122>

On Thu, 26 Sep 2002 11:39:04 -0400, Tim Peters <tim.one at comcast.net> wrote:

>[Robert Oschler]
>> Having written _many_ state machines in my life, which async programming
>> boils down to in one form or another, I feel it depends on the task.
>> ...
>> Contrast this with a simple procedure call where a linear series of
>> function calls are made, each with appropriate error handling and status
>> adjustment, in a pleasant centrally located manner.
>
>[Thomas Heller]
>> I wonder if generators can come to a rescue here?
>
>Absolutely.  Read the start of PEP 255.
>
>> Still trying to get them into my head...
>
>Think of them as resumable functions.  You're familiar with functions that
>remember their own data state across calls, via abusing globals, or via
>declaring private static vrbls, in C.  Methods on objects often do the same
>by stuffing their state into instance attributes between calls.  A generator
>remembers its data state across resumptions simply by virtue of that its
>local variables don't vanish, and also remembers its control-flow state
>across resumptions.  Remembering local data state by magic is a real
>convenience, but can be simulated in many other ways with reasonable effort;
>it's the ability to remember and restore control-flow state by magic that
>can make generators more than merely convenient.
>
Any thoughts on defining the next method with a generic *args optional arg list
and making it visible to an executing generator function g as g.__args__ ?

g.__args__ would then just be () for current usage, and could be
backwards-compatibly ignored as desired.

I'm thinking it would, e.g., provide a channel to pass messages etc to
a running task-implemented-as-generator.

Or you could pass control info to a tokenizer, e.g. telling it to back up
one and switch to profuse-info mode before yielding again, or whatever.

Another thought is to pass optional return args in the args tuple attribute
of the StopException object. This could be used variously, but might be used
to return a final result from a child task, where ordinary yields might be
reserved for control info for a dispatcher. Or it could be used to distinguish
exceptional return conditions without using other exceptions. Etc., etc.

This could also be reverse-compatibly ignored, but I think the two thoughts above
could open up a lot of possiblilities. Perhaps they have both been discussed before?

Regards,
Bengt Richter


From jb at cascade-sys.com  Tue Sep 17 01:26:30 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 16 Sep 2002 22:26:30 -0700
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de>	<pUdh9.9377$yt3.4604908@newssrv26.news.prodigy.com>	<mailman.1032184821.5180.python-list@python.org> 	<c%qh9.102794$ub2.2120685@news1.tin.it> <mailman.1032212180.11973.python-list@python.org> <3d868ccc$1@news.sentex.net>
Message-ID: <3D86BD06.3050106@cascade-sys.com>

   
Peter Hansen wrote:

> Copyright is actually a passive thing now, as opposed to active.
> That is, you don't have "to copyright your work", because unless
> you explicitly disclaim the copyright, you automatically get it. 

True.

> As I recall, this was not always the case in at least the U.S.,
> but has been the case for a number of years now.  

True.

Furthermore, even if you formally assign the copyright to somebody else, 
I understand ownership eventually reverts to the original author after a 
lengthy period (45 years or something).  Evidently it preserves the 
copyright for the heirs of the author.

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From jarober at gosmalltalk.com  Sun Sep 22 15:07:55 2002
From: jarober at gosmalltalk.com (James A. Robertson)
Date: Sun, 22 Sep 2002 19:07:55 GMT
Subject: Smalltalk vs. Python and COM was Re: Byte Magazine and Smalltalk
References: <3d87a686@tobjects.> <rd8fougqbgdivimn7vuest3ugoj57lvrr6@4ax.com> <Xns928CA588EF17Dtrippyparcplacecom@64.164.98.7> <3d87dd7c$1@tobjects.> <g9Vh9.313655$kp.1041470@rwcrnsc52.ops.asp.att.net> <Xns928D58DF09B6trippyparcplacecom@64.164.98.6> <20020918.115128.1605894428.1527@cybermesa.com> <amkpk0$or2$1@fe2.cs.interbusiness.it> <amkqfs$ldg$1@reader1.tiscali.nl> <amkup0$qtb$1@fe2.cs.interbusiness.it> <aml2a5$q19$1@reader1.tiscali.nl>
Message-ID: <k65sou8a06p8h3bckv4ror37ufjs9v61ul@4ax.com>

On Sun, 22 Sep 2002 20:29:27 +0200, "GerritM" <gmuller at worldonline.nl>
wrote:

>"Pino Gargiulo" <jk3380 at naida.org> schreef in bericht
>news:amkup0$qtb$1 at fe2.cs.interbusiness.it...
><...snip...>
>>
>What is VWNC? What end level application do you want to support and/or
>build?
>

VWNC is VisualWorks non-commercial.  To download it and try it out,
visit http://www.cincom.com/smalltalk, and follow the download links

>The ActiveState distribution I suggested is a Python distribution
>("batteries included") which allow you to do nearly everything. Additional
>packages can be found at many places a.o. in the vaults of Parnassus.
>
>regards Gerrit

<Talk Small and Carry a Big Class Library>
James Robertson
Product Manager, Cincom Smalltalk


From andrewm at object-craft.com.au  Sun Sep 15 22:48:23 2002
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Mon, 16 Sep 2002 12:48:23 +1000
Subject: Pickling slotted instances 
In-Reply-To: Message from loewis@informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=) 
   of "12 Sep 2002 12:12:45 +0200." <j4fzwf349u.fsf@informatik.hu-berlin.de> 
References: <mailman.1031805375.9030.python-list@python.org>  <j4fzwf349u.fsf@informatik.hu-berlin.de> 
Message-ID: <20020916024823.B64683C8FF@coffee.object-craft.com.au>

>> Instances using __slots__ can't be pickled - is there a fundamental reason
>> why this is so, or is it simply yet to be implemented?
>
>See python.org/sf/520644 for a longer discussion.

I'd actually already found several discussions like that while googling
and reading python-dev archives.  But I'm still no wiser - the closest
I can come to an answer is "it's not quite trivial and therefore hasn't
been done yet".

Guido says:

    The solution space is large, and I don't want to rush it, hence the
    decision to make it fail cleanly in 2.2.1 -- that's the best I can do.

Fair enough (I'm using the CVS 2.3, btw - it can't pickle slots either).
I'm really just wondering what makes the solution space large: my
superficial understanding suggests all the necessary information is
readily available for serialising the object's state (but I can see
metaclasses make it complex).

The SF bug you quote suggests to me that one problem is that internal
machinery that uses slots assumes that slots are not pickled, and would
get upset if slotted attributes suddenly resumed their old values.

>> At first glance, a __getstate__ for a slotted class could be as simple as:
>> 
>>     def __getstate__(self):
>>         attrs = [getattr(self, attr, None) for attr in self.__slots__]
>>         return self.__slots__, attrs
>> 
>> What am I missing?
>
>That won't work: it won't access the slots of the base type. Also, you
>are missing a __setstate__ implementation.

Indeed. The code snippet was intended as a discussion point, rather than a
serious implementation... 8-)

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/



From jb at yahoo.de  Tue Sep 10 06:34:56 2002
From: jb at yahoo.de (JB)
Date: Tue, 10 Sep 2002 12:34:56 +0200
Subject: Python 2.3
References: <3d7cfc19_5@news.newsgroups.com> <5t0qnu0dfbebm1uvk1v93msk4tespgp6bn@4ax.com>
Message-ID: <3d7dc8cd_1@news.newsgroups.com>

Gon?alo Rodrigues wrote:

> There are no plans to implement rationals. There is
> however a Rat class in the demo dir, implementing
> rationals in pure Python.

Thx.
But there is probably no way to write a function f Rat x Rat 
--> Rat and then at the input prompt write f(2/3,-3/2).

-- 
Janos Blazi


-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From max at malva.com.uaREMOVE.IT  Tue Sep 24 07:36:03 2002
From: max at malva.com.uaREMOVE.IT (Max Ischenko)
Date: Tue, 24 Sep 2002 14:36:03 +0300
Subject: Reloading code without shutting down the app
Message-ID: <3nipma.itu.ln@cvs>


Hello.

I have an application which processes some incoming data (recv. as UDP
packets). Currently it is implemented as two threads: first is recieving
data and "put" it into Queue.Queue while second thread "get" data from the
queue and process it.

The problem is when I restart my application to load new
changes some packets could be lost due to unreliable nature of the UDP
protocol.

The solution could be to use TCP-link instead of Queue to bind the
producer and the consumer programs. Then, if the consumer is down the
producer would accumulate data and send it when the link would be
restored.

While workable, this solution seems too complex to me.
Changing UDP transport to TCP is outside of my scope either.

Does python have some ways to reload source code of already running
modules? For example, when consumer recv. a SIGHUP it would delete
currently used objects/classes/modules, import all again, creates objects
and continues to process the data from the Queue.

Is it possible? And how?


-- 
Real Programmers Don't Use Pascal!


From Hobbes2176 at yahoo.com  Wed Sep 18 22:54:33 2002
From: Hobbes2176 at yahoo.com (Mark)
Date: Thu, 19 Sep 2002 02:54:33 GMT
Subject: Minimal debug/rep functionality
Message-ID: <J%ai9.893$1D.574@nwrddc01.gnilink.net>

Hello all,

In a nutshell, what I'd like to do is develop a utility that would print out 
every line in a python program as the program executes (I may even just 
want to print line numbers).  I can see hints of how to do this in pdb (and 
more directly in the underlying bdb class) and also in the code module.

So, 

1) Should I try and inherit a class and override some methods or should I 
just pull out the relevant code and hand spin a small interpreter.

2) Anyone have a 20 liner solution for this?

Thanks.

Regards,
Mark


From sholden at holdenweb.com  Wed Sep 25 07:35:05 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 25 Sep 2002 07:35:05 -0400
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amabpb$5rm$1@news.nems.noaa.gov> <3D8B1537.4010803@thinkware.se> <ad052e5c.0209201128.1a53108@posting.google.com> <mailman.1032664289.615.python-list@python.org> <amo0aq$5nf$1@nntp2-cm.news.eni.net>
Message-ID: <Sdhk9.50340$b%3.20967@fe03>

"Stuart D. Gathman" <stuart at bmsi.com> wrote in ...
> On Sat, 21 Sep 2002 23:10:09 -0400, Christian Tismer wrote:
>
> > Dan Bishop wrote:
>
> >  > Please name one other real-world situation in which decimal
> >  > arithmetic is actually *needed*.
> >
> > There is no reason to argue about that. Decimal arithmetic is exact for
> > most of the operations bankers require. They have been demanding for it
> > since decades, and they will not drop it.
>
Strangely enough, my experience has been somewhat different. You can perform
the arithmetic by mixing chickens' blood with short sticks if you want, the
thing that interests the financial guys is *correct answers*.

> Let me add a horror story of what happens when financial
> applications do *not* demand fixed point/decimal arithmetic.
>
> We took on a client whose accounting system used binary floats for all
> dollar amounts.  And I mean *dollar* amounts, not penney amounts.  Since
> all pennies were approximate, the total due at the bottom of the invoices
> would rarely match what the customer would get adding up the line items
> on a calculator.  It was only off by a few cents, but it tended to make
> customers lose faith in the system.  Statements also did not add up, and
> the Balance Sheet didn't balance.
>
In point of fact it is perfectly possible to use floating-point for such
purposes. The important thing is not to allow rounding errors to accumulate,
which effectively means rounding to 2DP or whatever is required for each
figure that will be presented as part of any total.

> We converted their system to use fixed point, and now everything balances
> and totals to the penney.  Unfortunately, their CPA keeps his own version
> of their General Ledger on Excel with binary floating point.  The CPA
> balance sheets do not balance.  The CPA General Journal transactions do
> not balance.  I demonstrated with a calculator that the CPA figures did
> not balance, but they insisted that we had to go with the CPA figures
> because he is a CPA, and I'm just a programmer.
> This is tricky since their new General ledger does not accept unbalanced
> transactions.
>
Although Excel will use binary floating point, the problem is precisely its
failure to round the figures it's presenting, which *will* allow rounding
errors to creep in. Excel has a calculation option to use "precision as
displayed", shich will help a lot. Unfortunately you are pissing in the wind
trying to get a CPA to admit there may be things they don't understand about
arithmetic. Make sure that if you *can* get him(?) to try this option he
uses a copy iof his spreadsheet, since it will change the precision of
constants permanently.

> My solution was to create a 'CPA adjustment' account to contain
> all the discrepancies, and manually decide how to allocate the
> extra/missing pennies for the CPA transactions.
>
I remember having to explain why it was never going to be possible to
reconcile a multi-currency sales ledger when daily rate changes meant that a
payment arrived having a different value than it had when it set off. At the
same time the accountant educated me as to why columns of figures such as
invoices really must "add up".

I discovered the funniest thing when a few years later I worked with a bunch
of five CPAs to build a combined sales/purchase/general ledger with
integrated job costing, job pricing and payroll. Obviously we had several
interesting discussions about what signs should have which meaning in every
possible context. I discovered that even experienced accountants frequently
use soem rule relating back to the physical circumstances of their training,
like "debits go in the column nearest the window". So they don't actually
think all the time in abstract concepts.

> There ought to be severe penalties for idiots that use floating point
> dollars for financial applications.  If forced to use floating point
> (e.g. because customer demands BASIC), then keep money amounts in
> whole pennies (or whatever the smallest currency unit for the country
> is), and divide by 100 (or whatever) for printing only (or just add the
> decimal point yourself).
>
I disagree. The penalties should be reserved for those who undertake this
kind of work without understanding what they are getting into. The methods
used must satisfy the customer, but the customer isn't usually bothered how
the results are achieved. In my particular case everything was done in BASIC
Plus, so any kind of decimal computation was right out. Floats
(appropriately rounded at every step) were perfectly acceptable.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From jeff at sasmor.com  Sat Sep 14 15:57:17 2002
From: jeff at sasmor.com (Jeff Sasmor)
Date: Sat, 14 Sep 2002 19:57:17 GMT
Subject: Python11
Message-ID: <xwMg9.1433$Dy4.464@nwrddc03.gnilink.net>

Is there any definitive information on the next Python conference,
Python 11?  I've read the conferences-discuss archive on
Python.org (http://mail.python.org/pipermail/conferences-discuss/)
and the discussion seems to be tapering off.  IIRC, by last year
at this time there was already a call for papers.

It would be a shame not to have a forum for things Pythonic in 2003.
Just my 2 cents, but Python needs more visibility and not less!
#--------------------------------
Jeff Sasmor
jeff at sasmor.com




From SBrunning at trisystems.co.uk  Wed Sep 18 10:34:26 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 18 Sep 2002 15:34:26 +0100
Subject: monetary applications [was: Python GUI app to impress the bos
	s?]
Message-ID: <31575A892FF6D1118F5800600846864DCBD817@intrepid>

> From:	Alex Martelli [SMTP:aleax at aleax.it]
> > Perhaps the lack of a FixedPoint type?
> 
> http://starship.python.net/crew/aahz/FixedPoint.py
 
A candidate for the standard library, perhaps? I use it all the time, and if
nothing else, it would be nice for DB-API modules to know that it was there,
so that they could return instances of it for NUMERIC and DECIMAL values.

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




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



From peter at engcorp.com  Fri Sep  6 08:52:18 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 06 Sep 2002 08:52:18 -0400
Subject: how could we know which module is imported already ?
References: <7153651c.0209060233.2a72caaa@posting.google.com>
Message-ID: <3d78a4e8$1@news.sentex.net>

black wrote:
> want to know how many and what the module is imported~

If you would explain why you want to do this, you might get
some even better suggestions than just "sys.modules".

-Peter



From lac at strakt.com  Thu Sep 26 10:07:06 2002
From: lac at strakt.com (Laura Creighton)
Date: Thu, 26 Sep 2002 16:07:06 +0200
Subject: off topic -- spam assassin
Message-ID: <200209261407.g8QE76Wo028975@ratthing-b246.strakt.com>

On the advice of several of you here, I installed spam assassin, and
had a pleasant six months of nearly spam free mail.  Thank you, each
and every one of you who recommmended it.  Alas, the spammers have
got smarter, and I am now getting 25+ pieces of spam that aren't
caught by spam assassin every day again. :-(  What's new for our
side of the arms race?  

Laura Creighton



From tjreedy at udel.edu  Thu Sep 19 11:01:37 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 19 Sep 2002 15:01:37 GMT
Subject: Nokia 3310 SMS Application, done:connecting to serial port(linux/win), done:reading some hex like \x00 . . ., problem:write/send command! noresponse
References: <52facf1d.0209190533.3048e62c@posting.google.com>
Message-ID: <kFli9.146453$5r1.5856420@bin5.nnrp.aus1.giganews.com>

"Polerio Babao Jr.II" <ptbjrii at hotmail.com> wrote in message
news:52facf1d.0209190533.3048e62c at posting.google.com...
> Greetings: Python Masters
>
> I wanted to create an SMS application using Nokia 3310 with
> DAU-9P(data cable),
<snip>
> When I connect to the port for the first time I get the ser.read()
> output like this \x00\x0c I think. But that's two response is what I
> get on my first connect. After that when I envoke ser.read() again,
I
> get nothing.
<snip>

I know nothing about Nokia but have had experience (and battle scars)
programming serial i/o.  Two suggestions.

First: make *sure* you have serial protocol set right --  baud rate,
data bits, stop bits, parity, handshaking?  Also make sure you have
proper pins active and properly connected.  It is easy to be off.

Second: if you do not have one, get a breakout box that lets you see
activity at the port.  Its the equivalent of having a
volt-ohm-ammeter.  You plug cable into box and run another cable from
box to port.  A simple one just has lights that flash when a signal
passes thru.  More complicated one allows some signal patching.  I got
mine long ago from Inmac.  Would check google for source now.

That you *seem* to be getting response from Nokia and then not is
puzzling.  Breakout box would give you better idea of what is
happening when.

Terry J. Reedy




From sholden at holdenweb.com  Thu Sep 12 11:49:19 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 12 Sep 2002 11:49:19 -0400
Subject: "str.contains(part)" or alternatives?
References: <mailman.1031787694.27175.python-list@python.org> <6g2g9.295995$aA.51455@sccrnsc02>
Message-ID: <CJ2g9.191683$l_4.98067@atlpnn01.usenetserver.com>

"Pete Shinners" <pete at shinners.org> wrote > Delaney, Timothy wrote:
> > In fact, this was the deciding factor in Guido changing his position on
'if
> > s1 in s2'. A string has sequence-like behaviour (e.g. it is iterable,
> > individual characters can be indexed), but also has its own behaviour
(e.g.
> > substring containment, all elements are single-character strings).
>
> does this mean we can also now do this? if so, nifty
>
>  >>> 1,2,3 in range(10)
> 1
>  >>> 3,2,1 in range(10)
> 0
>

No, you can't.

1,2,3 is in [1, (1, 2, 3), 2, 3]

but it won't be in [1, 2, 3, 4] - strings are a less general structure than
lists or tuples, and the "subsequence" implementation will apply to strings
*only*.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From info at voidhobo.de  Tue Sep 17 19:24:31 2002
From: info at voidhobo.de (John D. Boy)
Date: 17 Sep 2002 23:24:31 GMT
Subject: CGI question
References: <slrnaoerk7.mgo.jboy@otaku.freeshell.org> <mailman.1032290603.24910.python-list@python.org>
Message-ID: <slrnaofeco.m7n.jboy@otaku.freeshell.org>

Thanks to everyone who responded; this was more than I had even hoped
for! I will probably do what you recommended, Ian. I didn't think it
would be as easy as that. It is the most elegant and (even) more
aesthetically pleasing than my initial idea.

Unfortunately, I do not have access to my webserver's configuration to
use the rewrite module, but I will see what I can do within my
.htaccess.

Thanks again, and all the best,
--John

In article <mailman.1032290603.24910.python-list at python.org>, Ian Bicking wrote:
> On Tue, 2002-09-17 at 13:04, John D. Boy wrote:
>> I am working on a simple CGI script to use as a framework for a website
>> of mine. Basically, I want it to work as follows: When somebody accesses
>> the site, a CGI (called e.g. index.py) is run. It takes an html-template
>> and fills it with content from another file, called e.g. index.dat, and
>> returns the completed html-document. When I want to present another
>> page, I want the url to look like this: http://my.domain.name/?page. In
>> that case, index.py would fill the same template with the content of a
>> file called e.g. page.dat. 
> 
> BTW, I wouldn't recommend doing it this way.  If you have a cgi script
> at, say, /page.py, you can access URLs like /page.py/path/to/something
> -- "/path/to/something" will be stored in os.environ['PATH_INFO'].  
> 
> Then your URLs will look like real URLs, and not like queries.  And for
> mod_rewrite people, if you want to serve the root of your domain off
> that script, I believe this will work:
> 
> RewriteEngine On
> RewriteRule ^/(.*)$ /path/to/page.py/$1
> 
> (I don't think you can use Alias in this case, but I might be wrong)
> 
>   Ian


-- 
John D. Boy                               "Information Superhighway" is 
email      <mailto:info at voidhobo.de>       just an anagram for "I'm on 
hypertext  <http://jboy.chaosnet.org>       a huge wispy rhino fart."


From bokr at oz.net  Wed Sep  4 00:38:34 2002
From: bokr at oz.net (Bengt Richter)
Date: 4 Sep 2002 04:38:34 GMT
Subject: default tuple unpacking?
References: <slrnanaknb.1gh.huaiyu@gauss.almadan.ibm.com> <3d758180$1@news.sentex.net> <3d7584d2$1@news.sentex.net>
Message-ID: <al42oa$1n3$0@216.39.172.122>

On Tue, 03 Sep 2002 23:58:39 -0400, Peter Hansen <peter at engcorp.com> wrote:

>Peter Hansen wrote:
>> Huaiyu Zhu wrote:
>> 
>>> It is often very convenient to have default arguments such as
>>>
>>>     def func(a, b, *c): ...
>>>
>>> Is there a way to use this in statements too?  Examples:
>>>
>>>     a, (b, *c), *d = e
>
>Hmm... I think there might be an inconsistency or an unexpected
>result in this.  What if e already contains a tuple?
>
>   e = (1, 2, 3)
>   *a = e        # a is ((1, 2, 3), )
>
>That seems sensible, and it's just what happens if you pass
>"e" in to a function defined as "def func(*a)".
>
>Now what about this:
>
>   a, *b = e     # a is 1, b is (2, 3)
>
>That's one possible result, and presumably the one you want with
No, I think you'd want it to match the function effect
>this syntax, but it's not what you'd get with the function syntax:
>
>   def func(a, *b):
>     print a
>     print b
>
>   func(e)
>
>Here of course you get (1, 2, 3) in "a" and () in b.  To get the
Fine
>same as the proposed syntax above, you can't pass in a tuple, but
>have to pass the parameters separately:
but you can separate them by unpacking with '*'
>
>   func(1, 2, 3)  # now a is 1, b is (2, 3)
>
You can write the call as
    func(*e)
>The problem is that in function calling, you can pass in a single
>tuple, which is one argument, or you can pass the elements separately
>as multiple arguments, which is when the special * syntax kicks in.
>
Kick it in on the calling side if that's what you want.
>When doing assignment with tuples you don't have the same two options:
>
>   a, *b = 1, 2, 3
>
>is the same as
>
>   a, *b = (1, 2, 3)
>
>So maybe this won't fly.  Comments?
>

With function calling, '*' gets you packing or unpacking depending on context.
Why not also for the above? I.e.,
   a, *b = 1, 2, 3    # => a is 1, b is (2, 3)
and 
   a, *b = (1, 2, 3)  # => a is (1, 2, 3), b is ()
but
   a, *b = *(1, 2, 3) # => a is 1, b is (2, 3)
much as in

 >>> def foo(a,*b): print 'a is %s, b is %s' % (a,b)
 ...
 >>> foo(1,2,3)
 a is 1, b is (2, 3)
 >>> foo((1,2,3))
 a is (1, 2, 3), b is ()
 >>> foo(*(1,2,3))
 a is 1, b is (2, 3)

and

 >>> e=(1,2,3)
 >>> foo(e)
 a is (1, 2, 3), b is ()
 >>> foo(*e)
 a is 1, b is (2, 3)

but

 >>> f=(1,2)
 >>> foo(f,3,4)
 a is (1, 2), b is (3, 4)
 >>> foo(*f,3,4)
   File "<string>", line 1
     foo(*f,3,4)
            ^
 SyntaxError: invalid syntax
 >>> foo(3,4,*f)
 a is 3, b is (4, 1, 2)

Hm. Why no unpacking at the front? (Ok, it's easier not to ;-)

Regards,
Bengt Richter


From peter at engcorp.com  Sat Sep 14 02:39:50 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 14 Sep 2002 02:39:50 -0400
Subject: Part Time Help Needed
References: <uo5bbmrtuccc2d@corp.supernews.com>
Message-ID: <3d82d9a5@news.sentex.net>

(Cross-posting removed.)

Demi Vitalis wrote:
> Part Time Work at Home: PHP Programmers
> 
> First let me open by saying this;
> 
> I am a very seasoned programmer but have no friends or associates ...

Yeah.  Well.

In case it isn't obvious to someone here in comp.lang.python, this
is not likely a legitimate proposal.  I strongly advise against
anyone doing more than what Gerhard has already done.

-Peter



From gua81 at XXXyahoo.com  Mon Sep 23 05:34:21 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Mon, 23 Sep 2002 19:34:21 +1000
Subject: Python and cgi and <a> problem!
Message-ID: <ammn6e$lj7$1@lust.ihug.co.nz>

Hi Im currently making a cgi script in python which list all the
files&directory in a directory and list them into an html file with all the
files linked to the actual file.
..... did what I just said make any sense?

Currently I can list all the files and directories and print it out in the
html file.
my problem now is how to add the anchor tag <a>???

here is the line which prints the files&directories in the html file:
print crap

, where crap is a filename such as colour.txt







From tjreedy at udel.edu  Fri Sep 27 15:53:06 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 27 Sep 2002 19:53:06 GMT
Subject: Is Boost Python version 2 ready?
References: <MvZk9.12124$Lg2.3584877@news2.news.adelphia.net>
Message-ID: <CG2l9.263192$216.10040127@bin4.nnrp.aus1.giganews.com>

"Robert Oschler" <Oschler at earthlink.net> wrote in message
news:MvZk9.12124$Lg2.3584877 at news2.news.adelphia.net...
> Is Boost Python version 2 ready for primetime yet?

Not to exactly answer... from post on PyDev list, I believe they want
to test it with Python 2.2.2 (due in about 10 days?) before calling it
finished.

TJR




From jdavis at empires.org  Thu Sep 19 03:03:53 2002
From: jdavis at empires.org (Jeff Davis)
Date: Thu, 19 Sep 2002 07:03:53 GMT
Subject: Need advice on choosing the technology for intranet based application 
References: <3d88fd1e@news.ivm.net> <Gn7i9.1769$ik2.430@fe02> <3d89121f@news.ivm.net>
Message-ID: <tFei9.18265$V7.4672511@twister.socal.rr.com>

The advice I was about to give you followed what Steve had to say, so I 
have included my comments here.

> Whenever I read "CGI" I also read "a pain to maintain" some words
> later... ? It's a project with a (modest) possibility to turn into a
> product, so both portability and maintanability are issues.

Well, CGI isn't considered the best development environment. It isn't 
horrible though (unless performance is an issue...).The main advantage to 
CGI is that it's portable, and you don't want to go rewrite your app. One 
thing you might do is write your program in python, and use libraries to 
handle everything. That way if you do need to put it on something 
unexpected, you can just rewrite a few lines. 

> On the other hand, Mark also meant CGI/Apache, writting Python/Apache,
> right?

Not necessarily. If you think you'll at least be able to install some 
software on the target webserver, you could use mod_python. Use the 
publisher handler, and read the simple docs on the publisher handler (a 
part of mod_python) and the cgi module (thats a python module with 
routines that are useful for any web programming, not just CGI).

Or, you might want to stick with python CGIs. CGIs seem safer, because I 
don't see as many possible problems or stumbling blocks.If it get's turned 
into some other software later, it will be easy to change it from CGI to 
something else if you code defensively.

Regards,
        Jeff




From joconnor at cybermesa.com  Thu Sep  5 11:34:30 2002
From: joconnor at cybermesa.com (Fearless Freep)
Date: 5 Sep 2002 08:34:30 -0700
Subject: Getting parameters in a walkback
Message-ID: <7d3dc526.0209050734.65a0e85d@posting.google.com>

I was banging around with traceback and trying to drill down into
frame objects and such and I'm kinda stuck.

What I'd like to do is find out if, when an Exeption is raised in a
method or function, is it possible to get the values of the parameters
that were passed in?  Something like...

def divide (x,y):
    return x / y


def divideTwoNumbers (a,b):
    return divide (a,b)


try: 
    divideTwoNumbers (1,0)

except:
   # now what?

This is kinda a contrived example just to show calling a function that
calls a function that will raise an exception back out to the top
level.  I can determine in the except: block that a divide by zero
error was raised in devide(), but can I determine what the values for
the two parameters (x and y) were?

Like I said, a contrived example, but it could help in debugging
larger systems if, for example, I tried to write to a file and the
file didn't exist or something

Thanks

Take care,
Jay


From aleax at aleax.it  Fri Sep 20 03:08:53 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 07:08:53 GMT
Subject: monetary applications [was: Python GUI app to impress the bos s?]
References: <31575A892FF6D1118F5800600846864DCBD817@intrepid> <23891c90.0209190014.26c5bb9@posting.google.com> <s2fdma.pp3.ln@ix.netcom.com> <7xk7lhb943.fsf@ruckus.brouhaha.com>
Message-ID: <9Qzi9.117104$ub2.2497933@news1.tin.it>

Paul Rubin wrote:
        ...
> I'm finding it hard to imagine how any real-world financial
> calculation using IEEE doubles could be off by as much as a penny.

I suggest:

http://www2.hursley.ibm.com/decimal/decifaq.html

and other material involving M. Cowlishaw's long campaign for
decimal arithmetic, including debates in the IEEE committee.
E.g.:

"""
A more typical example is applying a 5% sales tax on an item (such as a 
$0.70 telephone call), rounded to the nearest cent. 

Using double binary floating-point, the result of 0.70 x 1.05 is 
0.73499999999999998667732370449812151491641998291015625; the result should 
have been 0.735 (which would be rounded up to $0.74) but instead the 
rounded result would be $0.73.
"""

That's typical Python behavior today, btw (of course, since it uses
the underlying floating binary representation):

>>> 0.70 * 1.05
0.73499999999999999
>>>

Should your business get tax-audited, inspectors may take a dim view
of your shaving a cent off the sales tax due on each and every item
you sell to the public at $0.70 -- "I found it hard to imagine" how
rounding effects might affect the result (when the law specifies
rounding to the nearest cent and always rounding half cents up) may
not be an acceptable defense.


Alex



From pyth at devel.trillke.net  Sun Sep  1 12:11:55 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 1 Sep 2002 18:11:55 +0200
Subject: yield keyword reserved
In-Reply-To: <770604FC-BDC3-11D6-8A4A-00039351FE6A@mac.com>; from erikprice@mac.com on Sun, Sep 01, 2002 at 11:57:06AM -0400
References: <770604FC-BDC3-11D6-8A4A-00039351FE6A@mac.com>
Message-ID: <20020901181155.D3893@prim.han.de>

Erik Price wrote:
> I'm sorry if this is an easily answered-question.  I have Python2.2 
> running on my system (it is installed on Darwin by default in Mac OS X 
> 10.2), and am learning about iterators from 
> <http://python.org/doc/current/whatsnew/node5.html>, but the keyword 
> doesn't appear to be present on my system:
> 
> Python 2.2 (#1, 07/14/02, 23:25:09)
> [GCC Apple cpp-precomp 6.14] on darwin
> Type "help", "copyright", "credits" or "license" for more information.
>  >>> def generate_ints(N):
> ...  for i in range(N):
> ...   yield i
> <string>:3: Warning: 'yield' will become a reserved keyword in the 
> future
>    File "<string>", line 3
>      yield i
>            ^
> SyntaxError: invalid syntax
>  >>>

With python2.2 you have to explicitely enable generators.  put a 

    from __future__ import generators

at the beginning of a script or in an interactive session.

With the future python2.3 you don't need this.

    holger



From loewis at informatik.hu-berlin.de  Mon Sep 30 13:23:42 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 30 Sep 2002 19:23:42 +0200
Subject: XMLRPC calls don't work in threads, or rather, they don't play well with others ! (help)
References: <3D97D190.9030003@bcm.tmc.edu> <m38z1kqb4r.fsf@mira.informatik.hu-berlin.de> <3D9885AD.2030603@bcm.tmc.edu>
Message-ID: <j4fzvrz90h.fsf@informatik.hu-berlin.de>

Steve Ludtke <sludtke at bcm.tmc.edu> writes:

> Already did this. It definitely seems to be the xmlrpclib calls that
> are  preventing the thread switching.    I'll have to see if I can
> come up with a simple demonstration script to show the problem.

No need to do this. Instead, I recommend that you find out which of
the functions in xmlrpclib is blocking, and then put print statements
into *this* function, and so on, until you cannot find Python code for
a function being called anymore.

Regards,
Martin



From python-list at hendel.net  Wed Sep 25 05:53:49 2002
From: python-list at hendel.net (Manuel Hendel)
Date: Wed, 25 Sep 2002 11:53:49 +0200
Subject: comparing all values of a list to regex
Message-ID: <20020925095348.GG57911@partagas.as.de.cw.net>

I have to compare all values of a list to a regex. If the regex
matches all list items the list should be added to a new list. If the
regex matches just one list item, the list should be added to another
new list. And last but not least, if the regex doesn't mutch at all,
the list should be added again to another new list.

Can someone help me with this? This is driving me crazy.

Thanks,
    Manuel

-- 
For ages I thought life was like fishing in a swimming pool. Now the water's 
all drained out! 



From gerhard.haering at gmx.de  Fri Sep 27 18:23:04 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sat, 28 Sep 2002 00:23:04 +0200
Subject: windows xp command prompt
In-Reply-To: <576c1752.0209271411.6d7bb4c4@posting.google.com>
References: <576c1752.0209271411.6d7bb4c4@posting.google.com>
Message-ID: <20020927222303.GC7781@lilith.ghaering.test>

* mike henley <rsrchstr at hotmail.com> [2002-09-27 15:11 -0700]:
> i have installed the activestate perl, python, tcl, and pragmatic
> programmer's ruby. when i go to the command prompt on windows xp and
> type perl for example... (or python, ruby... etc) i get the following
> message.
> 
> 'perl' is not recognized as an internal or external command,
> operable program or batch file.
> 
> in the control panel| system | advanced | environment variables | Path
> i have
> 
> C:\ruby\bin;C:\Python22;C:\Perl\bin;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem
> 
> does anyone know what i should do to fix this problem? (apart from the
> obvious of restarting the machine)

Close the console window and open a new one. Only upon starting the
shell, the environment is read.

-- Gerhard



From Bill.Scherer at VerizonWireless.com  Fri Sep  6 13:36:48 2002
From: Bill.Scherer at VerizonWireless.com (Scherer, Bill)
Date: Fri, 6 Sep 2002 13:36:48 -0400 (EDT)
Subject: Interapp communication under Python[Linux]?
In-Reply-To: <St5e9.25080$w51.7041171@news2.news.adelphia.net>
Message-ID: <Pine.LNX.4.44.0209061331200.14527-100000@localhost.localdomain>

On Fri, 6 Sep 2002, Robert Oschler wrote:

> My Python app will talk to 3 or 4 modules written in Java and C/C++.
> Fortunately I have the source for the Java and C/C++ modules and I have
> experience in both languages.  Currently the external modules have no
> provision for being accessed outside of the process they are running as.  So
> I'm wondering what the best protocol would be used to "glue it all
> together".  There a dizzying array of possibilities:
> 
> 1) Have everybody talk to each other via SOAP (web services).
> 2) Have everybody talk to each other via XML-RPC (Don't know much about this
> method).
> 3) Use Boost's SWIG interface for C++ to link the C/C++ modules to Python.
> How to link to Java in the same manner?
> 4) Use "sockets" and my own custom protocol for everybody.
> 5) Code in Jython instead and use various methods above to talk to C/C++
> modules from Jython app.
> 6) CORBA? (yikes!)
> 
> So as I said, a lot of ways to get things done.  I'd like to hear from those
> of you that have actually done a lot of cross-app communication with Python
> on what your experiences have been, and the pros and cons you encountered.

<posted and mailed>

Robert - 

Have a serious look at Spread (http://www.spread.org).  There's a 
Python module for it on python.org, and I believe the Java 
interface is included with the code from spread.org.  I've not 
used Spread from Java, but I have a bunch of different Python 
processes on half a dozen or so machines (Both Linux and Win2k), 
all communicating via spread.  It's quite simple to use.

The spread server daemon has been up since I started it, way back 
when the first Python spread module came up.  It's stability is, 
so far, as good as the linux box upon which it runs.

HTH, 

Bill

-- 
Bill.Scherer at Verizon Wireless
RHCE 807101044903581




From aleax at aleax.it  Thu Sep 19 02:26:22 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 19 Sep 2002 06:26:22 GMT
Subject: shared mem & semaphores module
References: <YWbi9.17569$V7.4455639@twister.socal.rr.com> <pan.2002.09.19.08.02.16.550254.723@thomas-guettler.de>
Message-ID: <i6ei9.112316$ub2.2394246@news1.tin.it>

Thomas Guettler wrote:
        ...
> AFAIK there is no module for shared memory.

mmap comes reasonably close.


Alex



From peter at engcorp.com  Sun Sep  8 08:54:27 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 08 Sep 2002 08:54:27 -0400
Subject: python program to watch a file - example.
References: <mailman.1031464092.30848.python-list@python.org> <ales54$2un$0@216.39.172.122> <3d7b3437$0$12727@echo-01.iinet.net.au>
Message-ID: <3d7b486c@news.sentex.net>

Top-posting Rob Hall wrote:
> Bengt Richter <bokr at oz.net> wrote in message
> news:ales54$2un$0 at 216.39.172.122...
>>On Sun, 8 Sep 2002 01:47:12 -0400, "Wayne R" <wayne at tbnets.com> wrote:
>>>I am trying to write a python program that will run as a deamon and
[snip]
>>>Any ideas or links to more info?
>>
>>Don't annoy people you want help from, for starters.
>>What system are you targeting? (Outlook in the headers says you're on
>>windows generating this abominable mix of actual message and useless cruft).

 > Get a fucking life Bengt!!!

He wasn't that far off, Rob.  The OP asks a question in a
way that suggests Linux ("daemon" is not a typical Windows
term) but posts in HTML from Windows.  Even if the HTML
itself was not as annoying as, say, top-posting is, the
OP still ought to provide a little background to help us
give a useful answer.

-Peter



From aleax at aleax.it  Thu Sep 26 08:35:26 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 26 Sep 2002 12:35:26 GMT
Subject: Getting started
References: <00a801c26511$4e084750$df7e4e18@brianspiv1700> <mailman.1033021239.13549.python-list@python.org>
Message-ID: <iaDk9.145545$ub2.3207724@news1.tin.it>

James J. Besemer wrote:
        ...
> Types are otherwise pretty secure in C++ except the "cast" mechanism
> "foils" the type system and thus allows type errors to occur.

Actually, unions are at least as bad -- they're not often used
for type-punning because casts are handier for that, but if you
took cast aways, people would (ab-)use unions (just like they
abuse EQUIVALENCE in Fortran for similar purposes).

Technically I believe that in C++ (not sure about the two ISO
C standards -- I'm more definite that this was the case in
Fortran-77 and I believe in Fortran-IV too) storing one union's
member then retrieving another one puts your program in
undefined-behavior land.  In practice, I know of no existing
compiler that goes to the bother of checking (e.g. by storing
a hidden tag when you store into a union and checking whenever
you retrieve from it -- even harder if you consider _pointers_
to the various members of the union).


Alex



From bokr at oz.net  Sun Sep 29 16:53:17 2002
From: bokr at oz.net (Bengt Richter)
Date: 29 Sep 2002 20:53:17 GMT
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05> <an7bvd$burft$1@ID-125932.news.dfncis.de>
Message-ID: <an7p7t$cpg$0@216.39.172.122>

On 29 Sep 2002 17:06:53 GMT, Christopher Browne <cbbrowne at acm.org> wrote:

>Quoth "Steve Holden" <sholden at holdenweb.com>:
>> <sismex01 at hebmex.com> wrote...
>>> >
>> [ ... ]
>> OK, so don't use whole pennies (or smallest currency) as your
>> base, use 10,000ths of a penny, but still use integer arithmetic
>> for your calculations.
>>
>> Floating point has too many rounding errors to be trustworthy;
>> or, looking at it from another angle, floating-point calculations
>> cannot be 100% accurate because of the way it's implemented.
Floating point IS EXACT over a large range of values and operations.
The problem is that people who don't understand exactly what is happening
want to be safe (fine) and substitute FUD or downright misinformation
(not fine) for fact in talking about it.
>
>The problem with FP is that FP involves the use of binary
>approximations, which commonly cannot exactly represent decimal
>values.
>
>Notably, the decimal fraction 0.1, which is /exactly/ 1/10, does not
>have any exact encoding in binary.  Its representation in binary is a
>repeating binary fraction.
How is that a significantly different problem from repeating decimal
fractions? (E.g., 1/3. or 1/7. or wxyz/9999. which will give you 0.wxyzwxzy..?)
>
>When the (not unreasonable!) expectation is that computers ought to be
>able to deal /exactly/ with plain simple decimal values turns out to
>be wrong, that ought to lead people to think.  
>
>But all too often it doesn't, as we get the typical reaction "Well,
>IEEE FP values can be sufficiently exact for this range of numbers;
>just use them right!"
>
>There's thirty years of experience now of programmers being able to
>use COBOL and PL/1 to manipulate decimal values without having to go
>into the numerical analysis effort to "just use FP right!"

There's thirty years of experience now of programmers being able to
use cars and trains to manipulate bodily location values without having to go
into the numerical analysis effort to "just use diesel compression ignition right!"

Many/most of them don't understand the details of the mechanisms they depend on,
and don't care whether they're using gas/sparks or steam/coal or diesel underneath.
Thirty years of commuting doesn't automatically give the commuter insight into
engine technology.

>
>The "FP way" is that once in a while, you've got to insert operations
>to round things appropriately.  The "old school" folks can
>legitimately gripe that they had decimal types that never required any
>of that nonsense, and that they had those types before many of the
>modern developers were even born.
I'm curious what those "old school" folks thought their programs did when
they had to divide or multiply, and how they conceived of the general rules
behind the results they got ;-)

IMO the big point re use of any kind of arithmetic in a program is that the
definitive rules come from the customer's requirements (whatever the process
of determining those ;-), and should be recorded as an unambigous specification.

Then it's either implemented correctly or not, and it doesn't matter if you are
internally manipulating decimals with digits represented in bi-quinary
or using floating point, other than optimization or real time issues (ok
possibly also understandability issues for the lucky maintainers ;-)

A customer who wants to convert between European currencies according to law
will probably have references to those laws in his spec. A customer who wants to
control machinery or compress music is going to give you other rules to get right
(not unlikely among them to watch out for patent and copyright law ;-)

Regards,
Bengt Richter


From Padraig at Linux.ie  Fri Sep  6 09:52:46 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Fri, 06 Sep 2002 13:52:46 GMT
Subject: tree data structure
References: <FM0e9.2860$cP3.6165@news.iol.ie> <Xns9281891D87291duncanrcpcouk@127.0.0.1> <Us1e9.2865$cP3.6131@news.iol.ie> <Xns92818FDA1D5F4duncanrcpcouk@127.0.0.1>
Message-ID: <Oq2e9.2876$cP3.6115@news.iol.ie>

Duncan Booth wrote:
> Padraig Brady <Padraig at Linux.ie> wrote in
> news:Us1e9.2865$cP3.6131 at news.iol.ie: 
> 
> 
>>>assert result==('L0', [('S2', [('S0', [('Q1', [])]), ('S1', [('Q2',
>>>[])]), ('S3', [])])])
>>>
>>>
>>
>>It certainly does!
>>Just one thing...Why use lists instead of tuples?
> 
> 
> Specifically because you can't append to tuples, but you can append to a 
> list.
> 
> However, a general rule of thumb is to use tuples where you want a record 
> (but don't want the overhead of a full-blown class). So each node here is a 
> record with the fields name,children. Use a list where you want to store a 
> collection of unknown length: so a list is appropriate for the children.

OK

> Note that neither your original list, nor the code I used actually force 
> this to be a tree, you could equally well have a graph with or without 
> cycles. You will find it impossible to represent a graph containing cycles 
> if you use tuples throughout.

True, but I am assured that the parts of the graph I'm interested
in will not have cycles.

thanks again,

P?draig.



From ludal at logilab.fr  Wed Sep 11 13:59:28 2002
From: ludal at logilab.fr (Ludovic Aubry)
Date: Wed, 11 Sep 2002 17:59:28 +0000 (UTC)
Subject: Anyone working on a line level python profiler?
Message-ID: <slrnanv19s.cr.ludal@gemini.logilab.fr>

It looks like it is possible to do a line level profiler with
the new profiling API in python 2.2
Has anyone started to work on such a tool ?

Ludovic



From aahz at pythoncraft.com  Mon Sep 16 11:25:09 2002
From: aahz at pythoncraft.com (Aahz)
Date: 16 Sep 2002 11:25:09 -0400
Subject: Are generator functions thread-safe?
References: <T%dh9.6427$Lg2.1180859@news2.news.adelphia.net>
Message-ID: <am4t4l$b33$1@panix1.panix.com>

In article <T%dh9.6427$Lg2.1180859 at news2.news.adelphia.net>,
Robert Oschler <Oschler at earthlink.net> wrote:
>
>I know that a call to a generator function actually returns an independent
>iterator object (mea culpa if I'm wrong on this), but are there any hidden
>gotchas that would cause the use of the same generator function from
>multiple threads, to result in a sharing (especially local storage) or
>memory collision issue?

The same gotchas for any other function: you're screwed if you share
mutable objects without locks.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From eppstein at ics.uci.edu  Wed Sep 18 17:27:24 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Wed, 18 Sep 2002 14:27:24 -0700
Subject: A list comprehension to do this?...
References: <3D88E4F5.B443918@quartercase.com>
Message-ID: <eppstein-1024F4.14272418092002@news.service.uci.edu>

In article <3D88E4F5.B443918 at quartercase.com>,
 joe sixpack <joe.sixpack at quartercase.com> wrote:

> Is there a list comprehension or other 'cleaner' or 'prettier' way 
> to do the following:
> 
> >>> xx = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10)
> >>> ww = []
> >>> for i in range(0, len(xx), 2):
> 	ww.append((xx[i+1],xx[i]))
> 	
> >>> ww
> [(2, 1), (4, 3), (6, 5), (8, 7), (10, 9)]
> >>>
> 
> Thanks in advance...

It's not much different, but it can easily be done with a list comp:

>>> xx = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10)
>>> ww = [(xx[i+1],xx[i]) for i in range(0,len(xx),2)]
>>> ww
[(2, 1), (4, 3), (6, 5), (8, 7), (10, 9)]

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From tim.one at comcast.net  Mon Sep 16 18:50:04 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 16 Sep 2002 18:50:04 -0400
Subject: difflib optimization -- wow (tim peters?)
In-Reply-To: <20020916152504.A7537@unpythonic.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEIPBEAB.tim.one@comcast.net>

[jepler at unpythonic.net]
> wow, what a speedup!   thanks, whoever it was (tim?)

Don't thank me yet -- it's in a semi-broken state now, although you won't
notice that when comparing text files containing text <wink>.  By default,
it's dynamically deciding what constitutes a "junk line" now, instead of
using the "this is junk" regexp argument that nobody bothers to set.  Given
C source input, for example, it's now likely to decide that lines consisting
solely of

		}

are junk.  Trying to synch up on what are really noise lines costs an
enormous amount of time, so getting a better idea of what *is* likely junk
saves an enormous amount of time.  OTOH, if you have two files consisting
solely of repetitions of "A\n" and "B\n" lines, the current code is going to
decide they're *all* junk lines.  That's the sense in which it's currently
broken.

>     $ time python2.3 -O ~/pyunidiff.py foo.c foo2.c > /tmp/pat
>
>     real    0m5.457s
>     user    0m4.640s
>     sys     0m0.110s
>     $ time python2.2 -O ~/pyunidiff.py foo.c foo2.c > /tmp/pat
>
>     real    3m14.042s
>     user    3m7.480s
>     sys     0m1.180s
>
> foo.c and foo2.c are ~300k long C files from different releases of a
> piece of software.  the diff produced is nearly 250k.
>
> pyunidiff is at http://unpythonic.net/~jepler/pyunidiff.py

Three attempts to connect to that timed out for me, so I don't know what
pyunidiff is doing.  Since ndiff outputs the full text of both input files,
I wouldn't consider it a good thing if it produced an output smaller than
its inputs (or, if that is a good thing, I'm on to a great new compression
scheme <wink>).

> tim, I only need another factor of 30x and it'll beat diff(1) -u in wall
> time (or 5x to beat diff(1) -u -d).  It already gives files about 5%
> shorter on the files I've tried (or on par with diff -u -d).

What is "it", and 5% shorter than what?

> I notice that the diff produced from 2.2 is marginally smaller, another
> 1.5% or so.

See above.  You *seem* to be saying that 2.3's output is both 5% shorter and
about 1.5% longer than 2.2's output.

> The price to pay for the optimization?

The difflib algorithm couldn't care less how long the diffs it produces are.
It's aiming at human comprehensibility, not at minimality.  Tell me that the
new output is 1.5% less intuitive than 2.2's and then I'd care -- provided
you could convince me you had an objective way to measure intuitiveness to
two significant digits <wink>.

It would have been a pointless exercise to recode diff's algorithm in
Python.

if-you-want-diff-you-know-where-to-find-it-ly y'rs  - tim




From Oschler at earthlink.net  Tue Sep 10 13:21:16 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Tue, 10 Sep 2002 17:21:16 GMT
Subject: Trouble upgrading to 2.2.1
References: <aXof9.29813$w51.7932884@news2.news.adelphia.net> <slrnans81k.jc.gerhard.haering@lilith.my-fqdn.de> <_qpf9.29862$w51.7937790@news2.news.adelphia.net> <slrnans9pd.jc.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <gSpf9.29910$w51.7941504@news2.news.adelphia.net>

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote in message
news:slrnans9pd.jc.gerhard.haering at lilith.my-fqdn.de...
>
> Your newsreader is broken. Please fix it or use something better than
> this crapware to post to usenet.

Yes, it's Outlook Express and I'm not love with it.  I'd like to switch out
but currently I have a zillion folders and sub-folders all organized so I'm
kind of stuck for the moment.  Strange, I've made hundreds of posts to
Usenet with OE and never received a complaint before.  If you choose not to
answer my posts in the future because of this I'll understand, you've been
extremely helpful in the past and I appreciate that.

>
> $ ln -s /usr/lib/libcrypto.so.0.9.6 /usr/lib/libcrypto.so.2
>

No I have the exact same versions as you.  I tried making the same symlinks
to no avail.  I even tried moving the symlinks from /usr/lib to a different
directory that I am sure is on the PATH (/sbin) and it still didn't work.
Perhaps it's a "kpackage" thing.  I'll try learning more about 'rpm'.

thx





From jingguoming at yahoo.com.cn  Fri Sep  6 04:03:36 2002
From: jingguoming at yahoo.com.cn (jgm2163)
Date: 6 Sep 2002 01:03:36 -0700
Subject: how can i make a executable program with a python2's script
Message-ID: <dd083388.0209060003.1188000c@posting.google.com>

os     :RedHat 7.2
Python :Python2.2.1

now,i had writen a python's script (extend with .py)
how can i make a executable program on RedHat 7.2?


From david at fielden.com.au  Tue Sep 17 18:34:23 2002
From: david at fielden.com.au (DG)
Date: Wed, 18 Sep 2002 08:34:23 +1000
Subject: Python equivalent to perldoc?
In-Reply-To: <%Avh9.7933$Lg2.1392332@news2.news.adelphia.net>
Message-ID: <002601c25e9a$5ff213b0$3101a8c0@borg.fielden.com.au>

The contents of my f:\python22\pydoc.bat file:

@f:\python22\python f:\python22\lib\pydoc.py %1 %2 %3 %4 %5 %6 %7 %8 %9

Rowdy

> Well...I found "pydoc". However I have to explicitly tell python to 
> execute it on Windows XP. I cannot type "pydoc sys" nor "python pydoc 
> sys" but I have to type "c:\python22\libs\pydoc sys" to get 
> it to work. 
> Shouldn't Python know its own libs? Or at least put the "libs" in the 
> ENV variable?
> 
> Bob




From boiko at demogr.mpg.de  Wed Sep 25 15:05:54 2002
From: boiko at demogr.mpg.de (Serge Boiko)
Date: Wed, 25 Sep 2002 21:05:54 +0200
Subject: pyqt 3.4 -- libqtcmodule.so undefined symbol
Message-ID: <m265wtubxp.fsf@boiko_linux.demogr.mpg.de>

I've just compiled pyqt v.3.4 on Mandrake Linux 9.0 (rc3) and any pyqt
applications I tried crashes with the message:

Traceback (most recent call last):
  File "/usr/lib/python2.2/site-packages/eric/eric.py", line 6, in ?
    from qt import QApplication, QTranslator, QTextCodec
  File "/usr/lib/python2.2/site-packages/qt.py", line 39, in ?
    import libqtc
ImportError: /usr/lib/python2.2/site-packages/libqtcmodule.so: 
undefined symbol: _ZNK9QSGIStyle9classNameEv

I have python 2.2.1 and gcc version 3.2-1mdk. Compilation ran
smoothly.
What's wrong?

-Serge


From emile at fenx.com  Tue Sep  3 21:36:39 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 3 Sep 2002 18:36:39 -0700
Subject: efficient list merging
References: <ced73313.0209031012.13624f3@posting.google.com>
Message-ID: <al3o2b$e5g$1@main.gmane.org>

"Peter Saffrey" <theoryboy at my-deja.com> wrote in message
news:ced73313.0209031012.13624f3 at posting.google.com...
> I have two lists of lists of strings which I would like to merge
> efficiently without repeats.

I like sets in current cvs:

Python 2.3a0 (#29, Aug 31 2002, 19:16:08) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import sets
>>> s1 = sets.Set('abcdefg')
>>> s2 = sets.Set('efghijk')
>>> s1^s2
Set(['a', 'c', 'b', 'd', 'i', 'h', 'k', 'j'])
>>> s1|s2
Set(['a', 'c', 'b', 'e', 'd', 'g', 'f', 'i', 'h', 'k', 'j'])
>>> s1&s2
Set(['e', 'g', 'f'])
>>>



Emile van Sebille
emile at fenx.com








From phr-n2002b at NOSPAMnightsong.com  Fri Sep 20 06:30:21 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 20 Sep 2002 03:30:21 -0700
Subject: monetary applications [was: Python GUI app to impress the bos s?]
References: <31575A892FF6D1118F5800600846864DCBD817@intrepid> <7xk7lhb943.fsf@ruckus.brouhaha.com> <9Qzi9.117104$ub2.2497933@news1.tin.it> <7xd6r9uj2i.fsf@ruckus.brouhaha.com> <HYBi9.117876$ub2.2506835@news1.tin.it>
Message-ID: <7xd6r9kl6q.fsf@ruckus.brouhaha.com>

Alex Martelli <aleax at aleax.it> writes:
> Say that you publish the erroneous information that each phone
> call is $0.70 plus $0.03 sales tax for a total of $0.73.  Sales
> tax is actually $0.04.  Now, a customer makes 100 calls.  Do
> you charge them $74 explaining that 100 times 73 cents is 74
> dollars?  Or do you charge $73 and eat the dollar loss yourself?
> And even then, your tax books won't balance without fiddling --
> even if the unbalance is against your interests, as in this case,
> it's _still_ illegal (at least around here)... double-entry tax
> book-keeping is supposed to balance to the last cent.

Now that I think about it, the way it's done around here is you're
supposed to look up the amount of tax in a table, not calculate it.  I
remember giving my local ice cream store a hard time when I was
little.  I used to go in for an ice cream and coke, and because of how
the tax table worked, it was one cent cheaper to buy them in two
separate transactions than in one transaction, so I did that.

At the end of the day, though, I'm sure they just took the amount that
the cash register said was collected pre-tax, multiplied by the tax
rate and sent that amount to the state.  If the amount actually
collected was slightly lower or higher, they paid the difference or
kept it themselves.

> But what you said were that you had a hard time imagining any
> commercially significant computation being off by a whole penny
> (which, I imagine, means a whole cent) by using IEEE double
> (binary floating point).  Surely this one example IS enough to
> jog your imagination?  

Yes, point taken.


From timr at probo.com  Fri Sep 20 00:30:05 2002
From: timr at probo.com (Tim Roberts)
Date: Thu, 19 Sep 2002 21:30:05 -0700
Subject: wxPython i18n, help wanted
References: <amd2hs$i1k11@news.vtx.ch>
Message-ID: <fo8lous4r0jrc458qfn5sl400hb8pnr69c@4ax.com>

Cedric Dutoit <dutoitc at hotmail.com> wrote:
>
>I'm trying to translate an application in Chinese with i18n, but
>I don't know how to display Chinese characters in wxPython.
>
>Can anybody help me ?

If you go here: http://www.wxpython.org/, you will find the mailing list
for wxPython.  That's the best place to ask wxPython-specific questions.

I think you will be disappointed with the i18n support in wxPython.  On the
other hand, perhaps you would be willing to help make it better!  What it
really needs is someone with the experience.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From msanchez at grupoburke.com  Fri Sep  6 11:15:21 2002
From: msanchez at grupoburke.com (Marcos =?ISO-8859-1?Q?S=E1nchez?= Provencio)
Date: 06 Sep 2002 17:15:21 +0200
Subject: [DB-SIG] Re: What database should I use
In-Reply-To: <1031323316.6785.9.camel@asimov>
References: <Pine.LNX.4.21.0209061629050.19270-100000@samba.crysberg.dk> 
	<1031323316.6785.9.camel@asimov>
Message-ID: <1031325321.6371.3.camel@cynar>

That's right, there are not enough arguments to judge. The next thing
is: How many concurrent users are you going to have? If it is one I'd go
for gadfly. Does it need to be portable? If it's Windows only, I'd say
Access (Jet). And don't forget about a text file or a cPickle to be
minimalistic...

El vie, 06-09-2002 a las 16:41, Tom Jenkins escribi?:
> On Fri, 2002-09-06 at 10:29, Leif Jensen wrote:
> > 
> >     PostgreSQL, no doubt, if you use Linux, WinNT or possible WinXP !
> > 
> >  Leif
> > 
> 
> I disagree.  Don't get me wrong, we use PostgreSQL EXTENSIVELY.  But for
> 2-3k records?  These "requirements" don't require anything that
> powerful.
> 
> I'd look at Access, pySQLite (which intrigues me), Gadfly.
> 
> > 
> > 
> > On Thu, 5 Sep 2002, A wrote:
> > 
> > > Hello,
> > > I am going to program a small application (in Python and wxPython) which consists about 
> > > 2000 - 3000 records . Can you please recommend which database is the best for that. I 
> > > use Windows.
> > > Maybe MS Access  or FoxPro or MySQL or better?
> > > I want to keep the program as small as possible.
> > > Thanks
> > > Ladislav
> > > 
> > > 
> > > _______________________________________________
> > > wx-users mailing list
> > > wx-users at lists.wxwindows.org
> > > http://lists.wxwindows.org/mailman/listinfo/wx-users
> > > 
> > 
> > 
> > _______________________________________________
> > DB-SIG maillist  -  DB-SIG at python.org
> > http://mail.python.org/mailman/listinfo/db-sig
> -- 
> 
> Tom Jenkins
> Development InfoStructure
> http://www.devis.com
> 
> 
> 
> _______________________________________________
> DB-SIG maillist  -  DB-SIG at python.org
> http://mail.python.org/mailman/listinfo/db-sig





From see_reply_address at something.invalid  Tue Sep 24 23:07:04 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Wed, 25 Sep 2002 15:07:04 +1200
Subject: Three dumb questions (ordered by dumbness descending)
References: <amnvit$7kqdt$1@ID-77524.news.dfncis.de>
Message-ID: <3D912858.90104@something.invalid>

Thorsten Kampe wrote:

> 1. Why is 'zip(zip(x)) != x' (should be the same because it's the 
> transposed)


zip(x) isn't the transpose of x. Try it and you'll see.

However, zip(*x) IS the transpose of x (or apply(zip, x) in
older Pythons).

Therefore,

    zip(*zip(*x)) == x

is true, provided you start with an x of the right type
(a list of tuples, as opposed to a list of lists or
tuple of tuples, etc.)

> 2. Peter Norvig mentions in "Python for Lisp Programmers" some 
> "don'ts": "[x] + y" and "x[1:]". Are there more things to avoid 
> (especially in a loop)?


The "for Lisp Programmers" is the important thing here.
He's pointing out that these operations are not as
efficient as the corresponding ones in Lisp, because
Python lists aren't linked lists like Lisp lists, so
these operations involve copying a whole list.

> So: can I use random.shuffle or do I have to write my own - slower - 
> version to get a random list?


It depends on how imporant it is to you that all
possible permutations have equal probability. If
you're writing a solitaire game, probably nobody
will notice the difference. If it's a nuclear
missile control system, you might have to take
more care...

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From aldo at nullcube.com  Sat Sep 14 23:38:33 2002
From: aldo at nullcube.com (Aldo Cortesi)
Date: Sun, 15 Sep 2002 13:38:33 +1000
Subject: confused with class inheritance
In-Reply-To: <3d83fa2c$0$24041@echo-01.iinet.net.au>
References: <3d83fa2c$0$24041@echo-01.iinet.net.au>
Message-ID: <20020915033833.GA21366@nullcube.com>

Thus spake Rob Hall (bloke at ii.net):

> In a sample app, I have the following:
> 
> >>> class a(sgmllib):
>  pass
> 
> >>> a
> <module '?' (built-in)>
> >>> dir(a)
> []
> >>> dir(sgmllib)
> ['SGMLParseError', 'SGMLParser', 'TestSGMLParser', '__all__',
> '__builtins__', '__doc__', '__file__', '__name__', 'attrfind', 'charref',
> 'commentclose', 'endbracket', 'entityref', 'incomplete', 'interesting',
> 'markupbase', 'piclose', 're', 'shorttag', 'shorttagopen', 'starttagopen',
> 'tagfind', 'test']
> >>> b=a
> >>> dir(b)
> []
> >>>
> 
> Why doesn't dir(b) show the same as dir(sgmllib) ?

sgmllib is a module, not a class. If you try to instantiate
your "class" a, you will get an exception. 


This should act the way you expect it to:

class Foo:
    def oink(self):
        pass

class Bar(Foo):
    pass

print dir(Foo)
print dir(Bar)



Cheers,



Aldo


-- 
Aldo Cortesi
aldo at nullcube.com
http://www.nullcube.com



From jslowery at hotmail.com  Fri Sep 13 19:38:56 2002
From: jslowery at hotmail.com (Jeremy Lowery)
Date: Fri, 13 Sep 2002 18:38:56 -0500
Subject: html template system?
References: <btug9.688$Ee4.1090@news-server.bigpond.net.au>
Message-ID: <Msug9.4334$AY5.523821@e3500-atl1.usenetserver.com>

":B nerdy" <thoa0025 at mail.usyd.edu.au> wrote in message
news:btug9.688$Ee4.1090 at news-server.bigpond.net.au...
> what are good templating systems avaliable for python?

http://www.cheetahtemplate.org/ is the one I use.

J Lowery





From brueckd at tbye.com  Wed Sep 18 17:05:17 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 18 Sep 2002 14:05:17 -0700 (PDT)
Subject: Python GUI app to impress the boss?
In-Reply-To: <96904b50.0209181128.63cd71f9@posting.google.com>
Message-ID: <Pine.LNX.4.44.0209181404380.32726-100000@ns0.tbye.com>

On 18 Sep 2002, Lee Gray wrote:

> Chris Liechti <cliechti at gmx.net> wrote in message news:<Xns928CED692541cliechtigmxnet at 62.2.16.82>...
> > lpgray at uop.com (Lee Gray) wrote in 
> > news:96904b50.0209171305.6d65f574 at posting.google.com:
> > > My boss is pretty open-minded, but also needs to see something working
> > > to be convinced Python is even a viable platform (he had never heard
> > > of it).  Otherwise, .Net is a given, whether it's any good or not,
> > > since corporate is MS all the way.
> > 
> > with python you have to choice of diffrent GUI libs. if you consider 
> > wxWindows you can play around with the wxPython demo application that comes 
> > with the download. on the other hand when you support win32 only the 
> > win32all package has support for all the windows APIs and it has some 
> > samples that get installed.
> > 
> > chris
> 
> I've been playing with wxWindows and Boa Constructor. wxWindows is
> pretty darn cool and will make a good demo for him.  Boa looks good
> but is still a bit rough (naturally, being alpha), but I'm looking
> forward to future releases.

wxDesigner is also pretty nice (http://www.roebling.de).

-Dave




From emile at fenx.com  Mon Sep  9 17:26:01 2002
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 9 Sep 2002 14:26:01 -0700
Subject: Binary input into NumPy.
References: <alil2o$mmk$1@news.tudelft.nl>
Message-ID: <alj3lu$kt5$1@main.gmane.org>

Maurice van de Rijzen:
> I just downloaded Numerical Python and I think it will be very usefull.
> However I have one major broblem:
> I don't know how to import binary data from a file on disk into an array!
> Can onyone help me with this problem?
>

look into Numeric.fromstring(open('filename', 'br').read(), numpytype)

HTH,

Emile van Sebille
emile at fenx.com








From mstenner at phy.duke.edu  Fri Sep  6 18:23:12 2002
From: mstenner at phy.duke.edu (Michael Stenner)
Date: Fri, 6 Sep 2002 18:23:12 -0400
Subject: how can i make a executable program with a python2's script
In-Reply-To: <3D7924BF.805@motorola.com>; from stephen.boulet@motorola.com on Fri, Sep 06, 2002 at 04:57:19PM -0500
References: <dd083388.0209060003.1188000c@posting.google.com> <3D7924BF.805@motorola.com>
Message-ID: <20020906182312.A15215@phy.duke.edu>

On Fri, Sep 06, 2002 at 04:57:19PM -0500, Stephen Boulet wrote:
> You can put
> 
>     #/usr/bin/env python
> 
> at the top of the file and do a chmod +x file. Is this the answer to the 
> right question?

Although since you're trying to use python 2.2.1 on a RHL 7.2 box, you
may need to make that

 #!/usr/bin/env python2

Otherwise, you'll get python 1.5.2, which can lead to confusing errors.

					-Michael
> jgm2163 wrote:
> > os     :RedHat 7.2
> > Python :Python2.2.1
> > 
> > now,i had writen a python's script (extend with .py)
> > how can i make a executable program on RedHat 7.2?
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

-- 
  Michael Stenner                       Office Phone: 919-660-2513
  Duke University, Dept. of Physics       mstenner at phy.duke.edu
  Box 90305, Durham N.C. 27708-0305



From bokr at oz.net  Sun Sep 29 21:05:17 2002
From: bokr at oz.net (Bengt Richter)
Date: 30 Sep 2002 01:05:17 GMT
Subject: watching mutables?
References: <an1p3c$jrd$1@news.hccnet.nl>
Message-ID: <an880d$ncp$0@216.39.172.122>

On Fri, 27 Sep 2002 16:12:05 +0200, anton at vredegoor.doge.nl (Anton Vredegoor) wrote:

>Hello all,
>
>here's a function I would like to have:
>
><hypothetical code>
>
>>>> m = [1,2]
>>>> m
>>>> [1,2]
>>>> from watcher import watch
>>>> watch(m)
>>>> m.append(3)
>>>> "hey, you changed variable m!, new value is [1,2,3]"
>
></hypothetical code>
>
>Is this possible? How would this be called?
>
I suspect you may have to adjust your concept of "variable"
to think about this in Python terms. I.e.,

    m = [1, 2]

doesn't result in a "variable" m with value [1,2]. It results in
an object (a list instance [1, 2]) and a name "m" in a particular
name space, and a relationship between the name and the object,
called binding.

This gives you several things that can change: the binding of "m"
and the mutable features of the object it is bound to.

Watching for a change in the binding of a name is possible, but you
will somehow have to give the watching mechanism information about the
actual name and name space (e.g., watch('m', vars()), not watch(m) )
and provide a trigger for the checking operation. If the name is an attribute
of an object that you can mess with, you may be able to monitor all
attribute name setting via __setattr__, but you might then miss changes
made some other way like obj.__dict__['m'] = whatever, so you probably
want to use some kind of tracing hook to get control for polling the status.

Watching for a change in the object being referred to by m is a different thing.
For that, you don't need to tell the watcher about the name per se ('m') unless
you want the name in a message like the (somewhat misleading) one you specified.
For change in the object, you also have to decide what "change" is supposed to mean.
I.e., is it based on '==' or what? Also, you have to get control to the checking
mechanism again, either by hooking into the object's mutation facilities, or an
independent polling trigger.

I'd be surprised if some solutions to your problem didn't already exist, and I
expect someone will chime in. I mainly wanted to point out that the "variable"
concept does not apply to Python the way it does to C or C++.

Nevertheless, here's something that appears to track a single named binding.
I use change in marshalled value as the criterion for change detection.
(not tested beyond what you see ;-):

----< Watcher.py >-----
class Watcher:
    from sys import settrace
    from marshal import dumps
    def __init__(self, name, namespace):
        self.name = name
        self.ns = namespace
        obj = namespace.get(name)
        self.objid = id(obj)
        self.objs = self.dumps(obj)
        self.settrace(self.watch)
    def watch(self, frame, event, arg):
        obj = self.ns.get(self.name)
        objid = id(obj)
        if objid != self.objid:
            self.objid = objid
            what = 'new'
        else:
            what = 'same but mutated'
        objs = self.dumps(obj)
        if self.objs != objs:
            print '"%s" is bound to %s object: %s' % (self.name, what, `obj`)
            self.objs = objs
        elif what=='new':
            what = 'new but equivalent'
            print '"%s" is bound to %s object: %s' % (self.name, what, `obj`)

        return self.watch
    def finished(self):
        self.settrace(None)
-----------------------
 >>> import Watcher
 >>> m=[1,2]
 >>> w = Watcher.Watcher('m',locals())
 >>> m.append(3)
 "m" is bound to same but mutated object: [1, 2, 3]
 >>> m=4
 "m" is bound to new object: 4
 >>> m=5
 "m" is bound to new object: 5
 >>> m=5
 >>> m=[6]
 "m" is bound to new object: [6]
 >>> m=[6]
 "m" is bound to new but equivalent object: [6]
 >>> m[0]=7
 "m" is bound to same but mutated object: [7]
 >>> m = m
 >>> x = m
 >>> m = [7]
 "m" is bound to new but equivalent object: [7]
 >>> m = x
 "m" is bound to new but equivalent object: [7]
 >>> x.append(8)
 "m" is bound to same but mutated object: [7, 8]

Notice that x.append(8) affected m's "value", but neither binding.

Regards,
Bengt Richter


From mmoales at fluent.com  Thu Sep  5 10:48:40 2002
From: mmoales at fluent.com (Mark Moales)
Date: Thu, 05 Sep 2002 10:48:40 -0400
Subject: How to force built-in commands over imported.
References: <3D761C39.4020802@akst.tn.tudelft.nl>
Message-ID: <3D776EC8.F84C8EF0@fluent.com>

Maurice,

How did you import the os module?  I'm guessing you did:

from os import *

If you just do:

import os

then when you invoke open, you'll be using the builtin function.  Of
course, all of your os function calls will need to be prefixed with
"os.", but you won't have anymore name clashes.

Mark

Maurice van de Rijzen wrote:
> 
> Dear all.
> I'm a newbie on the field of python, therefore the answer to my question
> may be simple.
> There is a built-in commans called open(fileName). There is also a
> open() command in the os-module open which needs more than the fileName.
>   Originally I didn't import the os-module and everything worked fine.
> However, when I imported the os-module I received an error that open
> needed more arguments. How can I force to take the built-in open()
> instead of the open() in the os-module.
> 
> Gr,
> Maurice
-------------- next part --------------
A non-text attachment was scrubbed...
Name: mmoales.vcf
Type: text/x-vcard
Size: 281 bytes
Desc: Card for Mark Moales
URL: <http://mail.python.org/pipermail/python-list/attachments/20020905/c1145556/attachment.vcf>

From aleax at aleax.it  Fri Sep 20 05:07:16 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 09:07:16 GMT
Subject: Avoiding referencing (was: Python sequences reference - problem?)
References: <mailman.1032443760.5578.python-list@python.org> <YZli9.114624$ub2.2439313@news1.tin.it> <amejvb$55qsi$1@ID-77524.news.dfncis.de>
Message-ID: <8zBi9.117743$ub2.2504724@news1.tin.it>

Thorsten Kampe wrote:

> * Alex Martelli
>> Just remembed that EVERYTHING works by reference, and what
>> could possibly be the problem...?
> 
> The problem is that we newbies sometimes don't want referencing:

Then you explicitly ask for a copy, just like an experienced
programmer would.

> #v+
> def part(seq, indices):
>     """ chop seq by indices """
>     partition = []
>     
>     for slice in indices:
>         partition.append(seq[:slice])
>         del seq[:slice]
>     partition.append(seq)
>     return partition
> #v-

If you don't want your function to affect its argument seq,
you start by making a copy.  Since you'd also like to ensure
you work with a list (whatever sequence or other iterable
you may have been passed), as you need a mutable sequence
for the "del seq[:slice]" statement to work, you can kill
two birds with one stone -- change the start to:

def part(seq, indices):
    """ chop seq by indices """
    partition = []
    seq = list(seq)

and leave the rest unchanged.  The function would be a little
bit clearer if the names involved made it more obvious that
the sequence you receive as an argument, and the local copy
you slice up, are different things, though -- e.g.:

def part(input_seq, chop_sizes):
    """ chop input_seq into a list of lists of given sizes """
    partition = []
    seq_copy = list(input_seq)

    for size in chop_sizes:
        partition.append(seq_copy[:size])
        del seq_copy[:size]
    partition.append(seq_copy)
    return partition

Choosing names of variables for maximum clarity and readability
is always a delicate issue, of course -- here, I've tried to
change things around a bit, and edit the docstring too, to make
it clearer that the second argument is not about indices at all,
but rather about sizes (lengths, if you will) of the sub-lists.

A completely different approach would be: 

def part(input_seq, chop_sizes):
    """ chop input_seq into a list of lists of given sizes """
    partition = []
    next_start = 0

    for size in chop_sizes:
        last_start, next_start = next_start, next_start + size
        partition.append(input_seq[last_start:next_start])
    partition.append(input_seq[next_start:])
    return partition

It's hard to choose between the two approaches, but I like
this second one a tiny little bit better, personally.  I'm
not sure why I do, to be honest.  I hope it's not because
of the slight possibility that its performance is superior
(e.g., it's O(N), not O(N squared), when chop_sizes degrades
to [1]*len(input_seq)...) -- I have not measured performance
nor do I have the slightest indication that this function
part has any need for optimization.  No, I just find it more
natural to avoid changing data structures except where such
changes make things more obvious to me, I guess.  Here, with
the second approach, I can get a _slightly_ more "FP" feel
(not all that much, since partition and the scalars do
change:-), which may have something to do with my preferences!-)

The second approach does lend itself to a slight optimization,
should it be needed, by preallocating the result to its
known final length rather than relying on append:

def part(input_seq, chop_sizes):
    """ chop input_seq into a list of lists of given sizes """
    partition = [None] * (1 + len(chop_sizes)]
    next_start = 0

    for i in range(len(chop_sizes)):
        last_start, next_start = next_start, next_start + chop_sizes[i]
        partition[i] = input_seq[last_start:next_start]
    partition[-1] = input_seq[next_start:]
    return partition

but this does seem to involve a tiny loss of clarity wrt the
second one (as "for i in range(len(..." so often does, sigh),
therefore I wouldn't use it unless performance WAS important
(and, of course, after measuring the various alternatives!!!).

A more FP alternative would involve building first a list of
indices from the list of sizes, e.g. (dirty trick warning...):

    start = 0
    chop_indices = [ start for x in [0]+chop_sizes 
                     for start in (start+x,)] + [None]

and then just using it:

    return [input_seq[chop_indices[i]:chop_indices[i+1]
            for x in range(len(chop_sizes))]

but this fails the base test -- it's not as obvious and
simple and the others, mostly because Python lacks an
"expand" function to get the chop indices from the chop
sizes (whence the "for start in (start+x,)" dirty trick).


But, back to your post...:


> Obviously I did not pass 'foo' "by value" to function 'part' but by
> reference and obviously not only local variable 'seq' was modified in
> the function but "outer" variable 'foo', too.

Rather, an OBJECT was modified -- and the two names are
referring to the same object (as you didn't ask for a copy).

> So how to deal with that?

By asking for a copy.

> 1. "Just don't call function 'part' by with a variable, call it by
> value - part([11, 22, 33, 44, 55, 66, 77], [2, 3]) - and everything's
> okay".
> 
> Obviously no solution to me.

Obviously!

> 2. "If you don't want referencing, call it with a copy of 'foo' like:
> part(foo[:], [2, 3])"
> 
> Yes, but how do I know in advance that 'part' modifies 'foo'? I might
> not be the author of 'part'.

Indeed, part should not modify its arguments UNLESS it says it does
in its docstring.  Unless there are very good reasons for a function
to modify its arguments (in which case the function should normally
return None -- mixing arg modification and significant return values
can be confusing!!!), avoiding such mods is the default case.


> 3. "Rewrite function 'part' so that it works with a tuple as argument.
> Thereby you're avoiding all possible modifying of existing objects."
> 
> Yes, but I'm also losing the "extended" possibilities of lists:
> "list.append", "list.reverse", "list.index" etc. My code will be twice
> as long as it would be with list methods.

I think it might be nice for part to accept as arguments sequences
that are not lists -- as you see in the first version I propose,
making a copy and ensuring you have a list can be fruitfully
combined in a single call to 'list'.  As you see in the second
version, working directly with the argument, without modifying
it, doesn't expand the code size AT ALL in this specific case,
anyway -- but the first version is slightly more general, in that
it will accept ANY argument that's acceptable to built-in 'list'
(e.g., a file -- list(afileobject) is like afileobject.readlines()),
while the second version does require a sequence (which must
accept being sliced AND supply copies when sliced -- a Numeric
array would accept being sliced, BUT would supply slices that
are arrays sharing data with the original array... this MAY be
a problem in some cases, although performance will surely be
enhanced by avoiding copies, if the data be truly huge).


> Is there a way to "bypass" referencing _without_ disadvantages? I'm

Absolutely not.  Making a copy takes time: there is absolutely
no way around this!  Therefore, there's no way to "bypass
_without_" ANY disadvantages at all: as a minimum, the time
and memory you spend for the copy are a disadvantage in some
cases.  But, of course, you DO pay those prices _whenever_ you
use a language with copy semantics.  That's part of why passing
of array arguments doesn't use copy semantics in most languages --
in Fortran because all argument passing is by reference (while
all assignment is by copy), in C because arrays in particular
"decay" to pointers when mentioned.  Implicitly copying huge gobs
of data each and every time an array is passed as an argument
would be an unmitigated performance disaster for most cases.

> sure there is a "canonical" way to do this in Python!

The canonical way to copy any object is function copy of module
copy (assuming a shallow copy suffices, as is often the case).

But often when you copy you ALSO want something else -- like
here, for example, in the first version (turn any sequence into
a list, as well as copying it).  Thus, don't blindly follow the
"canonical" way: THINK about what you want and how best to
obtain it -- sometimes copy.copy is OK, sometimes an all-sequence
slice x[:], but often calls to built-ins list or tuple, depending
on what, exactly, you're trying to accomplish with your copy.


Alex



From aleax at aleax.it  Thu Sep 19 03:31:47 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 19 Sep 2002 07:31:47 GMT
Subject: monetary applications (et al)
References: <mailman.1032383305.18515.python-list@python.org>
Message-ID: <D3fi9.112501$ub2.2399001@news1.tin.it>

<posted & mailed>

terry wrote:
        ...
> I do admit to being OO inhibited, so maybe you could take it from this
> short piece code and execution results, and, provide a brief
> description of what you would put into this money class that would
> eliminate this error message, while maintaining the code simplicity.
> 
> -------------------------------------------------
> class money:
> def __init__(self,name):
> self.name = name

It doesn't seem to me, from the example use below, that the
argument to money's constructor has anything to do with a
"name" -- rather, you seem to want to use it as an *amount*.
Therefore, call it 'amount', NOT 'name'.

Giving your variables (attributes, functions, classes,
methods, whatever) names that are widely at variance with
the variables' meaning is a horrible practice.  I'm not
sure what this has to do with your "being OO inhibited":
it would be a practice just as despicable in '60s-vintage
Cobol as in the spiffiest 21st-century language.

Then, it seems the money class needs to participate in
arithmetic as a number.  Thus, it clearly needs to define
the special methods that denote it's a number.  All in all:

class Money:
    def __init__(self, amount):
        self.amount = amount
    def __add__(self, other):
        return Money(self.amount + other)
    __radd__ = __add__
    def __mul__(self, other):
        return Money(self.amount * other)
    __rmul__ = __mul__

So far, class Money is just a wrapper over an 'amount' of
unknown type, simply delegating addition and multiplication
to the 'amount' (other arithmetic aspects work much the
same, except of course that they're not all commutative).
It's not hard for it to supply extra functionality that
you haven't requested but probably do want, e.g.:

    def __str__(self):
        return "$%.2f" % self.amount

so that, e.g.,
    print Money(100) * 0.5
would display
    $50.00

This still doesn't do anything particularly interesting
(but the __str__ method would of course in fact use the
locale settings to display monetary amounts in ways more
appropriate to the locale - NOW we're starting to add
some value, at last).  One probably also wants a __repr__:

    def __repr__(self):
        return "Money(%.2f)" % self.amount

More interesting may be doing something substantial
about the underlying arithmetic.  This class we're
sketching does arithmetic just like the underlying
type of amount -- it just pretties things up at output
time to display exactly two decimal digits, but there
is nothing about it that deals with the _accuracy_ of
those digits.  In any real financial application there
are constraints, sometimes legal ones, about what kind
of arithmetic must be computed when the true results
would be a fraction of a cent (or Eurocent or whatever
is the minimum legal fraction of money).

This class doesn't _know_ what the legal restrictions
are (as they'll no doubt differ from country to country,
at least, and more often than not by application area
too) -- but it makes it reasonably easy to insert such
knowledge when you DO have it!  For example, if you
work in an application area and jurisdiction where the
arithmetic must obey the standard rules of decimal
computation, you could simply fix the constructor to:

    def __init__(self, amount):
        import Decimal
        self.amount = Decimal.Decimal(amount)

and everything would fall in line as if by magic, thanks
to polymorphism.  It's pretty easy to customize class
Money to accept the underlying numerical class as a
parameter, too, e.g by inheritance:

    class FlexibleMoney:
        numberclass = float
        def __init__(self, amount):
            self.amount = self.numberclass(amount)
        # the rest is, again, unchanged

so now you could have, depending on jurisdiction and
application area, e.g.:

    import Decimal
    class Money(FlexibleMoney):
        numberclass = Decimal.Decimal

or else, again for example:

    import gmpy
    class Money(FlexibleMoney):
        numberclass = gmpy.mpq

(to respect the rules of unbounded-precision computation
via rational numbers), and so on, and so forth.


> unitprice = money(100)
> qty = 2
> amount = qty * unitprice
> 
> terry at terry:~> python money.py
> Traceback (most recent call last):
>   File "money.py", line 7, in ?
>     amount = qty * unitprice
> TypeError: unsupported operand type(s) for *: 'int' and 'instance'
> -------------------------------------------------

Right -- your class Money didn't supply __rmul__, so the
multiplication you tried to perform couldn't be performed.
The above examples do supply __rmul__, so this will work fine.


> Support for mixed number types in expressions involving this money
> class must come from somewhere - where?  And, it must be capable of

That's the nice thing about polymorphism: it doesn't matter where
the underlying arithmetic functionality is coming from, only that
it respect the rules you want it to respect (depending on legal
jurisdiction and application area).  gmpy wraps the GMP library,
so that's where the functionality comes from; Decimal is pure
Python, and you should import it from Aahz's side and inspect it
to gain some knowledge; and so on, and so forth.

> determing and setting amount to be of class money to be consistent with
> the rest of Python's complexity hierarchy for the handling of numbers.

That's reasonably trivial -- particularly now that (thanks be) you
don't NEED to support coercion any more (though gmpy's current
version still does support it, because it's meant to work with
older versions of Python too -- one day I'll do a "gmpy in a tie"
release and clean it up enormously, I promise:0).


> The rest of the problem is that the above is a trivial calculation.
> What happens to your recommendation when faced with an additional
> number type?

Nothing strange, if your underlying code is sound.

> taxrate = 0.06
> amount = qty * unitprice + (qty*unitprice)*taxrate

Works fine with the trivial class I sketched here, of course.

 
> As I see it, if Python had an intrinsic number type of 'money', the
> logic for determing the results of the above calculation would be
> easily determined and consistent with expectations, as in VB.

You see it wrong.  The law establishes what rules arithmetic must
follow in a lot of hairy cases, and no single intrinsic type can
follow all the rules for every jurisdiction and application, since
the rules DO differ.  In theory, you could go to jail in some
places because of the way you round certain fractions of a cent
(or whatever) in some application -- and the judge should not
accept "but VB does it that way" as covering your responsibility:
an application can be used to compute (e.g.) worker-compensation,
or tax liability, or actual annualized interest rates to be
posted as the law requires (in the EU), etc, etc, only if the
application is certified and carries the needed forms of sworn
affidavits about it.  When you sign a piece of paper swearing that
the application abides by whatever weirdness the law wants it to
abide by, you have to KNOW that it does.  In practice, white collar
crims is generally dealt with very leniently by the courts -- though
that might change in the current climate (but not in Italy, where
the government's main political purpose is ensuring that nobody
can possibly go to jail for financial shehanigans far darker than
these, since the prime minister and his closest associates would
be the prime candidates for any such jail terms if applicable...
so, ME, being an Italian citizen and resident, I'm reasonably
well covered in this regard... US citizens and residents should be
far more wary these days:-).

Note: my knowledge of such issues is second- and third-hand: the
student in financial economics in the family is my son, I'm an
engineer and the computation-constraining laws I know anything
about are rather those about how you're allowed to compute
tensile resistance of reinforced concrete and so forth.  But it
DOES make sense that the law should specify such issues, even
though it may not seem so when you're *not legally allowed* to
do arithmetic the same way for any of three different purposes.


> Thanks for your time,

You're welcome.  Better to SPEND some time considering such
issues in all needed detail, than to SERVE time later because
one didn't, right?-)


Alex



From emile at fenx.com  Mon Sep 16 22:58:48 2002
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 16 Sep 2002 19:58:48 -0700
Subject: Not sure how to phrase this question
References: <000501c25dde$79a8be60$9865fea9@arthur>
Message-ID: <am65m7$un6$1@main.gmane.org>

Arthur:
> Something that I guess can be called a simple class factory.  I have a
> routine that will analyze *args sufficiently to provide information as
to
> which class to be called from the ClassFactory to which *args is
> passed.
>
> In the end I want to create an instance of the called/returned class,
not of
> the ClassFactory class.

I don't think you need __new__, or if you do, you'll want to read up on
the existing documentation and prior discussions.  Oh... I couldn't find
the docs.  There are some usage examples in the demo/newmetaclasses, but
they only seem to exist in cvs.  All this probably has a bit to do with
the lack of response you mention.

Anyway, it sounds like you want to create an interface to, eg, allow
students to experiment with geometric properties.  To support this, you
want to create class instances that provide a flexible interface and do
not require arguments passed in an expected sequence.  ISTM a wrapper
class could do this trick.

class _Intersection:
    def __init__(self, plane=None, line=None):
        self.line = line
        self.plane = plane
class Plane:
    def __init__(self):
        self.ima = 'plane'
class Line:
    def __init__(self):
        self.ima = 'line'
class Intersection(_Intersection):
    def __init__(self, *args, **kwargs):
        for ii in args:  kwargs[ii.ima] = ii
        _Intersection.__init__(self, **kwargs)

#test
l = Line()
p = Plane()
ii = Intersection(l,p)
jj = Intersection(p,l)

print ii.line
print jj.plane

HTH, and if you keep trying metaclass stuff, keep us up on your progress
and findings.  ;-)

--

Emile van Sebille
emile at fenx.com

---------








From see_reply_address at something.invalid  Thu Sep 19 21:13:20 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Fri, 20 Sep 2002 13:13:20 +1200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <uokoorcumdrmc6@corp.supernews.com>
Message-ID: <3D8A7630.9090108@something.invalid>

A.M. Kuchling wrote:

> Note that the Parrot VM is already stackless, so a hypothetical Python
> running on Parrot would be stackless from the start.


How does it (or how is it envisaged that it will)
handle callbacks from C to interpreted code? Will
it remain stackless under those conditions?

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From spam at fisher.forestry.uga.edu  Sat Sep  7 15:22:16 2002
From: spam at fisher.forestry.uga.edu (Chris Fonnesbeck)
Date: 7 Sep 2002 12:22:16 -0700
Subject: returning values from FORTRAN with f2py
Message-ID: <c3835e5f.0209071122.338a787d@posting.google.com>

I am wondering how to persuade some fortran code that I have compiled
with f2py to return its calculated values to me in python.  I have
some code that does gaussian elimination (mostly from Numerical
Reciopes), and is supposed to return an inverted matrix so that I can
solve a system of linear equations.  However, when run in python, the
function runs without error but returns no values:

>>> a = [[2,1,3],[1,-1,1],[3,6,-1]]
>>> b = [7,2,10]
>>> import gauss
>>> gauss.gauss(a,3,b,1)
>>>

Is there something that I have to change in my fortran code to make
this work?

Many thanks,
cjf


From ark at research.att.com  Mon Sep  2 10:41:57 2002
From: ark at research.att.com (Andrew Koenig)
Date: Mon, 2 Sep 2002 14:41:57 GMT
Subject: stylistic question -- optional return value
References: <mailman.1030608587.4698.python-list@python.org> <30292041.0209012344.ab2d4@posting.google.com>
Message-ID: <yu99r8gco3m2.fsf@europa.research.att.com>

Sebastian> Why not create an additional optional *argument* (say a list, or a
Sebastian> function), to which you optionally provide your debug information?

It's not debug information.  I used that term only as an analogy.

Anyway, I don't want to separate the optional information from the
rest of it because everything needs to go into the same data structure
eventually.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From fperez528 at yahoo.com  Sat Sep 28 21:54:44 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Sat, 28 Sep 2002 19:54:44 -0600
Subject: Lambdaizing Python Code
References: <3d9559c4.14281625@news.t-online.de>
Message-ID: <an5mh4$7pr$1@peabody.colorado.edu>

Gerson Kurz wrote:

> A really usefull study ... NOT.
> 
> http://p-nand-q.com/e/quicksort_lambda.html
> 
> also known as "making Guido regret he put lambda in THE BEST
> PROGRAMMING LANGUAGE IN THE WORLD(TM)".

Now, as an excersise in sillyness it's funny. However, a more reasonable 
rewriting as a lambda can be found at:

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66473

And is a simple one-liner:

q=lambda x:(lambda o=lambda s:[i for i in x if cmp(i,x[0])==s]:len(x)>1 \
  and q(o(-1))+o(0)+q(o(1)) or x)()

Granted, for a quicksort I'd _much_ rather use a normal functional form. But 
this doesn't mean that a lambda form has to be the monstrosity depicted in 
that page :) I do take my hat off to whoever wrote it: just the fact that it 
works at all is an accomplishment in and of itself.

Cheers,

f.


From aglyportREMOvethispart at nospam.yahoo.com  Wed Sep  4 15:09:49 2002
From: aglyportREMOvethispart at nospam.yahoo.com (Anton Graph)
Date: Wed, 04 Sep 2002 12:09:49 -0700
Subject: Memory fault - core dumped   while compiling 2.2.1 on SCO
References: <3D757EFA.9050401@nospam.yahoo.com> <yu99sn0qqw1r.fsf@europa.research.att.com>
Message-ID: <3D765A7D.7000804@nospam.yahoo.com>

Andrew Koenig wrote:
> Anton> Any ideas?
> 
> Looks like the same symptom I had.
> 
> If you are using binutils 2.13, try going back to 2.12.1 and
> see if that fixes the problem.
> 

I don't think I use binutils:


Wed Sep  4 11:37:53 PDT 2002 bongo:/u5/src/Python-2.2.1 $ as --version
Assembler: (null)
  aline 1        : Warning: Illegal flag (-) - ignored
  aline 1        : Warning: Illegal flag (v) - ignored
  aline 1        : Warning: Illegal flag (e) - ignored
  aline 1        : Warning: Illegal flag (r) - ignored
  aline 1        : Warning: Illegal flag (s) - ignored
  aline 1        : Warning: Illegal flag (i) - ignored
Nonexistent file



Wed Sep  4 11:37:48 PDT 2002 bongo:/u5/src/Python-2.2.1 $ ar --version
ar: ERROR: Illegal option -- -
Usage: ar [-V?] -key[arg] [posname] afile [name ...]
   where key[arg] is one of the following: r[uabi], m[abi], d, q, t, p, x


Can't think of other utils in binutils.



From marklists at mceahern.com  Thu Sep 19 07:40:11 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 19 Sep 2002 06:40:11 -0500
Subject: StructuredText -> HTML module??
In-Reply-To: <wbii9.19869$0p1.359879@news2.ulv.nextra.no>
Message-ID: <JHEOKEOOLIGLDHCMAHMOCENLDPAA.marklists@mceahern.com>

[Thomas Weholt]
> Need a module to parse StructuredText/Wiki into HTML. Looked at
> PyWiki, but the generated code was not what I expected. Does anybody
> have module or lib to do this?

docutils has reStructuredText:

  http://docutils.sf.net/

// m

-




From sholden at holdenweb.com  Thu Sep 12 12:05:59 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 12 Sep 2002 12:05:59 -0400
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid> <alln71$1s2s$1@nntp6.u.washington.edu> <b16e4ef7.0209110616.48dfe04c@posting.google.com> <3D7FFC43.10002@something.invalid> <b16e4ef7.0209120504.62d3c5bd@posting.google.com>
Message-ID: <eZ2g9.191989$l_4.17068@atlpnn01.usenetserver.com>

"Stephen Ferg" <steve at ferg.org> wrote
> Thanks for the tips on where to find Human Interface Guidelines.
>
> > It's a fairly universal principle in Tk that the
> > "command" of a widget is invoked according to the
> > standard behaviour of that widget, so I think it
> > deserves treatment as a fundamental principle
> > on the same footing as event binding.
>
> Absolutely!  This observation is GOLD! This is just the kind of thing
> I'm trying to get at in "Thinking in Tkinter".
>
> Let me make sure I've got it right.  We have two fundamental, and
> different, approaches to specifying the behavior of a widget.
>
> (1) event binding, in which you get to "roll your" own bindings by
> binding callbacks to particular events that you specify.
>
> (2) command binding, in which a standard behavior for a widget is
> already defined for you by Tkinter (and Tk), and you bind a callback
> to that standard behavior.
>
> This explains why "command" binding doesn't automatically pass an
> event object the way that event binding does.  The standard behavior
> of a widget  might be defined across multiple kinds of events  (e.g.
> Button's standard behavior is defined across both a button-press and a
> button-release), so there is no *single* event that could be passed.
>

Correct. When you *want* the event data, of course (cursor position, etc.)
then bindings are a godsend. But the average programmer of end-user
applications gets most of what they need from a command callback.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From fgranger at alussinan.org  Fri Sep 27 12:49:56 2002
From: fgranger at alussinan.org (=?ISO-8859-1?Q?Fran=E7ois_Granger?=)
Date: Fri, 27 Sep 2002 18:49:56 +0200
Subject: combining mailbox and email modules
References: <NPck9.581867$UU1.104088@sccrnsc03> <slrnap41re.1mh.matt@happy-hour.t-island.com> <LAxk9.79207$gA4.27843@sccrnsc02>
Message-ID: <1fj6jb2.62mx08f47amsN%fgranger@alussinan.org>

Bob Horvath <nospam at horvath.com> wrote:

> The problem is that it looks good to me.  I've played around with 
> it, and if I take out the MIME gunk (It actually is a Klez virus), 
> it parses ok.

I have had difficulties with the handling of multipart message with the
email package. I did not found a solution yet.

-- 
Ce qu'il y a d'emmerdant dans la morale, c'est que c'est toujours
la morale des autres - L?o F?rr? - Cit? par No?lle Adam


From cjw at sympatico.ca  Wed Sep  4 12:27:21 2002
From: cjw at sympatico.ca (Colin J. Williams)
Date: Wed, 04 Sep 2002 12:27:21 -0400
Subject: MLab in Numeric.py a possible bug
Message-ID: <3D763469.F51A60B2@sympatico.ca>

Line 296 in the function cov is:

    val = squeeze(dot(transpose(m) * conjugate(y)) / fact)

and gives the error message:
    
    ValueError: frames are not aligned

dot is expecting two arguments, but does not complain when one is
provided.

m and y are arrays, replacing them by matrices does not appear to
resolve
the problem as, with matrix m, transpose(m) is an array, not a matrix.

Providing dot with two arguments, array or matrix, seems to resolve the
problem.

    val = squeeze(dot(transpose(m), conjugate(y)) / fact)



From jafo at tummy.com  Wed Sep 18 13:30:46 2002
From: jafo at tummy.com (Sean Reifschneider)
Date: Wed, 18 Sep 2002 11:30:46 -0600
Subject: New RPMs at python.org
In-Reply-To: <JHEOKEOOLIGLDHCMAHMOOEJHDPAA.marklists@mceahern.com>
References: <20020918171900.GB13272@tummy.com> <JHEOKEOOLIGLDHCMAHMOOEJHDPAA.marklists@mceahern.com>
Message-ID: <20020918173046.GL1910@tummy.com>

On Wed, Sep 18, 2002 at 12:24:06PM -0500, Mark McEahern wrote:
><quote>Installs "python2.2" as well as "python2.2".</quote>
>                       ^^^                    ^^^
>
>They are the same, no?

That should be "python2.2" and "python2".

Also, I had to stop the presses for a moment, because of an issue I
noticed.  The files are all mode 000 right now, they'll be back to normal
in a few minutes.

Sean
-- 
 I find that a great part of the information I have was acquired by looking
 up something and finding something else on the way.  -- Franklin P. Adams
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com, ltd. - Linux Consulting since 1995.  Qmail, Python, SysAdmin



From peter at engcorp.com  Fri Sep  6 23:06:35 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 06 Sep 2002 23:06:35 -0400
Subject: Interrupting Python
References: <mailman.1031345649.9840.python-list@python.org> <Ef9e9.7309$rU2.218110@news4.srv.hcvlny.cv.net> <albel2$7te$0@216.39.172.122> <R3ce9.9745$rU2.313350@news4.srv.hcvlny.cv.net>
Message-ID: <3d796d22@news.sentex.net>

Bob Easton wrote:
> Using catchCtlC.py in a command line environment works as you expect. When I
> hit Ctrl-Break, Windows displays "^C" and then a command prompt.  If I hit
> Ctrl-C, the exception handler catches it and displays the results as
> specified in catchCtlC/py
> 
> However, my preferred execution environment is PythonWin, not a command
> prompt. There, neither Ctrl-break, not Ctrl-C have any effect.  That could
> easily be a bug with PythonWin.

Not a bug, I think.  The whole Ctrl-C/Ctrl-Break concept is only
meaningful for command-line ("console") programs, not GUI apps.
I don't use PythonWin, but my guess is it does not bother passing
ctrl-c on to the application, possibly by design.

Can't you just run it under the console?  What advantages do
you find by running it under PythonWin other than during
development?

Also, have you looked at the msvcrt module and kbhit() ?

-Peter



From tebeka at cs.bgu.ac.il  Wed Sep 18 02:47:25 2002
From: tebeka at cs.bgu.ac.il (Miki Tebeka)
Date: 17 Sep 2002 23:47:25 -0700
Subject: Loading modules from several directories
References: <a1247c5b.0209171610.243f3c02@posting.google.com>
Message-ID: <33803989.0209172247.2e6afc1f@posting.google.com>

Hello Eugene,
> Trying to reuse the code I found a problem (with Python or me) where
> if I have for example, a foo package in two separate directories
> (where foo/ contains common files and foo/project_a/ contains
> project-specific files) I can't load from both.
> 
> If I put the common path first in the sys.path it tries to load the
> project-specific stuff from there too. If I try putting the
> project-specific one first it can't find the common stuff.
> 
> Anyone know of a good solution for this problem?
Doesn't

sys.path.append('/somedir/foo')
sys.path.append('/somedir/foo/project_a')
import foostuff
import projectstuff

works?

Miki


From aleax at aleax.it  Wed Sep 25 06:18:01 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 10:18:01 GMT
Subject: general thoughts
References: <ifjloug0dkla5o3lvli464mmou5mbh47d5@4ax.com> <L6Gi9.119475$ub2.2533804@news1.tin.it> <3D8B3250.BA964BCF@ObjectWay.it> <amq6e5$85gpo$1@ID-139865.news.dfncis.de>
Message-ID: <t3gk9.163574$pX1.5850119@news2.tin.it>

Dave Kuhlman wrote:
        ...
>>> I think you need substantial capacity and appetite for abstract thought
>>> to really _enjoy_ SICP -- it doesn't spoon-feed you, nor does it use
        ...
> This book is a huge time sink.  One section shows you how to fake object
> oriented programming in a dialect of lisp with no O-O support.  After
> working through that section, I asked myself, Why do I need to know this?
> What good is it going to do me, unless I'm using a lisp dialect that does
> not support object-oriented programming?

People who like to think abstractly generally also like to have
multiple switchable mental models of how something like method
dispatching could be implemented.  People who prefer concrete
thinking generally want ONE such mental model, preferably the
exact one they're dealing with at a given time.

That's NOT a put-down of either kind of people - we NEED both
kinds, and there's a synergy between the two types when put in
the same time, although clashes are quite possible too.

I'd never suggest SICP to somebody who prefers concrete thought.
It's not a matter of more or less intelligence, please note, but
of the WAY one's mind prefers to work.


> And, a large part of the book describes how to program without variables
> and
> state.  That's weird and interesting, just like the Elephant Man was weird
> and interesting.  But, after that, then what?  Am I supposed to try to
> program that way?

You can choose to -- or you can choose to add another capability
of abstraction to your panoply of such.  You probably already
have a mental model of computation WITH side effect -- now you've
added another, of computation WITHOUT side effect.  Python happens
to have just enough functional flavor (though GvR doesn't like
that:-) that this specific capability has some use.  Actually
(see Coplien's "Multi-paradigm DESIGN for C++") a reasonable
collection of mental models for highly varied paradigms can come
in useful even when you might not think so.  Procedural, OO,
functional, relational, logical/constraint ... they all have their
place even if you're forced to code in machine code, and so do
the various subflavors (e.g., the many ways in which OO can be
approached -- with or without multiple dispatching, the many
ways to model and implement dispatching, etc).

For example, often you build little languages for task specific
issues, embedded in whatever other languages you're programming
with.  Scheme and Lisp are quite good for that, but you can and
do perform such activities in any language.  Then, having some
model of paradigms you can use and how to implement them is nice.


> What is the lesson to be learned from "Structure and interpretation of
> Computer Programs"?  Maybe it's main purpose is to provide an enjoyable
> read for those who have still not given up on lisp.

It's not about Lisp -- it's about mental models of computation.

SICP's purpose is definitely not to give concrete, usable-today
tips.  It's to broaden your horizons, enriching your panoply of
mental models about computation.  That's why it's no good for
people who like to think concretely, and why the "gurus like this
book" diagnosis is unfair to many brilliant people who _don't_.


> Seriously, what was I supposed to have gained from all that time I spent
> on
> this book.  As you may suspect, it captivated me, too.

Aha -- this sounds like somebody WITH appetite for abstract thought
(else you wouldn't have liked it) subject to cultural / peer pressures
towards concreteness (else you wouldn't feel that broadening your
horizons, without learning anything immediately-usable, might be
a waste of time).


> It did not teach me good Python programming style.  I could not hardly
> read my own code with all those many layers of parentheses.

True, alas -- lispish concrete syntax [expletive deleted] -- expect
mild flames from those who like it, of course:-).

> It did not teach me how to use Python data structures.  I found myself
> doing too many screwy things with lists.

Python's lists are definitely not what you want as a direct
replacement for lisp/scheme ones -- you can build the latter
with two-item tuples (if immutable) or lists (if mutable) a la:

class lispnode(object):
    __slots__ = '_car', '_cdr'
    def __init__(self, car=None, cdr=None):
        self.car = car
        self.cdr = cdr
cons = lispnode

and any embellishments you want to iterate on lispish lists made
of such nodes, form such a lispish list from any Python iterable, &c.


> I was supposed to learn something more theoretical, right?  What was it?

See above.


> Or, was this book just a devious way to trick more people into using lisp?

Naah... not in the concrete-syntax way.  It's just that some
concepts are fruitful and usable even in other languages.


Alex



From steve at ferg.org  Fri Sep 13 08:43:01 2002
From: steve at ferg.org (Stephen Ferg)
Date: 13 Sep 2002 05:43:01 -0700
Subject: ANN: Thinking in Tkinter
References: <b16e4ef7.0209062009.26e218e2@posting.google.com> <3D798461.B3FE0E06@earthlink.net> <b16e4ef7.0209070401.16087fd@posting.google.com> <3D7C210F.7090708@something.invalid> <b16e4ef7.0209090552.3ad7691b@posting.google.com> <V57f9.77078$l_4.29066@atlpnn01.usenetserver.com> <b16e4ef7.0209101725.6f508364@posting.google.com> <3D7FF299.2020509@something.invalid>
Message-ID: <b16e4ef7.0209130443.1f282bd7@posting.google.com>

Greg Ewing <see_reply_address at something.invalid> wrote in message news:<3D7FF299.2020509 at something.invalid>...
> It might be better to start with some object
> other than a button for the purpose of introducing the
> concepts of event binding. Otherwise you're teaching
> people an incorrect way of using buttons that they're
> going to have to unlearn later.

Thanks for your help in other messages in this thread.  

After thinking it over, my current plan is still to start with Button,
just because it supports both event binding and command binding.  And
also, of course, because Button is the obvious choice for the first
widget to learn.

So my plan now is first to explain event binding on Button.  Then I'll
explain command binding on button, and do a "compare and contrast"
with event binding:  discussing why command binding is better than
event binding, noting that it does NOT automatically pass an event
object, that it may be available with other widgets than Button, etc. 
And I'll be careful in the earlier discussion to note that even though
we're looking at event binding first as a way to introduce the concept
of binding, we'll also soon be looking at a better way (i.e. command
binding).


From ant at mongrel.dogpound  Sun Sep  1 17:57:39 2002
From: ant at mongrel.dogpound (Anthony Ventimiglia)
Date: Sun, 01 Sep 2002 21:57:39 GMT
Subject: Converting Python app to C++ completely
References: <aklm6c$r1r$1@news.swman.net.uk> <mailman.1030644868.1322.python-list@python.org> <aknnbt$212$2@news.swman.net.uk>
Message-ID: <slrnan53ak.11n.ant@mongrel.dogpound>

On Fri, 30 Aug 2002 13:09:48 +0100, Alan James Salmoni 
<salmonia.nospam.please at cardiff.ac.uk> wrote:
> Sean 'Shaleh' Perry wrote:
>> In general C++ will be more verbose than Python.  Using references and the STL 
>> it is possible to code fairly complex applications without using pointers but 
>> most likely you will have to learn about them and use a couple.
> 
> I think that's what worries me - luckily, my code is fairly basic so I 
> might be able to avoid them as much as possible, but I should learn 
> about them really.


Don't be afraid of pointers, they're really not that bad. The STL really
allows you to do quite a bit without having to directly do any memory
allocation.

I'll also add another recommendation for reading Stroustrup. 


From magnus at thinkware.se  Fri Sep 20 06:37:19 2002
From: magnus at thinkware.se (=?ISO-8859-1?Q?Magnus_Lyck=E5?=)
Date: Fri, 20 Sep 2002 12:37:19 +0200
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <r58ama.lkf.ln@demonseed.net>
Message-ID: <3D8AFA5F.4030205@thinkware.se>

jason petrone wrote:
> A few other polished programs written in python:
> 
>   Visual Python: http://www.activestate.com/Products/Visual_Python/
...
> Visual Python probably has the prettiest screenshot.

Stop there! Visual Python is a python plugin for Microsoft
Visual Studio, right? So what you see is Visual Studio,
which is AFAIK written in C++ with MFC. Or am I in deep
waters here? I think it's written _for_ python, not _in_
python.



From amk at amk.ca  Sat Sep 21 20:32:51 2002
From: amk at amk.ca (A.M. Kuchling)
Date: Sun, 22 Sep 2002 00:32:51 -0000
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <uokoorcumdrmc6@corp.supernews.com> <3D8A7630.9090108@something.invalid>
Message-ID: <uoq3tjg4tnn945@corp.supernews.com>

In article <3D8A7630.9090108 at something.invalid>, Greg Ewing wrote:
> How does it (or how is it envisaged that it will)
> handle callbacks from C to interpreted code? Will
> it remain stackless under those conditions?

I have no idea, and would imagine that it hasn't been designed yet.
Looking at how Perl5 works might provide some guidance, because it'll
probably look similar in Parrot.

(Looking at the perlcall.pod document, which describes how to call
Perl5 code from C, it looks like callbacks from C are not stackless,
because you just call the call_sv() function which will return the
result.)

--amk


From aleax at aleax.it  Thu Sep 26 07:54:54 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 26 Sep 2002 11:54:54 GMT
Subject: CGI server by Python
References: <mailman.1033036294.12279.python-list@python.org>
Message-ID: <iACk9.145383$ub2.3204425@news1.tin.it>

Brian Lee wrote:

> Hi,
> 
> I tested CGIHTTPServer.py in Python 2.2 and found that cgi files in
> sub directory does not run. Does anybody know about this?
        ...
> avante.xxx.co.kr - - [26/Sep/2002 19:20:24] code 403, message CGI script
> is not a plain file ('/cgi-bin/a') avante.xxx.co.kr - 

Exactly as it says, /cgi-bin/a is not a plain file (it's a directory!)
and CGIHTTPServer.py is written to support plain files right under
cgi-bin only.  See the parsing in function run_cgi around line 100
and following of that file in your Python library directory.  If you
don't like the way it parses things, it ain't hard to modify...


Alex



From zopestoller at thomas-guettler.de  Tue Sep 24 06:41:00 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Tue, 24 Sep 2002 12:41:00 +0200
Subject: convention for Pickler file extension
References: <918bc22f.0209230208.4eaca702@posting.google.com> <svEj9.226013$z91.9600123@bin3.nnrp.aus1.giganews.com>
Message-ID: <ampfbe$rq0$07$2@news.t-online.com>

Terry Reedy schrieb:
> "Donnal Walter" <donnal at donnal.net> wrote in message
> news:918bc22f.0209230208.4eaca702 at posting.google.com...
> 
>>When one uses Pickler.dump(object) to serialize an object to a file
>>associated with the Pickler, is there a Python convention for the
> 
> file
> 
>>extension name?
> 
> 
> If there is not currently, I think .pyp is available.

I call files *.pickle

  thomas



From bogle at ihug.co.nz  Fri Sep  6 19:32:57 2002
From: bogle at ihug.co.nz (Gib Bogle)
Date: Sat, 07 Sep 2002 11:32:57 +1200
Subject: Python from C/SDL
Message-ID: <3D793B29.3198AB0E@ihug.co.nz>

Hi,

I am a complete beginner with Python.  My interest in it is focussed on
a great piece of open-source software called pymol, a system for
visualization of molecules that uses Python and OpenGL.  I'd like to get
access to pymol's graphical functions from my own program, which is
built from C, SDL and OpenGL.

I'm wondering if (with a bit of surgery) I'd be able to get access to
the lower-level functionality of pymol, by which I mean the file I/O and
rendering functions.  I will not want to use the GUI at all.  SDL would
create the surface(s), and handle the buffer swapping.  My goal is to be
able to send a request to the pymol API asking for a frustum view to be
rendered on the provided surface.

Sorry if my question is one that gets asked and answered frequently.  If
there is a place where this has been dealt with please direct me.

thanks
Gib

--------------------------------------------
Gib Bogle              bogle at ihug.co.nz
1/44 Arthur St         Tel: (64-9) 525-6878
Ellerslie, N.Z.        Fax: (64-9) 525-6878
--------------------------------------------


From eric.brunel at pragmadev.com  Tue Sep  3 03:51:58 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Tue, 3 Sep 2002 09:51:58 +0200
Subject: what about things like __*** ?
References: <7153651c.0209021927.16a3efaf@posting.google.com>
Message-ID: <al1p8n$12d$1@wanadoo.fr>

black wrote:
> Hi, all~
> 
> when reading the tutorial I met some trouble.(Maybe trifle to you all
> but really trouble to me)
> 
> That is I couldnt figure out what __*** is. Here is the description of
> that article but even confused me:
> 
> There is limited support for class-private identifiers. Any identifier
> of the form __spam (at least two leading underscores, at most one
> trailing underscore) is now textually replaced with _classname__spam,
> where classname is the current class name with leading underscore(s)
> stripped.
> 
> What's it trying to say please ???

As a good practice in object-oriented programming, the structure of a 
class, i.e. its actual attributes, are often hidden from the outside world. 
It's called "encapsulation". It helps to separate the interface of the 
class from the actual implementation of the methods, that shouldn't be 
known outside the class itself.

Languages like C++ or Java have a native support for encapsulation, by 
allowing to declare attributes "private" or "protected". A value of a 
private or protected attribute can't be read or set by another class than 
the owner class (or one of its sub-classes for protected attributes).

IIRC, before Python 2.0, there were no support at all for encapsulation in 
Python. So you could always do:

class C:
  def __init__(self):
    self.a = 12
o = C()
o.a = 13
print o.a

The attribute "a" of instances of class C can be read or modified outside 
the class, and there's no (simple...) means to prevent that.

>From Python 2.0, there's a (limited) support for encapsulation: if the name 
of an attribute begins with a double-underscore, it's just as if it had 
been declared private:

class C:
  def __init__(self):
    self.__a = 12
  def get_a(self):
    return self.__a
o = C()
o.__a = 13       # Does not work
print o.__a      # Again, does not work
print o.get_a()  # Works

class CC(C):
  def __init__(self):
    C.__init__(self)
    print self.__a    # Does *not* work

The attribute __a is private: it can be seen only in the class C. It's 
impossible to read or set it from outside the class, even in one of its 
sub-classes. This sounds like a big restriction, but is actually very often 
used in OO programming.

The actual mechanism involves a "name-mangling", and that's what the 
comment tried to explain. But it's not that important: the final result is 
just that any attribute beginning with a double-underscore is private.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From loewis at informatik.hu-berlin.de  Mon Sep  9 03:50:57 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 09 Sep 2002 09:50:57 +0200
Subject: Embedding - Local dictionaries for local scripts
References: <1691b372.0209060746.55b8cf1@posting.google.com>
Message-ID: <j4k7lvbnz2.fsf@informatik.hu-berlin.de>

chris.walsh at real-sense.com (Chris Walsh) writes:

>     PyObject* PyRun_String(
>         char *str,
>         int start,
>         PyObject *globals,
>         PyObject *locals);
> 
> I have seen some stuff in comp.lang.python about creating a dict and
> adding the __builtin__ module but I have been searching now for an
> hour and cannot find it.
> 
> Can anyone point me to some code which creates this dictionary?

The globals/locals dictionary is not really different from any other
dictionary; you create them with PyDict_New. Python should
transparently add an __builtins__ into the globals dictionary if there
is none, but you can copy the __builtins__ entry just as well yourself
(via PyDict_SetItemString).

HTH,
Martin



From sschwarzer at sschwarzer.net  Thu Sep 12 16:01:32 2002
From: sschwarzer at sschwarzer.net (Stefan Schwarzer)
Date: Thu, 12 Sep 2002 22:01:32 +0200
Subject: "str.contains(part)" or alternatives?
References: <alo8jr$l04$1@news.online.de> <mailman.1031776875.28219.python-list@python.org>
Message-ID: <alqrnc$kg7$1@news.online.de>

Hi Skip

Skip Montanaro wrote:
>     >>> try:
>     ...     s.index('ll')
>     ...     do_something_if_found()
>     ... except ValueError:
>     ...     do_something_if_not_found()
> 
>     Stefan> This has the problem that a ValueError raised in
>     Stefan> do_something_if_found() may give the false impression that the
>     Stefan> substring isn't contained in the string s. 
> 
> I try to minimize the amount of code executed in try blocks.

Me too. That above was only a (bad) example. ;-)

 > Use this instead:
> 
>     try:
>         s.index('ll')
>     except ValueError:
>         do_something_if_not_found()
>     else:
>         do_something_if_found()

I just read today in the Python Cookbook that this syntax is valid. I never saw
it in Python code (not meaning that I believe it doesn't exist :) ).

As a side note, I remembered then that there are also "else"-Forms of the while
and the for loop which are executed when the loop is left via break (IIRC).

Thank you (and all others) for their answers. I didn't expect to get so many.
Btw, Greg, the .count approach is much nicer than ".find != -1".

Stefan



From fredrik at pythonware.com  Wed Sep  4 18:13:43 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 04 Sep 2002 22:13:43 GMT
Subject: How to force built-in commands over imported.
References: <3D761C39.4020802@akst.tn.tudelft.nl> <slrnanc7jg.24k.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <rAvd9.6092$e5.1068320@newsb.telia.net>

Gerhard H?ring wrote:

> The technical answer is to use __builtins__["open"].

that should be:

    import __builtin__
    file = __builtin__.open(...)

__builtins__ is an implementation detail.

</F>




From lu_gio at hotmail.com  Tue Sep  3 19:19:00 2002
From: lu_gio at hotmail.com (Max)
Date: Tue, 3 Sep 2002 23:19:00 +0000 (UTC)
Subject: Database BSD on Mac?
Message-ID: <3144be3c280c062e5e71a98ab191f664.98737@mygate.mailgate.org>

Hi!
I'm implementing a simple data manager to store few images and sounds
(<5Mb).
I'm developing a game on windows platform (so, I don't need a client
server db like mysql) but I'd like to port the software on every python
supported platform (well, principally on win,linux and mac).

I was looking on BSD but I'm not sure it is ported to all platforms.
Wich db should I utilize?


-- 
Posted via Mailgate.ORG Server - http://www.Mailgate.ORG


From cbarber at curl.com  Thu Sep 19 13:01:05 2002
From: cbarber at curl.com (Christopher Barber)
Date: 19 Sep 2002 13:01:05 -0400
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <bd751532.0209082142.54236307@posting.google.com> <allrrs$6hm$2@solaris.cc.vt.edu> <ambpgq$dre$1@nntp9.atl.mindspring.net> <mailman.1032437557.24458.python-list@python.org>
Message-ID: <psovg51sylq.fsf@bandit.curl.com>

"Chris Gonnerman" <chris.gonnerman at newcenturycomputers.net> writes:
> No smiley this time... I really hate typing curly after curly.
> It makes me wonder that anyone *likes* the freakin' things so
> much that they named a LANGUAGE after them.

The Curl language was indeed named after curly braces, but not because we like
them so much but because the nature of the language required some sort of
bracketing delimiter punctation and the curly brace seemed a good choice for
this.  One of the points of the language is to make it easy to do powerful
text markup and otherwise mix code and text.  To support this, you really need
some way to distinguish code from text.  For instance:

   {paragraph Hello {bold Bold} World!}

Although it might be annoying to have to type the curly braces, I really don't
know what the alternative would be. Parens and square brackets are used much
more frequently in text than curly braces and would have to be escaped more
frequently, and SGML syntax is much worse.  I don't think there is a way to do
this kind of thing using Pythonic syntax.

- Christopher


From chironsw at swbell.net  Thu Sep 19 14:06:57 2002
From: chironsw at swbell.net (Greg Goodman)
Date: Thu, 19 Sep 2002 18:06:57 GMT
Subject: monetary applications (et al)
References: <mailman.1032390744.1503.python-list@python.org> <bsei9.130231$pX1.4674970@news2.tin.it> <mailman.1032443727.4727.python-list@python.org>
Message-ID: <3D8A161E.8060906@swbell.net>

> Out of curiousity, does python have any standing in the field of
> business accounting?  Perhaps those with experience can chime in as to
> how this has been handled effectively.  We learned long ago to keep all
> amounts in whole pennies.


The original poster was asking about "currency", which implies a need to 
  store, and possibly to convert between, different currencies.  A 
generic currency class could reasonably store values as long integers, 
representing the smallest indivisible unit within a given currency (i.e. 
American cents, English pence, French centimes, Israeli agorot, Afghani 
puls, etc), along with the number of subunits in the whole (usually 100, 
but there are exceptions, such as 1000 fils in an Iraqi dinar, or 5 
khoums in a Mauritanian ouguiya).

On the other hand, cents aren't really indivisible in the American 
marketplace, where stock prices often change by fractions of a dollar, 
as in "General Motors down 2 and 1/8".

As always, the class you build is determined by the use you to which you 
intended to put it.

Greg Goodman



From jb at cascade-sys.com  Fri Sep 20 00:51:25 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Thu, 19 Sep 2002 21:51:25 -0700
Subject: e vs exp()?
References: <3a838998.0209191933.1f22e977@posting.google.com>
Message-ID: <3D8AA94D.1080801@cascade-sys.com>

Buckshot wrote:

>Can anyone here explain why I'm getting different results from
>e**2 and exp(2)?  Which is more accurate?  
>
>Python 2.2 (#1, Mar 27 2002, 14:56:58)
>[GCC 2.95.3 20010315 (release)] on sunos5
>Type "help", "copyright", "credits" or "license" for more information.
>  
>
>>>>from math import *
>>>>e**2
>>>>        
>>>>
>7.3890560989306522
>  
>
>>>>exp(2)
>>>>        
>>>>
>7.3890560989306504
>  
>
Mathematica says e**2 to 30 digits is:

    7.3890560989306502 272304274606

in which case the second form would happen to be more accurate.

Since Python math uses your machine's native floating point formats, 
precision is necessarily limited and one should not be surprised to 
encounter minor errors in the least significant bits or digits.

Although Mathematica recognizes that the two forms are identical, Python 
uses different formula for computing the two results.  The fact that the 
base is "e" is of no consequence to the computation.  exp() is computed 
from C's exp() runtime library routine, whereas e**2 is evaluated using 
C's "pow()" function.  

In general, I would expect pow() to be less accurate, as pow(x,y) 
probably is computed by exp(log(x)*y).

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From observer at NOSPAM.space.pl  Sun Sep 22 15:30:50 2002
From: observer at NOSPAM.space.pl (Johann)
Date: Sun, 22 Sep 2002 21:30:50 +0200
Subject: Problems with starting Python IDE
Message-ID: <e86souol60mh0eddeffbjcfsa47gk85t92@4ax.com>

I cannot I "Python IDE" on my PPC 7600 because of exception errors as
the following:

"Python 2.2.1:Mac.Tools:IDE:Splash.py, line 51, in Update.Splash
splash.GetDialogWindow().ValidWindowRect(splash.GetDialogPort().portRect)
NotImplementedError: Not available in this shared library/OS version"

What is wrong? I am using clean install of os 8.1. PPC7600/200MHz,
256MB RAM. It should work.... I got no warnigs nor errors during
instalation.

--
Johann


From neal at metaslash.com  Wed Sep  4 17:43:23 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Wed, 04 Sep 2002 21:43:23 GMT
Subject: Strange threads/stack size interaction on FreeBSD
References: <3D765A95.6090308@tu-clausthal.de>
Message-ID: <pan.2002.09.04.21.43.19.620760.2752@metaslash.com>

On Wed, 04 Sep 2002 15:10:13 -0400, Stefan Schwarzer wrote:

> OS: FreeBSD 4.6.2
> Python: 2.2.1 (built from the ports collection) Hardware: Dual Processor
> Athlon, 1 GB RAM
> 
> I've searched the Python FAQ, Google Groups and the bug database of the
> Python project on this, but without success, so I have some questions.
> 
> - Has anybody here experienced something similar, perhaps on other
>    platforms?

Doesn't sound familiar.

> - Is Python's threading code dependent on a stack limit set at compiled
>    time (seems so)?

Not that I know of.

> - What are your recommendations to deal with the problem?

You can try reducing the recursion limit:

  import sys
  sys.setrecursionlimit(100) # default is 1000

> - Should I enter a bug report on this topic?

Probably, although, there are few people who can test with your
configuration.  So you may be stuck debugging this yourself. 

If you have stack traces, are they very long?  where does the crash
occur?  What compiler are you using?  Can you report all the tests
that don't pass?

Have you tried with the latest and greatest source?
  2.2.1+: cvs update -r release22-maint
  2.3:    cvs update -A

Neal


From see_reply_address at something.invalid  Wed Sep 25 22:12:18 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Thu, 26 Sep 2002 14:12:18 +1200
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com> <3D8E6353.7030303@something.invalid> <96zj9.150553$pX1.5399868@news2.tin.it> <3D911CFA.4060405@something.invalid> <Wefk9.163287$pX1.5841874@news2.tin.it>
Message-ID: <3D926D02.4000503@something.invalid>

Alex Martelli wrote:

> (take it up with KDE KNode's maintainers whether their news
> reader is "dumb"... I like it, though it doesn't show tabs:-).


Apologies -- I didn't mean to denigrate any piece of
software with the term. I was just using it as
shorthand for (in this context) "not having any
features provided for customising the display of
tabs".

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From pereira at cis.upenn.edu  Tue Sep  3 20:09:33 2002
From: pereira at cis.upenn.edu (Fernando Pereira)
Date: Tue, 03 Sep 2002 20:09:33 -0400
Subject: Would you like a Snobol pattern matcher? (Re: Making regex
	suck less)
References: <3d725881.345921@news.t-online.de> <un7n06qrl1gm93@news.supernews.com> <3D740DFC.1040009@something.invalid> <un9bkdrqeh305c@news.supernews.com> <al30tq$aud$1@solaris.cc.vt.edu>
Message-ID: <B99AC77D.11E96%pereira@cis.upenn.edu>

On 9/3/02 2:13 PM, in article al30tq$aud$1 at solaris.cc.vt.edu, "Carl Banks"
<imbosol at vt.edu> wrote:
> I think there is a good reason to have a higher level syntax.  It
> serves a purpose heretofore unserved: that complex regular expressions
> no longer have to be unbearably unreadable.
The stronger version of this point is that at the moment there are no
methods for composing complex expressions from simpler ones that do not
involve manipulating strings, like building complex Python programs by
taking their pieces as strings, gluing them together and passing the result
to eval. Clearly a mess. For a good example of an alternative approach (in
Scheme) see

http://www.ai.mit.edu/people/shivers/sre.txt

The beauty of this kind of approach is that the expressive power of the host
language is fully available in constructing complex expressions.

-- F



From gtaylor at lowebroadway.com  Fri Sep 27 07:34:37 2002
From: gtaylor at lowebroadway.com (Garry Taylor)
Date: 27 Sep 2002 04:34:37 -0700
Subject: Unicode (inc. Japanese) in Python and TK
Message-ID: <f0fd5987.0209270334.47cbd9c0@posting.google.com>

Hi,
I'm just seeking a little reassurance regarding use of unicode in
Python, before I write my program, and then later find out I'm going
to have to re-write in Java or something. Basically, am I going to run
into any brick walls by doing the following:

1) Parsing XML with unicode characters.
2) Measuring unicode characters using TK
3) Printing unicode to a file

Really, I just want someone to pat me on the back any say 'everythings
going to be alright'. The reason I am sceptical is I have used Perl
for similar antics, and now now I wake up screaming, lying in a pool
of my own sweat.

Thanks, sorry if this is a FAQ

Garry


From blarson at crary.com  Thu Sep 12 12:08:26 2002
From: blarson at crary.com (Bradley D. Larson)
Date: Thu, 12 Sep 2002 11:08:26 -0500
Subject: Coding Style: Quotes
References: <bt1g9.17429$Pf7.590136@news1.west.cox.net> <20020912105305.A6508@phy.duke.edu>
Message-ID: <3D80BBFA.8F1122A1@crary.com>

Since most SQL statements utilize single quotes to designate 'strings' I have gone
pretty much to double on the outside and single on the inside.

    ie.
          sql = "update mytable set firstname = 'Brad' where idx = 1;"
          execute (sql)

 Otherwise the decision is pretty much arbitrary.

Michael Stenner wrote:

> On Thu, Sep 12, 2002 at 02:25:11PM +0000, John Waycott wrote:
> > I'm curious if others have adopted any standards for choice of single vs.
> > double quotation marks to delimit strings. A look through the standard
> > library reveals the choice of one over the other is rather arbitrary. I
> > suspect it really makes no difference, but the question has come up during a
> > code review.
>
> Since there is no "physical" difference between them, I can't imagine
> why it would matter.  The only reason there "needs" to be two is for
> nested quoting.  Sure, it would be annoying to alternate for each
> string, but beyond that...
>
> Personally, my use tends to be c/perl influenced.  I use single quotes
> for short static things and double quotes for longer things and format
> strings.
>
> f['key'] = "This is a long string, but it's just _my_ style..."
>
>                                         -Michael
>
> --
>   Michael Stenner                       Office Phone: 919-660-2513
>   Duke University, Dept. of Physics       mstenner at phy.duke.edu
>   Box 90305, Durham N.C. 27708-0305
>
> --
> http://mail.python.org/mailman/listinfo/python-list
-------------- next part --------------
A non-text attachment was scrubbed...
Name: blarson.vcf
Type: text/x-vcard
Size: 400 bytes
Desc: Card for Bradley D. Larson
URL: <http://mail.python.org/pipermail/python-list/attachments/20020912/fcf4373d/attachment.vcf>

From lbrannma at cablespeed.com  Sun Sep 22 14:10:06 2002
From: lbrannma at cablespeed.com (Lance)
Date: Sun, 22 Sep 2002 11:10:06 -0700
Subject: sys.ps2 assignment has no effect
Message-ID: <uos1np8t53k396@corp.supernews.com>

I'm running the latest Python on Windows XP. The default installed secondary
string appears to be blank spaces rather than an ellipse. The following has
no effect:

import sys
sys.ps2 = '... '

Any suggestions?

Thanks,
Lance




From missive at frontiernet.net  Sun Sep  1 17:19:14 2002
From: missive at frontiernet.net (Lee Harr)
Date: Sun, 01 Sep 2002 21:19:14 -0000
Subject: del(instance.attr) -- no attribute
Message-ID: <un512ii7hjgq22@corp.supernews.com>

Is this a bug?

>>> class C:
...  a = 1
... 
>>> c = C()
>>> c.a
1
>>> del(c.a)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: C instance has no attribute 'a'
>>> hasattr(c, 'a')
1
>>> 



I can see that maybe del(c.a) should not be allowed to delete
the attribute, but I think the error message is misleading.



From vetter at lincom-asg.com  Wed Sep  4 11:40:22 2002
From: vetter at lincom-asg.com (Keith)
Date: 4 Sep 2002 08:40:22 -0700
Subject: Regexp Over Multiple Lines
Message-ID: <565e316f.0209040740.36573b56@posting.google.com>

In Python can I have a regular expression that spans multiple lines...
e.g. In Perl I may define a regular expression like:

$depends_def =   qr/
         [\w_\.]+\s*:                     # simobject:
         \s*
         [\w_\.]+                         # job name
         \(\s*(?:[\d]*)?\s*\)             # arg list for job (optional
num)
         \s+
         depends\s+on                     # keywords "depends on"
         \s+
         (?:
          [\w_\.]+\s*:                    # simobject:
          \s*
            (?:
              [\w_\.]+                    # job name
              \(\s*(?:[\d]*)?\s*\)        # arg list for job (optional
num)
              (?:
                (?:
                 \s+
                 and\s*[\w_\.]*\s*:       # "and simobject:"
                 \s*
                 [\w_]*                   # job name
                 \(\s*(?:[\d]*)?\s*\)     # arg list for job (optional
num)
                )+
              )?                          # match optional "and" 1 or
more times
            )
         )
        /sx;

then later I can use that regular expression in another regular
expression.

It helps to span multiple lines for commenting what the heck the ugly
regular expression is trying to do...

BTW: I am investigating Python and have no knowledge of it.  Our Perl
code has just gotten out-of-hand.

Thanks,
Keith


From fgranger at alussinan.org  Fri Sep 27 12:49:57 2002
From: fgranger at alussinan.org (=?ISO-8859-1?Q?Fran=E7ois_Granger?=)
Date: Fri, 27 Sep 2002 18:49:57 +0200
Subject: Length of file extension Was: convention for Pickler file extension
References: <918bc22f.0209230208.4eaca702@posting.google.com> <svEj9.226013$z91.9600123@bin3.nnrp.aus1.giganews.com> <ampfbe$rq0$07$2@news.t-online.com> <smm0pus5t41l783ffo8ms6uur2umob0ftf@4ax.com> <amuns8$cc2$03$1@news.t-online.com>
Message-ID: <1fj6mww.zdjb2wpp3dogN%fgranger@alussinan.org>

Thomas Guettler <zopestoller at thomas-guettler.de> wrote:

> Richie Hindle schrieb:
> >>I call files *.pickle
> > 
> > Are there any non-medieval platforms that don't let you have arbitrary
> > length file extensions?
> 
> I think it works for Win98, WinNT, Win2000 and of corse Unix.
> 
> Don't know about Mac

MacOS 9 does not know of file extension and allow dots(.) inside a name.
So, no limit.

> or MacOSX

is Unix.

-- 
"Le plus difficile au monde est de dire en y pensant ce que tout le
monde dit sans y penser."
- Alain (?mile Chartier, dit), Histoire de mes pens?es
- cit?e par Natacha


From kathuw at swbell.net  Thu Sep 26 01:27:31 2002
From: kathuw at swbell.net (Kat & Huw)
Date: Thu, 26 Sep 2002 00:27:31 -0500
Subject: curses troubles
Message-ID: <000001c2651d$7f2b5b40$0100a8c0@mshome.net>

Ok, problem.

I'm trying to take user entered data input and store it as a var.

inside an __init__ I have curses set up a window, and display some 
options. Then I use:
while 1:
    self.k = self.stdscr.ch()
    if self.k == ord('G'): do_this()
    elif self.k == ord('P'): do_another()
    elif self.k == ord('Q'): break

Now, when running this, if I hit either G or P, the applications closes as if I'd hit Q

any idea what I'm doing wrong here?

yours,
Huw
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020926/ba4a9832/attachment.html>

From jhorneman at pobox.com  Sun Sep 29 12:39:28 2002
From: jhorneman at pobox.com (Jurie Horneman)
Date: 29 Sep 2002 09:39:28 -0700
Subject: Python's import: why doesn't it work?
Message-ID: <9ac02e81.0209290839.4f4d6475@posting.google.com>

Once again my attempt to profit from Python's vaunted productivity
increase has come to a screeching halt because some innocent thing
that one would reasonably expect to work in one line, intuitively, and
without even having to think about, turns into a nightmare already
lasting more than an hour, requiring numerous experiments, searching
through the documentation - both manually and using my editor's
find-in-files - finally leaving me no other option but to write a post
to comp.lang.python, knowing that my programming flow for today is
well and truly gone. I hate doing this, because I know I cannot keep
my frustration out of my writing, and also that someone will reply and
gently or not so gently point out something which actually IS written
down in some obscure corner of the documentation, despite my attemps
to RTFM.

So, what is it this time? No, it's not Unicode, another aspect that
frequently trips me up, causing my face to meet the rough asphalt of
the infobahn. It's importing modules. Again.

(This is an obscure reference to an attempt on my part to implement
MBCS conversion functions as Python codecs. This failed miserably and
despite the helpful advice I received here, I never did try again. Too
bad, because those conversion functions are pretty good. And since I'm
doing some parenthesized ranting here: does anyone outside of the
Python world use 'codec' to refer to anything except audio / video
codecs? I actually had to post here just to find out that what I was
looking for were *called* codecs. Sheesh.)

Anyway.

I have this script, let's call it 'blah.py'. It sits in a directory
far, far away from where my Python is installed. It wants to include a
module called 'intro' from a subdirectory called 'templates'.

This is evidently impossible.

Among my vain, futile attempts were:

*** Solution 1 ***
The World is a Beautiful Place

>>> import templates.intro

Ha! ha! ha! OK, even I knew this was a bit optimistic.

*** Solution 2 ***
.PTH file

Put a templates.pth containing the single line 'templates' in the same
place as 'blah.py', then

>>> import templates.intro
or
>>> import intro

No dice: 'ImportError: No module named...' etc.

*** Solution 3 ***
IMPorting by hand

>>> import imp
>>> f = open('templates\intro.py')

>>> imp.find_module('intro.py')
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
ImportError: No module named intro.py

>>> imp.find_module('intro')
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
ImportError: No module named intro

>>> imp.find_module('intro', '\\templates\\')
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
ImportError: No frozen submodule named .\templates\.intro

(No WHAT?)

>>> imp.find_module('intro', '\\templates\\')
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
ImportError: No frozen submodule named \templates\.intro

>>> imp.find_module('intro', ['\\templates\\'])
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
ImportError: No module named intro

>>> imp.find_module('intro.py', ['\\templates\\'])
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
ImportError: No module named intro.py

>>> imp.find_module('intro', ['.\\templates\\'])
(<open file '.\templates\intro.py', mode 'r' at 0x01511110>,
'.\\templates\\intro.py', ('.py', 'r', 1))

Whoa! I finally managed to call imp.find_module without raising an
exception! OK, now to call imp.load_module, because the documentation
says:

"Load a module that was previously found by find_module()"

Alright, so there seems to be some connection between find_module and
load_module... The latter probably takes the return value of the
former as a parameter, right? Right? Wrong.

Here are my attemps to call imp.load_module:

x = imp.find_module('intro', ['.\\templates\\'])

>>> y = imp.load_module('intro', f, 'intro.py', x)
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
TypeError: load_module() argument 4, item 0 must be string, not file

>>> y = imp.load_module('intro', x)
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
TypeError: load_module() takes exactly 4 arguments (2 given)

Okay, wait, my assumption was obviously naive, better read the doc...
aha. Almost the same parameters as find_module... wonder why I even
needed to call find_module in the first place. OK:

>>> y = imp.load_module('intro', f, 'intro.py', imp.get_suffixes())
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
TypeError: load_module() argument 4 must be sequence of length 3, not
5

What? But... but... *splutter* the doc says:

"The description argument is a tuple, as would be returned by
get_suffixes(), describing what kind of module must be loaded."

So why doesn't this work? OK, I have no idea why I'm bothering with
these kinds of arcana, but for the heck of it I'll try:

>>> y = imp.load_module('intro', f, 'intro.py', ('.py', 'r', 1))
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
  File "intro.py", line 29, in ?
ImportError: No module named Base

Yay! What?! Now my IMPORTED module can't find the modules IT is trying
to import? (It doesn't matter if I put Base next to Blah.py or in
templates - Intro can't find it.)

Arrrgghhh!!!! OK, enough time spent on this dead end. (Oh, and by the
way not all the functions of the imp module are documented.)

*** Solution 4 ***
Packages

OK, more searching, browsing and scouring my hard drive for the word
'import'. Of course! I forgot that in the Python docs, important
information is cunningly hidden in the Tutorial section! (Really, am I
being unreasonable for assuming that once one has reached a certain
level of knowledge, one shuns the Tutorial and goes straight for the
main documentation, because that's where one can find everything there
is to know? That's only how it works in every language or piece of
software in general I've ever encountered.)

So there are these things called 'packages'. Wow! Handy, that! All one
needs to do is put an empty file called '__init__.py' in a directory,
and hey! That directory becomes a package and Bob's your uncle.

So I put an empty file called '__init__.py' in 'templates' and try the
following:

>>> import templates.intro
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
ImportError: No module named templates.intro

>>> from templates import intro
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
ImportError: No module named templates

This is *exactly* how it's written in the Tutorial, and
it-does-not-work.

*** Solution 5 ***
Abject Defeat, Ask on Internet, Be Ridiculed by Peers for Not Managing
Basic Operation

Amusing little anecdote: I've performed all of the above steps twice -
once while growing increasingly frustrated, and a second time while
writing this post. Twice, while typing in the code above, I at some
point entered:

import intro

... and it would work. For real. The symbols were there. Not a peep
out of the interpreter.

Because I don't trust PythonWin (don't get me started on PythonWin, I
haven't found a better IDE yet, I may actually succumb and PAY for one
soon *shock horror*) I quit and restarted and lo and behold! the
interpreter complained that there was no such module, as it had been
doing all along, the horrible beast.

In other words, at random, unreproducable moments, Python perversely
imported my module, whereas normally it refused to.

To cut a long story... hm... Please please help me! Stop me before I
whine again. And please don't be offended by my ranting, I just can't
hide this much frustration. I'm using Python 2.2 by the way.

Thanks in advance,

Jurie.


From phr-n2002b at NOSPAMnightsong.com  Fri Sep 20 00:03:40 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 19 Sep 2002 21:03:40 -0700
Subject: monetary applications [was: Python GUI app to impress the bos s?]
References: <31575A892FF6D1118F5800600846864DCBD817@intrepid> <mailman.1032362843.10509.python-list@python.org> <23891c90.0209190014.26c5bb9@posting.google.com> <s2fdma.pp3.ln@ix.netcom.com>
Message-ID: <7xk7lhb943.fsf@ruckus.brouhaha.com>

Dennis Lee Bieber <wlfraed at ix.netcom.com> writes:
>         The only language I've ever encountered that had a truely monetary 
> safe fixed point data type is COBOL -- which has always used packed BCD 
> formats (though where the sign "bit" is stored may vary). COBOL was 
> created from the beginning for "business" applications (heck: COmmon 
> Business Oriented Language) and was spec'd by a government team. My 
> college mainframe required /four/ of the sixteen registers to hold 
> /one/ BCD number (32 BCD digits, 2 digits/byte, 16bytes @ 4/register).

I'm finding it hard to imagine how any real-world financial
calculation using IEEE doubles could be off by as much as a penny.

BCD doesn't guarantee accuracy either.  If you do a compound interest
calculation in BCD and in IEEE floating point, the floating point
answer will probably be more accurate.


From sa at hogia.net  Mon Sep  9 05:24:27 2002
From: sa at hogia.net (Sebastian Andersson)
Date: Mon, 9 Sep 2002 09:24:27 +0000 (UTC)
Subject: How to embed python ?
Message-ID: <slrnanoq4l.s40.sa@gamegirl.hogia.net>

I've got an old game server where I would like to make the server
extensible via python code which will be called when things happen in
the game. I would like to call different code depening on the event
and I would like to pass on different data.

In the game server I've got level files where I intend to put the
python code. It will look something like this:

triggers {
  0,0 79,17 (enter) {
    def my_helper_function():
      rch2.message(m, NOW, "Hello there")

    my_helper_function()
    rch2.message(m, NOW, " and bye now!)
  };
};

(The first two lines and the last two lines are part of the level file
format).

I use Py_CompileString to compile the python code. As it is now I have
to remove the extra white space before each line while building the
string. Is there a way around that to let Py_CompileString do it by
itself? Or even better, is there a way to pass each line by itself
to the python parser and let it signal when the white space count
has become too small (ie on the "  };" line) ?

How should I organize the python code?

The first argument to all "triggers" is an integer called m.
Then follows one, two or three arguments that vary depending on
the trigger.

Should I make each piece of code into a function with the correct
arguments or should I just put the triggers' arguments into variables
before calling the code with PyImport_ExecCodeModule ?

Should I use the code as above or should I put it inside a function
definition and later call that function instead of using
PyImport_ExecCodeModule ?

I want each trigger call to be as fast as possible.

TIA,
/Sebastian


From sholden at holdenweb.com  Mon Sep  9 16:04:06 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 9 Sep 2002 16:04:06 -0400
Subject: transform a list of lists in a lit of strtings???
References: <mailman.1031507832.24527.python-list@python.org> <algah9$1p0kp3$1@ID-71831.news.dfncis.de> <algb9t$1q0jg0$1@ID-71831.news.dfncis.de>
Message-ID: <za7f9.77274$l_4.5493@atlpnn01.usenetserver.com>

"James Kew" <james.kew at btinternet.com> wrote in message
news:algb9t$1q0jg0$1 at ID-71831.news.dfncis.de...
> "James Kew" <james.kew at btinternet.com> wrote in message
> news:algah9$1p0kp3$1 at ID-71831.news.dfncis.de...
> > labels1 = [["D1"], ["D2"], ["D3", "D4"]]
> > --> labels2 == ["D1", "D2", "D3", "D4"]
>
> > labels2 = reduce(lambda x, y: x+y, labels1)
> >
> > Can this be written in list comprehension form?
>
> Emile's google link suggests (a) flattening is a hoary old chestnut, and
(b)
> yes it can:
>
> labels2 = [i for item in labels1 for i in item]
>
> (I find this oddly hard to read.)
>

This will work if your list structure can contain data items and lists with
data items in them, but it won't flatten a nested list structure of general
complexity. For example:

>>> labels1 = [ [1, 2], [[3, 4, 5], 6], [[[7], 8], 9]]
>>> labels2 = [i for item in labels1 for i in item]
>>> labels2
[1, 2, [3, 4, 5], 6, [[7], 8], 9]

For a complete solution to such a problem you have to use recursion, testing
whether each item you process is composite and recursing to flatten it if
so. This has been discussed so many times before that you should use Google
to find the best way.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From aleax at aleax.it  Thu Sep 26 07:50:46 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 26 Sep 2002 11:50:46 GMT
Subject: Getting last char in string.
References: <mailman.1033027776.19152.python-list@python.org> <amumk5$p52$1@lust.ihug.co.nz> <amuna6$erv$1@news1.wdf.sap-ag.de> <amuq6t$sv9$1@lust.ihug.co.nz>
Message-ID: <qwCk9.145363$ub2.3203712@news1.tin.it>

CheapSkate wrote:
        ...
> if s[-1] returns a single element, why woulnd't it return " ?
> Single character are strings of length 1(as u said) and " is single
> character right?

Nope: '', the empty string, is 0 characters, not 1 character.


Alex



From jepler at unpythonic.net  Sat Sep 28 15:25:26 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Sat, 28 Sep 2002 14:25:26 -0500
Subject: Python threading?
In-Reply-To: <an4qud$rmm$0@216.39.172.122>
References: <mailman.1033054839.28121.python-list@python.org> <an4qud$rmm$0@216.39.172.122>
Message-ID: <20020928142517.A3027@unpythonic.net>

On Sat, Sep 28, 2002 at 06:03:57PM +0000, Bengt Richter wrote:
> Any thoughts on defining the next method with a generic *args optional arg list
> and making it visible to an executing generator function g as g.__args__ ?
> 
> g.__args__ would then just be () for current usage, and could be
> backwards-compatibly ignored as desired.

I think you can get something like this using a class.  For this solution,
you'll need to make the list visible as a parameter to the generator, or
else write it so that you derive a class for each generator, making the
generator a method.  You may also be able to work something up involving
nested functions that lets the inner function see the argument list as
a variable in the outer scope.

class ArgsGenerator:
    def __init__(self, f, *args, **kw):
	self.args = []
	self.__iter = f(self.args, *args, **kw)

    def next(self, *args):
	self.args[:] = list(args)
	return self.__iter.next()

def primes(x):
    n = 2
    while 1:
        if x: n=x[0]
        if n == -1: return
        while 1:
            composite = 0
            for d in range(2, n-1):
                if n%d == 0:
                    composite = 1
                    break
            if not composite:
                yield n
                break
            n=n+1
        n=n+1

x = ArgsGenerator(primes)
print x.next()   # 2
print x.next(8)  # 11
print x.next()   # 13
print x.next(-1) # raises StopIteration

Jeff



From nhycqb at yahoo.com  Sat Sep 14 01:03:09 2002
From: nhycqb at yahoo.com (nhycqb at yahoo.com)
Date: Sat, 14 Sep 2002 05:03:09 +0000 (UTC)
Subject: $$$$$ secret realease for the first time  5484
Message-ID: <alufud$7go$3887@nobel.pacific.net.sg>

http://myweb.ecomplanet.com/PETE7323/
http://myweb.ecomplanet.com/PETE7323/


 - Instant Download
 - Instant Download
 - Instant Download

Free Download
The Absolute Beginner's Guide to Starting a Web Site
Totally Free Web Resources
Unlimited Profits
7 Secrets To Unlimited Traffic
Scientific Advertising
How to Start Your Own Traffic Virus
Marketing Warriors Tips
eBookoMatic
eBook Submitter
The Best of Web Gold
On Line Stealth Marketing
and more 

http://myweb.ecomplanet.com/PETE7323/
http://myweb.ecomplanet.com/PETE7323/


 - Instant Download
 - Instant Download
 - Instant Download


http://myweb.ecomplanet.com/PETE7323/
http://myweb.ecomplanet.com/PETE7323/
klmbteirptjjozqulyehheueoosxviecjcdneirgijgvnsjbtdkphdvecnqzdtuclpjjucnmftwmkdhvliqcbdby



From jason at mastaler.com  Thu Sep  5 16:01:37 2002
From: jason at mastaler.com (Jason R. Mastaler)
Date: Thu, 05 Sep 2002 14:01:37 -0600
Subject: New `spambayes' project on SourceForge
References: <15735.40259.117828.402419@anthem.wooz.org>
Message-ID: <hhofbc19zy.fsf@hrothgar.la.mastaler.com>

barry at zope.com (Barry A. Warsaw) writes:

> There are also two mailing lists related to the spambayes project.
> The first is a general discussion list:
>
>     http://mail.python.org/mailman-21/listinfo/spambayes
>
> and the other is a list for cvs checkin message notices:
>
>     http://mail.python.org/mailman-21/listinfo/spambayes-checkins

These lists have now been added to Gmane (http://gmane.org) as well:

spambayes at python.org <==> news://news.gmane.org/gmane.mail.spam.spambayes.general

spambayes-checkins at python.org <==> news://news.gmane.org/gmane.mail.spam.spambayes.cvs

-- 
(http://tmda.net/)





From pan-newsreader at thomas-guettler.de  Thu Sep 19 02:02:16 2002
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Thu, 19 Sep 2002 08:02:16 +0200
Subject: shared mem & semaphores module
References: <YWbi9.17569$V7.4455639@twister.socal.rr.com>
Message-ID: <pan.2002.09.19.08.02.16.550254.723@thomas-guettler.de>

On Thu, 19 Sep 2002 05:57:44 +0200, Jeff Davis wrote:

> I found a module online by "INRIA" (I don't know who that is) that seems
> to wrap the C functions for shared memory management and semaphores.
> 
> http://gigue.peabody.jhu.edu/~mdboom/omi/source/shm_source/shm.html
> 
> My question is: are these functions included elsewhere?


The threading module has a semaphore class.

AFAIK there is no module for shared memory.

 thomas


From henrik.motakef at web.de  Sun Sep  1 16:37:34 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 01 Sep 2002 22:37:34 +0200
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de>
Message-ID: <874rd9phtd.fsf@pokey.henrik-motakef.de>

gerson.kurz at t-online.de (Gerson Kurz) writes:

> So basically, couldn't one come up with a *human readable* syntax
> for re, and compile that instead? Python prides itself on its clean
> syntax, and human readability, an bang - import re, get perl-ish
> code instantly!

There is an Emacs Lisp package called "symbolic regexps" (or
sregexp.el) that lets you write regular expressions in standard lisp
syntax, like

(sregexq bol (or "abc" "def"))

instead of "^(abc|def)" (bol meaning "beginning of line"). I don't see
how that maps elegantly to Python syntax, however.

Regards
Henrik


From kennethg at pd.jaring.my  Sat Sep 21 12:09:43 2002
From: kennethg at pd.jaring.my (Kenneth Gomez)
Date: Sat, 21 Sep 2002 16:09:43 GMT
Subject: Local Python Server Installation
References: <3d8b47c6.277669@news.jaring.my> <slrnaomn6m.39s.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <3d8c993c.2190426@news.jaring.my>

Dear Gerhard

I tried your method. But instead of getting a cgi listing of my
environmental listings, all I saw on my browser was the two input
lines ..i.e. import cgi and cgi.test()

My url was http://localhost/test.py

Any advise.
Regards,
Kenneth.

On 20 Sep 2002 17:36:21 GMT, Gerhard =?iso-8859-1?Q?H=E4ring?=
<gerhard.haering at opus-gmbh.net> wrote:
>
>Add this line at the end of your .../Apache Group/Apache2/conf/httpd.conf:
>
>    ScriptInterpreterSource Registry
>
>Then you can write Python CGI scripts that have the .py extension. I
>recommend to create a test.py script with the following two lines:
>
>import cgi
>cgi.test()
>
>After you verify that this works (http://localhost/cgi-bin/test.py) you can
>continue reading the documentation about the cgi module and writing your
>own CGI scripts. The cgitb and Cookie modules should come in useful, too.
>
>If you experience the limits of CGI one day, you can think about using a
>full-blown appliation server like WebWare.
>
>-- Gerhard

Cheers,
Kenneth Gomez.


From claird at phaseit.net  Mon Sep 30 15:18:19 2002
From: claird at phaseit.net (Cameron Laird)
Date: Mon, 30 Sep 2002 19:18:19 -0000
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Sep 30)
Message-ID: <uph8rr1ins94fa@corp.supernews.com>

QOTW:  "One can always write very sparse or very dense lines of code that
break the model, but in-the-large programmers do not normally do that.
They tend to write code in lines that fit easily in their head and, because
of this, one line of code is pretty much comparable to another."
Jonathan Hogg

"Why are there always jobs for salesmen?"  Stephen Satchell 


    Jack Diederich makes high-performance Permutations, Combinations,
    and other probability and statistics constructs available through
    SourceForge.
	http://probstat.sourceforge.net/

    Puffin makes automation nice.
        http://www-106.ibm.com/developerworks/opensource/library/os-puffin2.html 
    Pexpect fills in Puffin's most severe lack.
	http://sourceforge.net/projects/pexpect/

    Uche Ogbuji introdues Python's XML and XML-RPC capabilities, and PyXML.
	http://www.xml.com/pub/a/2002/09/18/py.html
	http://www-106.ibm.com/developerworks/library/ws-pyth10.html 
	http://www.xml.com/pub/a/2002/09/25/py.html

    September brought another round of attention to CP4E in the EDU-SIG group
        http://mail.python.org/pipermail/edu-sig/2002-September/thread.html

    Think in Tkinter.
	http://home.att.net/~stephen_ferg/thinking_in_tkinter/index.html

    Martin v. L?wis gives background to management of patches and
    contributions in the Python project.
        http://mail.python.org/pipermail/python-list/2002-September/122988.html
    
    Blog with Python.
	http://www.awaretek.com/nowak/kaa.html        
    

========================================================================

Everything you want is probably one or two clicks away in these pages:

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
         http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Brett Cannon continues the marvelous tradition established by 
    Andrew Kuchling and Michael Hudson of summarizing action on the
    python-dev mailing list once every other week.
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

    Much of Python's real work takes place on Special-Interest Group
    mailing lists
        http://www.python.org/sigs/

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
        http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
        http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
        http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

    Links2Go is a new semi-automated link collection; it's impressive
    what AI can generate
        http://www.links2go.com/search?search=python

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning
(approximately), ask <claird at phaseit.net> to subscribe.  Mention
"Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.



From DennisR at dair.com  Sun Sep 29 15:38:26 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Sun, 29 Sep 2002 19:38:26 GMT
Subject: Python and GUI question
References: <3d96b808_3@news.bluewin.ch>
Message-ID: <SEIl9.465$%L1.56153240@newssvr13.news.prodigy.com>

Python could also be used for writing CGI interfaces so that HTML pages are
the GUI.  HTML has well characterized strengths, weaknesses, and tools
support.  My web site at http://www.autorunning.com/arpyexe.cgi permits
building AutoRunning Windows executables using HTML CGI interfaces from
Python source code.
--

Dennis Reinhardt

http://www.dair.com




From pereira at cis.upenn.edu  Thu Sep  5 22:59:50 2002
From: pereira at cis.upenn.edu (Fernando Pereira)
Date: Thu, 05 Sep 2002 22:59:50 -0400
Subject: Graham's spam filter
References: <mailman.1029992314.27908.python-list@python.org> <3D647B4B.28D9BF04@alcyone.com> <mailman.1030050253.11330.python-list@python.org> <3D655AE6.A097AA25@alcyone.com> <ak44j1$1f2ft4$2@ID-125932.news.dfncis.de> <c2ef5940.0209041912.16e102fb@posting.google.com> <3D770156.2E55864C@alcyone.com> <al7tg4$1o5a9k$5@ID-125932.news.dfncis.de> <3D77AC75.B9B7F27@alcyone.com> <al8ehh$1nt6ua$1@ID-125932.news.dfncis.de> <3D77D657.DE280373@alcyone.com>
Message-ID: <B99D9266.121F4%pereira@cis.upenn.edu>

On 9/5/02 6:10 PM, in article 3D77D657.DE280373 at alcyone.com, "Erik Max
Francis" <max at alcyone.com> wrote:

> Christopher Browne wrote:
> 
>> Quoth Erik Max Francis <max at alcyone.com>:
>> 
>>> Spammers are hitting upon the strategy, though, of sending emails in
>>> which the body consists of nothing but a completely encoded base64
>>> MIME part.  So in that case, the entire body of your message would
>>> consist solely of your "base64encoded" token.  So in the general
>>> case of any kind of spam filter (not just limited to a Graham
>                                    ^
>>> filter), it's questionable how useful this will be, unless you plant
>>> to always filter against that token, presuming it to always indicate
>>> spam.
>> 
>> I've been using naive Bayesian filtering for years; I don't assume
>> that _any_ particular token indicates _any_ particular result.
> 
> I thought I made it clear that I was discussing spam filters in general,
> not just Graham/Bayesian filters.
That is, you overstated the case. Your point may be true of rule-based
filters that look for particular words presumed to be spam indicators. But a
machine-learning approach using a rich set of message features will pick on
*anything* that is correlated with spamminess, not just the obvious
features. In an arms race, the spammer may gain temporary advantage by
changing message format and thus which features are significant, but after
receiving a few new spams, retraining will pick on other appropriate
features, for instance fraction of the message bytes that are base-64
encoded. Occasionally, filter designers may want to add new features based
on some exploratory analysis of spam, but the decision of which of those are
relevant is left to the learning algorithm.

-- F



From nospam at bigfoot.com  Wed Sep 11 08:08:35 2002
From: nospam at bigfoot.com (Gillou)
Date: Wed, 11 Sep 2002 14:08:35 +0200
Subject: IDLE
References: <wpbf9.264320$aA.45691@sccrnsc02>
Message-ID: <alnb0g$2tl8$1@norfair.nerim.net>

Can't you be more explicit ?
What do you exactly observe when you try to run IDLE ?
What's your OS, Python version ?
How do you try to run IDLE ?
Did you switch the computer power on ?
...

--Gilles

> Why won't my IDLE work?
>
>




From aleax at aleax.it  Wed Sep 18 02:59:15 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 18 Sep 2002 06:59:15 GMT
Subject: Python GUI app to impress the boss?
References: <mailman.1032314723.10878.python-list@python.org>
Message-ID: <7vVh9.108027$ub2.2286654@news1.tin.it>

David LeBlanc wrote:
        ...
> What prevents the creation of a decimal money type? I agree, it would be

Nothing prevents it.  In particular, nothing prevents implementing such
a type in Python itself.  There are various implementations of decimal
numbers around, including one (not sure how mature and complete) that
meets the standards for decimal arithmetic behavior (as per the IBM Rexx
implementation).  It seems to me a relatively simple problem, compared
with many others that are routinely solved by Python programmers.  If
there's no single "best of breed" Currency extension that dominates the
scene (as, for example, Lemburg's mxDateTime dominates the date-time
handling scene) I suspect it must be because of a lack of demand, strange
as it may seem.  Defining new types is SO easy...

So prove my suspicion wrong: pick an almost-there implementation,
complete it per your specs, release it, and see it take over the
world.  How much effort can it be...?

> nice if python could do something akin to calculator aritmetic with types
> like scientific (covered) engineering (improper scientific with the
> exponent as a power of 3 and the (oops, forgot the term - mantissa?) more

That's just an issue of display, isn't it?  Computations are just the
same whether the calculator is set to always display exponents as a
power of three, or not -- at least, that's how I recall my calculator
worked: a single button switched display mode from scientific to
engineering and back, but all numbers in the calculator's stack and
registers were supremely unaffected by it.

> then a single digit as in 12.2e3), fixed precision (1..n to the right of
> the decimal). The forthcoming fractions are going to be interesting...
> (chinese interpretation).

If by "fractions" you mean Rationals, that's another thing yet.


> Be nice too if math could do rads, grads and decimal degrees.

That's a completely different issue: math is a module, not a type.
In this case, what you want is simply a different module that wraps
math and multiplies/divides some arguments or results by appropriate
conversion factors.  Still a pretty modest programming effort (all
of, what, 10 minutes...?).  Modifying the math module to behave this
way upon some switch set in it should not be much more.


> Just my $0.02 worth.

If you really think, just to take a simple example, that having
a sin function that accepts degrees rather than radians is
important for Python, why not just take the 10 minutes to
implement that (or 20 minutes, if you're keep to modify rather
than wrap the math module) and thus prove I'm wrong in suspecting
that such a project would be pretty much useless...?

I'll even volunteer to perform the programming part of the task
(not a huge commitment:-) if you undertake to deal with testing,
documentation, and advocacy (to get it widely used, and in the
long run to get it into the core by popular demand).


Alex



From jorgencederberg at hotmail.com  Fri Sep 13 03:46:27 2002
From: jorgencederberg at hotmail.com (Jørgen Cederberg)
Date: Fri, 13 Sep 2002 09:46:27 +0200
Subject: Tkinter experts out there?
References: <mOfg9.461067$UU1.75649@sccrnsc03>
Message-ID: <nJgg9.22$6b7.514@news.get2net.dk>

"Marcus" <noone at nowhere.com> wrote in message
news:mOfg9.461067$UU1.75649 at sccrnsc03...
> I am trying to get some answers on using Tkinter, does anyone know if
there
> are channels or forums being actively used?

Hi Matthew,

you can ask away here in comp.lang.python. People are happy to answer your
questions regarding Tkinter. Other resources of interest are:

An introduction to Tkinter by Fredrik Lundh:
http://www.pythonware.com/library/tkinter/introduction/index.htm

Tkinter: GUI programming with Python
http://www.nmt.edu/tcc/help/lang/python/tkinter.html

Python Faqts
http://www.faqts.com/knowledge_base/index.phtml/fid/264


Regards
Jorgen




From jbublitzNO at SPAMnwinternet.com  Sat Sep  7 02:44:12 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Sat, 07 Sep 2002 06:44:12 GMT
Subject: PyQT picture display
References: <albp27030dq@enews4.newsguy.com>
Message-ID: <3D799F42.2020903@SPAMnwinternet.com>

Jim wrote:
> Hello to all,
> 
> I've a widget with three buttons and a label. I want to display a JPEG 
> picture when either of the first two buttons is clicked. When I run the 
> code below, the widget displays on screen but when either button is clicked,
> the picture does not appear( the text in the label disappears as it should 
> ). Can anyone tell me what I'm doing wrong here?

As far as I can tell, nothing's wrong. It works fine
for me. Are you sure the jpg files actually exist and
the path to them is correct? The only thing I changed
in your program was the filenames for the image files.

I used PyQt-3.3.2 with Qt 2.3.0 on Linux.

(another) Jim



From narnett at mccmedia.com  Wed Sep  4 19:15:01 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Wed, 4 Sep 2002 16:15:01 -0700
Subject: Bind threads to addresses -- Windows & urllib?
In-Reply-To: <oTvd9.6$7z6.716@castor.casema.net>
Message-ID: <PHEDLIMPMIAGAEPEFBGCAEFHCHAA.narnett@mccmedia.com>


> i think you mean something like this:
>
> import thread
>
>
> def spiderStuff(Socket):
>     # do the spider stuff here
>     pass
>
> ...
>
> thread.start_new_thread(SpiderStuff, (,newSocket))

I would like to avoid going to the low-level modules -- I guess I was hoping
for something along the lines of urllib or urllib2 that would have already
have this exposed.  Perhaps I can hack one of them.  It looks to me as
though I could at least hard-code the addresses into my code, then pass them
to a hacked urllib that uses them instead of the localhost and thishost
methods.  Of course, hard-coding the addresses would be very ugly and
non-portable.

Any further thoughts out there?

Nick




From CousinStanley at HotMail.com  Fri Sep 27 13:20:19 2002
From: CousinStanley at HotMail.com (Cousin Stanley)
Date: Fri, 27 Sep 2002 10:20:19 -0700
Subject: Data Plotting Library DISLIN 8.0
References: <3D940A37.92FD45D5@linmpi.mpg.de>
Message-ID: <an23s1$ambss$1@ID-130333.news.dfncis.de>

Helmut ... 

The  DISLIN  Plot Library from http://www.dislin.de/ 
looks  VERY  impressive  and  widely diversified ... 

I'm interested in the Python 2.2 version 
but a bit confused about the various compiler version listings ... 

Are the compilers listed for the various distributions
the ones that have been used to pre-compile a particular version
or is the compiler listed the one needed to compile
a particular version once  DISLIN  has been downloaded ???  

In otherwords, can I download  DISLIN  
and run  __out_of_the_box__   without compiling anything ???

Cousin Stanley


 


From ckasso at sprynet.com  Thu Sep 26 18:11:59 2002
From: ckasso at sprynet.com (Chris Kassopulo)
Date: Thu, 26 Sep 2002 18:11:59 -0400
Subject: How do you pronounce "__" (double underscore)?
References: <525eaba.0209260956.72bf6786@posting.google.com>
	<20020926164641.1753ebbb.ckasso@sprynet.com>
	<873crwfowo.fsf@pokey.henrik-motakef.de>
Message-ID: <20020926181159.30ec41f3.ckasso@sprynet.com>

On 26 Sep 2002 22:59:19 +0200
Henrik Motakef <henrik.motakef at web.de> wrote:

> Chris Kassopulo <ckasso at sprynet.com> writes:
> 
> > How about:
> > 
> >  _  = "un"
> >  __ = "dun"
> 
> "dun-init-dun"? Sounds too much like "done" for me... :-)

Then how about:

 _  = "unce" (rhymes with)
 __ = "dunce"
 
dunce-init-dunce has a nice ring to it. no? :-)


From ponderor at lycos.com  Tue Sep 17 14:03:37 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: Tue, 17 Sep 2002 18:03:37 -0000
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (May 28)
Message-ID: <uoerjpf9q2dk55@corp.supernews.com>

QOTW:  "that was the original poster's point.  When you start side
scrolling it i=s=20 time to refactor."
    http://mail.python.org/pipermail/python-list/2002-September/121710.html

Robert Oschler: "If Peter Piper pickles a peck of Python packages, can
he pack the pickled Python peckers in a proper packing place?"
Greg Ewing: "Will such suitably squashed SQLed strings satisfy?"
    http://mail.python.org/pipermail/python-list/2002-September/122479.html

"The changes to Python are also so small now, that it doesn't matter if
it will be merged or not."  Christian Tismer


Threads
    Dealing with whitespace in XML node attribute values:
        http://mail.python.org/pipermail/python-list/2002-September/121947.html
    
    NumPy two-fer: Binary Input and exponentiate zero solutions:
        http://mail.python.org/pipermail/python-list/2002-September/121978.html
        http://mail.python.org/pipermail/python-list/2002-September/121981.html
    
    Rational Numbers in Python? Demo, but no target for Python 2.3, and PEP 239:
        http://mail.python.org/pipermail/python-list/2002-September/121998.html
    
    Web Crawler iresources noted:
        http://mail.python.org/pipermail/python-list/2002-September/121591.html
    
    Python on WinCE issues summarized:
        http://mail.python.org/pipermail/python-list/2002-September/122424.html
        http://www.murkworks.com/Research/Python/PocketPCPython/Overview
    
    Pickling objects for an object-relational adapter:
        http://mail.python.org/pipermail/python-list/2002-September/122491.html
    
    Searching for sub-strings alternatives discussed:
        http://mail.python.org/pipermail/python-list/2002-September/122270.html
    
    How to properly call a Python COM component from C#/.Net:
        http://mail.python.org/pipermail/python-list/2002-September/122630.html
    
    Tim Peters follows up on potential speed issues parsing large data files: 
        http://mail.python.org/pipermail/python-list/2002-September/121646.html
    
    Stackless Python merge into Python's core trunk possibilities explained:
        http://mail.python.org/pipermail/python-list/2002-September/122734.html
    
Software
    pyPgSQL v.2.2 has been released. This is a bug fix release for v.2.1.
    PyPgSQL is a Python DB-API 2.0 compliant interface to PostgreSQL
    databases. 
        http://pypgsql.sourceforge.net
    
    PyQuante 1.03 available. PyQuante is a suite of programs for writing
    quantum chemistry software. 
        http://pyquante.sourceforge.net
    
    The (standalone) Python email package version 2.3 is available.  It
    is compatible with Python 2.1.3 and Python 2.2.1.
        http://sf.net/projects/mimelib/
    
    SC-Track Roundup issue tracking system, v. 0.5 beta2 announced. 
        http://mail.python.org/pipermail/python-announce-list/2002-September/001686.html
    
    Pymqi .4 released. Pymqi is a Python extension for IBM's Messaging &
    Queuing middleware, MQSeries.
        http://www.hare.demon.co.uk/pymqi/
    
    omniORB 4.0.0 and omniORBpy 2.0 release candidates available, with
    the intention to make a full release in a week or so:
        http://mail.python.org/pipermail/python-announce-list/2002-September/001683.html
    
    Leo (Literate Editor with Outlines) gains momentum and moves to v. 3.6:
        http://mail.python.org/pipermail/python-announce-list/2002-September/001684.html
    
    The first beta release of ZODB3 (3.1b1), formerly known as StandaloneZODB :
        http://mail.python.org/pipermail/python-announce-list/2002-September/001685.html
    
    Plone 1.0 goes Beta, and will be presented at OSCOM.
        http://plone.org/Members/runyaga/news/Plone-beta1.news
        http://plone.org/Members/runyaga/news/oscom2002
    
    Python 2.2.1 for DOS test release 
        http://www.delorie.com/archives/browse.cgi?p=djgpp/2002/05/19/15:30:25
    
    



    

========================================================================

Everything you want is probably one or two clicks away in these pages:

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
         http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Brett Cannon continues the marvelous tradition established by 
    Andrew Kuchling and Michael Hudson of summarizing action on the
    python-dev mailing list once every other week.
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

    Much of Python's real work takes place on Special-Interest Group
    mailing lists
        http://www.python.org/sigs/

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
        http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
        http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
        http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

    Links2Go is a new semi-automated link collection; it's impressive
    what AI can generate
        http://www.links2go.com/search?search=python

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning
(approximately), ask <claird at phaseit.net> to subscribe.  Mention
"Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.



From grabiller at 3dvf.net  Wed Sep 25 11:14:32 2002
From: grabiller at 3dvf.net (Guy Rabiller)
Date: Wed, 25 Sep 2002 17:14:32 +0200
Subject: Newbie References Question -> Exemple
References: <3d91c72f$0$8558$79c14f64@nan-newsreader-02.noos.net>
Message-ID: <3d91d2cf$0$1484$79c14f64@nan-newsreader-01.noos.net>

Ok,

thanks for your answers, I understand now that a variable name is 'just' a
name in a dictionary, and not a container.

What I want to do is simple, and I suppose that this is my 'way of thinking'
regarding Python that is wrong.

I have a list of points, and a list of polygons that should contain
'references' to points.

So for me:
-> points list:
p = []
p.append([0,0])
p.append([0,1])
p.append([1,1])
p.append([1,0])
-> polygon:
quad = []
quad.append(p[0])
quad.append(p[1])
quad.append(p[2])
quad.append(p[3])

But now, if I change the coordinate of one point:
p[0] = [5,5]
I will have:
>>> p
[[5, 5], [0, 1], [1, 1], [1, 0]]
but:
>>> quad
[[0, 0], [0, 1], [1, 1], [1, 0]]

For me the quad list of point is not 'updated' with the change of one of its
point.

What I want to do is that, quad keeps a list of 'references' to points
coordinates.
Then if I change one point coordinates, the change is reflected in the quad
list of points.

I have no clue how to do this.

--
guy rabiller
3d animator / td
grabiller at 3dvf.net
http://grabiller.3dvf.net

Guy Rabiller wrote:
> Hi,
>
> let say I have:
> i1 = 1
> i2 = 2
> and
> p = [i1,i2]
>
> How can I have:
> p=[*i1,*i2]
> rather than
> p=[**i1,**i2] as it is curently ?
>
> ( Sorry for this nasty hybrid syntax )
>
> Textualy, how if I want that p[0] refere to the 'i1' container, and
> not to the refererence it contains ?
>
> What I want is that if now I set:
> i1 = 4
> that automaticaly:
> p -> [4,2]
> and not keeping [1,2]
>
> Thanks in advance.




From max at alcyone.com  Sun Sep 15 05:07:26 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 15 Sep 2002 02:07:26 -0700
Subject: html template system?
References: <btug9.688$Ee4.1090@news-server.bigpond.net.au>
Message-ID: <3D844DCE.3C1C3B05@alcyone.com>

:B nerdy wrote:

> what are good templating systems avaliable for python?

In addition to the ones already mentioned, there's also empy:

	http://www.alcyone.com/pyos/empy/

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ An undevout astronomer is mad.
\__/ Edward Young
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From werner.merkl at fujitsu-siemens.de  Fri Sep 27 09:05:27 2002
From: werner.merkl at fujitsu-siemens.de (Werner Merkl)
Date: Fri, 27 Sep 2002 15:05:27 +0200
Subject: Python, unicode, Windows32 and NTFS...
Message-ID: <an1l2o$fvc$1@dackel.pdb.sbs.de>

Hi,

in my job I have a lot of differnt language to deal with, now including
japan. And we have all these languages on our server (Windows 2000)...
So NTFS uses UNICODE (I think it's 'utf-16').

Python -- at least since version 1.6 -- is supposed to understand
and handle UNICODE. But my experience is, it doesn't support
UNICODE for NTFS file names.

eg:
PythonWin 2.2.1 (#34, Apr 15 2002, 09:51:39) [MSC 32 bit (Intel)] on win32.
Portions Copyright 1994-2001 Mark Hammond (mhammond at skippinet.com.au) - see
'Help/About PythonWin' for further copyright information.
>>> abc = u'a:\u03b1\u03b2\u03b3.txt'; print abc
a:???.txt # if you use Pythonwin you should see 'A:alpha beta gamma .txt'
>>> open(abc,'w').write('test')
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
IOError: [Errno 2] No such file or directory: 'a:a\xdf?.txt'
>>>

Oops! Does this mean, I have to use VisualBasicScript to handle this????
Or are there any swiches or tools or libs to use UNICODE filesystems?

BTW: It would be fine, to use all that good stuff like walk, isdir, isfile
...


Thanks in advance

Regards/Mir freundlichen Gr??en
Werner Merkl

--
----------------------------------------------------------------------
Werner Merkl
Senior Engineer Software Pre-Installation
FSC VP BP PRO PI
Fujitsu Siemens Computers
Buergermeister-Ulrich-Str. 100
86199 Augsburg
Germany

Telephone: +49(0)821 804-3548
Telefax: +49(0)821 804-3835
E-mail: mailto: werner.merkl at fujitsu-siemens.com
Internet: http://www.fujitsu-siemens.com




From joorce at hotmail.com  Tue Sep 24 04:30:08 2002
From: joorce at hotmail.com (Jose Ortega)
Date: 24 Sep 2002 01:30:08 -0700
Subject: [wxPython] ListCtrl: How can I change the column header text at runtime?
References: <al5jek$eah$1@rex.ip-plus.net>
Message-ID: <25f7fd21.0209240030.3c99d137@posting.google.com>

Hi
Maybe this will be useful.

http://wiki.wxpython.org/index.cgi/ListAndTreeControls

Bye.


From dsavitsk at e-coli.net  Tue Sep 17 18:10:05 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Tue, 17 Sep 2002 22:10:05 GMT
Subject: Test Post
References: <am8656$3npc1@ncc1701.polaroid.com>
Message-ID: <1LNh9.10302$yt3.4826468@newssrv26.news.prodigy.com>

perhaps in the future people wanting to test could test w/ a real (or even
fake) question of some relevance.  Even a post saying "I am new to Python
and I just love it." would be more appropriate.  In your heart you would
know it was a test, but we could all be slightly happier thinking that the
knowledge disseminated per bit ratio for c.l.p was that much higher.

welcome. your test worked.

-d


"David B. Margossian" <margosd at polaroid.com> wrote in message
news:am8656$3npc1 at ncc1701.polaroid.com...
> Please excuse the intrusion.  Trying to establish a feed for this group...
>
>




From sandorlevi at yahoo.com  Fri Sep 13 23:12:39 2002
From: sandorlevi at yahoo.com (Levente Sandor)
Date: 13 Sep 2002 20:12:39 -0700
Subject: transform list of int in a list of string??
References: <mailman.1031945716.2586.python-list@python.org>
Message-ID: <8b5e42a6.0209131912.3c598acb@posting.google.com>

jubafre at brturbo.com wrote in message news:<mailman.1031945716.2586.python-list at python.org>...
> Traceback (most recent call last):
>   File "C:\Documents and Settings\jubafre\Desktop\montador\teste.py", line 
> 78, in ?
>     t=str(x[i])
> TypeError: 'list' object is not callable

The error message says that a list object in your code is identified
by 'str'. You must use a different name for your list, or the str()
function will no longer be available.
It's generally a good idea to avoid using built-in identifiers for
your own objects.

levi


From ianb at colorstudy.com  Wed Sep  4 12:16:49 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 4 Sep 2002 11:16:49 -0500
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Sep  4)
Message-ID: <448A0F008E0AFE11.19B9EF86BEE2B191.107D5020B234F743@lp.airnews.net>

QOTW:  The canonical, "Python is a great first language", elicited,
"Python is a great last language!" -- Noah Spurrier
<c9d82136.0208251844.63a6d187 at posting.google.com>

"We C++ programmers have developed tricks to help us deal with this
sort of thing, in much the same way that people who suffer severe
childhood trauma develop psychological mechanisms to insulate themselves
from those experiences :-) " - Joseph A.  Knapka
http://groups.google.com/groups?th=b16d1daabdacd1de

"A growing number of utilities is a sign that a format is popular, not
that it is good." -- Oren Tirosh

	       
Threads:
    A thoughtful note on one person's decision to use Python in a
    scientific environment:
        <3D60A1E2.7FD246E0 at anansispaceworks.com>

    A Lisp implementation of a statistical spam filter described at
        http://www.paulgraham.com/spam.html
    is converted to Python, debated, and discussed:
        http://groups.google.com/groups?th=163046edb9100ee8

    A newbie's question about the behavior of lists and references is answered:
        <001601c24750$c5490d10$0effa8c0 at john>

Sites and stuff:
    gamasutra.com has an article on using Python to script games:
        http://www.gamasutra.com/features/20020821/dawson_01.htm
    (free but intrusive registration required)

    New PEP Format: reStructuredText
        http://aspn.activestate.com/ASPN/Mail/Message/python-dev/1341652

    'Seen the new Python book?
	http://click.unixreview.email-publisher.com/maaavMbaaTpIEa2scaRb/

Software:
    Psyco 0.4.1 is released.  Psyco is an extension to the standard Python
    interpreter to perform "specialized compiling" -- giving Python some of
    the speed advantages of statically typed languages like C:
        http://psyco.sourceforge.net/
    Jimmy Retzlaff posts benchmarks of standard Python, Pyrex, and Psyco
    which show impressive performance:
        http://groups.google.com/groups?th=d85d63a710c53607

    XMail 1.00 is released, both a suite of applications, and a framework
    to extend.  It is both an SMTP and POP server:
        http://xmail.marketmix.com/products/xqm/xqm_lib.php

    The Albatross web application framework reaches 1.00:
        http://www.object-craft.com.au/projects/albatross/

    SimpleParse 2.0.0 is released.  SimpleParse is a parser generator --
    for parsing programming-language-like syntaxes:
        http://simpleparse.sourceforge.net/

    Python In PHP: a Python interpreter embedded in PHP, for the
    linguistically indecisive (alpha code):
        http://www.csh.rit.edu/~jon/projects/pip/


========================================================================

Everything you want is probably one or two clicks away in these pages:

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
         http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Michael Hudson continued Andrew Kuchling's marvelous tradition
    of summarizing action on the python-dev mailing list once every
    other week, into July 2001.  Any volunteers to re-start this
    valuable series?
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

    Much of Python's real work takes place on Special-Interest Group
    mailing lists
        http://www.python.org/sigs/

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
        http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
        http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
        http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

    Links2Go is a new semi-automated link collection; it's impressive
    what AI can generate
        http://www.links2go.com/search?search=python

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning
(approximately), ask <claird at phaseit.net> to subscribe.  Mention
"Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

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


From phil at river-bank.demon.co.uk  Fri Sep 13 04:25:22 2002
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Fri, 13 Sep 2002 09:25:22 +0100
Subject: Extending a C++ App
References: <996eade8.0209120912.3a653d7@posting.google.com> <n0qnhvfn.fsf@boeing.com> <996eade8.0209122352.45721938@posting.google.com>
Message-ID: <3D81A0F2.5090808@river-bank.demon.co.uk>

Bwuce_Wee wrote:

> Thanks for your help
> 
> I looked at SWIG
> However I have a few concerns (forgive me if Im being dense)
> 
> (1) All the examples are created as DLLs (ie APIs)
>     This is fine - except how would my C++ App use the same object instances?
>     (ie. share the memory and API objects)
> 
> (2) SWIG does not do a nice job of object wrapping (in my opinion)
>     For instance instead of saying
>       c = example.Circle
>       c.x = 10
>       c.y = 20
>       del c
> 
>     I need to do this:
>       c = example.new_Circle()
>       example.Shape_x_set(c,10)
>       example.Shape_y_set(c,20)
>       delete_Circle(c)
> 
>     Isn't there a cleaner library somewhere to do this?
>     SIP has no documentation - so I don't know what to expect of it


A .sip file containing...

%Module example

class Circle
{
%HeaderCode
#include "circle.h"
%End
public:
     int x, y;
};

...would generate a module where you...

import example

c = example.Circle()
c.x = 10
c.y = 20
del c

While documentation is obviously a problem, there are enough people on 
the PyKDE mailing list who have made the necessary blood sacrifices to 
be able to help.

Phil




From duncan-news at grisby.org  Tue Sep 24 04:40:42 2002
From: duncan-news at grisby.org (Duncan Grisby)
Date: Tue, 24 Sep 2002 08:40:42 GMT
Subject: ANN: omniORB 4.0.0 and omniORBpy 2.0
Message-ID: <eyVj9.87$8d4.368@news13-win.server.ntlworld.com>

I am pleased to announce the release of omniORB 4.0.0 and omniORBpy
2.0.

omniORB is a robust, high performance CORBA ORB for C++. omniORBpy is
a version for Python. They are freely available under the terms of the
GNU Lesser General Public License (for the libraries), and GNU General
Public License (for the tools). For more information, see

  http://omniorb.sourceforge.net/

Highlights of the new features since omniORB 3.0 and omniORBpy 1:

 - Updated to CORBA 2.6.
 - Support for GIOP 1.1 and 1.2.
 - Wide string and codeset negotiation.
 - Unix domain and SSL transports.
 - Bidirectional GIOP.
 - Flexible thread pool mode.
 - PortableServer::Current.
 - Interceptors.
 - Fixed point.
 - Complete freeing of all heap allocations.
 - Efficient in-process calls between C++ and Python.
 - Python / C++ object reference translation API.
 - New comprehensive configuration mechanism.
 - Commercial support. See http://www.omniorb-support.com/

Release notes are available from

  http://sourceforge.net/project/shownotes.php?release_id=112373

and

  http://sourceforge.net/project/shownotes.php?release_id=112375


You can download the releases themselves via the release notes pages,
or from

  http://sourceforge.net/project/showfiles.php?group_id=51138


Enjoy!

Duncan.

-- 
 -- Duncan Grisby         --
  -- duncan at grisby.org     --
   -- http://www.grisby.org --


From drifty at bigfoot.com  Mon Sep 16 00:00:11 2002
From: drifty at bigfoot.com (Brett C.)
Date: 15 Sep 2002 21:00:11 -0700
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Sep  9)
References: <D5F9F6F5A0C46CAC.9DCA346C4331EBAF.B9AA42F7FC35A33A@lp.airnews.net>
Message-ID: <f85b5ab1.0209152000.6b04ff67@posting.google.com>

<big snip>
>     Michael Hudson continues Andrew Kuchling's marvelous tradition
>     of summarizing action on the python-dev mailing list once every
>     other week.
<snip>

Uh, actually, I (Brett Cannon) am doing them now.  Michael has decided
to not come out of retirement.  The last thing we would need is him
blowing his fingers out dealing with the summary twice a month.  =)

-Brett C.


From aleax at aleax.it  Fri Sep 27 06:57:37 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 27 Sep 2002 10:57:37 GMT
Subject: comparing all values of a list to regex
References: <mailman.1032947731.3297.python-list@python.org> <mailman.1033039492.1772.python-list@python.org> <0lUk9.149240$ub2.3291527@news1.tin.it> <mailman.1033118078.9248.python-list@python.org>
Message-ID: <BQWk9.150207$ub2.3303056@news1.tin.it>

Manuel Hendel wrote:

> Sorry for forgetting some details.
> 
> On Fri, Sep 27, 2002 at 08:07:24AM +0000, Alex Martelli wrote:
>> 
>> def classify(login_email):
>>     l_e = login_email.split(',')
> 
> I think that I nearly understand what you are doing here. But I don't
> understand the assert part. I already checked at the documentations
> but this doesn't helped me at all.
> 
>>     assert 1 <= len(l_e) <= 2, "More than one comma in (%s)" %
>>     login_email


assert is just a sanity check.  I want the program to bail out
with a clear error message if it's ever presented with an input
file that does NOT respect the specs you've given (the only
ones which the program is prepared to handle), rather than
keep chugging and producing meaningless, misleading results.

assert is nice to use for this purpose, because when you run
the program with python -O it silently goes away and you pay
no performance price whatsoever for having assert there.

By the same token, assert may not be the right statement in
this case, because I'm validating sanity of *input*, NOT
sanity of my program.  Even when the program is fully valid
it may still be presented with invalid input, and it needs
to scream when that happens.

So, you may prefer to change the assert into, e.g.:

    L = len(l_e)
    if L<1 or L>2:
        raise ValueError, "%s commas in (%r)" % (L-1, login_email)


Alex



From claird at starbase.neosoft.com  Thu Sep  5 20:24:26 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 5 Sep 2002 19:24:26 -0500
Subject: newbie problem: use socke lib to retrieve one web page:
References: <mailman.1031240588.2234.python-list@python.org>
Message-ID: <D613C81A8A3A93B8.5464F563C49215DF.E1058876104EFE5E@lp.airnews.net>

In article <mailman.1031240588.2234.python-list at python.org>,
Erik Price  <erikprice at mac.com> wrote:
>
>On Wednesday, September 4, 2002, at 11:53  PM, koko wrote:
>
>> I write this to retrieve one web page using socket lib.
			.
			.
			.
>Works for me:
>
> >>> import socket
> >>> host = 'www.uic.edu'
> >>> port = 80
> >>> s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
> >>> s.connect((host, port))
> >>> header = """HEAD /home/events.shtml HTTP/1.0
>... From: hh at uic.edu
>... User-Agent: test/1.0
>...
>... """
> >>> s.send(header)
>72
> >>> data = s.recv(4096)
> >>> print data
>HTTP/1.1 200 OK
>Date: Thu, 05 Sep 2002 15:40:00 GMT
>Server: Apache/1.3.26 (Unix) PHP/4.1.2 mod_perl/1.27 mod_ssl/2.8.10 
>OpenSSL/0.9.6
>Connection: close
>Content-Type: text/html
>
>
> >>> s.close()
>
>
>I used the HEAD method instead of GET for brevity.  But GET works too.
			.
			.
			.
To criticize code that's already giving satisfaction
is nearly beyond me.  I'll point out, though, that
the original questioner might consider this alterna-
tive which has, I believe, evident advantages for
long-term maintenance:
  from urllib import urlopen

  URL = "http://www.uic.edu"
  page = urlopen(URL).read()
  print page
Note that urllib is one of the batteries the standard
Python distribution includes.

My summary:  use higher-order facilities when applicable.
-- 

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


From mal at lemburg.com  Mon Sep 23 06:44:29 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Mon, 23 Sep 2002 12:44:29 +0200
Subject: ANN: eGenix.com mx BASE Extension Package 2.0.4
Message-ID: <3D8EF08D.204@lemburg.com>

________________________________________________________________________

ANNOUNCING:

              eGenix.com mx BASE Extension Package for Python
                              Version 2.0.4

        Open Source Python extensions providing important and useful
                    services for Python programmers.


________________________________________________________________________

WHAT IS IT ?:

The eGenix.com mx BASE Extensions for Python are a collection of
professional quality software tools which enhance Python's usability
in many important areas such as fast text searching, date/time
processing and high speed datatypes.

Python is an object-oriented Open Source programming language which
runs on all modern platforms (http://www.python.org/). By integrating
ease-of-use, clarity in coding, enterprise application connectivity
and rapid application design, Python establishes an ideal programming
platform for todays IT challenges.

The tools have a proven record of being portable across many Unix and
Windows platforms. You can write applications which use the tools on
Windows and then run them on Unix platforms without change due to the
consistent platform independent interfaces.

All available packages have proven their stability and usefulness in
many mission critical applications and various commercial settings all
around the world.

________________________________________________________________________

WHAT'S NEW ?

The RPM packages were updated to install to /usr/ instead
of /usr/local/ to comply with the Linux Standard Base (LSB)
and to be compatible with the Python RPMs which are available
from python.org.

As always we are providing precompiled versions of the package for
Windows and Linux as well as sources which allow you to install the
package on all other supported platforms.

________________________________________________________________________

EGENIX.COM MX BASE PACKAGE OVERVIEW:

mxDateTime - Generic Date/Time Types

       mxDateTime is an extension package that provides three new object
       types, DateTime, DateTimeDelta and RelativeDateTime, which let
       you store and handle date/time values in a much more natural way
       than by using ticks (seconds since 1.1.70 0:00 UTC; the encoding
       used by the time module).

       You can add, subtract and even multiply instances, pickle and
       copy them and convert the results to strings, COM dates, ticks
       and some other more esoteric values. In addition, there are
       several convenient constructors and formatters at hand to greatly
       simplify dealing with dates and times in real-world applications.

       In addition to providing an easy-to-use Python interface the
       package also exports a comfortable C API interface for other
       extensions to build upon. This is especially interesting for
       database applications which often have to deal with date/time
       values (the mxODBC package is one example of an extension using
       this interface).

mxTextTools - Fast Text Processing Tools

       mxTextTools is an extension package for Python that provides
       several useful functions and types that implement
       high-performance text manipulation and searching algorithms in
       addition to a very flexible and extendable state machine, the
       Tagging Engine, that allows scanning and processing text based on
       low-level byte-code "programs" written using Python tuples. It
       gives you access to the speed of C without the need to do any
       compile and link steps every time you change the parsing
       description.

       Applications include parsing structured text, finding and
       extracting text (either exact or using translation tables) and
       recombining strings to form new text.

mxStack - Fast and Memory-Efficient Stack Type

       mxStack is an extension package that provides a new object type
       called Stack. It works much like what you would expect from such
       a type, having .push() and .pop() methods and focusses on
       obtaining maximum speed at low memory costs.

mxTools - Collection of Additional Builtins

       mxTools is an extension package that includes a collection of
       handy functions and objects giving additional functionality in
       form of new builtins to the Python programmer.

       The package auto-installs the new functions and objects as
       builtins upon first import. This means that they become
       instantely available to all other modules without any further
       action on your part. Add the line import NewBuiltins to your
       site.py script and they will be available to all users at your
       site as if they were installed in the Python interpreter itself.

mxProxy - Generic Proxy Wrapper Type

       mxProxy is an extension package that provides a new type that is
       suitable to implement Bastion like features without the need to
       use restricted execution environments.

       The type's main features are secure data encapsulation (the
       hidden objects are not accessible from Python since they are
       stored in internal C structures), customizable attribute lookup
       methods and a cleanup protocol that helps in breaking circular
       references prior to object deletion.

       The latest version adds a very interesting new feature: weak
       references which help you work with circular references in a way
       that doesn't cause memory leakage in a Python system. Note that
       even though Python 2.1+ has its own weak reference implemetation,
       this package can be used to write applications which also work on
       Python 1.5.2 and 2.0.

mxBeeBase - On-disk B+Tree Based Database Kit

       mxBeeBase is a high performance construction kit for disk based
       indexed databases. It offers components which you can plug
       together to easily build your own custom mid-sized databases (the
       current size limit is sizeof(long) which gives you an address
       range of around 2GB on 32-bit platforms).

       The two basic building blocks in mxBeeBase are storage and
       index. Storage is implemented as variable record length data
       storage with integrated data protection features, automatic data
       recovery and locking for multi process access. Indexes use a high
       performance optimized B+Tree implementation built on top of
       Thomas Niemann's Cookbook B+Tree implementation
       (http://epaperpress.com/).

________________________________________________________________________

WHERE CAN I GET IT ?

The download archives and instructions for installing the packages can
be found at:

       http://www.egenix.com/

________________________________________________________________________

WHAT DOES IT COST ?

The BASE package comes with a Python 2.0 style license, which means
that you can use it in both commercial and non-commercial settings
without fee or charge. The package comes with full source code.


________________________________________________________________________

WHERE CAN I GET SUPPORT ?

Commercial quality support for these packages is available from
eGenix.com Software GmbH. Please see

http://www.egenix.com/files/python/eGenix-mx-Extensions.html#Support

for details about the eGenix support offerings.

________________________________________________________________________


Enjoy,
-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/







From brian at sweetapp.com  Wed Sep  4 18:05:14 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Wed, 04 Sep 2002 15:05:14 -0700
Subject: str() of a tuple
In-Reply-To: <20020904145034.S7096-100000@bisquick.cs.washington.edu>
Message-ID: <000a01c2545f$2567ef70$df7e4e18@brianspiv1700>

> > (Before I tell you, could you please tell me what this tuple
contains? I
> > bet that you can't figure it on in less than, say, 5, days)
> 
> Huh?  The second sentence especially does not parse.

I'm trying to cleverly say that, if str() were used recursively, item
boundaries would be ambiguous.

Cheers,
Brian




From max at alcyone.com  Mon Sep 23 22:37:39 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 23 Sep 2002 19:37:39 -0700
Subject: confusion about opening files
References: <3d8fcda4$0$198$75868355@news.frii.net>
Message-ID: <3D8FCFF3.FD52AFE5@alcyone.com>

robie1373 wrote:

	...
>     file = os.open("e:\\d12", "r")
> TypeError: an integer is required
> 
> I get this error if the mode is r, r+, a+ or whatever.  I also get it
> whether the file exists or not. Can someone explain what I am doing
> wrong?

You meant open, not os.open.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Insight -- the titillating knack for hurting!
\__/ Colette
    Erik Max Francis' bookmarks / http://www.alcyone.com/max/links/
 A highly categorized list of Web links.


From jiba at tuxfamily.org  Sat Sep 28 16:24:31 2002
From: jiba at tuxfamily.org (Jiba)
Date: Sat, 28 Sep 2002 22:24:31 +0200
Subject: watching mutables?
References: <an1p3c$jrd$1@news.hccnet.nl>
Message-ID: <3d96f25a$0$498$7a628cd7@news.club-internet.fr>

Anton Vredegoor wrote:

> Hello all,
> 
> here's a function I would like to have:
> 
> <hypothetical code>
> 
>>>> m = [1,2]
>>>> m
>>>> [1,2]
>>>> from watcher import watch
>>>> watch(m)
>>>> m.append(3)
>>>> "hey, you changed variable m!, new value is [1,2,3]"
> 
> </hypothetical code>
> 
> Is this possible? How would this be called?
> 
> Anton.

Try EditObj (http://oomadness.tuxfamily.org/en/editobj) and look at the 
module editobj.eventobj.

It does exactely what you need, but it is quite a hack...

Jiba


From ianb at colorstudy.com  Tue Sep 17 02:51:17 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 17 Sep 2002 01:51:17 -0500
Subject: Stackless Python, eventual merge?
In-Reply-To: <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de>
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com>
	<mailman.1032107006.6411.python-list@python.org>
	<am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net>
	<j4znuio0kr.fsf@informatik.hu-berlin.de>
	<mailman.1032204706.14737.python-list@python.org>
	<m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de>
	<1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> 
	<m34rcpcfuk.fsf@mira.informatik.hu-berlin.de>
Message-ID: <1032245477.2444.1465.camel@dsl-65-184-205-5.telocity.com>

On Tue, 2002-09-17 at 01:03, Martin v. Loewis wrote:
> Ian Bicking <ianb at colorstudy.com> writes:
> 
> > But it doesn't seem like complexity was ever the fundamental problem --
> > the complexity could have been dealt with, but I don't believe there was
> > ever any indication that if it was simplified or documented that it
> > would get into standard Python.  
> 
> That is not true. I have indicated this atleast once, namely when I
> studied the code. For me, the complicatedness was always the problem -
> complexity is fine, since this is a complex problem.

Was there some indication that Stackless would be integrated if the code
was cleaned up, documented, or otherwise simplified?  I don't believe
there was, but I may be wrong.  

Without such an indication, there is no reason to think that mere
technical changes to Stackless will lead to it being included in
standard Python.

  Ian





From pinard at iro.umontreal.ca  Sat Sep  7 19:07:38 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Sat, 07 Sep 2002 19:07:38 -0400
Subject: Larry Wall's comment on python...
In-Reply-To: <200209071429.26401.shalehperry@attbi.com> ("Sean 'Shaleh'
 Perry"'s message of "Sat, 7 Sep 2002 14:29:26 -0700")
References: <mailman.1031377930.4225.python-list@python.org>
	<1549197.0PZNOebDTH@cartman>
	<200209071429.26401.shalehperry@attbi.com>
Message-ID: <oqit1hv1ol.fsf@titan.progiciels-bpi.ca>

[Sean 'Shaleh' Perry]
> [yka]
>> [Kow Kuroda]

>> > [...] For example, vi and emacs indent by 8 characters by default

>> [...] Well at vim :set ts=4 makes more fitting

> actually emacs defaults to 4 space indenting [...]

The Python mode writers for Emacs managed to automatically recognise the
indentation in use in a file, and continue editing that file with the same.
For new files, I get four spaces per level.

However, for Python files indented with TABs only, Emacs cannot successfully
guess if the author's intent was to expand those TAB with four or eight
spaces.  By default, I vaguely remember it picks eight.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



From =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?=  Thu Sep 12 18:21:42 2002
From: =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?= (=?iso-8859-5?Q?=B4=D0=DC=F8=D0=DD_=B3=2E?=)
Date: Fri, 13 Sep 2002 00:21:42 +0200
Subject: [ANN] PySQLite 0.3.0 released
References: <mailman.1031850702.838.python-list@python.org>
Message-ID: <3d811376$1@news.mt.net.mk>

>> for row in cursor:
>>        print "%14s, %15s, %19s, %8s, %25s" % tuple(row)
>> ?
>> 
>> Terry J. Reedy
>> 

> You can already do it, using iter():

> for row in iter(cursor, None):
> 	print ....

you mean:

for row in iter(cursor.fetchone, None):
	print ...
	


-- 
??????

	In a world without walls and fences,
	do we care about Windows and Gates ?



From pereira at cis.upenn.edu  Fri Sep  6 20:57:29 2002
From: pereira at cis.upenn.edu (Fernando Pereira)
Date: Fri, 06 Sep 2002 20:57:29 -0400
Subject: Python class ( Prolog ) ???
References: <20020905234600.11264.60328.Mailman@mail.python.org> <3D7914FD.5D713A48@anansispaceworks.com> <mailman.1031353570.17391.python-list@python.org>
Message-ID: <B99EC739.1232B%pereira@cis.upenn.edu>

On 9/6/02 7:06 PM, in article
mailman.1031353570.17391.python-list at python.org, "Terry Hancock"
<hancock at anansispaceworks.com> wrote:
> Can
> anyone fluent in both compare what sort of applications
> you might apply a prolog-python solution to?  Where
> would you draw the line between the two? What would
> using prolog help me with? What's a good problem to
> solve as a learning project?  Any more book or website
> recommendations for a learner?
While a superficial connection between Python and Prolog is possible, I
don't think that's particularly interesting. It's the kind of thing you
would do if you had code in both languages that you wanted to glue together.
I think that you are looking for closer connections. That's not easy, but
there are some interesting examples of logic-programming ideas being married
with other languages for mutual benefit. A good example are the constraint
programming engines of ILOG (www.ilog.com) that mesh with OO languages like
C++ and Java. A more researchy example are hybrid languages like Oz/Mozart
(http://www.mozart-oz.org/) that combine logic, OO and functional features
in an interesting package. Finally, there are several nice embeddings of
logic or constraint programming in Lisp-family languages, starting from the
tutorial examples in SICP (http://www-mitpress.mit.edu/sicp/) and including
systems like Screamer (ftp://ftp.ecn.purdue.edu/qobi/ircs-93-03.ps.Z).
There's a lot more along these lines, but these are just those that came to
mind right away. I think that Python would be a good language to host
something along those lines, but of course someone would have to do the
work, and if you want reasonable efficiency as well as power you'll have to
work pretty hard. 

-- F
 



From loewis at informatik.hu-berlin.de  Tue Sep 17 03:13:03 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 17 Sep 2002 09:13:03 +0200
Subject: file mode attributes...
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E2019201253349@admin56.narex.com>
References: <60FB8BB7F0EFC7409B75EEEC13E2019201253349@admin56.narex.com>
Message-ID: <j4bs6xks1s.fsf@informatik.hu-berlin.de>

"Bjorn Pettersen" <BPettersen at narex.com> writes:

> Thanks. Do you think there would be any interest in adding these
> kinds of utility functions to e.g. the os.path module?

I doubt that. Your implementation is valid only on Windows, on Unix,
you need to determine whether you are interested in user, group, or
other bits, and you also have not only read and write, but also
execute. The current API is as portable and convenient is it can get.

Also, finding out that way that a file is writable does not mean you
can write to it:
- the file might be on a read-only file system (will report EROFS, on
  Unix),
- the Win32 ACL might prevent writing,
- on Unix, you need to get your owner and group id,
- the Unix euid/fsuid might prevent writing, as it might differ from
  your user id,
- the file system might be full.

So the only way to truly find out whether you can read is to attempt
to read or write, and prepare for an exception.

> win32all isn't standard on Windows <wink>? 

No; that's a huge code base, and nobody has proposed to incorporate it
into Python proper.

Regards,
Martin




From aleax at aleax.it  Tue Sep 17 06:05:21 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 17 Sep 2002 10:05:21 GMT
Subject: faster way for adding many strings
References: <am6tgt$lqm$01$1@news.t-online.com>
Message-ID: <B7Dh9.119326$pX1.4297177@news2.tin.it>

Ulli Stein wrote:

> What is the faster way: appending many strings to a list and then joining
> them, or writing to
> StringIO and then getvalue()?

Best ways to answer such questions is to try!

For example, put in timit.py:

import time, cStringIO, array

manystrings = [str(x) for x in xrange(10000)]
repeater = [None] * 100


def directly(manystrings):
    return ''.join(manystrings)

def withappend(manystrings):
    alist = []
    for x in manystrings: alist.append(x)
    return ''.join(manystrings)

def withwritelines(manystrings):
    aux = cStringIO.StringIO()
    aux.writelines(manystrings)
    return aux.getvalue()

def withwrite(manystrings):
    aux = cStringIO.StringIO()
    for x in manystrings: aux.write(x)
    return aux.getvalue()

def witharray(manystrings):
    aux = array.array('c')
    for x in manystrings:
        aux += array.array('c', x)
    return aux.tostring()

def timit(func):
    start = time.clock()
    for x in repeater:
        func(manystrings)
    stend = time.clock()
    return '%.2f %s' % (stend-start, func.__name__)

for func in directly, withappend, withwritelines, withwrite, witharray:
    print timit(func)


Now run it:

[alex at lancelot examples]$ python -O timit.py
0.21 directly
1.43 withappend
0.26 withwritelines
1.11 withwrite
19.51 witharray
[alex at lancelot examples]$ python -O timit.py
0.22 directly
1.48 withappend
0.22 withwritelines
1.15 withwrite
19.67 witharray

>From this we see that, on my machine, if I have the strings
to join already, joining them directly is a bit faster; but
if I have to append them as they come, then cStringIO is a
bit faster -- nothing much to it one way or another, but may
perhaps be meaningful in a hot-spot of a program.  The array
solution, as coded above, is clearly out of the fray (but
might be interesting if you had some estimation of total
size and could size the result array directly, rather than
have to use += -- I'll leave that to you).

Having a way to measure means you can try directly on the
platforms and python versions that matter to you.  Such timing
issues may often change between releases and/or platforms,
after all!


Alex



From uwe.schmitt at procoders.net  Fri Sep 27 10:46:10 2002
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 27 Sep 2002 14:46:10 GMT
Subject: __getitem__() in python2.2
Message-ID: <an1qvi$mpi29$1@hades.rz.uni-sb.de>

Hi,

in Python 2.1 one could define

   class X:
       def __getitem__(self, par): print par

and

   a=X()
   a[1,2,3]

resulted in

   (1,2,3)

If I change to newstyle classes, __getitem__ does not accept
tuples anymore. Is there a good reason for this behaviour ?
Can i simmulate the v2.1 functionality ?? It was quite usefull
to access multidimensional matrices...

Greetings, Uwe.


-- 
Uwe Schmitt                    Computer science is no more about Computers,
dr.schmitt at procoders.net       than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           


From python at rcn.com  Tue Sep 10 01:49:06 2002
From: python at rcn.com (Raymond Hettinger)
Date: Tue, 10 Sep 2002 01:49:06 -0400
Subject: IDLE
References: <wpbf9.264320$aA.45691@sccrnsc02>
Message-ID: <alk170$pai$1@bob.news.rcn.net>

"Ali K" <akakakhel at attbi.com> wrote in message
news:wpbf9.264320$aA.45691 at sccrnsc02...
> Why won't my IDLE work?

This is the documented behavior according
to the American Heritage Dictionary:

idle (intr. verb):  To pass time without working
                            or while avoiding work.






From aleax at aleax.it  Tue Sep 24 10:26:23 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 24 Sep 2002 14:26:23 GMT
Subject: Three dumb questions (ordered by dumbness descending)
References: <amnvit$7kqdt$1@ID-77524.news.dfncis.de> <82Zj9.136094$ub2.2966822@news1.tin.it> <m3smzzo97h.fsf@intevation.de>
Message-ID: <jC_j9.158631$pX1.5703764@news2.tin.it>

Bernhard Herzog wrote:
        ...
>> def shuffle_ian(x):
>>     newList = [(random(), i) for i in x]
>>     newList.sort()
>>     x[:] = [i[1] for i in newList]
> 
> What's missing, though, is a proof that shuffle_ian produces each
> possible permutation with the same probability (assuming random() has
> suitable properties)

Among the "suitable properties" would be to have random() return
real numbers -- thus the probability of two equal numbers in N
calls becomes 0 -- or come from a pseudo-random generator arranged in a
plain/usual way so that it will never return two identical numbers
until you use up its cycle (e.g. a typical linear-congruential one).


> One problem I can see is that there's a non-zero chance that there will
> be at least two equal random numbers in newList in which case the
> corresponding values of x are used to sort newList. This skews the
> probabilities and furthermore means that shuffle_ian can only shuffle
> lists whose items can be ordered which among others exludes complex
> numbers.

Good point, of course -- an arbitrary pseudo-random (or HW random)
generator returning output with a finite number of bits DOES have a
non-0 chance of a duplicate, in general -- e.g., (1-2**53)**N prob of
no problems if 53 bits differ significantly and you generate N+1
truly-random numbers.


Alex



From g0w1s0 at yahoo.ca  Sun Sep  1 20:52:16 2002
From: g0w1s0 at yahoo.ca (Cesar)
Date: Sun, 01 Sep 2002 20:52:16 -0400
Subject: curses: AttributeError: initscr
Message-ID: <un5dhs4f7qlte8@corp.supernews.com>


I'm running python 1.5/2.2.1 under redhat linux 7.2.

I'm trying to use the curses module like this :

1: import curses
2: 
3: screen = curses.initscr()
4:

and I get this error msg:

Traceback (innermost last):
  File "./curses_test.py", line 6, in ?
    screen = curses.initscr()
AttributeError: initscr

any idea ??


From deblnononospammmmmyyy at theworld.com  Thu Sep 19 02:19:19 2002
From: deblnononospammmmmyyy at theworld.com (David Lees)
Date: Thu, 19 Sep 2002 06:19:19 GMT
Subject: How to build Python 2.2.1 on Windows?
References: <mailman.1032402712.14232.python-list@python.org>
Message-ID: <3D896C68.39591E76@theworld.com>

Excellent.  I just built the core and python.exe out of the box with my
antique installation of VC++ Version 6.  And it runs on my Win98 box.

Thanks guys.

David


David LeBlanc wrote:
> 
> I have built Python on Windows using both the distribution and CVS download.
> The necessary VC++ project files are included in both. You will have to also
> download some other components as described in the readme file, including a
> special version of BSDDB version 1.85, zlib, Tcl/Tk (Tk needs Tcl to build),
> expat and maybe some other things. These are for extensions, and are not
> required to build the core Python.
> 
> WRT BSDDB, there are projects out there to build Sleepycat BSDDB version 4.0
> and that's a better choice then the default version 1.85 used by Python (due
> to bugs). The downside is that Sleepycat BSDDB is free for personal use
> only.
> 
> See \dist\Python-2.2.1\PCbuild for complete details.
> 
> Oh yes, put your python source distro or cvs dump under a directory called
> 'dist'. I found the directions to do that unclear in the readme and it gave
> me a bit of trouble. The other packages you'll need also go in 'dist' as
> peers of the python source tree. Those should be built before Python.
> 
> David LeBlanc
> Seattle, WA USA
> 
> > -----Original Message-----
> > From: python-list-admin at python.org
> > [mailto:python-list-admin at python.org]On Behalf Of David Lees
> > Sent: Wednesday, September 18, 2002 18:43
> > To: python-list at python.org
> > Subject: How to build Python 2.2.1 on Windows?
> >
> >
> > I have been playing with Python for awhile and would like to know how to
> > try building Python on a Windows machine that has MS C++ version 6.  Is
> > there a standard download that is configured to do this?  If not could
> > someone point me in the direction of information on builds?  I
> > downloaded Python-2.2.1.tgz.tar on my Win98 box and WinZip 8 does not
> > even seem to know how to unzip it.  I just transfered the tar file over
> > to a linux box and it unzips without and problem using tar.  The README
> > seems to only talk about linux/unix type systems.
> >
> > David Lees
> > --
> > http://mail.python.org/mailman/listinfo/python-list


From wlfraed at ix.netcom.com  Sat Sep 21 14:32:16 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Sat, 21 Sep 2002 11:32:16 -0700
Subject: Decimal arithmetic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <7xn0qcfafp.fsf@ruckus.brouhaha.com> <amgpm4$5si22$1@ID-125932.news.dfncis.de> <7xwupf52jx.fsf@ruckus.brouhaha.com>
Message-ID: <hvdima.ng3.ln@ix.netcom.com>

Paul Rubin fed this fish to the penguins on Friday 20 September 2002 
10:34 pm:


> 
> I do know that financial modelling software I worked on that was sold
> to banks used floating point.  Also, spreadsheets use floating point.
>
        There's probably a subtle distinction between "modelling" (to me that 
implies something that approximates expected behavior for rapid "what 
if" scenarios) and accounting type applications.
 
> But it's possible (likely even) that some bookkeeping applications use
> BCD.  One that I worked on actually used exact, multi-precision
> rational arithmetic, but I think that's unusual.
>
        If it is done in COBOL that's pretty much a given... Though most COBOL 
compilers /have/ added things like "usage is computation" -- wherein 
the programmer explicitly has to request a binary value rather than 
normal BCD.

        Though I must be mis-interpreting the book I have -- it implies that 
the default for Fujitsu Cobol v4 is "Display" (8-bit characters), and 
one would have to ask for "packed-decimal" in a usage clause. COMP for 
floating point, and BINARY for integer.

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From plusk at radford.edu  Thu Sep 19 20:45:10 2002
From: plusk at radford.edu (Paul D. Lusk)
Date: 20 Sep 2002 00:45:10 GMT
Subject: Cookbook - multireplace
Message-ID: <Xns928ED324AE142ID134675userdfncisd@130.133.1.4>

This is more an exercise in sub-classing a dict than a speed optimization.
It may not even be a speed optimization, since I didn't time it.


# Original algorithm by Xavier Defrang.
# http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/81330
# and modified by alane at sourceforge.net
# This implementation by plusk at radford.edu

import re
 
class NotImplemented(Exception):
     pass
 
class textsub(dict):
    __slots__ = ('pat','regex','flags')
    def __init__(self, *ds):
        self.pat = None
        self.regex = None
        self.flags = None
        dict.__init__(self)
        for d in ds:
            self.update( d )
    def popitem(self):
        """popitem doesn't make sense for this class"""
        raise NotImplemented
    def setdefault(self,key,default):
#       print 'set default 1 %s %s' % (key, default)
        new = not dict(self).has_key(key)
        val = dict(self).get(key,default)
#       print 'set default 2 %s %s %s' % (key,default, val)
        if new:
            dict.__setitem__(self,key,default)
            self.compile()
        return val
    def update(self,newdict):
        dict.update(self,newdict)
        self.compile()
    def __setitem__(self,key,val):
        new = not dict(self).has_key(key)
        dict.__setitem__(self,key,val)
        if new:
            self.compile()
    def __delitem__(self,key):
        dict.__delitem__(self,key)
        self.compile()
    def compile(self):
#        print 'compile called'
        if dict(self) and len(dict(self)) > 0:
            tmp = "(%s)" % "|".join(map(re.escape,
                                        dict(self).keys()))
            if self.pat != tmp:
                self.pat = tmp
                if self.flags:
                    self.regex = re.compile(self.pat,self.flags)
                else:
                    self.regex = re.compile(self.pat)
                    
    # not called from outside, but is needed by
    # the sub below
    #  self.regex.sub(self,s) is regex.sub(func=self.__call__,s)
    def __call__(self, match):
        return dict(self)[match.group()]

    def sub(self, s):
        if len(dict(self)) == 0:
            return s
        return self.regex.sub(self, s)


if __name__ == "__main__": 
  text = "As you know, Bob, Larry Wall is the creator of Perl"
 
  adict = {
    "Larry Wall" : "Guido van Rossum",
    "creator" : "Benevolent Dictator for Life",
    "Perl" : "Python"
  } 
  print 'start'
  xlat = textsub(adict)
  print 'Fred for Bob'
  xlat["Bob"] = "Fred"
  print 'set default is'
  xlat.setdefault('is','really is')
  print xlat.sub(text)
  print """set default Bob, xlat['Bob'] is 'Fred'"""
  print xlat.setdefault('Bob','Jay')
  print xlat.sub(text)
  print """del xlat['Bob'], setdefault"""
  del xlat['Bob']
  print xlat.setdefault('Bob','Jay')
  print xlat.sub(text)
  print 'del creator'
  del xlat["creator"]
  print xlat.sub(text)
  print 'popitem'
  try:
      dummy1,dummy2 = xlat.popitem()
  except NotImplemented:
      print 'popitem not implemented'
  else:
      print 'oops'
  print 'clear'
  xlat.clear()
  print xlat.sub(text)



From tjreedy at udel.edu  Mon Sep 30 13:34:16 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 30 Sep 2002 17:34:16 GMT
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <ofLk9.304267$z91.13145351@bin3.nnrp.aus1.giganews.com> <dITk9.175519$pX1.6209886@news2.tin.it> <t_3l9.334924$AR1.14880141@bin2.nnrp.aus1.giganews.com> <oOZl9.162526$ub2.3640508@news1.tin.it>
Message-ID: <sW%l9.47746$Ii4.2360264@bin2.nnrp.aus1.giganews.com>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:oOZl9.162526$ub2.3640508 at news1.tin.it...
> Terry Reedy wrote:
> > If b is a non-null constant, there is no such risk.  Consider
>
> If it's a constant, you don't need short-circuiting!

If the alternative is *also* a constant, correct; indexing works fine.
If it is not, the short-cicuiting of 'or' may be crucial.  Example:
y==0 and NAN or x/y.  Reversing the test and alternatives won't work
if x might be 0.

(Yes, I might prefer to let x/y raise an exception. Or I might write a
function fdiv(x,y).   But for one time use or in the innermost loop of
a calculation, the quickly written condtional might be just what I
want.)

> def plur(*ns):
>     for n in ns:
>         yield n
>         yield 's'[n!=1:]

Using the conditional in a slice rather than an index to get a null
alternative is a cute trick I had not thought of.

Posting a concrete usage for and/or an pulled out nice alternatives
from 3 or 4 different people.  Let's try a couple of others:

In the middle of an expression, I want the 7th or 3rd item in seq
depending on <cond> (an expression yielding 0 or 1).  Five
alternatives (where ... ... is usage context):

...seq[<cond> and 2 or 6]...
...seq[(6,2)[<cond>]]...
...seq[6-4*<cond>]...
...seq[2+(not<cond>)*4]...  # this generalizes better -- see below

if <cond>: tem = 2
else         : tem = 6
...tem...

Anything else?  Among these four I would probably prefer the first,
but don't care if others choose elsewise.  Now for a three-way split:

...seq[<cond1> and 2 or <cond2> and 6 or 3]... # abbreviating

if     <cond1>: tem = 2
elif  <cond2>: tem = 6
else               : tem = 3
...tem...

Assuming <cond2> doesn't need short-cicuiting when <cond1> is true and
doesn't take significantly long to calculate when not needed, we can
again substitute tuple indexing (or direct index calculation), but I
think it would take most people longer to correctly write any
alternative below, at least the first time.  Is it worth it to avoid
the easy-to-write and easy-to-read-and-understand conditional
expression?  Not for me, I think, but others are welcome to differ.

...seq[((3,6)[<cond2>],2)[<cond1>]]... #or
...seq[(3,6,2,2)[2*<cond1> + <cond>]]...
...seq[2 + (not<cond1>)*(1+3*<cond2>)]...
...seq[2 + (not<cond1>)*(4-(not<cond2>)*3)]...

Terry J. Reedy





From shalehperry at attbi.com  Thu Sep 12 01:31:20 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 11 Sep 2002 22:31:20 -0700
Subject: [FEEDBACK] Is this script efficient...is there a better way?
In-Reply-To: <HPRf9.1780$Lg2.247316@news2.news.adelphia.net>
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <mailman.1031781254.1680.python-list@python.org> <HPRf9.1780$Lg2.247316@news2.news.adelphia.net>
Message-ID: <200209112231.20706.shalehperry@attbi.com>

On Wednesday 11 September 2002 18:09, Bob X wrote:
> > 2) do you expect to find more than one keyword in a particular line?  If
> > not you could save some iterations by stopping the inner line.find() loop
> > as soon as one item is found.
>
> I could but it doesn't matter after the first on a line. How would I
> stop it?
>

use 'break'

for i in range(1,20):
    if i == 6:
        break
    print i

and you get:
1
2
3
4
5

so your code becomes:

for line in inFile.readlines():
     for badword in kw:
         if line.find(badword) > -1:
             result = '%s %s' % (badword, line)
             print result            # Print the result
             outFile.write(result)   # Write the result
             break                   # only need one keyword on a line

there is also the keyword 'continue' which says "I am done with this loop 
iteration, go back to the top of the loop".

for i in range(1,20):
    if (i % 2) == 0:
        continue
    print i

this loop will only print the odd numbers.



From macdict at optushome.com.au  Wed Sep 18 08:34:58 2002
From: macdict at optushome.com.au (Richard Tardif)
Date: Wed, 18 Sep 2002 22:34:58 +1000
Subject: minimizing Tk programs to the taskbar
Message-ID: <3D8872F1.87850856@optushome.com.au>

Hello all,

Is there any way to minimize programs to the taskbar? I'm talking about
windows here. By taskbar I mean the area down in the bottom right
corner, near the clock, that shows icons for what proggies you've got
running. Thanks for any help.



From a_salnikov at yahoo.com  Wed Sep 25 19:26:01 2002
From: a_salnikov at yahoo.com (Andy Salnikov)
Date: Wed, 25 Sep 2002 16:26:01 -0700
Subject: a python puzzle
References: <Pine.LNX.4.44.0209251736470.16806-100000@hydrogen.sabren.com> <mailman.1032992492.8970.python-list@python.org>
Message-ID: <amtgmd$c00$1@usenet.Stanford.EDU>

"Gary Herron" <gherron at islandtraining.com> wrote in message
news:mailman.1032992492.8970.python-list at python.org...
> On Wednesday 25 September 2002 02:42 pm, Michal Wallace wrote:
> > Here's a python puzzle. Solve it and you win a prize. :)
> >
>
> I decrypted (building the strings up character by character) with:
>
>    tr hokdrawfezxmylqcubijHpvsIFgt importysngalejufzvchIwxbCSkd < puzzle
>
> and got:
>
>
> The fact that this puzzle was easy to decrypt says something (good)
> about Python: It is so readable, that even encrypted, one can
> recognize significant portions.  Imagine decrypting a similar puzzle
> in C or, worse yet, Perl!
>

  C'mon, "the fact that this puzzle was easy to decrypt says" that
"encription" was really trivial. Indeed the most "important" spaces,
and non-alpha characters were not encripted at all. With such
simple encription you'll be able as easily decrypt C and Perl (or
any English text of suffitient length)

  Cheers,
  Andy.




From tjreedy at udel.edu  Thu Sep 12 09:26:10 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 12 Sep 2002 13:26:10 GMT
Subject: "str.contains(part)" or alternatives?
References: <B43D149A9AB2D411971300B0D03D7E8BF0A592@natasha.auslabs.avaya.com> <mailman.1031804415.15646.python-list@python.org>
Message-ID: <SB0g9.593193$m91.23940817@bin5.nnrp.aus1.giganews.com>

"Joseph A. Knapka" <jknapka at earthlink.net> wrote in message
news:mailman.1031804415.15646.python-list at python.org...
> Ah. So we have:
>
> not(X in Y <==> X in [Z for Z in Y])
>
> But only if Y is a string. That strikes me as a trifle odd,
> but nevertheless I agree that the "substring in string"
> behavior seems more useful.

As Tim said, this was debated on the dev list.  In the end, Guido
decided that giving 'in' slightly expanded semantics for strings and
strings only, though not 'pure', seemed very natural and practical and
no wierder than some other type-specific differences for other
operators.  While strings are sequences, they are different in that
they are *not* recursive, unlike tuples and lists.   A string of len>1
can be a substring of a string but it cannot be a 'member' of the
string.  On the other hand, [1,2,3] is both a sublist and item of
[1,2,3,[1,2,3]].  So "[1,2,3] in alist" would be ambiguious unless
restricted to one meaning or the other.

Terry J. Reedy





From maxm at mxm.dk  Fri Sep  6 03:47:57 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 06 Sep 2002 09:47:57 +0200
Subject: Python-dev summary for 2002-08-16 - 2002-09-01
References: <mailman.1031253706.7156.clpa-moderators@python.org>
Message-ID: <3D785DAD.5050905@mxm.dk>

Brett Cannon wrote:
> This is a summary of traffic on the python-dev mailing list between August
> 16, 2002 and September 1, 2002 (exclusive).  It is intended to inform the
> wider Python community of ongoing developments.  To comment, just post to
> python-list at python.org or comp.lang.python in the usual way.


Great stuff!!!

> =================================
> New PEP Format: reStructuredText
> =================================
> It has been suggested that the summaries try using reST; I am considering
> it.


LOL!


regards Max M



From wade at lightlink.com  Mon Sep 16 12:17:58 2002
From: wade at lightlink.com (Wade Leftwich)
Date: 16 Sep 2002 09:17:58 -0700
Subject: [xml] convert funny chars to char entites?
References: <vNah9.10101$Ee4.23987@news-server.bigpond.net.au>
Message-ID: <5b4785ee.0209160817.64b3842f@posting.google.com>

":B nerdy" <thoa0025 at mail.usyd.edu.au> wrote in message news:<vNah9.10101$Ee4.23987 at news-server.bigpond.net.au>...
> ive got a string "Il Est N?" and i want to put it into a xml document
> 
> is there a quick and easy way to conver the funny characters to char
> entities so i can store them?
> or is there existing libraries?
> 
> cheers

I've got a content syndication application that sends out XML, and
some of the recipients have a surprisingly hard time coping with
UTF-8. So I supply them with ASCII, after escaping everything over 7
bits to a character reference. This also guarantees proper display in
Internet Explorer.

>>> s = 'Il est n\x82!'
>>> u = unicode(s, 'iso-8859-1')
>>> L = []
>>> for char in u:
	val = ord(char)
	if val > 127:
		L.append('&#%s;' % val)
	else:
		L.append(char)

		
>>> u2 = ''.join(L)
>>> u2
u'Il est n&#130;!'
>>> s2 = u2.encode('ascii')
>>> s2
'Il est n&#130;!'


-- Wade Leftwich
Ithaca, NY


From tjreedy at udel.edu  Fri Sep  6 08:49:49 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 06 Sep 2002 12:49:49 GMT
Subject: Python "compiler" is  too slow for processing large data files???
References: <mailman.1031285708.5878.python-list@python.org>
Message-ID: <Nv1e9.208900$On.8191536@bin3.nnrp.aus1.giganews.com>

> list1 = [
>     (323, 870, 46, ),
>     (810, 336, 271, ),
>     (572, 55, 596, ),
>     (337, 256, 629, ),
>     (31, 702, 16, ),
> ]

 > Anyway, as my data files went from just a few lines, up to about
8000 lines
 > (with 10 values in each line for total of about 450KB of text), the
time to
 > 'exec' the file became too slow (e.g. 15 seconds) and used too much
memory
 > (e.g. 50MB) (using ms-windows, python 2.2.1).  It is the "compile"
phase,
 > because if I re-run, and there is *.pyc file available, the import
goes very
 > fast (no compilation required).

I am curious whether any C/C++ compilers (or those for other
languages) would have any problems with literal array of 10000
10-member structs?

Would the following go faster?  Leave off outer list lines (first and
last) to make more data like.  Then read file twice: once to count
lines and allocate list; second to populate list.

f = file('data')
lenf = len(f.readlines())
data = [0]*lenf
f.rewind()
for i in range(lenf):
  data[i] = maketuple(f.readline()) # substitute code for maketuple

Terry J. Reedy






From =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?=  Mon Sep 23 20:53:04 2002
From: =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?= (=?iso-8859-5?Q?=B4=D0=DC=F8=D0=DD_=B3=2E?=)
Date: Tue, 24 Sep 2002 02:53:04 +0200
Subject: list problem
References: <mailman.1032818129.8126.python-list@python.org>
Message-ID: <3d8fb76f@news.mt.net.mk>

> i have:	
> x=[['0020', '0000x9'], ['0030', '0000xa'], ['00B4', '0000x8'], ['0030', '0000xb'], ['00F0'], ['0001'], ['0003'], ['0005']]

> and i want:
> y=['0020', '0000x9', '0030', '0000xa', '00B4', '0000x8', '0030', '0000xb','00F0', '0001', 
> '0003', '0005']

> how to do this?

hmm, x[0]



-- 
??????

   If Bill Gates had a dime for every time a Windows box crashed...
                ...Oh, wait a minute, he already does.



From debl2nonospammmmyy at bellatlantic.net  Wed Sep 18 21:43:04 2002
From: debl2nonospammmmyy at bellatlantic.net (David Lees)
Date: Thu, 19 Sep 2002 01:43:04 GMT
Subject: How to build Python 2.2.1 on Windows?
Message-ID: <3D892BBD.5B38DD49@bellatlantic.net>

I have been playing with Python for awhile and would like to know how to
try building Python on a Windows machine that has MS C++ version 6.  Is
there a standard download that is configured to do this?  If not could
someone point me in the direction of information on builds?  I
downloaded Python-2.2.1.tgz.tar on my Win98 box and WinZip 8 does not
even seem to know how to unzip it.  I just transfered the tar file over
to a linux box and it unzips without and problem using tar.  The README
seems to only talk about linux/unix type systems.

David Lees


From lutz at rmi.net  Tue Sep 17 17:54:59 2002
From: lutz at rmi.net (Mark Lutz)
Date: Tue, 17 Sep 2002 15:54:59 -0600
Subject: Ann: Python training, Colorado, Oct 28-30
Message-ID: <005b01c25e94$e4b47e40$88133b41@PythonToy>

I will be holding another 3-day Python training 
session in Colorado, on October 28-30.  This is
a public class, open to individual enrollments.
  
For more details about this session, please visit
this page:

http://www.rmi.net/~lutz/oct02-longmont-class.html

General course details are maintained here:

http://www.rmi.net/~lutz/mytrain.html

Cheers,
--Mark Lutz  (http://www.rmi.net/~lutz)





From gerhard.haering at opus-gmbh.net  Wed Sep 18 10:19:39 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 18 Sep 2002 14:19:39 GMT
Subject: BibTeX module?
References: <Xns928D9F1FBAA22mhueningzedatfuberli@130.133.1.4>
Message-ID: <slrnaoh2i3.25g.gerhard.haering@haering.opus-gmbh.net>

Matthias Huening wrote:
> Hi,
> 
> I am looking for a simple BibTeX module, allowing me to read and
> manipulate/convert a bibliography file. 

Did you try Google? It suggests http://canvas.gnome.org:65348/pybliographer/

-- Gerhard


From max at alcyone.com  Wed Sep 25 23:50:37 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 25 Sep 2002 20:50:37 -0700
Subject: a python puzzle
References: <mailman.1032999133.8121.python-list@python.org>
Message-ID: <3D92840D.3D422522@alcyone.com>

"Delaney, Timothy" wrote:

> I suspect it took Holger somewhat less than half an hour to solve ...
> 5
> minutes at most, and most of that is writing the decryption code ;)

It only took two or three minutes for me to solve with an emacs window
open and tr on an interactive shell.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Gods are born and die, but the atom endures.
\__/ Alexander Chase
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From teaandbikkie at aol.com  Wed Sep 11 20:49:06 2002
From: teaandbikkie at aol.com (TeaAndBikkie)
Date: 12 Sep 2002 00:49:06 GMT
Subject: file upload using msvcrt
References: <bb2665bf.0209110457.62d78bfe@posting.google.com>
Message-ID: <20020911204906.00669.00000156@mb-fh.aol.com>

>I am trying to upload msword docs to my webserver.

I suspect the following lines...

data = fileinfo.file.read( )
if not data: break

This will break on the "not data" condition...
ie. when data is zero, "", None

So your binary file may have a zero...

read() returns empty string for EOF, so this would be better:

if data == "": break

Kind regards...



From vincent_a_primavera at netzero.net  Fri Sep 27 04:53:42 2002
From: vincent_a_primavera at netzero.net (Vincent A. Primavera)
Date: 27 Sep 2002 08:53:42 +0000
Subject: comp.lang.python...
In-Reply-To: <an1ip7$rgs$00$1@news.t-online.com>
References: <mailman.1033054480.24506.python-list@python.org> 
	<an1ip7$rgs$00$1@news.t-online.com>
Message-ID: <1033116822.1570.3.camel@del-vap-bos.ralphpill.com>

Hello,
	I've been subscribed to python-list at python.org for at least a couple of
years I would guess.  I have been using a mail client to receive all of
the messages, but I figure that using a news client would be a more
organized method.  However, I can't seem to find what I need to enter
for a server & user/password combination...

	Thank you,

	Vincent A. Primavera

On Fri, 2002-09-27 at 12:28, Thomas Guettler wrote:
> Vincent A. Primavera schrieb:
> > Hello,
> > 	Stupid question...  What information do I need to set up
> > comp.lang.python with my news client(Pan)?
> > 
> 
> You need to subscribe this newsgroup and
> you need a newsserver.
> 
> But how could you post this message without it?
> 
>   thomas
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list







From aleax at aleax.it  Fri Sep 27 02:43:05 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 27 Sep 2002 06:43:05 GMT
Subject: http://www.python-in-business.org
References: <cGNk9.237$2Q6.169@newssvr16.news.prodigy.com>
Message-ID: <Z5Tk9.175376$pX1.6206330@news2.tin.it>

Conrad Schneiker wrote:

> Chris,
> 
> Do you have anything to do with these people?

Not quite sure who you are addressing, but, anyway, I do have
something to do with them -- I was elected to the PBF board.


> Is there any sort of analogous North American organization?

The PBF can accept member firms from anywhere in the world: it
is not limited to Europe... North Americans are all right (our
Treasurer is Canadian, for example).


Alex



From janeaustine50 at hotmail.com  Thu Sep 26 04:38:39 2002
From: janeaustine50 at hotmail.com (Jane Austine)
Date: 26 Sep 2002 01:38:39 -0700
Subject: Is mmap usable for IPC?
Message-ID: <ba1e306f.0209260038.37ab940c@posting.google.com>

Is mmap module usable for inter-process communication?

Does it provide lock, or semaphore? Is it safe to write
from a process and read from multiple other processes
simultaneously?

Any information appreciated.


From sebastian.wangnick at eurocontrol.int  Sat Sep 21 04:32:26 2002
From: sebastian.wangnick at eurocontrol.int (Sebastian Wangnick)
Date: 21 Sep 2002 01:32:26 -0700
Subject: Trying the impossible?
Message-ID: <30292041.0209210032.6e07d9f3@posting.google.com>

Dear all,

I'm trying to overload, using dynwin/calldll under Windows, the
Tkinter drawing routines to perform some additional drawing on top of
Tk.

My first attempt was to bind to "<Expose>", but this binding seemingly
is triggered before Tk performs its drawing, so my stuff was erased by
Tk :(

Now I'm trying to overload the WindowProc itself. To test the
mechanism, I started by simply passing everything on to
CallWindowProc:

from Tkinter import *
import dynwin
import dynwin.gencb
import dynwin.windll
import dynwin.winwin
user32 = dynwin.windll.module ('user32')

def my_winproc (hwnd, message, wparam, lparam):
    return user32.CallWindowProc (tk_winproc,
        hwnd, message, wparam, lparam)
my_winproc_cb = dynwin.gencb.generated_callback ('llll', my_winproc)

def map (event):
    global tk_winproc
    hwnd = frame.winfo_id()
    tk_winproc = user32.GetWindowLong (hwnd,
        dynwin.winwin.GWL_WNDPROC)
    user32.SetWindowLong (hwnd, dynwin.winwin.GWL_WNDPROC,
        my_winproc_cb.address)

root = Tk()
frame = Frame(root,width=100,height=100)
frame.pack(fill=BOTH,expand=1)
frame.bind("<Map>",map,"+")
root.mainloop()

However, when starting this it crashes with 'The instruction at
"0x1e0381fa" referenced memory at "0x00000008". The memory could not
be "read".'

Any idea what I'm doing wrong? I'm using XP, btw.

Regards,
Sebastian


From robin at execulink.com  Sat Sep 14 21:57:06 2002
From: robin at execulink.com (robin at execulink.com)
Date: Sat, 14 Sep 2002 21:57:06 -0400
Subject: threads or queue for this task
References: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com> <GmCg9.2354$Ee4.6690@news-server.bigpond.net.au>
Message-ID: <cqp7ou8d0jjcdtrqetu9v9ls9gst2cp4qi@4ax.com>

Mark Hammond <mhammond at skippinet.com.au> wrote:

>threads = []
>for i in range(number_of_tasks):
>   thread.append(thread_for_task)
># let the threads run - just wait for shutdown
>for t in threads:
>   t.join()
>
>Each thread itself contains a loop which continually look for more work 
>to do, or exit.

Er, quite. For some reason I had not thought of simply having the
worker threads look for work. I was thinking more in terms of a Queue,
where the workers are fed work. But this is likely not even needed. 

In my case I would have them loop indefinitely. Something simple like:

nap = 1
while 1:
    try:
        LookForWork()
		time.sleep(nap)
	except MyInterrupt:
		break
	except:
		raise

Having individual threads sleeping will not adversely effect
performance I imagine. Or is there some reason to avoid mixing the
mechanisms?

>Look for Aahz's threading tutorial - URL not at hand, but it can't be 
>too hard to find.  Indeed, I predict he will tell you exactly where it is ;)

Yes he has! But this slideshow is a little sketchy. Has someone done a
more thorough tutorial covering all the different mechanisms
(semaphore, events, etc.)? 'Twould be useful for a newbie like me.

-- robin


From sismex01 at hebmex.com  Sat Sep 21 11:51:33 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Sat, 21 Sep 2002 10:51:33 -0500
Subject: if...else on the same line?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F0623302532A59@mail.hebmex.com>

> 
> i'm sorry for this terrorizingly basic question, but is it 
> possible at all to write an if ... else statement on a single
> line in python?
> 
> --
> dave
> 

Depends on what you want it for.

If you're trying to do the equivalent of C's  a?b:c  expression,
then you "kinda" can, but if it's for statements, you can not.

-gus




Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From abuse at nospam.riot.com.au  Fri Sep  6 02:31:31 2002
From: abuse at nospam.riot.com.au (Glen Murphy)
Date: Fri, 6 Sep 2002 16:31:31 +1000
Subject: about parse a link
References: <cVUd9.6962$yt3.3362358@newssrv26.news.prodigy.com>
Message-ID: <al9i48$286h$1@otis.netspace.net.au>

Prefix them with http://www.uic.edu/

That's what your browser does.


"koko" <kokohh at hotmail.com> wrote in message
news:cVUd9.6962$yt3.3362358 at newssrv26.news.prodigy.com...
> if I have extracted the links on the page:
> e.g:
>
> http://www.uic.edu/index.htm
> on this page: there are
> a.htm
> b.htm
> c.htm
> http://www.uic.edu/home/e.htm
>
> how can I log the a.htm, b.htm, c.htm with the full web address?
>
> thx
>
>




From paul at boddie.net  Mon Sep 16 04:50:29 2002
From: paul at boddie.net (Paul Boddie)
Date: 16 Sep 2002 01:50:29 -0700
Subject: AspectP?
References: <jOQg9.6057$Lg2.882928@news2.news.adelphia.net> <mailman.1032095664.13848.python-list@python.org> <7cl9oucm65a4qr2tolm74gltqn7morbbm5@4ax.com>
Message-ID: <23891c90.0209160050.7ba0e11@posting.google.com>

Lothar Scholz <llothar at web.de> wrote in message news:<7cl9oucm65a4qr2tolm74gltqn7morbbm5 at 4ax.com>...
>

[Aspect Oriented Programming for Python]

> I think from this posting he has no problems in finding instable 
> implementations.

What about the Zope acquisition stuff, or does that not go far enough?

Paul


From frithiof.jensen at removethis.ted.ericsson.dk  Tue Sep 17 05:06:44 2002
From: frithiof.jensen at removethis.ted.ericsson.dk (Frithiof Andreas Jensen)
Date: Tue, 17 Sep 2002 11:06:44 +0200
Subject: More on Protecting Source Code
References: <mailman.1032232999.22661.python-list@python.org>
Message-ID: <am6qv1$7q8$1@newstoo.ericsson.se>

"David LeBlanc" <whisper at oz.net> wrote in message
news:mailman.1032232999.22661.python-list at python.org...

> * Why invest a substantial amount of time and money developing in a
language
> that makes it trivial to gain access to the work product?
>

Why indeed - why do you bother?

Why is existing Copyright law not sufficient for you?

> beyond those that
> are fairly closely licensed and/or have substantial parts of the app
written
> in C/C++.

Do that then - there is nothing intrinsic in the Python licensing that
restricts what you can do with it; I seem to recall that you can pretty much
do whatever you bloody like, including wrapping Python in your own product
and selling it!

> A better solution is to make a .pyc file approximately as hard as a binary
> .exe file to decompile - however that could be done.

And according to the generous license awarded to you by the very people
whoms work you are critisizing, *you* can indeed do that; you are in fact
*encouraged* to make you own enhancements to Python and release them.

Simple Darwinism will show whether others feels the same way.

One could, in theory, package the application with a custom build of the
Python interpreter so that each interpreter uses it's own unique set of
bytecodes. Performance will likely suck and you will have a real issue with
upgrades, extension modules and maintenance of the many unique applications
shipped - but again this is your choice to make.

Personally, I do not care - I like Python exactly the way it is, including
the "openess", since that is an essential part of simplifying the
development task. If I was really, really worried about some "IP-leakage" I
would stick the core algorithm in an extension library - that would not stop
someone from just ripping the software though.





From mwh at python.net  Thu Sep 26 13:35:39 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 26 Sep 2002 17:35:39 GMT
Subject: Is there a portable way to copy files in Python?
References: <3D80A78B.5030206@impathnetworks.com> <Xns9287A1A862C3Aduncanrcpcouk@127.0.0.1>
Message-ID: <lkd6r0wte5.fsf@pc150.maths.bris.ac.uk>

Duncan Booth <duncan at NOSPAMrcp.co.uk> writes:

> Pierre Rouleau <prouleau at impathnetworks.com> wrote in 
> news:3D80A78B.5030206 at impathnetworks.com:
> 
> > I could write code that identifies the operating system (the os.path 
> > way) and use os.system("cp " + source_filename + " " + destination) for 
> > Unix-type and os.system("copy " + source_filename + " " + destination) 
> > but i would prefer to use a simple call.
> > 
> > Is this available in the Python library?  If not, why is it not part of 
> > the library?
> 
> Try the shutil module.

Or distutils.file_copy.  I don't like shutil much; somewhere on my
todo list is doing something about that...

Cheers,
M.

-- 
  > I'm a little confused.
  That's because you're Australian!  So all the blood flows to
  your head, away from the organ most normal guys think with.
                        -- Mark Hammond & Tim Peters, comp.lang.python


From peter at engcorp.com  Thu Sep 19 07:40:23 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 19 Sep 2002 07:40:23 -0400
Subject: where is the __dict__ ?
References: <mailman.1032430104.28760.python-list@python.org>
Message-ID: <3d89b79b$1@news.sentex.net>

black wrote:
>   my book said all dictionaries have __dict__ method but couldnt find it by using dir(), where is it  ?

I think I already saw a response to this posting, but it just showed up
in c.l.p on my system...

Your book is probably wrong, since dictionaries don't have a __dict__ 
method.  In 2.2 their *classes* do, although dir() still doesn't show
it.

Perhaps you misread it.  What exactly did it say?  Any more importantly,
what goal do you have in mind.  Sometimes there are alternatives that
aren't the first thing that comes to mind.

-Peter



From martin at v.loewis.de  Tue Sep  3 16:16:54 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 03 Sep 2002 22:16:54 +0200
Subject: PyListObject & C Modules
References: <bdade4c9.0209031040.4cf4b948@posting.google.com>
Message-ID: <m3bs7esua1.fsf@mira.informatik.hu-berlin.de>

jggramlich at yahoo.com (Joshua Gramlich) writes:

> I am trying to find examples of this type of activity.  I've found the
> documentation at www.python.org (the C api), but I'm a bit green (new)
> and cannot read the List Objects page with any, well, usefulness. 

Here is a function that gets a list as an argument, and returns a list
that has the same first element:

PyObject*
get_first_elem(PyObject *unused, PyObject *args)
{
  PyObject *list;
  PyObject *first;
  PyObject *result;
  if(!PyArg_ParseTuple("O", &list))
    return NULL;

  if(!PyList_Check(list)) {
    PyErr_SetString(PyExc_TypeError, "get_first_elem expects a list");
    return NULL;
  }

  if(PyList_Size(list) == 0){
    PyErr_SetString(PyExc_ValueError, "empty lists not allowed");
    return NULL;
  }

  first = PyList_GetItem(list, 0);
  result = PyList_New(1);
  if (!result) return NULL;
  PyList_SetItem(result, 0, first);
  return result;
}

The Python equivalent of this is

def get_first_elem(_list):
  if type(_list) != list: raise TypeError, "expects a list"
  if len(_list) == 0: raise ValueError, "must not be empty"
  first = _list[0]
  return [first]

HTH,
Martin


From alessandro.riolo at sen.it  Thu Sep 26 09:15:36 2002
From: alessandro.riolo at sen.it (Alessandro Riolo)
Date: Thu, 26 Sep 2002 15:15:36 +0200
Subject: Decimal arithmetic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amgpm4$5si22$1@ID-125932.news.dfncis.de> <7xwupf52jx.fsf@ruckus.brouhaha.com> <TXOVyjACLDj9EwHM@jessikat.fsnet.co.uk> <WMzj9.150692$pX1.5406331@news2.tin.it>
Message-ID: <amv1kk$9gl8j$1@ID-44839.news.dfncis.de>

Alex Martelli wrote:
> The two ways of computing interest use the acronyms TAEG
> and _TAEV_ in Italy -- I'm not sure what they refer to, exactly:

They are TAEG ("Tasso annuo effettivo globale") and TAN ("Tasso annuo
nominale"); I never heard of TAEV.
The TAN is ever lower than the TAEG, 'cause the TAN is just the
yearly nominal interest rate, it is useful only to compare it with the
national infation rate or similar general purpose index, while the
TAEG
is a rate calculated following a specific law, and in TAEG are
comprised also the effective costs (all of the most hidden ones as
commissions and insurances), and it is the rate we have to use to
compare different loans.
When there is a large discrepance betweeen TAN and TAEG, I would not
entrust that so much, and even when the TAN is higher than the TAEG I
would be really worried (it would mean the bank is paying money to me
to open a loan ..).

p.s.: Alex, I'm really having funny time with Python ;-)

--
ale
www.sen.it




From pyth at devel.trillke.net  Mon Sep 30 17:45:54 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 30 Sep 2002 23:45:54 +0200
Subject: Question about __getattribute__/__getattr__
In-Reply-To: <Pine.LNX.4.44.0209301625470.12980-100000@mercury.sabren.com>; from jpm@papercut.org on Mon, Sep 30, 2002 at 04:32:22PM -0400
References: <20020930215327.X30315@prim.han.de> <Pine.LNX.4.44.0209301625470.12980-100000@mercury.sabren.com>
Message-ID: <20020930234554.Y30315@prim.han.de>

Joao Prado Maia wrote:
> On Mon, 30 Sep 2002, holger krekel wrote:
> 
> > Probably you are interested in the Memoize recipe at 
> > 
> >     http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52201
> > 
> 
> That looks interesting, and it is indeed similar to what I'm trying to 
> acomplish, but on my case I want to keep each specific return value in a 
> separate cache file.
> 
> The first Memoize class keeps every cached item in a dictionary, which it 
> may be fine for some applications, but I need it for a network server that 
> may be running for several days. It's not the best thing to keep 
> everything in memory ;)
> 
> The second class keeps everything on the same pickled file (please let me 
> know if I'm wrong), which will be very slow after a couple hundred megs of 
> data in there.
> 
> Anyway, I'm still going back to my original question: how do I get the 
> full list of arguments passed to a function ? This can't be that hard :)

The important lines from the above link are:

def __call__(self, *args):
        if not self.memo.has_key(args):
            self.memo[args] = self.fn(*args)
        return self.memo[args]

and 'args' is a tuple of arguments.  
'args[0]' would give you the first argument, for example.  

HTH,

    holger



From hancock at anansispaceworks.com  Sun Sep  8 13:27:28 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Sun, 08 Sep 2002 10:27:28 -0700
Subject: horizontal scrolling (was Re: Larry Wall's comment on python...)
References: <20020908044502.16742.69743.Mailman@mail.python.org>
Message-ID: <3D7B8880.A089BC76@anansispaceworks.com>

Hi all,

I wanted to point out a good reason why Python often
led me to overly-wide source files before I learned
more about it. I suspect this is one thing that trips
up a lot of newbies:

In C or Perl (any non-whitespace significant language),
you can easily code long if conditions (or other 
expressions), like this:

if ( condition_one_which_has_a_nice_long_expression ||
   ((condition_two_which_also_is_pretty_long)       &&
    (condition_three_wich_needs_two_to_apply))      ||
     condition_four_also_pretty_darn_long)
	{
	/* Do what I need to do in this oddball case	*/
	}

But the Python intro books usually don't put much emphasis
on how to do this, so you wind up thinking you have to write:

if condition_one_which_has_a_nice_long_expression or
(condition_two_which_also_is_pretty_long and
condition_three_wich_needs_two_to_apply) or
condition_four_also_pretty_darn_long:
	# Do what I need to do in this oddball case
	pass

which is indeed pretty obnoxious. The newbie has been taught
that you don't need () on the if-clause (they may think they
will cause a syntax error  -- which they *do* in csh/tcsh,
which they may have previous experience in, BTW).  So they
take a long time before they get around to trying:

if ( condition_one_which_has_a_nice_long_expression or
     ( condition_two_which_also_is_pretty_long      and
       condition_three_wich_needs_two_to_apply)     or
     condition_four_also_pretty_darn_long ):
	# Do what I need to do in this oddball case
	pass

which is a fairly simple trick that results from the Python
rules for parsing expressions, but isn't usually mentioned
when introducing the 'if' statement.   Maybe it ought to
be obvious, but some of us don't make the connection right
away. ;-D

BTW, I think the wording used at one point was more like
Python will continue any expression it knows to be unfinished,
rather than that is contained in delimiters.  This led me
to wonder why I can't just type:

if condition_one_which_has_a_nice_long_expression  or
   ( condition_two_which_also_is_pretty_long        and
     condition_three_wich_needs_two_to_apply)       or
   condition_four_also_pretty_darn_long:
	# Do what I need to do in this oddball case
	pass

which would seem to be unambiguous, but which causes a
traceback on the first line.  It seems to me that the
dangling operator is enough of a clue that more expression
is to follow. (I think there was a thread about this
a month or two ago, so I won't pursue it -- at least until
after I've reviewed it ;-D).

Anyway, I actually had a Perl instructor make a big deal
about being able to spread an if statement over several
lines like this in Perl, as if it were a big win over
Python.  I didn't have the heart to tell him that it
could be done exactly the same way in Python.

From: Peter Hansen <peter at engcorp.com>
> My point is that indentation (the way I've done it anyway) looks
> *exactly* the same in C, Pascal, Python, or any other block-structured
> language, so I still have no idea why Python would feel any different
> in this respect for someone.
> 
> Clearly I'm missing some point, but I still don't know what it is.
> [... examples ...]
>
> The only thing that would make me exceed 80 columns with a language
> like this is if I couldn't define subroutines and call them, to
> keep the nesting from growing too large.  Since all these languages
> allow that, I have no idea what is special about Python that leads
> to excessive indentation.

You can avoid this problem by always using a function
call to evaluate the expression, but IMHO it often
is bad style (fails the "explicit is better than implicit"
test).

This probably depends on the application domain,
I suspect -- in some types of projects, the expression
will almost always have some kind of logical meaning
in itself, in which case a function makes sense -- but
in other places, it's just a lump of conditions with
only one obvious place to use it.  You could define a
function for one use, but it just requires more typing
without really making anything clearer, IMHO.

BTW -- I always used to write for 132-columns because
that's what my printer printed (way back when).
Nowadays, I hardly ever use a terminal or printer
that is narrower than this.  I find wide formatted
code easier to comprehend (on the whole I don't like
splitting expressions over multiple lines if I can
help it, even though it's sometimes necessary, as I
was trying to point out here). I suppose this is
probably some kind of open-source sin.  Oh well.

Cheers,
Terry


-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
Anansi Spaceworks                 
http://www.anansispaceworks.com 
P.O. Box 60583                     
Pasadena, CA 91116-6583
------------------------------------------------------



From scherrey at proteus-tech.com  Mon Sep 16 23:12:56 2002
From: scherrey at proteus-tech.com (Benjamin Scherrey)
Date: Mon, 16 Sep 2002 23:12:56 -0400
Subject: Point of Sale Software in python
Message-ID: <am671t0ftu@enews2.newsguy.com>

I'm looking to develop some point of sale software (initially for
restaurants) that integrates with POS hardware that's out there and was
wondering if there might be something already developed along these lines.
My hope is for a Postgres database backend. Appreciate any pointers...

	Ben Scherrey


From rjones at ekit-inc.com  Sat Sep 14 10:00:56 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Sun, 15 Sep 2002 00:00:56 +1000
Subject: html template system?
In-Reply-To: <btug9.688$Ee4.1090@news-server.bigpond.net.au>
References: <btug9.688$Ee4.1090@news-server.bigpond.net.au>
Message-ID: <200209150000.56803.rjones@ekit-inc.com>

On Sat, 14 Sep 2002 9:24 am, :B nerdy wrote:
> what are good templating systems avaliable for python?
>
> ive been at http://zebra.sourceforge.org and the documentation is kinda
> sus. zope DocumentTemplate is included with zope. but i want a smaller
> footprint system.
>
> what are my options? cheers

I've had good results integrating Zope's PageTemplates with Roundup:

   http://www.zope.org/Members/4am/ZPT
   http://roundup.sf.net/

A remarkalbly well-designed templating system...


      Richard




From andrew.thompson at ashecastle.com  Mon Sep 23 13:22:29 2002
From: andrew.thompson at ashecastle.com (Andrew Thompson)
Date: Mon, 23 Sep 2002 18:22:29 +0100
Subject: parsing tabs in a string
In-Reply-To: <006201c26324$4148ae90$025e10ac@calibredd.com>
Message-ID: <012001c26325$cc10fb80$4802a8c0@blair>

Andrew,
 
You might find it easier to split up the line first into a collection of
fields using the tab-delimiter, and then just access the collection
As an array :
 
'
import string
 
aString='abc \t def \t xyz  \t'
 
anArray = string.split ( aString , '\t')
 
anArray[0] now equals
 
'abc '
 
 
 
Andrew
 
-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]
On Behalf Of Andrew Alzner
Sent: 23 September 2002 18:11
To: python-list at python.org
Subject: parsing tabs in a string
 
Hello,
 
I'm getting to know Python and have a quick question for something I
don't understand.
 
I'm trying to parse information extracted from a spreadsheet which is
tab separated. Here is a simplified version of what I have:
 
list=['"103-01a-17"\t2\t33\t256\t227\n']
y=list[0]
print y.find("\t")
print y.find("\t",1)
 
I thought the second .find should return the second occurence of "\t".
To get the index of the second tab I have to type:
 
print y.find("\t",13)
 
This doesn't make sense to me. Can someone help me please.
 
Thanks,
Andrew
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020923/b6861d29/attachment.html>

From op73418 at mail.telepac.pt  Wed Sep 18 17:24:16 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Wed, 18 Sep 2002 22:24:16 +0100
Subject: multiple index inconsistency
References: <TP5i9.9293$Rf3.1003108@newsfeed.slurp.net>
Message-ID: <4arhouona0ts9l2pur0p2scrar4h7ohnq6@4ax.com>

There is no bug here.

On Wed, 18 Sep 2002 13:59:58 -0700, bert <kilaruba at copper.net> wrote:

>The following line of code
>>>> a = 3*[range(3)]
>produces
>[[0, 1, 2], [0, 1, 2], [0, 1, 2]]

What this does is create a list, named a, where each element references
*the same* list range(3) - That is the semantics of the * operator, it
only makes shallow copies.

>
>If I then write, say, 
>>>> a[0][1] = 3.7
>I get
>[[0, 3.7, 2], [0, 3.7, 2], [0, 3.7, 2]]
>and not
>[[0, 3.7, 2], [0, 1, 2], [0, 1, 2]]

So when you change one of them, since they are all aliases to the same
list, you in fact are changing all the elements.
>
>Now if I write
>>>> a = [[0, 1, 2], [0, 1, 2], [0, 1, 2]]

Here no copies involved, since the list is built out of literals. They
are all different objects (different id's) - albeit equal.

>instead, and then 
>>>> a[0][1] = 3.7
>I get the result I expected (and wanted):
>[[0, 3.7, 2], [0, 1, 2], [0, 1, 2]]

So changing only one element changes... well, only one element.

>Why the inconsistency? Is this a bug or am I missing something?

There is no inconsistency. You just have to watch out that for mutable
objects (such as lists) this is the semantics - You will get the hang of
it.

All the best,
Gon?alo Rodrigues

P.S: BTW, I think there is a FAQ entry on this. You should read it.



From gherron at islandtraining.com  Wed Sep 25 18:20:19 2002
From: gherron at islandtraining.com (Gary Herron)
Date: Wed, 25 Sep 2002 15:20:19 -0700
Subject: a python puzzle
In-Reply-To: <Pine.LNX.4.44.0209251736470.16806-100000@hydrogen.sabren.com>
References: <Pine.LNX.4.44.0209251736470.16806-100000@hydrogen.sabren.com>
Message-ID: <200209251520.19397.gherron@islandtraining.com>

On Wednesday 25 September 2002 02:42 pm, Michal Wallace wrote:
> Here's a python puzzle. Solve it and you win a prize. :)
>
>
>
> """
> Ha'f ahoy ad ayfa wdqr srxhekdpyr! H phmm zhby
> fhv odeajf dc cryy pys jdfahez ad ajy chrfa kyrfde
> ad fdmby ajhf irwkadzrxo. Ixe wdq chzqry ha dqa?
>
> Fheiy wdq jxby axgye ow ijxmmyezy xet tyirwkayt
> ajy fyirya oyffxzy, H phmm ybye zhby wdq ajy
> irwkadzrxo-oxghez fdqriy idty ad xoxuy wdqr
> crhyetf!
> """
> hokdra fwf, farhez, rxetdo
> ihkjyr = mhfa(farhez.myaayrf[:26])
> rxetdo.fjqccmy(ihkjyr)
> ihkjyr = "".ldhe(ihkjyr)
> ihkjyr += ihkjyr.qkkyr()
> oxk = {}
> cdr kmxhe, irwka he uhk(farhez.myaayrf, ihkjyr):
>     oxk[kmxhe] = irwka
> cdr mhey he dkye(fwf.xrzb[1]):
>     krhea "".ldhe([oxk.zya(ij,ij) cdr ij he mhey])[:-1]
>

I decrypted (building the strings up character by character) with:

   tr hokdrawfezxmylqcubijHpvsIFgt importysngalejufzvchIwxbCSkd < puzzle

and got:

"""
It's time to test your brainpower! I will give
six months of free web hosting to the first person
to solve this cryptogram. Can you figure it out?

Since you have taken my challenge and decrypted
the secret message, I will even give you the
cryptogram-making source code to amaze your
friends!
"""
import sys, string, random
cipher = list(string.letters[:26])
random.shuffle(cipher)
cipher = "".join(cipher)
cipher += cipher.upper()
map = {}
for plain, crypt in zip(string.letters, cipher):
    map[plain] = crypt
for line in open(sys.argv[1]):
    print "".join([map.get(ch,ch) for ch in line])[:-1]



The fact that this puzzle was easy to decrypt says something (good)
about Python: It is so readable, that even encrypted, one can
recognize significant portions.  Imagine decrypting a similar puzzle
in C or, worse yet, Perl!

Dr. Gary Herron




From marklists at mceahern.com  Wed Sep  4 10:15:13 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 4 Sep 2002 09:15:13 -0500
Subject: efficient list merging
In-Reply-To: <ced73313.0209040601.64edac88@posting.google.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOGEEKDMAA.marklists@mceahern.com>

> Here's an example:
> 
> lol1 = [['a'], ['a', 'b', 'c'], ['d', 'e', 'f'], ['g', 'h', 'i', 'j']]
> lol2 = [['d', 'e', 'f'], ['k', 'l', 'm'], ['d'], ['f', 'g'], ['a']]
> 
> I want a function which returns
> 
> mergelol = [['a'], ['a', 'b', 'c'], ['d', 'e', 'f'], ['g', 'h', 'i',
> 'j'], ['k', 'l', 'm'], ['d']]

why isn't ['f', 'g'] in mergelol?

// m
-




From aahz at pythoncraft.com  Mon Sep  2 16:58:13 2002
From: aahz at pythoncraft.com (Aahz)
Date: 2 Sep 2002 16:58:13 -0400
Subject: Threads
References: <kIBb9.85641$mj7.1268320@twister.rdc-kc.rr.com>
Message-ID: <al0jd5$f5p$1@panix1.panix.com>

In article <kIBb9.85641$mj7.1268320 at twister.rdc-kc.rr.com>,
Shashank Date <ADATE at kc.rr.com> wrote:
>
>I am trying to write a simple threading app in Python 2.2 (on Windows) which
>will spawn mutiple threads.
>Each thread will go out on the web (various web sites) and fetch certain
>documents. The main thread will wait for all the threads to terminate OR
>some external event (like deletion of a file). If the external event fires
>before termination of child-threads it kills all the spawned threads and
>terminates.
>
>Any ideas how to go about doing this.

See my web page.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From akineko at pacbell.net  Sun Sep  8 16:27:49 2002
From: akineko at pacbell.net (Aki Niimura)
Date: 8 Sep 2002 13:27:49 -0700
Subject: Creating a Text widget in a Dialog  (Tkinter)
Message-ID: <b714b8de.0209081227.448fd2f2@posting.google.com>

Hi,

I'm trying to create a simple dialog which has a Text with
scrollbars(H and V)
in a dialog. (I'm inherting Tkinter Dialog class)

I'm using a grid to attach two scrollbars to the Text widget.
I don't know why but the text widget is disappeared (not mapped) while
two
scrollbars are displayed correctly. 

If I use pack instead of grid then all of them appear (but Horizontal 
scrollbar are not placed correctly). I used the same code in the main
window
and it worked fine.

Is there anything I'm overlooking?

Any insights are highly appreciated.

Thanks,
Aki-

###     
###     Class HelpDlg
###         
                
class HelpDlg(Dialog):
    def __init__(self, parent, excerpt, title = None):
        self.lines = excerpt
        Dialog.__init__(self, parent, title)
 
    def body(self, master):
        self.usage = Text(master, height=26, width=80)
        self.grid = Frame(master)
        self.yscroll = Scrollbar(master, command=self.usage.yview)
        self.xscroll = Scrollbar(master, orient=HORIZONTAL,
command=self.usage.xview)
        self.usage.configure(xscrollcommand=self.xscroll.set,
yscrollcommand=self.yscroll.set)
        
        for line in self.lines:
            self.usage.insert(END, line+'\n')
        
        self.grid.pack(expand=YES, fill=BOTH)
        self.usage.grid(in_=self.grid, row=0, column=0)
        self.yscroll.grid(in_=self.grid, row=0, column=1,
sticky='news')
        self.xscroll.grid(in_=self.grid, row=1, column=0,
sticky='news')


From bh at intevation.de  Wed Sep  4 09:46:25 2002
From: bh at intevation.de (Bernhard Herzog)
Date: 04 Sep 2002 15:46:25 +0200
Subject: default tuple unpacking?
References: <slrnanaknb.1gh.huaiyu@gauss.almadan.ibm.com> <3d758180$1@news.sentex.net> <3d7584d2$1@news.sentex.net>
Message-ID: <6qd6rtdg0e.fsf@thetis.intevation.de>

Peter Hansen <peter at engcorp.com> writes:

> Hmm... I think there might be an inconsistency or an unexpected
> result in this.  What if e already contains a tuple?
> 
>    e = (1, 2, 3)
>    *a = e        # a is ((1, 2, 3), )
> 
> That seems sensible, and it's just what happens if you pass
> "e" in to a function defined as "def func(*a)".

No it isn't. If you call func as func(e) the tuple being unpacked into
the function's parameters is (e,) not e. So the situation is more like
calling func as func(*e) so that *a = e would be the same as a = e or
perhaps a = tuple(e)


> Now what about this:
> 
>    a, *b = e     # a is 1, b is (2, 3)
> 
> That's one possible result, and presumably the one you want with
> this syntax, but it's not what you'd get with the function syntax:
> 
>    def func(a, *b):
>      print a
>      print b
> 
>    func(e)

Again the tuple being upacked to a, *b is (e,)...

> Here of course you get (1, 2, 3) in "a" and () in b. 

... so this is obviously what you should get.

   Bernhard

-- 
Intevation GmbH                                 http://intevation.de/
Sketch                                 http://sketch.sourceforge.net/
MapIt!                                           http://www.mapit.de/


From cmkleffner at gmx.de  Thu Sep  5 03:14:25 2002
From: cmkleffner at gmx.de (cmkl)
Date: 5 Sep 2002 00:14:25 -0700
Subject: make informixdb1.3
References: <aknb6n$evi3@news-1.bank.dresdner.net> <3D71300E.8010403@sschwarzer.net> <akv5ee$evv2@news-1.bank.dresdner.net>
Message-ID: <3b091a1c.0209042314.3b99bb80@posting.google.com>

"Farhad" <Farhad.Fouladi at Dresdner-Bank.com> wrote in message news:<akv5ee$evv2 at news-1.bank.dresdner.net>...
> Hi Stefan,
> 
> thank you for your response. I start "make" in correct directory. The point
> is, there is no "_informixdb.c" file in downloaded files (from python.org).
> I suppose the make file should generate one but in my case it couldn't.
> 
> Farhad
> 

Hi there,

there should be _informixdb.ec which has to be prepprocessed by esql - a 
tool supplied with the informix sdk on your machine. 
Locate esql and your informix libraries. 
run:   esql -e -G  -thread  -libs _informixdb.ec  to get a list 
of the libraries and potientially object files which has to be linked 
to your _informixdb module. 
Be aware, that on some systems the informix sdk is supplied as 
32 bit and/or 64 bit version. 
Instead of using the Makefile I compiled informixdb manually (I didn't 
patched the makefile). I use a patch to _informixdb.ec to get rid of 
the most common pitfalls (DATE problem, ignore warnings instead treat 
them as errors).
Send me a private email in the case of interest.

cmkl


From niemeyer at conectiva.com  Thu Sep 12 12:59:27 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Thu, 12 Sep 2002 13:59:27 -0300
Subject: ANN: NINZ 1.2.1
Message-ID: <20020912135926.A7584@ibook.distro.conectiva>

NINZ stands for NINZ Is Not ZSI. This name was used to leave it clear
that NINZ is a fork from ZSI. I have done this because I wanted a few
features in ZSI which were not available at the moment, and at the same
time, I don't wanted to distribute a hacked version of ZSI, since I
could cause harm to Rich Salz's work on ZSI. My plans are to keep in
sync with ZSI, always submitting what I change back to Rich, so he can
keep his great work, and use in ZSI any code he likes. I don't want to
takeover any credits of his work, and that's why I leave it clear here
that my changes are insignificant compared to the comprehensive work he
has done on ZSI. Indeed, my expectatives is that most of these features
will be included in ZSI itself someday, and NINZ will die.

More information about NINZ, including a comprehensive "NINZ vs. ZSI"
session at:

http://moin.conectiva.com.br/NINZ

More information about ZSI at:

http://pywebsvcs.sourceforge.net

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]



From danb_83 at yahoo.com  Wed Sep 25 16:15:20 2002
From: danb_83 at yahoo.com (Dan Bishop)
Date: 25 Sep 2002 13:15:20 -0700
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032322762.20239.python-list@python.org> <amabpb$5rm$1@news.nems.noaa.gov> <3D8B1537.4010803@thinkware.se> <ad052e5c.0209201128.1a53108@posting.google.com> <mailman.1032664289.615.python-list@python.org> <amo0aq$5nf$1@nntp2-cm.news.eni.net>
Message-ID: <ad052e5c.0209251215.45d2d317@posting.google.com>

"Stuart D. Gathman" <stuart at bmsi.com> wrote in message news:<amo0aq$5nf$1 at nntp2-cm.news.eni.net>...
>...
> There ought to be severe penalties for idiots that use floating point
> dollars for financial applications.  If forced to use floating point
> (e.g. because customer demands BASIC), then keep money amounts in
> whole pennies (or whatever the smallest currency unit for the country
> is), and divide by 100 (or whatever) for printing only (or just add the
> decimal point yourself).

That's exactly how things were done at my old workplace, except that
cents were also used for user input.  Most of the time.


From margosd at polaroid.com  Tue Sep 17 17:17:23 2002
From: margosd at polaroid.com (David B. Margossian)
Date: Tue, 17 Sep 2002 17:17:23 -0400
Subject: Test Post
Message-ID: <am8656$3npc1@ncc1701.polaroid.com>

Please excuse the intrusion.  Trying to establish a feed for this group...




From mbonig at hotmail.com  Sun Sep 22 03:45:03 2002
From: mbonig at hotmail.com (Matthew Bonig)
Date: Sun, 22 Sep 2002 07:45:03 GMT
Subject: Jython file manipulation?!
Message-ID: <2yej9.32454$gA4.4053@sccrnsc02>

I am trying to rename a file in Jython...
the command I am using is just
os.rename(filename, newFilename)
I also tried doing in in java with
f = File(filename)
f.renameTo(File(newFilename)

but that didn't work either!
Any ideas?
thanks
matthew




From joconnor at cybermesa.com  Wed Sep  4 11:11:56 2002
From: joconnor at cybermesa.com (Fearless Freep)
Date: 4 Sep 2002 08:11:56 -0700
Subject: Why Python?
References: <uafbnugdli4c3f8ns6qr33mna5urdno78g@4ax.com> <unbp765rbsn809@corp.supernews.com>
Message-ID: <7d3dc526.0209040711.6b10d5aa@posting.google.com>

"RPM1" <rpm1deletethis at nospamfrontiernet.net> wrote in message news:<unbp765rbsn809 at corp.supernews.com>...

> If you really want to get a feel for a language, read a fairly large
> program that somebody else has written in that language, (not
> just little snippets).


Amen to that.  I now hate TCL for that very reason

Take care,
Jay


From skip at pobox.com  Tue Sep  3 17:18:22 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 3 Sep 2002 16:18:22 -0500
Subject: Distributed Shared Memory
In-Reply-To: <20020903203102.88032.qmail@web11401.mail.yahoo.com>
References: <20020903203102.88032.qmail@web11401.mail.yahoo.com>
Message-ID: <15733.10014.586464.763881@12-248-11-90.client.attbi.com>

    Carlos> I need a Python bind for some lib that supports
    Carlos> "Distributed Shared Memory". Is there?

There is PyBrenda, which supports Gelerntner's tuple space concept (aka
Linda).  It only runs on Unix systems and there are some serious
restrictions (shared network drive, etc).  Google for "PyBrenda".  You might
find some similar packages by searching the Vaults of Parnassus for
"PyBrenda" and browsing the category in which it resides.

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html



From thomas at xs4all.net  Thu Sep 12 08:57:22 2002
From: thomas at xs4all.net (Thomas Wouters)
Date: Thu, 12 Sep 2002 14:57:22 +0200
Subject: Fatal Python error: PyThreadState_Get: no current thread
In-Reply-To: <F257a97F0epIncnz9h50000d230@hotmail.com>
References: <F257a97F0epIncnz9h50000d230@hotmail.com>
Message-ID: <20020912125722.GB797@xs4all.nl>

On Thu, Sep 12, 2002 at 09:45:39PM +0930, Ben Gerblich wrote:

> My code has just started crashing. I get the message:

> Fatal Python error: PyThreadState_Get: no current thread
> Aborted (core dumped)

> Is there a way to turn on debug mode or something that will allow me to see
> where/why my app is crashing. "python2 -d main.py" does not shed any light.

This is an internal Python error, not one related to your program (other
than that it seems to want threads, directly or indirectly.) If memory
serves, the problem is that Python is compiled without threads, but one or
more of the extention modules you're using were compiled _with_ threads.
(Or, possibly, the other way 'round, but I don't think so.) Testing to see
if python is compiled with threads is fairly easy: run 'python2' and type
'import thread'. Perhaps you need to get threaded (or non-threaded) versions
of some of your RPMs ?

*cough*debian*cough*-ly y'rs, :)
-- 
Thomas Wouters <thomas at xs4all.net>

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



From paul at boddie.net  Tue Sep 10 04:10:11 2002
From: paul at boddie.net (Paul Boddie)
Date: 10 Sep 2002 01:10:11 -0700
Subject: Larry Wall's comment on python...
References: <mailman.1031579787.7378.python-list@python.org>
Message-ID: <23891c90.0209100010.2c08c949@posting.google.com>

brueckd at tbye.com wrote in message news:<mailman.1031579787.7378.python-list at python.org>...
> 

["I need braces" complainants]

> But just because people consistently complain about something does not
> mean the answer is to bow to their wishes.

Especially if the adamant complainants in question are the usual
screaming 14 year old Slashdot "contributors" who think they know it
all after "hanging out" on IRC for a bit.

Paul


From michela_rossi66 at hotmail.com  Mon Sep 23 12:15:49 2002
From: michela_rossi66 at hotmail.com (michela rossi)
Date: 23 Sep 2002 09:15:49 -0700
Subject: UK ZOPE Contract.
Message-ID: <1ed550fd.0209230815.2ec6304a@posting.google.com>

Hi,

We have a ZOPE contract up and coming soon. Around two months work. If
there are any ZOPE contractors available, please get in contact with
me.

Regards, Michela.


From sholden at holdenweb.com  Mon Sep 16 08:03:52 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 16 Sep 2002 08:03:52 -0400
Subject: IDLE
References: <wpbf9.264320$aA.45691@sccrnsc02> <3D7D4F84.318CDB0F@bellatlantic.net> <v4Pg9.428631$me6.49396@sccrnsc01> <bs70f7ks.fsf@online.no> <am1efh$1ohti$1@ID-77524.news.dfncis.de>
Message-ID: <EOjh9.2033$Cf.81@fe03>

"Thorsten Kampe" <thorsten at thorstenkampe.de> wrote in ...
> * syver-en at online.no
> >"Ali K" <akakakhel at attbi.com> writes:
> >> Hardware: 386 enhanced, 75MHz, 32MB RAM.
> >> OS: Win98
> >> Copied files from other computer to mine.
> >> It doesn't open when I click on ide.pyw or when I choose to edit a py
> >> file
> >>
> >> with it.
> >
> > I suggest you install the python distro from www.python.org.
>
> You think this will work on medieval computers?
>

Depends what you mean by mediaeval. The latest release should work fine on
anything that runs Windows 95 or a more recent Microsoft OS, and pretty much
any old Unix implementation you care to choose.

Otherwise, installer are still available for version back to 1.5.2, I seem
to remember. Someone even recently revived the DOS implementation, though
I'm not sure I'd recommend that unless absolutely necessary.

You certainly shouldn't expect your Windows system to understand what a .pyw
file is supposed to do without being told (in some installation process, or
manually). ".pyw" files should run pythonw.exe (no console), ".py" files
should run python.exe.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From bosahvremove at netscape.net  Fri Sep  6 03:19:49 2002
From: bosahvremove at netscape.net (Bo)
Date: Fri, 06 Sep 2002 07:19:49 GMT
Subject: Zope Status?
Message-ID: <Xns9281221A07167bosahvnetscapenet@66.185.95.104>

I apologise if this isn't the proper forum to do this in.

1) Can anyone tell me roughly if\when Zope is moving to python 2.2?

2) Also any idea when they might release Zope 3?

3) Anyone close to the Zope team who is willing to shed some light on the 
near future for things Zope?


Great thanks in advance.

Bo


From tim.one at comcast.net  Wed Sep 11 20:20:43 2002
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 11 Sep 2002 20:20:43 -0400
Subject: SpamBayes PIK file?
In-Reply-To: <7KQf9.1760$Lg2.232748@news2.news.adelphia.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEJCBDAB.tim.one@comcast.net>

[Robert Oschler]
> I want to look at the SpamBayes project so I downloaded the zip file from
> source forge.  Inside there's just a PIK file.  How do I read a PIK file?

It's a binary pickle of a trained instance of class GrahamBayes in the code
base.  If what to do next isn't instantly obvious, you should return the
.pik file for a full refund <wink> -- unless you're a developer on this
project, you don't stand much chance of getting any good out of it.

For more read the associated mailing list:

    http://mail.python.org/mailman-21/listinfo/spambayes

Nothing in this project is end-user usable yet.




From jhorneman at pobox.com  Sun Sep 29 16:05:03 2002
From: jhorneman at pobox.com (Jurie Horneman)
Date: Sun, 29 Sep 2002 22:05:03 +0200
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com> <874rc8d9c9.fsf@nospam.eml.cc> <mailman.1033319977.7828.python-list@python.org>
Message-ID: <3d975c8f$0$18866$91cee783@newsreader02.highway.telekom.at>

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote in message
news:mailman.1033319977.7828.python-list at python.org...

> It's also highly recommended to start Python with "python -vv", then
> trying to import the module in question. This will show you where Python
> tries to find the module.

Thanks, that's good advice!

> Btw. sys.modules contains the module search path.

I did look at sys.path, but it didn't help me much...

Jurie.





From tismer at tismer.com  Wed Sep 18 09:16:56 2002
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 18 Sep 2002 15:16:56 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <3D881B66.1080200@something.invalid> <j4wupjafun.fsf@informatik.hu-berlin.de> <3D8868C3.9090005@one.net>
Message-ID: <3D887CC8.5040306@tismer.com>

Michael Schneider wrote:
...

>    "stackless python behaves the same as 'normal' python, until you 
> import the
>    stackless module.  At that point, it becomes stackless python.  This 
> allows you
>     to run 'stackless unsafe' modules with the same interpretor as 
> stackless modules,
>     in a predictable manner (as long as you don't import stackless, and 
> an unsafe module)"
> 
> 
> Is this the correct understanding????

Almost. It is slightly different (and slightly better):

Part I: stack slicing
---------------------
Stackless Python has built-in support for stack slicing.
This is by default off. Stackless Python behaves exactly
like Python as long as you don't activate things.

There is a couple of new functions in the sys module
which control behavior:
old_flag = sys.enable_stackless(new_flag)
sets stackless behavior to on or off and returns the
old state.
When it is enabled, then the C stack will be hijacked
every nth interpreter recursion, in order so save
stack space. This behavior is controlled by the two
functions
sys.setslicinginterval(n)
sys.getslicinginterval()

If this interval is set very high, you will also get no
stack slicing.

Part II: Multitasking
---------------------
Completely apart from the features of part I, there is
an extension module named stackless.
It can be imported only if the Python interpreter was built
for it, since it does in fact make use of the above
stack slicing technique.
But instead of slicing stacks in order to save space
in deep recursion, the stackless module uses it to
move stacks forth and back. This is like switching threads,
just not by stack switching but moving.
These multitasking features are independant from the
sys options of part I. If a tasklet is asked to switch
context to another one, it will do that, since you demanded
it. There will be no switching if you don't ask for it.

Both parts can be used in any combination.
If small stack size is desirable, turn stack slicing on.
If you want to use tasklets, use them.
In combination, you will of course xperience the best
performance.

About extension modules:
------------------------
Stack slicing has shown up as a problem with tkinter.
I had to modify the _tkinter module to make this safe
(and I really should include that in the distro now).
The problem with stack slicing is that it occours in
a context where an extension might expect certain
data to be still on the C stack.

Usually, using tasklets is much more safe than stack
slicing. It depends on your application. But a tasklet
switch, while technically the same, has the semantics
of a context switch. You are moving out of a context
where some extensiom module might be involved, into
a different context, where this extension module
is not active.

But still some care needs to be taken, since tasklets
do microthreading, which is scheduling at a finer grain
than real threads. An extension module (or even Python
module) might be thread safe, but not microthread safe.
The reason is that the extension does not have any knowledge
about this context switching, it still thinks of the
one existing thread.
For that reason, one should avoid to use certain
resources more than once in a real thread. Some
locking should be used to protect data structures.

At the moment the above problems are rather easy to
handle, since context switching is explicitly done
by the user.
I am going to support pre-emptive scheduling, soon,
which makes things more difficult, and I will provide
a mechanism that inhibits auto-scheduling unless
the user declares a module as definativly microthread
safe.

so much so far - ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From kowk at earthlink.net  Wed Sep 25 01:22:35 2002
From: kowk at earthlink.net (Kow K)
Date: Tue, 24 Sep 2002 22:22:35 -0700
Subject: how to change font and its size in IDLE (in UNIX)?
Message-ID: <CCBB7E70-D046-11D6-B43E-00050287C0A6@earthlink.net>

Hi all,

I'm trying to figure out how to change the font and its size in IDLE.
It's Courier in 10pt. I don't like Courier, and what's worse,
10pt is too small for me. I wish I could change it to Fixed 7x13.

Any input will be appreciated.

Kow




From aleax at aleax.it  Fri Sep 27 03:41:59 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 27 Sep 2002 07:41:59 GMT
Subject: readlines() and "binary" files
References: <3d90cacf$0$30457$afc38c87@sisyphus.news.be.easynet.net> <3d91de2d$0$30457$afc38c87@sisyphus.news.be.easynet.net> <9yqk9.143972$ub2.3161265@news1.tin.it> <3d936f05$0$30456$afc38c87@sisyphus.news.be.easynet.net>
Message-ID: <bZTk9.175574$pX1.6211960@news2.tin.it>

Justin wrote:

>> Pity you ruled out defining your own iterator --
> 
> I had ruled it out by laziness, but since you were kind enough to
> provide an implementation, I'll definitely use it! Thanks a lot,

You're welcome!  I'm lazy too -- that's why I like generators
so much... they do LAZY evaluation, after all:-).

> it works nicely. Gee I can't wait for that nutshell to crack !-}

Delay's not my fault _this_ time -- the complete, technically
reviewed and revised draft is now in the editor's hands -- she's
probably struggling to make my prose (a mix of verbose, enthusiastic
Mediterranean exhuberance [where I let myself go] and tight, precise, 
dry, maybe overly-technical conciseness [where I carefully revised
to minimize text down to the number of characters absolutely needed
to convey the essential information]) into a more reasonable,
Nutshell-style medietas (OK, so some of that _is_ my fault, I
guess, but there's nothing I can do about it right now...).


Alex



From ajw126NO at SPAMyork.ac.uk  Wed Sep 11 07:09:38 2002
From: ajw126NO at SPAMyork.ac.uk (Andrew Wilkinson)
Date: Wed, 11 Sep 2002 12:09:38 +0100
Subject: Abstract Syntax Trees
Message-ID: <GuFf9.1028$XE6.54057@newsfep1-win.server.ntli.net>

Hi,

I'm having a lot of trouble trying to get the parser and compiler modules to
properly parse some files.

At the moment the trouble seems to be with any loop construct (i.e. for or
while).

If you create a test file like this...

---
for i in range(10):
    print i
---
That'll run fine, but type this into python...

>>> import compiler
>>> compiler.parseFile(<insert name of file here>)

and you get...
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "C:\PROGRA~2\Python22\lib\compiler\transformer.py", line 43, in
parseFile
    return parse(src)
  File "C:\PROGRA~2\Python22\lib\compiler\transformer.py", line 47, in parse
    return Transformer().parsesuite(buf)
  File "C:\PROGRA~2\Python22\lib\compiler\transformer.py", line 115, in
parsesuite
    return self.transform(parser.suite(text))
parser.ParserError: could not parse string

I get the same thing if I replace the for loop with a while loop. A couple
of times it has parsed the for loop correctly, but that just seems to be
down to which way the wind was blowing a the time.

Does anyone know what's going on? I've only been able to test it on my
computer, so it could just be me...

I'm running Python 2.2.1

Thanks in advance,
Andrew Wilkinson




From gumuz at looze.net  Wed Sep  4 18:33:56 2002
From: gumuz at looze.net (Guyon Morée)
Date: Wed, 04 Sep 2002 22:33:56 GMT
Subject: Bind threads to addresses -- Windows & urllib?
References: <mailman.1031157680.4128.python-list@python.org>
Message-ID: <oTvd9.6$7z6.716@castor.casema.net>

i think you mean something like this:

import thread


def spiderStuff(Socket):
    # do the spider stuff here
    pass

...

thread.start_new_thread(SpiderStuff, (,newSocket))


"Nick Arnett" <narnett at mccmedia.com> wrote in message
news:mailman.1031157680.4128.python-list at python.org...
> I can't see an obvious easy way to bind threads to IP addresses on a Win2K
> machine.  I have a multi-threaded Web spider, which uses urllib to
retrieve
> pages, and would like to bind each thread to a different address on the
> machine, if possible.  Alternatively, I'd be willing to run separate
> processes to accomplish the same.  So far, my searches aren't coming up
with
> any solution... And it would be wonderful if such an approach were also
> portable to Linux...  Pointers?
>
> --
> Nick Arnett
> Phone/fax: (408) 904-7198
> narnett at mccmedia.com
>
>
>




From Henrik.Weber at sys.aok.de  Mon Sep 30 04:19:11 2002
From: Henrik.Weber at sys.aok.de (Henrik Weber)
Date: 30 Sep 2002 01:19:11 -0700
Subject: DBI cursor behaviour with multiple statements?
References: <7e964d1d.0209270137.568a3f3@posting.google.com> <mailman.1033143038.7112.python-list@python.org>
Message-ID: <7e964d1d.0209300019.361eea62@posting.google.com>

Gerhard H?ring <gerhard.haering at gmx.de> wrote in message news:<mailman.1033143038.7112.python-list at python.org>...
> 
> Ok. Then the Database SIG's mailing lists is the appropriate place to
> ask. You'll find that almost all module authors and many users are
> subscribed there.

Well, their archive is going to make interesting reading.

> 
> > Now I'm not quite sure how to interpret the DBI2 specification when it
> > comes to the execution of multiple data returning statements with the
> > same cursor.
> 
> The behaviour is undefined for more than one DQL statement. And
> .nextset() isn't meant for this case, either.

So I can take the easiest way out. I can live with that.

> 
> In fact, I asked the same question on the SIG's mailing list some time
> ago:
> 
> Date: Tue, 28 May 2002 21:36:47 +0200
> To: Python DB-SIG Mailing List <db-sig at python.org>
> Subject: [DB-SIG] Multiple Statements
> Message-ID: <20020528193647.GA902 at lilith.my-fqdn.de>
> 
> ... and was referred to a discussion about a year ago (then), where the
> consensus was what I summarized above.
> 
> Here's the SIG's mailing list: http://python.org/sigs/db-sig/
> 
> Btw. an updated form of the specification several interesting "optional
> features" is available as PEP 0249.
> 
> -- Gerhard

That sounds interesting. Thank you.

Henrik


From aldo at nullcube.com  Sat Sep 14 23:11:03 2002
From: aldo at nullcube.com (Aldo Cortesi)
Date: Sun, 15 Sep 2002 13:11:03 +1000
Subject: html template system?
In-Reply-To: <0ro7ouskb8ctv6f846gb8cfpqtl5vs1avc@4ax.com>
References: <btug9.688$Ee4.1090@news-server.bigpond.net.au> <0ro7ouskb8ctv6f846gb8cfpqtl5vs1avc@4ax.com>
Message-ID: <20020915031103.GA21206@nullcube.com>

":B nerdy" <thoa0025 at mail.usyd.edu.au> wrote:

>what are good templating systems avaliable for python?

A quick search on Google will turn up dozens. 

Start by reading the following interesting (but, sadly,
Perl-centric) article:

http://www.perl.com/pub/a/2001/08/21/templating.html

Perrin Harkins gives a very good summary of the broader
issues involved in picking a templating system. After that,
have a look at some of these:

	Cheetah - http://www.cheetahtemplate.org/
	YAPTU - http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52305
	Cubictemp - http://www.nullcube.com/software/cubictemp.html

They're all quite good, and your choice will depend mostly
on your coding style and sense of aesthetics. I use
Cubictemp for much of what I do, but since it was written by
my company, my preference for it may simply be habit or
bias...



Cheers,



Aldo



-- 
Aldo Cortesi
aldo at nullcube.com
http://www.nullcube.com



From mhammond at skippinet.com.au  Tue Sep 24 00:37:52 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 24 Sep 2002 04:37:52 GMT
Subject: Outlook 2002 and Python
References: <mailman.1032839910.12131.python-list@python.org>
Message-ID: <A_Rj9.13335$6g7.40581@news-server.bigpond.net.au>

Paul Sage wrote:
> What:  I saw an idea for a way to create a good SPAM filter for email. 
> 
> How: I would like to try to implement the algorithm in Python.
> 
> Why:  I am trying to get familiar with Python, as I will be using it at 
> work quite a bit, and would like to keep my skills honed.
> 
> Where:  Windows XP and Outlook 2002
> 
> Okay, so smart people would say, ?just go use Visual Studio or Visual 
> Basic,? I suppose.  I don?t know those two languages either, and they 
> might be the right tools for the job.  However, that would defeat the 
> point of trying something in Python.  Doing something else probably 
> wouldn?t be as useful to me personally.
> 
> Does anyone know how and where in the world I would even start to try to 
> write some kind of addin for Outlook 2002 in Python?   I don?t even know 
> how to start the project currently, and I am turning here first.

This would be an interesting project ;)

You would need to use the new Universal Gateway code (see 
win32com\test\test_pycomtest.py) and create a server that implements 
IDTExtensibility2.  From there, you should be able to use any of the VB 
samples to move forward - all the params passed are standard COM 
objects, so you should be able to do whatever you want.  You would also 
need to get your head around COM events and how to connect to them from 
Python, as addins are very much event based.

I'm doing some work on a plugin now - but not using Python :(  And I 
would *love* to see this work, especially in the face of the Bayes work 
being done in Python.  Mail me if you need a hand.

Mark.



From glibc at videotron.ca  Sun Sep  1 03:29:06 2002
From: glibc at videotron.ca (Pascal Gauthier)
Date: Sun, 01 Sep 2002 03:29:06 -0400
Subject: way to pickling locals()
Message-ID: <_Ljc9.154$LR1.3474@wagner.videotron.net>

Hi,
    I am currently writing an application that uses the 
InteractiveInterpreter (PyCrust to be exact). I want to be able to save 
the state by pickling locals() of the interpreter. The problem is that 
pickle doesn't support module pickling.

To make a long story short, I want to :

    l = locals()
    InteractiveInterpreter(locals=l)
    pickle.dump(l, file)
    ...
    InteractiveInterpreter(locals=pickle.load(file))

The only solution I found is to remove __builtins__ and marshal modules 
from the locals() dictionnary. But this is not really clean...

Any other suggestion ???
Thanks.



From mfranklin1 at gatwick.westerngeco.slb.com  Wed Sep 11 07:25:51 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Wed, 11 Sep 2002 11:25:51 +0000
Subject: function pointer
In-Reply-To: <ffcedbc9.0209110215.644c1788@posting.google.com>
References: <ffcedbc9.0209110215.644c1788@posting.google.com>
Message-ID: <200209111125.51483.mfranklin1@gatwick.westerngeco.slb.com>

On Wednesday 11 Sep 2002 10:15 am, Christina wrote:
> Hi!
>
> Does Python have function pointers? If not, is there a possibility to
> emulate them?


Excuse my ignorance but what is a function pointer....  
could this be the answer:-


def aFunction():
    print "aFunction was called"

## call it
aFunction()

## create a reference to it
af=aFunction

## call the reference
af()

## at the python command line:-

[dev]$ python
Python 2.2.1 (#3, May 29 2002, 20:32:44)
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> def aFunction():
...     print "aFunction was called"
...
>>> aFunction()
aFunction was called
>>> af=aFunction
>>> af()
aFunction was called
>>>


Regards
Martin




From phr-n2002b at NOSPAMnightsong.com  Fri Sep 20 05:05:57 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 20 Sep 2002 02:05:57 -0700
Subject: monetary applications [was: Python GUI app to impress the bos s?]
References: <31575A892FF6D1118F5800600846864DCBD817@intrepid> <23891c90.0209190014.26c5bb9@posting.google.com> <s2fdma.pp3.ln@ix.netcom.com> <7xk7lhb943.fsf@ruckus.brouhaha.com> <9Qzi9.117104$ub2.2497933@news1.tin.it>
Message-ID: <7xd6r9uj2i.fsf@ruckus.brouhaha.com>

Alex Martelli <aleax at aleax.it> writes:
> Using double binary floating-point, the result of 0.70 x 1.05 is 
> 0.73499999999999998667732370449812151491641998291015625; the result should 
> have been 0.735 (which would be rounded up to $0.74) but instead the 
> rounded result would be $0.73.

Interesting.

> Should your business get tax-audited, inspectors may take a dim view
> of your shaving a cent off the sales tax due on each and every item
> you sell to the public at $0.70 -- "I found it hard to imagine" how
> rounding effects might affect the result (when the law specifies
> rounding to the nearest cent and always rounding half cents up) may
> not be an acceptable defense.

I don't think they ask or care whether you round up or down on
individual 70 cent items.  They want to know the total value of the
stuff you sold, and then the tax you pay is X percent of that total,
for whatever your local value of X is.


From aglyportREMOvethispart at nospam.yahoo.com  Wed Sep  4 19:44:14 2002
From: aglyportREMOvethispart at nospam.yahoo.com (Anton Graph)
Date: Wed, 04 Sep 2002 16:44:14 -0700
Subject: Is it okay to compile 2.2.1 with gcc 2.96?
Message-ID: <3D769ACE.7000801@nospam.yahoo.com>

I've copied over the 2.2.1 tree from RH7.2 box to a RH7.3 box
and run gmake install. For some reason it decided to recompile everything.

My question is this. Is gcc 2.96:

Reading specs from /usr/lib/gcc-lib/i386-redhat-linux/2.96/specs
gcc version 2.96 20000731 (Red Hat Linux 7.3 2.96-112)

considered to be a stable compiler for the purpose of compiling 2.2.1?



From tismer at tismer.com  Thu Sep 19 09:10:05 2002
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 19 Sep 2002 15:10:05 +0200
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <mailman.1032329366.6813.python-list@python.org> <j4ptvbaapm.fsf@informatik.hu-berlin.de> <3D890FB0.4030408@something.invalid> <m3admeo4u1.fsf@mira.informatik.hu-berlin.de> <ambso1$4g8vh$1@ID-59885.news.dfncis.de> <j4fzw6tofh.fsf@informatik.hu-berlin.de>
Message-ID: <3D89CCAD.1020406@tismer.com>

Martin v. L?wis wrote:
> Thomas Heller <theller at python.net> writes:
> 
> 
>>>That is not true. It is more reliable to inspect the source code.
>>>In general, it is very easy to do so: if there are no callbacks to
>>>Python, the module is stackless-safe.
>>
>>So win32all probably wxPython are incompatible with stackless?
> 
> 
> No. If a module has callbacks, it is thread safe if it does not put
> data on the stack whose addresses are used outside the function, and
> it does not use setjmp, and it does not use exception handling.

Maybe sufficient for not crashing on the C level.
But this is not enough for micro-thread-safety.
See the random module for example, it needs a lock
or it will give unpredictable results.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From grabiller at 3dvf.net  Wed Sep 25 12:00:20 2002
From: grabiller at 3dvf.net (Guy Rabiller)
Date: Wed, 25 Sep 2002 18:00:20 +0200
Subject: Newbie References Question -> Exemple
References: <3d91c72f$0$8558$79c14f64@nan-newsreader-02.noos.net> <3d91d2cf$0$1484$79c14f64@nan-newsreader-01.noos.net> <yu99hegejcrw.fsf@europa.research.att.com>
Message-ID: <3d91dd96$0$14626$79c14f64@nan-newsreader-03.noos.net>

Yes, exactly.

I forgot that 'p' refere to nested lists :/

How shame..

Thanks a lot.

--
guy rabiller
3d animator / td
grabiller at 3dvf.net
http://grabiller.3dvf.net

Andrew Koenig wrote:
> Guy> But now, if I change the coordinate of one point:
> Guy> p[0] = [5,5]
> Guy> I will have:
>>>>> p
> Guy> [[5, 5], [0, 1], [1, 1], [1, 0]]
> Guy> but:
>>>>> quad
> Guy> [[0, 0], [0, 1], [1, 1], [1, 0]]
>
> That is because you are not changing the coordinates of the point; you
> are causing p[0]--the 0th element of p--to refer to a completely new
> point.
>
> If you wanted to change the coordinates of the point, you should
> have written this:
>
>         p[0][0] = 5
>         p[0][1] = 5
>
> or, more succinctly:
>
>         p[0][:] = [5, 5]
>
> If you try this, I think you will get the behavior you want.




From tjreedy at udel.edu  Thu Sep  5 18:12:21 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 05 Sep 2002 22:12:21 GMT
Subject: Python Source Code Documentation
References: <al69st$n1f$1@driftwood.ccs.carleton.ca> <G5Bd9.427560$m91.16660188@bin5.nnrp.aus1.giganews.com> <m3bs7cubp3.fsf@mira.informatik.hu-berlin.de> <al8k02$186$1@driftwood.ccs.carleton.ca>
Message-ID: <9FQd9.324581$Aw4.13464627@bin2.nnrp.aus1.giganews.com>

"Sean Ross" <sross at connectmail.carleton.ca> wrote in message
news:al8k02$186$1 at driftwood.ccs.carleton.ca...
> Terry: Yes, I did mean the Python interpreter. And I will give your
> suggestion a try.
>
> Martin: I will check out both of those documents.
>
> Thank you both for your suggestions.

When I looked a few years ago, I mentally divided the source files
into four groups:
  compilation phase (a few files)
  interpretation phase (1 file, the main loop -- ceval.c?)
  builtin datatypes (one for each -- int, long, tuple, etc -- each
with similar structure)
  other

Hope this helps a bit to get started

TJR






From rond at ixsys.com  Thu Sep  5 22:47:10 2002
From: rond at ixsys.com (Ron Darling)
Date: Thu, 5 Sep 2002 21:47:10 -0500
Subject: What database should I use
In-Reply-To: <F3EF14447A98D311B981009027B0FB380172469E@IMACS_EXCHANGE>
Message-ID: <BAEILLEBFGFMOHOGFDACEEIJCJAA.rond@ixsys.com>

Below is what I had sent to the original question off forum (replied
direct).  The description of VFP was not a sales pitch.  It was in response
to obsolete.

===============================

I program in Foxpro/VFP and have since 1989.  The reason I do that is not
because of the tables but because of the language and runtime enviroment
that is available to me.

In my opinion...

2000 - 3000 records is not a decision point for most DB table types.  Too
small they will all handle very well.

As far as I know in all three you will have to deal with ODBC so that does
not make a difference.

One of the advantages of the Fox enviroment (not Python -> Fox) is the table
can be native or not.  Meaning in VFP I can pull data from VFP tables as
well as Access and MySQL all at the same time but the VFP table is native.
Easy distribution.  I know this is not part of the answer you are looking
for but if you used VFP table and for some reason you needed more horse
power you have a cleaner option moving into VFP as far as the data is
concerned.  Most people are surprised to see the volume of data VFP is
handling out there in the business world once they find out.

MySQL - You have to install MySQL somewhere.  Seems a little much for 3000
records especially if you are running MySql on a separate box.

Access - Everything is stored in the MDB.  If it corrupts well you risk
loosing everything you stored in the .MDB file.  I have always just hated
Access but that is me.  Too many reasons to discuss here.

In my opinion VFP has the strongest most organized community mainly due to
http://www.universalthread.com.  I suspect you could even get support there
with Python -> VFP tables.  I see that discussed there often.

I think your choice comes down to what are you interested in learning and
living with or maybe what looks best on your resume with your specific
career path.

=============================

-----Original Message-----
From: Ganz, Will [mailto:wganz at imacs.com]
Sent: Thursday, September 05, 2002 11:18 AM
To: 'Ron Darling'; Sathish Chandrasekaran (S); printers at sendme.cz;
python-list at python.org; tutor at python.org;
activepython at listserv.ActiveState.com; python-help at python.org
Cc: db-sig at python.org; wx-users at lists.wxwindows.org
Subject: RE: What database should I use


I grant that FoxPro is still a formidable competitor in the marketplace but
given the size that the original poster stated of 2K-3k records; the 2gig
Access limitation shouldn't come into play. In this instance, Visual FoxPro
would be like using a supersonic fighter jet as a crop duster. 8-)

-----Original Message-----
From: Ron Darling [mailto:rond at ixsys.com]
Sent: Thursday, September 05, 2002 10:30 AM
To: Sathish Chandrasekaran (S); printers at sendme.cz; python-list at python.org;
tutor at python.org; activepython at listserv.ActiveState.com;
python-help at python.org
Cc: db-sig at python.org; wx-users at lists.wxwindows.org
Subject: RE: What database should I use


>> (though it's it considered obsolete now-a-days).

This is only a consideration for the uni-formed or those that have never
really used it.  MS is still dumping lots of money into it.  Just shipped
v7.0 will be going into Beta with 8.0 in a couple of months.

Amazing String capabilties.  As a matter of fact there is now a VFP style
.net class that moves a lot of the wonderful commands internal to VFP to
.NET.  I think the Fox Dev team created.

Cursor Technology, Rushmore Speed and many other things have moved there way
into not only Access but SQL Server and ADO.

Due to the fact that each part of a table is its own file.  .DBF is the
table.  .FPT is the Memo field  .CDX is the index.  The file size limitation
of the OS of 2 gig is at the individual file level.  In Access everything is
stored in the same file (you could break up but not really meant to) so the
2 gig file size is limited to the app and data which is more than likely in
the same file.  I have dealt with lots of situations where the .DBF was 1.8
gig, the index was 1.2 gig and the FPT was another 1+ gig size.

Indexes are stored in a .CDX file.  A .CDX file can have as many as 255
indexes (take into account 2 gig file size).  Each open table can have 10
open CDX's so you can have 2550 indexes on a table.

Very Robust SQL pass Through for back end databases.

There is nothing out there as fast as VFP for pure raw speed.  SQL 2000 gets
close but not quite there, the advantage to SQL as far speed goes is it does
not bog down under load.  Load means 25 + heavy users.

As far as speed goes I have seen the return of queries on tables as large as
21 million records within seconds and that was pulling data across a 10 meg
network with Foxpro For DOS.  United We Stand (remember Ross Perot)
member/prospect database.  Sounds a little dated but the point is its been
that way in Fox for a long time.

Fox and VFP managed/tracked the military assets to the Persian Gulf and is
till in use. Manages the flow of info for Chunnel. It is far more prevalent
than most realize.

Very good OOP.  It seems that VFP is becoming known for the best language
that can really implement the "Abstract Factory Pattern"

Very good interface. Rich Native Controls (No OCX) and does very well with
OCX's.

Easy Distribution.  (Drop a couple of DLLS into system directory not reg)

Rich Debugger.

Can access any data ODBC, ADO or ADO.Net can access.

Internal XML.  Select Result set from SQL into a local cursor then use
CURSORTOXML().  With the XMLTOCCURSOR() and
CURSORTOXML() commands it is very easy to build thin wire apps.  Bind the
interface to the cursor.

I can try any and all commands from the command prompt and even from within
a debugging suspend.  Pretty much interrogate the entire system at hand very
easily from a suspend.

Many options for hanging VFP behind IIS and ISAPI.

You can create com objects and run behind other languages even from within
SQL Server or visa versa.

One of the most robust SOUNDEX/Fuzzy Search add-ons available with the
PHDbase Library.  You can index memo fields and do 2 - 3 second searches on
huge amounts of memo data.

Best Table Repair Libraries that I am aware of are available.

Error trapping has been far superior to VB and a lot of languages since the
Fox 2.0 days (1992).

Last but not least http://www.universalthread.com.  Great support from all
over the world.


obsolete...hmmmmm maybe lack of exposure or conflicting marketing problems
with MS owning Access, SQL and Fox but certainly not obsolete.

-----Original Message-----
From: activepython-admin at listserv.ActiveState.com
[mailto:activepython-admin at listserv.ActiveState.com]On Behalf Of Sathish
Chandrasekaran (S)
Sent: Thursday, September 05, 2002 9:24 AM
To: printers at sendme.cz; python-list at python.org; tutor at python.org;
activepython at listserv.activestate.com; python-help at python.org
Cc: db-sig at python.org; wx-users at lists.wxwindows.org
Subject: RE: What database should I use


In that case MS-access will be better on the size issue. But on the other
hand if u want ur application to be flexible and ease of use I would rather
suggest FoxPro (though it's it considered obsolete now-a-days).


Here Is A Test To Know Whether The Mission Of Your Life Is Over . If You Are
Alive It Isn't .

Sathish C
Infra-Amdocs


-----Original Message-----
From: A [mailto:printers at sendme.cz]
Sent: 05 September 2002 15:21
To: python-list at python.org; tutor at python.org;
activepython at listserv.ActiveState.com; python-help at python.org
Cc: db-sig at python.org; wx-users at lists.wxwindows.org
Subject: What database should I use


Hello,
I am going to program a small application (in Python and wxPython) which
consists about 2000 - 3000 records . Can you please recommend which database
is the best for that. I use Windows. Maybe MS Access  or FoxPro or MySQL or
better? I want to keep the program as small as possible. Thanks Ladislav

_______________________________________________
ActivePython mailing list
ActivePython at listserv.ActiveState.com
To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
Other options: http://listserv.ActiveState.com/mailman/listinfo/ActivePython

_______________________________________________
ActivePython mailing list
ActivePython at listserv.ActiveState.com
To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
Other options: http://listserv.ActiveState.com/mailman/listinfo/ActivePython


_______________________________________________
ActivePython mailing list
ActivePython at listserv.ActiveState.com
To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
Other options: http://listserv.ActiveState.com/mailman/listinfo/ActivePython





From cliechti at gmx.net  Sat Sep  7 16:32:57 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 7 Sep 2002 22:32:57 +0200
Subject: getting clients ip address
References: <r0te9.25654$6W3.867995@weber.videotron.net>
Message-ID: <Xns9282E55E6E95Bcliechtigmxnet@62.2.16.82>

"Leader" <leader730 at hotmail.com> wrote in
news:r0te9.25654$6W3.867995 at weber.videotron.net: 

> Hello, I am using Python for my website, more specifically
> Spyce+mod_python, what I want to do is print out the clients ip
> address on the site so it can say something like "your ip
> is:xxx.xxx.xxx.xxx" basicly the python version of <?php echo
> getenv("REMOTE_ADDR"); ?> I'm sure I saw something a while ago while
> browsing through the module docs but I can't find it anymore.

something like:

>>> import os
>>> print os.environ["REMOTE_ADDR"]

HTH, 
chris

-- 
Chris <cliechti at gmx.net>



From Padraig at Linux.ie  Wed Sep 11 07:04:15 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Wed, 11 Sep 2002 11:04:15 GMT
Subject: regular expression for space seperated quoted string
Message-ID: <PqFf9.3627$cP3.7118@news.iol.ie>

Hi, I'm trying to split a string that is seperated
by spaces and also contains double quoted words which
can contain spaces:

For e.g. using: re.split('[ ]*"?([^"]*)"?[ ]*', s1)
on this string: s1='1 "2" "thre e"'
gives:          ['', '1 ', '', '2', '', 'thre e', '']

Problem with this is the '' entries, but this isn't too bad.

However using the above re with: s2='1 2 "th ree"'
I get:                           ['', '1 2 ', '', 'th ree', '']

any ideas?

thanks,
P?draig.



From aahz at pythoncraft.com  Mon Sep 23 21:55:59 2002
From: aahz at pythoncraft.com (Aahz)
Date: 23 Sep 2002 21:55:59 -0400
Subject: Can't suspend a thread?
References: <QZ4j9.10630$Lg2.2393215@news2.news.adelphia.net>
Message-ID: <amognf$r6g$1@panix1.panix.com>

In article <QZ4j9.10630$Lg2.2393215 at news2.news.adelphia.net>,
Robert Oschler <Oschler at earthlink.net> wrote:
>
>I was hoping to have a master thread that woke up and suspended several
>other threads dynamically, as needed.  From the books I have, that seems to
>be impossible in Python?  Is this true?  If so, is there a workaround?

Threads suspend themselves; they can be woken up by other threads.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From anthony at computronix.com  Mon Sep 30 12:55:36 2002
From: anthony at computronix.com (Anthony Tuininga)
Date: 30 Sep 2002 10:55:36 -0600
Subject: redhat8
In-Reply-To: <873crr4fy6.fsf@flibuste.net>
References: <873crr4fy6.fsf@flibuste.net>
Message-ID: <1033404937.1946.25.camel@chl0151.edmonton.computronix.com>

According to the release documentation:

Python 2.2.1 is on the CD.

Finally.... :-)

On Mon, 2002-09-30 at 10:09, William wrote:
> hi,
> 
> anybody know wich version of python is /usr/bin/python on the new redhat
> release ?
> 
> -- 
> William Dod? - flibuste.net
> http://wikipython.tuxfamily.org
> -- 
> http://mail.python.org/mailman/listinfo/python-list
-- 
Anthony Tuininga
anthony at computronix.com
 
Computronix
Distinctive Software. Real People.
Suite 200, 10216 - 124 Street NW
Edmonton, AB, Canada  T5N 4A3
Phone:	(780) 454-3700
Fax:	(780) 454-3838
http://www.computronix.com




From nm at bolas.clara.co.uk  Thu Sep 26 06:12:49 2002
From: nm at bolas.clara.co.uk (Nick B)
Date: Thu, 26 Sep 2002 11:12:49 +0100
Subject: Multiple file open dialog
Message-ID: <3d92dce4$0$1286$cc9e4d1f@news.dial.pipex.com>

Hi,

Can anyone advise on how to get a list of multiple file names from a GUI
dialog.  I am working exclusively in Win32 but so far only have GUI
experience with Tkinter.  What I would like is access to the standard Win32
shift-click and control-click conventions.

Tkinter's askopenfilename can only return a single filename (?).
Pythonwin's win32gui.GetOpenFileName looks like it could do the job, but a
simple example would really help. Other ideas ?

Any help appreciated,

Nick




From chris.walsh at real-sense.com  Fri Sep  6 11:46:41 2002
From: chris.walsh at real-sense.com (Chris Walsh)
Date: 6 Sep 2002 08:46:41 -0700
Subject: Embedding - Local dictionaries for local scripts
Message-ID: <1691b372.0209060746.55b8cf1@posting.google.com>

Hi,

I am embedding Python into a 3D renderer and for each "object" in the
environment, I want some local script which can get run.  Each object
will have it's own local variables which must persist between calls to
it's script there I believe I want to create a local dictionary for
each object which gets passed into the function:

    PyObject* PyRun_String(
        char *str,
        int start,
        PyObject *globals,
        PyObject *locals);

I have seen some stuff in comp.lang.python about creating a dict and
adding the __builtin__ module but I have been searching now for an
hour and cannot find it.

Can anyone point me to some code which creates this dictionary?

Kind Regards,

Chris.
"God is Real, unless declared Integer"


From wesc at deirdre.org  Thu Sep  5 02:25:35 2002
From: wesc at deirdre.org (wesc at deirdre.org)
Date: Wed, 4 Sep 2002 23:25:35 -0700 (PDT)
Subject: ANN: BayPIGgies mtg Wed 9/11 7:30pm
Message-ID: <200209050625.g856PZx20521@alpha.ece.ucsb.edu>

BayPIGgies: Silicon Valley-San Francisco Bay Area Python Users Group

Agenda:   Using a New MVC Architecture to Create a
	  Pythonic XML-based Web Application Framework
When:     September 11, 2002 @ 7:30pm
Where:    Stanford University, Palo Alto, CA
Speakers: Paul McGavin, Donovan Preston, Sam Penrose from InterSight

We will discuss how we are applying a new Python-based MVC
architecture to create an efficient, Pythonic web application
framework that separates presentation templates from the Python
source code with DOMTemplate. The framework componentizes behavior
into reusable objects with DOMWidgets and DOMHandlers. 

After creating web applications in three separate Python frameworks
-- Webware for Python, Zope and Apache/Python cgi -- InterSight
decided to create its own MVC framework, based on some of the ideas
from IBM and Lutris Enhydra's Barracuda project, to power their
automated publishing applications. Their applications enable scalable
content: the ability to produce, from a single set of managed data
and images, many different marketing communication pieces: press-
ready Adobe InDesign and QuarkXPress files, web pages, order sheets,
sample stickers, price tags and more. 

We are actively seeking speakers for BayPIGgies, esp. October 9th!
If you would like to give a talk at one of our meetings (any Python-
related topic), please contact us! 

more info including directions:    http://www.baypiggies.net

hope to see some of you on Wednesday!

-wesley

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

"Core Python Programming", Prentice Hall PTR, ? 2001
    http://starship.python.net/crew/wesc/cpp/

Silicon Valley-San Francisco Bay Area Python Users Group (BayPIGgies)
    http://www.baypiggies.net

wesley.j.chun :: wesc at deirdre.org
cyberweb.consulting : henderson, nv : cyberweb at rocketmail.com
http://roadkill.com/~wesc/cyberweb/



From see_reply_address at something.invalid  Wed Sep 18 02:21:26 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Wed, 18 Sep 2002 18:21:26 +1200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3D881B66.1080200@something.invalid>

Martin v. Loewis wrote:

> 
> They aren't really dangerous, just unconventional.


Last I heard, Christian was moving pieces of the C
stack while there could be pointers to it in use.
That's definitely dangerous, as he discovered when
it broke Tk.

I'd be very nervous about using Stackless until
it stops doing that.

 > And that cannot be
 > fixed, as it is essential to operation of Stackless.

Seems to me it could be fixed by just copying the
C stack as a whole, instead of piecemeal.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From aleax at aleax.it  Fri Sep 20 14:45:15 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 20 Sep 2002 18:45:15 GMT
Subject: Is there a widget or program  for a database browsing
References: <mailman.1032510447.21991.python-list@python.org>
Message-ID: <%0Ki9.140164$pX1.4973236@news2.tin.it>

A wrote:

> In my program users will search  a database and I need to display choosen
> record(s ) in a nice graphical way. Is there a module( program), that I
> could use together
> with Python, that  could easily  show the database record(s) and take care
> of all thinks during browsing? Thanks for help.

The latest version of PyQt comes close to these requirements, I think,
but I haven't tried out the "SQL-enabled widgets" yet.


Alex



From lac at strakt.com  Thu Sep 26 12:55:26 2002
From: lac at strakt.com (Laura Creighton)
Date: Thu, 26 Sep 2002 18:55:26 +0200
Subject: off topic -- spam assassin 
In-Reply-To: Message from Skip Montanaro <skip@pobox.com> 
   of "Thu, 26 Sep 2002 10:49:01 CDT." <15763.11373.573554.830138@12-248-11-90.client.attbi.com> 
References: <200209261407.g8QE76Wo028975@ratthing-b246.strakt.com>  <15763.11373.573554.830138@12-248-11-90.client.attbi.com> 
Message-ID: <200209261655.g8QGtQo9029577@ratthing-b246.strakt.com>

2.20 is was.  Thank you for the recommendation, 2.41 it is now.  We'll
see how long that holds them.

Thanks again,
Laura



From eric.brunel at pragmadev.com  Tue Sep  3 05:16:17 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Tue, 3 Sep 2002 11:16:17 +0200
Subject: what about things like __*** ?
References: <7153651c.0209021927.16a3efaf@posting.google.com> <al1p8n$12d$1@wanadoo.fr> <3d746df6$0$222$4d4ebb8e@news.nl.uu.net>
Message-ID: <al1u6n$j3u$1@wanadoo.fr>

Gumuz wrote:

> ah great, I found asking this myself last night, too.
> 
> isn't this the same for:   __methodName(self, parm) which makes it a
> private method?

Yep!

class C:
  def __m(self):
    print 'spam'
  def __init__(self):
    self.__m()          # Works
o = C()
o.__m()                 # Does not work

-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From gumuz at looze.net  Wed Sep  4 05:29:15 2002
From: gumuz at looze.net (Gumuz)
Date: Wed, 4 Sep 2002 11:29:15 +0200
Subject: encryption (passwords)
Message-ID: <3d75d115$0$221$4d4ebb8e@news.nl.uu.net>

I am quite a newbie regarding encryption stuff. There are some issues
boggling my head and I suppose these come from the fact that I have no clue
how to apply encryption in a real-life situation.

I am using network-sockets to communicate between two(or more) python
applications. I want to encrypt these messages, or at least a part of it
e.g. password and such.

My limited understanding of encryption tells me that I need a 'key' to
decrypt stuff. So I figure that this key will bey written(hard-coded?)
somewhere in my Python-script and it needs to be known at both sides in
order to encrypt and decrypt. Somehow, I can't believe this is true. I can't
get over the idea that if my application becomes open-source anyone could
see the key and decrypt my messages.

I am fairly certain that it probably doesn't work this way, but I have
really no idea. Cann someone clarify this for me?


thanx a lot,
Guyon






From mgerrans at mindspring.com  Fri Sep 13 02:12:05 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Thu, 12 Sep 2002 23:12:05 -0700
Subject: IDLE
References: <wpbf9.264320$aA.45691@sccrnsc02> <alk170$pai$1@bob.news.rcn.net>
Message-ID: <aluedr$rln$1@slb0.atl.mindspring.net>

Amazing that such an inane question could engender so many entertaining
answers, but I think this one wins the prize for best chuckle.   (...so
far...)

"Raymond Hettinger" <python at rcn.com> wrote in message
news:alk170$pai$1 at bob.news.rcn.net...
> > Why won't my IDLE work?
>
> This is the documented behavior according
> to the American Heritage Dictionary:
>
> idle (intr. verb):  To pass time without working
>                             or while avoiding work.





From sdeibel at wingide.com  Tue Sep 17 10:12:13 2002
From: sdeibel at wingide.com (Stephan R.A. Deibel)
Date: Tue, 17 Sep 2002 10:12:13 -0400 (EDT)
Subject: O'Reilly Python Success Stories
Message-ID: <Pine.LNX.4.30.0209170950530.1509-100000@pangolin.wingide.com>

Hi,

O'Reilly Associates has agreed to do a Python Success Stories book similar
to those that already exist for Perl.  This book will be given away for
free at bookstores and conferences, and is a great way to showcase what
Python can do.

If you have a Python success story to tell, please consider contributing
to this effort.  O'Reilly will do the formatting, cover design, and
printing, but the collection and editing of stories is left up to
volunteers in the Python community.

More information on how to contribute can be found here:

http://pythonology.org/successguide

While we prefer completed stories, we can also accept responses to the
list of questions given in the above URL.

Thanks,

- Stephan

------------------------------------------------------------------------
Wing IDE for Python                          Archaeopteryx Software, Inc
www.wingide.com                              Take Flight!





From jhorneman at pobox.com  Tue Sep  3 06:14:40 2002
From: jhorneman at pobox.com (Jurie Horneman)
Date: 3 Sep 2002 03:14:40 -0700
Subject: Which XML parser under Python can handle Unicode?
Message-ID: <9ac02e81.0209030214.515d6c0@posting.google.com>

<whine>
XML is complicated (if you dig deep).
Python libraries are complicated (if what you want to know is
apparently not in the documentation).
Unicode is complicated (if you dig deep).
</whine>

Aaahhhh....

Could someone please recommend an XML parser under Python which can
handle Unicode? Specifically, most European languages and Asian
languages such as Chinese (traditional, simplified), Japanese, Korean
and Thai? (Please don't make me look up the actual codetables.)
Ideally the parser would use Python's codec system.
I'd prefer an event-based parser, although tree-based would not be a
huge problem.

Thank you thank you

Jurie.


From gmuller at worldonline.nl  Tue Sep 24 15:56:33 2002
From: gmuller at worldonline.nl (GerritM)
Date: Tue, 24 Sep 2002 21:56:33 +0200
Subject: How to get memory usage on Windows?
References: <amilhk$bi8$1@reader1.tiscali.nl> <mailman.1032715528.24198.python-list@python.org> <3D8EEB7C.C1B2F01F@philips.com>
Message-ID: <amqg5i$iaq$1@reader1.tiscali.nl>

"Gerrit Muller" <gerrit.muller at philips.com> schreef in bericht
news:3D8EEB7C.C1B2F01F at philips.com...
>
>
> Emile van Sebille wrote:
> >
> > GerritM:
> > [snip]
> > > how much memory is used per phase and/or per function. Is there any
> > function
> > > returning this information. I noted that on Unix the resource module
> > gives
> > > access to this kind of information, but I could not find any Windows
> > > alternative.
> > >
<...snip...>
 In mxTools a sizeof(object) function is available, which should allow me to
reconstruct most memory consumption data via a procedure, although some care
should be taken wrt referenced data.

http://www.egenix.com/files/python/mxTools.html

regards Gerrit

> http://www.extra.research.philips.com/natlab/sysarch/index.html




From hancock at anansispaceworks.com  Tue Sep  3 01:06:00 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Mon, 02 Sep 2002 22:06:00 -0700
Subject: Would you like a Snobol pattern matcher? (Re: Making regex suck 
 less)
References: <20020903015907.15062.14810.Mailman@mail.python.org>
Message-ID: <3D744338.D2E223CC@anansispaceworks.com>

From: Greg Ewing <see_reply_address at something.invalid>
> Maybe it's time for me to resurrect the Snobol-style
> pattern matching module that I started on a while
> back and never got around to releasing.
> 
> Would anyone be interested in this? Its
> syntax is similar to that of Plex REs, except that
> the primitives are Snobol-like, and it uses a
> backtracking matching algorithm that's much more
> powerful than a DFA (you can write entire parsers
> in it, for example).

Some interesting references, possibly? --

http://sourceforge.net/projects/pystemmer/
http://snowball.sourceforge.net

These may be more specialized -- Snowball is
specifically for algorithmically stemming words,
and PyStemmer is an interface to it. I haven't
really looked into how it works.  The name is
related to SNOBOL, but I'm not sure how much
Snowball actually resembles it (if at all).

I'm using pystemmer as part of a function which
converts object titles to (hopefully) mnemonic
file names (ids) in Zope.  I haven't really
looked into how it works.

But I though it might be relevant to you.

Cheers,
Terry

-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
Anansi Spaceworks                 
http://www.anansispaceworks.com 
P.O. Box 60583                     
Pasadena, CA 91116-6583
------------------------------------------------------



From tjreedy at udel.edu  Fri Sep  6 09:00:16 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 06 Sep 2002 13:00:16 GMT
Subject: Python and Jython inconsistencies when encoding strings
References: <ala00d$dij$1@wanadoo.fr>
Message-ID: <AF1e9.209079$On.8198440@bin3.nnrp.aus1.giganews.com>

"Andre Michel Descombes" <amdescombes at qualicontrol.com> wrote in
message news:ala00d$dij$1 at wanadoo.fr...

> when I do the following in Python :
>
> >>> s = u'?' # Alt-0163, pound sign
> >>> print s
>
> I get the following exception, which is a normal exception:
>
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> UnicodeError: ASCII encoding error: ordinal not in range(128)
...

On Python2.2, Windows

>>> s=u'?' # Alt-163: u with acute accent
>>> s
u'\xa3'
>>> print s
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
UnicodeError: ASCII encoding error: ordinal not in range(128)

What surprises me is that I thought that raw expression
(interactively) was abbreviation for print expression.  Bug here?

Terry J. Reedy





From pinard at iro.umontreal.ca  Tue Sep  3 20:22:38 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: Tue, 03 Sep 2002 20:22:38 -0400
Subject: Pyrex 0.4.4
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8BF0A563@natasha.auslabs.avaya.com> ("Delaney,
 Timothy"'s message of "Wed, 4 Sep 2002 09:27:10 +1000")
References: <B43D149A9AB2D411971300B0D03D7E8BF0A563@natasha.auslabs.avaya.com>
Message-ID: <oqheh6ob75.fsf@titan.progiciels-bpi.ca>

[Delaney, Timothy]

>> From: Greg Ewing 
>> 
>> Pyrex 0.4.4 is now available:
>> 
>>     http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/
>
> The Plex package appears to be a shortcut/link, rather than the package ...

I noticed too.  To get going, I got Plex installed separately, and removed the
'Plex' line in `setup.py' for Pyrex.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



From loewis at informatik.hu-berlin.de  Mon Sep  9 07:48:31 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 09 Sep 2002 13:48:31 +0200
Subject: xml.minidom is stripping out my CRLF's in attrib values!!
References: <3D7C80C4.9070204@NOSPAMbitbuilder.com>
Message-ID: <j4n0qr9yeo.fsf@informatik.hu-berlin.de>

Ahmad Baitalmal <ahmad at NOSPAMbitbuilder.com> writes:

> What's the deal, is there a setting I can set?

XML specifies that line endings must be normalized by a processor, see

http://www.w3.org/TR/REC-xml#sec-line-ends

There is no option to report anything but #xA to the application.

Regards,
Martin



From see_reply_address at something.invalid  Thu Sep 12 21:33:42 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Fri, 13 Sep 2002 13:33:42 +1200
Subject: Piper Pickling?
References: <dg9g9.2373$Lg2.427869@news2.news.adelphia.net>
Message-ID: <3D814076.9050509@something.invalid>

Robert Oschler wrote:

> If Peter Piper pickles a peck of Python packages, can he pack the pickled
> Python peckers in a proper packing place?
> 
> Translated: Anybody seen a module that can override the standard pickling
> process and store and retrieve Python objects to a MySQL database?


Well, you can use pickle.dumps() to pickle your objects
into strings, and then store the strings in your database.

Will such suitably squashed SQLed strings satisfy?

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From peter at engcorp.com  Mon Sep  2 21:59:38 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 02 Sep 2002 21:59:38 -0400
Subject: RedHat 7.3 and Python2.2.1
References: <mailman.1031014083.14649.python-list@python.org>
Message-ID: <3d74176d@news.sentex.net>

Kow Kuroda wrote:
> 
> On Monday, September 2, 2002, at 11:43 , Peter Hansen wrote:
> 
>> Doug Farrell wrote:
>>> I'm trying to install the latest version of Python
>>> (python2-2.2.1-1.i386.rpm) from the www.python.org download page, and
>>> I'm having problems. When I run
>>>
>>> rpm -i python2-2.2.1-1.i386.rpm
>>>
>>> I get all kinds of conflict messages like this:
>>>
>>> file /usr/bin/pydoc2 from install of python2-2.2.1-1 conflicts with
>>> file from package python2-2.2-16
>>
>> Sounds like you already have a previous version of Python2.2 installed.
>> You should probably be using the "-U" option to update instead of
>> the "-i" option to install... or uninstall the previous one first.
> 
> I just wanted to make a comment on Peter's post.
> 
> I think -U or --upgrade option works, but it's dangerious to uninstall 
> Python 2.1, which is now part of Red Hat Linux 7.3, unlike in prevous 
> releases where Python 1.5.2 or something is used for installation scripts.

(Note he had a conflict with Python 2.2, not 2.1.  It should work fine.)

Your comments about RH7.3 don't quite fit with my experience.  I just
installed it (posting from the machine it's on, in fact) and "python"
executes 1.5.2 still, while "python2" executes 2.2, not 2.1.  What am
I missing?

-Peter



From aahz at pythoncraft.com  Fri Sep 20 10:58:47 2002
From: aahz at pythoncraft.com (Aahz)
Date: 20 Sep 2002 10:58:47 -0400
Subject: Destructor never called ???
References: <76710511.0209111450.40999dd8@posting.google.com> <4e8lou018f9uosl5p9mb0tt6691mfivamj@4ax.com> <ame8es$ppc$1@panix2.panix.com> <buzi9.136158$pX1.4848554@news2.tin.it>
Message-ID: <amfd37$r1a$1@panix2.panix.com>

In article <buzi9.136158$pX1.4848554 at news2.tin.it>,
Alex Martelli  <aleax at aleax.it> wrote:
>Aahz wrote:
>>
>> There are several different issues.  First of all, __del__ does not get
>> called if an object is destroyed by GC (GC gets used to clean up cyclic
>> garbage).  Secondly, Jython does not use refcounting; it's pure GC, just
>
>Right, but, I'd put it differently: GC is inhibited from destroying
>objects whose types define a __del__ -- rather, it collects them into
>gc.garbage, because it can't know that finalizing them is safe.  

Thanks (and also to Tim) for correcting my sloppy writing.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From see_reply_address at something.invalid  Tue Sep 24 23:23:13 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Wed, 25 Sep 2002 15:23:13 +1200
Subject: Strange values in Accept-Language header?
References: <f72dac0d.0209231303.12349f2e@posting.google.com> <8FWj9.39597$V7.10466682@twister.socal.rr.com>
Message-ID: <3D912C21.30109@something.invalid>

Jeff Davis wrote:

>>Accept-Language: en-us,x-ns1rDGeT4e2FpA,x-ns2f67971hgDw1
>>Accept-Language: x-ns13V8cN8S1Xz9,x-ns2T329gfxKa7d
>>
>>I have no idea what these could mean


Oh, boy, you've done it now. These refer to secret code
languages used by the PSU. Now that you've drawn attention
to them, the PSU will cut off all your



From ehagemann at comcast.net  Mon Sep 30 17:28:32 2002
From: ehagemann at comcast.net (Eric Hagemann)
Date: Mon, 30 Sep 2002 21:28:32 GMT
Subject: help with writing extension
References: <inHl9.24658$Ii4.986202@bin2.nnrp.aus1.giganews.com> <pQYl9.190868$pX1.6828893@news2.tin.it>
Message-ID: <4m3m9.331653$216.13979015@bin4.nnrp.aus1.giganews.com>

Alex
    Thanks for another excellent answer.  I am going to dig into the
Py_TPFLAGS_CHECKTYPES flag as that might just fix my problem.  I did not
really want to use the __coerce__function but I did not know how to get
around it. If adding this flag to the definition will skip the __coerce__
function, that should save me a bunch-o-code.  BTW I see references to the
CHECKTYPES flag in the Python 2.1 source (Yes - I am still using 2.1) any
reason why I need to move to 2.2 ?

I eventually (after studying the sources) did find the reason for my
"problem".

Summarized for any that might want to know.....

    When you do a binary_op between two objects that are not equal in type
the __coerce__ function is called for the first object.  If that does not
'coerce' the different objects into compatable ones then the __coerce__
function is called for the second object.  If that does not return
sucessfully then you are done for.  However if that returns sucess then the
operation is called using the method defined for the first of the two
objects

    As it turns out I already skipped throught __coerce__ function by just
incrementing the references on the two objects and returning success.  This
worked well when I was doing operations of <obj>+int  as I would just pass
both on to my 'add' function and I delt with the integer in my function.
However when working with the arguments in reverse, this method would not
work as after the second coerce function I would still have an  int+<obj>
and it would call the 'add' function for the int, which even after two
attempted coercions, would not know what <obj> was and the TypeError was
generated.

    What was refreshing was that I could find the root of the problem by
studying the source -- says something for how the code was written, nice --
very nice.

Cheers



"Alex Martelli" <aleax at aleax.it> wrote in message
news:pQYl9.190868$pX1.6828893 at news2.tin.it...
> <posted & mailed>
>
> Eric Hagemann wrote:
>
> > I have been handcrafting an extension module creating a new type
> >
> > I am making use of the numeric methods (+,-,* etc)
> >
> > I solved the problem of the object interating with other types by
working
> > through the __coerce__() function and have no issue with operations like
> > <obj>+2 but I am getting an error on 2-<obj>.
>
> As I understand, __coerce__ is not recommended for new code.  You should,
> I believe, add Py_TPFLAGS_CHECKTYPES to your tp_flags slot, and be ready
> in your numeric functions to deal with types as they come.  Sometimes
> coercion can be handier, but in general avoiding it is better.
>
>
> > In Beazley's book there are descriptions of functions like __radd__()
when
> > defining a new type in python, but I cannot find reference to these
> > functions in C ?
>
> If you grep Python's C sources (2.2 or better), you'll find some such
> references.  Slots nb_add of the PyNumberMethods structure to which
> slot tp_as_number of your PyTypeObject structure point supplies both
> __add__ and __radd__ (when you call PyType_Ready on your type, suitable
> wrappers on that slot are exposed with those two names).
>
>
> > I have checked the doc's (not to say that I did not miss something ;) )
>
> I think you still need to complement existing docs with some little
> study of the Python sources when you want to design rich and complete
> extension types -- and I think that dealing with __radd__ etc in the
> right way is part of a "rich and complete" (rather than just a
> really minimal) extension type.  I think this will still be true
> when my "Python in a Nutshell" is finally out, because the chapter
> on extending and embedding with C is limited to 40-50 pages (can't
> be more precise right now because the book's being edited and I'm
> not sure how many of my examples and details the editor will want
> to keep -- or if even more will be added), and doing complete justice
> to the subject would require about 10 times as much space.
>
> But I also think that asking here about such issues is just right!
> Best place to get help -- and this way the help gets archived (thanks
> to google groups and other archivers) and can be pointed to in the
> future (which alas doesn't work for other mailing lists such as
> python-help, nor, of course, for private mail:-).
>
>
> An example can help -- sorry if it's not very concise, but the
> PyTypeObject structure is so large that any examples of its use
> are also large, even though most fields are 0.  Say I have a
> directory ~/extype with a setup.py file such as:
>
> from distutils.core import setup, Extension
>
> setup(name = "samp1",
>     version = "1.0",
>     description = "Sample extension type",
>     maintainer = "Alex Martelli",
>     maintainer_email = "aleaxit at yahoo.com",
>
>     ext_modules = [ Extension('samp1', sources=['samp1.c']) ]
> )
>
> and a samp1.c file such as:
>
> #include "Python.h"
>
> typedef struct {
>     PyObject_HEAD
> } samp1;
>
> static void
> samp1_dealloc(PyObject *op)
> {
>     op->ob_type->tp_free(op);
> }
>
> static PyObject *
> samp1_amethod(PyObject *self)
> {
>     Py_INCREF(Py_None);
>     return Py_None;
> }
>
> static PyObject *
> samp1_add(PyObject *self, PyObject *other)
> {
>     PyObject* result = PyTuple_New(2);
>     if(!result)
>         return NULL;
>     Py_INCREF(self);
>     PyTuple_SET_ITEM(result, 0, self);
>     Py_INCREF(other);
>     PyTuple_SET_ITEM(result, 1, other);
>     return result;
> }
>
> static PyMethodDef samp1_methods[] = {
>     {"amethod", (PyCFunction)samp1_amethod, METH_NOARGS},
>     {NULL,  NULL}           /* sentinel */
> };
>
> static PyNumberMethods samp1_as_number = {
>     samp1_add,
> };
>
> static PyTypeObject samp1_t = {
>     PyObject_HEAD_INIT(0)
>     0,
>     "samp1",
>     sizeof(samp1),
>     0,
>     samp1_dealloc,
>     0,
>     0,
>     0,
>     0,                                  /* tp_compare */
>     0,                                  /* tp_repr */
>     &samp1_as_number,                   /* tp_as_number */
>     0,                                  /* tp_as_sequence */
>     0,                                  /* tp_as_mapping */
>     0,                                  /* tp_hash */
>     0,                                  /* tp_call */
>     0,                                  /* tp_str */
>     0,                                  /* tp_getattro */
>     0,                                  /* tp_setattro */
>     0,                                  /* tp_as_buffer */
>     Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE|Py_TPFLAGS_CHECKTYPES,
>     "sample extension type",
>     0,                                  /* tp_traverse */
>     0,                                  /* tp_clear */
>     0,                                  /* tp_richcompare */
>     0,                                  /* tp_weaklistoffset */
>     0,                                  /* tp_iter */
>     0,                                  /* tp_iternext */
>     samp1_methods,                      /* tp_methods */
>     0,                                  /* tp_members */
>     0,                                  /* tp_getset */
>     0,                                  /* tp_base */
>     0,                                  /* tp_dict */
>     0,                                  /* tp_descr_get */
>     0,                                  /* tp_descr_set */
>     0,                                  /* tp_dictoffset */
>     0,                                  /* tp_init */
>     PyType_GenericAlloc,                /* tp_alloc */
>     PyType_GenericNew,                  /* tp_new */
>     _PyObject_Del,                      /* tp_free */
> };
>
> static PyMethodDef no_methods[] = { {0} };
>
> void
> initsamp1(void)
> {
>     PyObject* self;
>     PyType_Ready(&samp1_t);
>     self = Py_InitModule("samp1", no_methods);
>     PyObject_SetAttrString(self, "samp1", (PyObject*)&samp1_t);
> }
>
>
> Now of course running python setup.py in this dictionary builds
> and installs extension module samp1.  Let's look at it in action:
>
> [alex at lancelot extype]$ python
> Python 2.2.1 (#2, Jul 15 2002, 17:32:26)
> [GCC 2.96 20000731 (Mandrake Linux 8.1 2.96-0.62mdk)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> import samp1
> >>> examp = samp1.samp1()
> >>> examp + 'ciao!'
> (<samp1 object at 0x8160f28>, 'ciao!')
> >>> 'salute...' + examp
> ('salute...', <samp1 object at 0x8160f28>)
> >>>
>
> See what's going on?  Despite the (traditional) argument names,
> samp1_add is being called with the LEFT operand of + as the first
> argument ('self'), the RIGHT operand as the second argument
> ('other') -- your C code can test the types directly to do the
> right thing in both the __add__ and __radd__ cases...!
>
> Sure, you CAN make do with coercion (so have all Python extension
> types until about last year), but bypassing it makes things a bit
> clearer, I believe.  If you do want coercion in some but nor all
> of your numeric-methods you can deal with it yourself -- call your
> own "coercion-like" function from your methods' code as and where
> appropriate.  I do think it's simpler than having Python do it for
> you and sometimes having to "bypass" it, so to speak... this way
> you can still supply a function to be called when the user does an
> explicit coerce call, and yet have whatever precise behavior you
> want for all operators that look like numeric operators ("look like",
> because, for example, sequence concatenation isn't really numeric --
> it just LOOKS LIKE it is, and has to live in PyNumberMethods, but
> that's not quite where one might like to place it...).
>
>
> Feel free to keep posting such questions to this list if there is
> anything unclear in the above, and/or if you need any more help!
>
>
> Alex
>




From megabytemonster at hotmail.com  Tue Sep 24 15:42:23 2002
From: megabytemonster at hotmail.com (Stuart)
Date: 24 Sep 2002 12:42:23 -0700
Subject: Using internet modules to download files
Message-ID: <3ab719f7.0209241142.1bf5987@posting.google.com>

I have a slight problem.  The college which I attend has broadband
access for everyone, but unfortunately insists upon placing a firewall
between the pupils and downloads, therefore this fast connection is
wasted upon retrieving HTML files.  I merely want to use the broadband
to download a few more python libraries, such as win32all, as my home
pc doesn't have internet access.

I have been experimenting with Python's internet protocols, in
particular with trying to retreive files while bypassing the firewall.
 I am not sure, however, which modules to use (i am a relative python
neophyte).  I first tried with urllib, but that unforunately merely
routes you throught the webbrowser, which of course, promptly blocks
any attempt.  Are then any modules which will bypass the
browser/firewall, and allow you to download a file directly through
the script?  Any advice (especially samples of code) would be greatly
appreciated.

Thanks,
Stuart


From tjreedy at udel.edu  Tue Sep 17 18:31:21 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 17 Sep 2002 22:31:21 GMT
Subject: subclassing tuple
References: <H2LMGE.1JBF@beaver.cs.washington.edu>
Message-ID: <Z2Oh9.101815$z91.3948137@bin3.nnrp.aus1.giganews.com>

"Douglas Zongker" <dougz at cs.washington.edu> wrote in message
news:H2LMGE.1JBF at beaver.cs.washington.edu...
> I could do this with lists instead, but that means changing a lot of
> existing code that expects tuples (and I don't want these objects to
> be mutable!)

You might get what you want by starting with a list, doing your custom
modifications, and then 'return tuple(mylist)' or subclass thereof.

TJR




From Padraig at Linux.ie  Fri Sep 13 09:23:35 2002
From: Padraig at Linux.ie (Padraig Brady)
Date: Fri, 13 Sep 2002 13:23:35 GMT
Subject: [newbie] Buying an introductory/reference text
References: <Mu5g9.1702$U_.1021@news04.bloor.is.net.cable.rogers.com>
Message-ID: <3D81E6B0.3070404@Linux.ie>

Brad Fonseca wrote:
> Greetings!
> 
> I'm looking for a few opinions on the above subject.  I'm a computer 
> science student who has had some introductory lectures on Python and I've 
> done some very basic mucking around with.  I'm looking for a clear 
> reference and/or introductory text.

I've only been doing python for a few days myself
with just online docs, which I find excellent.
If you want a treeware reference, have a look @:
http://www.accu.org/bookreviews/public/reviews/0sb/python.htm

P?draig.



From jdavis at empires.org  Tue Sep 24 05:41:30 2002
From: jdavis at empires.org (Jeff Davis)
Date: Tue, 24 Sep 2002 09:41:30 GMT
Subject: Downloading huge files via urllib
References: <mailman.1032822990.23973.python-list@python.org>
Message-ID: <erWj9.39469$V7.10455397@twister.socal.rr.com>

Hi,

It's certainly possible. urllib is meant as a high-level interface that 
hides those kind of details. Perhaps you could find a way to make it work, 
but I would suggest using the socket module and create a socket and use 
SSL. Then you can read an arbitrary number of bytes from the socket at a 
time. That way, you can be sure to never take in more than, say, 500 bytes 
at a time, preventing your MemoryError.

Note that there is a warning in the docs for socket.ssl():
"Warning: This does not do any certificate verification! "

I think that just means that it doesn't check with a certificate authority, 
so you'll have to do your own verification of the certificates involved. 
But read up if security is a concern of yours (which it is, otherwise you 
wouldn't be using SSL).

Regards,
        Jeff

VanL wrote:

> Hello,
> 
> For various reasons, I have to use https to download large (20+ MB) text
> files which I then parse.  I set up a basic function to do this using
> urllib:
> 
> response = urllib.urlretrieve(serverURL, 'run.log')
> 
> However, I then get a MemoryError.  Tracking down the source of the
> error, I see the offending function in httplib:
> 
>     def makefile(self, mode, bufsize=None):
>         """Return a readable file-like object with data from socket.
> 
>         This method offers only partial support for the makefile
>         interface of a real socket.  It only supports modes 'r' and
>         'rb' and the bufsize argument is ignored.
> 
>         The returned object contains *all* of the file data
>         """
> 
> I think the problem is that bufsize argument that is ignored.  Does
> anyone know if this is correct, and what I can do about it?  I would
> like to automate the process of downloading this file, but is it
> possible?
> 
> Thanks,
> 
> VanL



From mal at lemburg.com  Thu Sep 19 13:30:20 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 19 Sep 2002 19:30:20 +0200
Subject: ANN: eGenix.com mxODBC Python Database Interface Version 2.0.5
Message-ID: <3D8A09AC.5060203@lemburg.com>

________________________________________________________________________

ANNOUNCING:

     eGenix.com mxODBC Database Interface for Python 1.5.2 - 2.2
                            Version 2.0.5

      Full Source Python extension providing ODBC connectivity
         to Python applications on Windows, Unix abd Linux


________________________________________________________________________

WHAT IS IT ?:

The mxODBC Database Interface allows users to connect from Python
to just about any database on the market today, on Windows, Unix
and Linux -- using just one interface to program against for all
supported databases and platforms. This makes mxODBC the ideal
basis for writing cross-platform database programs and
utilities.

mxODBC is included in the eGenix.com mx COMMERCIAL Extension Package
for Python, the commercial part of the eGenix.com mx Extension Series,
a collection of professional quality software tools which enhance
Python's usability in many important areas such as ODBC database
connectivity, fast text processing, date/time processing and
web site programming. See http://www.egenix.com/ for details.

________________________________________________________________________

WHAT'S NEW ?

The 2.0.5 version introduces work-arounds for bugs in various
ODBC drivers to enhance the compatibility of mxODBC with
Unicode-aware ODBC drivers such as the latest MS Access and
MS SQL Server drivers.

________________________________________________________________________

EGENIX.COM MX COMMERCIAL PACKAGE OVERVIEW:

mxODBC - Generic ODBC 2.0-3.5 interface for Python

       mxODBC is an extension package that provides a Python Database
       API compliant interface to ODBC capable database drivers and
       managers.

       In addition to the capabilities provided through the standard DB
       API it also gives access to a rich set of catalog methods which
       allow you to scan the database for tables, procedures,
       etc. Furthermore, it uses the mxDateTime package for date/time
       value interfacing eliminating most of the problems these types
       normally introduce (other in/output formats are available too).

       mxODBC allows you to interface to more than one database from one
       process, making inter-database interfacing very flexible and
       reliable.

       The source version includes a varity of preconfigured setups for
       many commonly used databases such as MySQL, Oracle, Informix,
       Solid, SAP DB, Sybase ASA and ASE, DBMaker and many more. The
       precompiled versions for Windows and Linux include the interfaces
       to the standard ODBC manager on these platforms to allow for a
       more easily configurable setup.

       More details are available at:

            http://www.egenix.com/files/python/mxODBC.html

________________________________________________________________________

WHERE CAN I DOWNLOAD IT ?

The download archives and instructions for installing the package can
be found at:

http://www.egenix.com/files/python/eGenix-mx-Extensions.html#Packages

Note that in order to use the eGenix.com mx COMMERCIAL package you
will first need to install the eGenix.com mx BASE package which can
be downloaded from the same location.

________________________________________________________________________

WHERE CAN I BUY LICENSES ?

mxODBC is free for non-commercial use. Commercial users have to
buy licenses for continued use after a 30-day evaluation period.
Special licensing setups are available for commercial product
developers. Please see

http://www.egenix.com/files/python/eGenix-mx-Extensions.html#BuyLicenses

for details.

________________________________________________________________________

WHERE CAN I GET SUPPORT ?

Commercial quality support for these packages is available from
eGenix.com. Please see

http://www.egenix.com/files/python/eGenix-mx-Extensions.html#Support

for details about our support offerings.

________________________________________________________________________


Enjoy,
-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
_______________________________________________________________________
eGenix.com -- Makers of the Python mx Extensions: mxDateTime,mxODBC,...
Python Consulting:                               http://www.egenix.com/
Python Software:                    http://www.egenix.com/files/python/






From fperez528 at yahoo.com  Thu Sep 12 12:09:43 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 12 Sep 2002 10:09:43 -0600
Subject: Anyone got a logfile script? - Help a n00b
References: <mailman.1031821818.15125.python-list@python.org> <3D806151.9020707@Linux.ie> <mailman.1031824892.20563.python-list@python.org>
Message-ID: <alqe87$t56$1@peabody.colorado.edu>

Daniel Stuart Kelly wrote:

> Thanks for the quick answer P?draig.
> 
> The context grep will pull back all the instances of the variable.. The
> variable in this case is a unique user ID. The ID will be in the log file
> multiple times for different actions that the user performs which are being
> logged....
> 
> I want to be able to search for specific actions.. Ie connections,
> disconnections, config changes. and not for every action that the user has
> performed.
> 
> Will Grep do this for me?

$ grep ID file | grep action

Think of it as an 'AND' grep implemented with a pipe.

Cheers,
f


From marklists at mceahern.com  Fri Sep 13 11:43:21 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 13 Sep 2002 10:43:21 -0500
Subject: Restricting methods in derived classes
In-Reply-To: <JHEOKEOOLIGLDHCMAHMOEEIFDOAA.marklists@mceahern.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOKEIGDOAA.marklists@mceahern.com>

Huaiyu,

I feel compelled to gently ask, "What's the point of this rigamarole?"

Cheers,

// m



From dsavitsk at e-coli.net  Sat Sep 14 22:30:34 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Sun, 15 Sep 2002 02:30:34 GMT
Subject: representing a literal single slash
References: <dd5454fb.0209141828.1ad24260@posting.google.com>
Message-ID: <ehSg9.8464$yt3.4491656@newssrv26.news.prodigy.com>

"Jonathan Claggett" <hellen at claggetts.net> wrote in message
news:dd5454fb.0209141828.1ad24260 at posting.google.com...
> Here is an easy question (I'm hoping).
>
> I'm trying to process a string so that all doubled up backslashes (\\)
> are replaced with a single backslash (\).
>
> The obvious approach doesn't work since a single backslash can't end a
> string:
>
> str = str.replace('\\', '\') # broken

in the above you are replacing a single slash with, well, with nothing.  try
>>> str = str.replace('\\\\', '\\') # broken

-d




From eugene1977 at hotmail.com  Sun Sep 15 04:36:30 2002
From: eugene1977 at hotmail.com (eugene kim)
Date: Sun, 15 Sep 2002 08:36:30 +0000
Subject: python + java for web development?
Message-ID: <am0nea$mhd$1@newsreader.mailgate.org>

hi..
i'm a python beginner, 
used java for about an year

we will develop a website for a semester(i'm a student)
our group members are all familiar with java, but not python..
probably, our main language will be java..

but java is really painful sometime , like database access
and i'd like to use python when python is much easier 

since java passes objects , not just strings among servlets and jsps 
i'm wondering if it is possible to mix java and other language..

for example
if servlet processes database work, and passes the result as ArrayList
type and jsp takes ArrayList to display the results.

how python can interact with these servlet/jsp..?

thank you a lot


From eric.brunel at pragmadev.com  Tue Sep  3 03:57:16 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Tue, 3 Sep 2002 09:57:16 +0200
Subject: getch in python
References: <75ca4e69.0209022326.4cde72e8@posting.google.com>
Message-ID: <al1pii$cul$1@wanadoo.fr>

Venkat Venkataraju wrote:

> Hi All
> 
> Is there any function that behaves like getch() function in C?
> 
> Venkat

Seems like this is what you want:

import sys
c = sys.stdin.read(1)

On EOF, c is just ''.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From michael at stroeder.com  Wed Sep 18 12:00:49 2002
From: michael at stroeder.com (=?ISO-8859-1?Q?Michael_Str=F6der?=)
Date: Wed, 18 Sep 2002 18:00:49 +0200
Subject: Python-style license?
Message-ID: <jikm5-p5l.ln1@junker.stroeder.com>

HI!

In some Python modules there's only a note about a "Python-style 
license". Now I wonder what that means. Any clues?

Ciao, Michael.



From pedronis at bluewin.ch  Wed Sep  4 17:29:44 2002
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Wed, 4 Sep 2002 23:29:44 +0200
Subject: jython memory usage
References: <hVm5ydAY2hd9Eww$@jessikat.demon.co.uk><3d76366a$1_5@news.bluewin.ch> <rnLbuKAv5jd9Ewzz@jessikat.demon.co.uk><3d76494d$1_5@news.bluewin.ch> <YR13DVAUNld9EwXH@jessikat.demon.co.uk><3d765ce6_4@news.bluewin.ch> <zzKd7KAqhnd9EwxT@jessikat.fsnet.co.uk>
Message-ID: <3d767c4a$1_1@news.bluewin.ch>

Robin Becker <robin at jessikat.fsnet.co.uk> wrote in message
zzKd7KAqhnd9EwxT at jessikat.fsnet.co.uk...
> .....
> it, test_pdfbase_ttfonts.py doesn't look as though it's doing anything
> particularly odd, certainly I see no references to system/spawn in
> there.
> --

have you tried my little program with your JVM.

If my program work as for me, that means that your py plus jython trigger a
bug of the JVM, because anyway and in any case  the JVM ought to scream "out
of memory" well before eating a 1GB of memory.

regards.




From gerhard.haering at opus-gmbh.net  Thu Sep 19 04:20:21 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 19 Sep 2002 08:20:21 GMT
Subject: Code analyser + testing
References: <mailman.1032420325.9623.python-list@python.org>
Message-ID: <slrnaoj28t.28c.gerhard.haering@haering.opus-gmbh.net>

Julien Anguenot wrote:
> Hi !,
> 
> I'm looking for a code analyser for Python 2.2.

PyChecker!

> I'm student and I have to find a software (in Python of course) for setting
> tests. (unit testing, etc....)

Python has two integrated unit testing frameworks in the 'unittest' and
'doctest' modules. I personally use unittest.

-- Gerhard


From nospam at bigfoot.com  Fri Sep  6 19:23:31 2002
From: nospam at bigfoot.com (Gillou)
Date: Sat, 7 Sep 2002 01:23:31 +0200
Subject: old ODBC module
References: <f5ae9.2146$Qx1.35346@news1.telusplanet.net>
Message-ID: <albcmn$2ia2$1@norfair.nerim.net>

Google for "Mark Hammond win32 python extensions".
It's in the package.

--Gilles

mailto:gilles at pilotsystems.net
"ryan" <user at domain.invalid> a ?crit dans le message de news:
f5ae9.2146$Qx1.35346 at news1.telusplanet.net...
> Does anyone know where I can find the old free odbc module? Not the
> mxODBC module, but the other (apprently not very well supported) one.
> I've been looking for a while without success. Thanks!
>
> ryan.
>





From hwlgw at hotmail.com  Sun Sep  1 16:47:46 2002
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 1 Sep 2002 13:47:46 -0700
Subject: Dead Man Timer?
References: <fE6c9.145308$aA.31724@sccrnsc02> <pan.2002.08.26.17.48.00.759341@effectif.com>
Message-ID: <cb035744.0209011247.81b42dc@posting.google.com>

"Graham Ashton" <graham at effectif.com> wrote:
> > Need to detect a stalled process.
> > 
> There may be a better way, but you could try using signal.alarm() to set
> an alarm. Your process should be interrupted by a signal when it is
> received. If your write operation works you can cancel the alarm
> immediately.
> 
> This will probably work much better on *nix than on windows (for which
> there will be better alternatives, I'm sure).

I am interested in a Windows solution!  Anybody?

::

Who does not trust enough will not be trusted.
                -- Lao Tse


From dsavitsk at e-coli.net  Wed Sep 11 22:35:12 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Thu, 12 Sep 2002 02:35:12 GMT
Subject: ntpath bug or expected?
References: <xXRf9.7957$yt3.4120228@newssrv26.news.prodigy.com> <mailman.1031795954.4770.python-list@python.org>
Message-ID: <A3Tf9.7966$yt3.4129854@newssrv26.news.prodigy.com>

<jepler at unpythonic.net> wrote in message
news:mailman.1031795954.4770.python-list at python.org...
> On Thu, Sep 12, 2002 at 01:18:21AM +0000, dsavitsk wrote:
> > python 2.1.3.  this is pretty minor ... os.path.join() on win32 fails if
the
> > first path component is C: (or any other letter).  that is, if the first
> > component is a drive letter, it must contain either '\\' or '/' or else
the
> > path is wrong.  this makes it the only component which must have
slashes.
> >
> > example
> > >>> os.path.join('C:', 'path', 'to', 'file.txt')
> > C:path\\to\\file.txt
>
> Sure.  This is the file on drive C which is in the relative path
> "path\\to\\file.txt" from the current directory of drive C.
>

it seems to me, i guess, that being inconsistent here and in os.getcwd in a
consistent way is confusing.  if getcwd returned C: instead of C:\\ (as it
returnes C:\\path and not C:\\path\\) then
os.path.join(os.getcwd(), 'path', 'file.txt')
with my modification would do the same thing as it now does, but it would
have the added bonuses of returning a full file path and removing 2 bits of
necessary documentation from the language.

i'll go find something better to do now.

-doug


> Remember, DOS maintains a separate working directory for each drive.
>
> A first element of C:/ would make the path absolute, and a first element
of
> C: would make the path relative to the current directory of C.  I think
> that is why
>     >>> ntpath.join("C:", "/path", "to", "file.txt")
>     'C:/path\\to\\file.txt'
> is "correct".  This is the one I wonder about, though:
>     >>> ntpath.join("C:", "/a", "/b", "x.txt")
>     '/b\\x.txt'
>     >>> ntpath.join("C:/a", "/b", "x.txt")
>     '/b\\x.txt'
> if the drive letter is preserved in my first example, why isn't it in the
> last two?
>
> Jeff
>




From mark at mceahern.com  Thu Sep 26 20:08:48 2002
From: mark at mceahern.com (Mark McEahern)
Date: Thu, 26 Sep 2002 19:08:48 -0500
Subject: well-formed xml
Message-ID: <JHEOKEOOLIGLDHCMAHMOAEOPEBAA.mark@mceahern.com>

I'm obviously missing something because this seemingly innocent chunk of
xhtml:

  from xml.dom import minidom

  s = "<a href='http://google.com/search?hl=en&q=foobar'>search</a>"
  #                                             ^
  #                                             - seems to be the problem
  #
  # maybe it thinks I'm trying to reference the &q entity?

  doc = minidom.parseString(s)

Exception traceback follows.

Is there a way for me to tell it to ignore apparent entity references inside
attribute values?

// m

$ python junk.py
Traceback (most recent call last):
  File "junk.py", line 5, in ?
    doc = minidom.parseString(s)
  File "/usr/lib/python2.2/site-packages/_xmlplus/dom/minidom.py", line 965,
in
parseString
    return _doparse(pulldom.parseString, args, kwargs)
  File "/usr/lib/python2.2/site-packages/_xmlplus/dom/minidom.py", line 952,
in
_doparse
    toktype, rootNode = events.getEvent()
  File "/usr/lib/python2.2/site-packages/_xmlplus/dom/pulldom.py", line 256,
in
getEvent
    self.parser.feed(buf)
  File "/usr/lib/python2.2/site-packages/_xmlplus/sax/expatreader.py", line
148,
 in feed
    self._err_handler.fatalError(exc)
  File "/usr/lib/python2.2/site-packages/_xmlplus/sax/handler.py", line 38,
in f
atalError
    raise exception
xml.sax._exceptions.SAXParseException: <unknown>:1:41: not well-formed
(invalid
token)

This is with Python 2.2.1 without PyXML installed separately.  The same
thing happens with PyXML 0.8.1:

$ python junk.py
Traceback (most recent call last):
  File "junk.py", line 5, in ?
    doc = minidom.parseString(s)
  File "/usr/local/lib/python2.2/site-packages/_xmlplus/dom/minidom.py",
line 16
05, in parseString
    return expatbuilder.parseString(string)
  File
"/usr/local/lib/python2.2/site-packages/_xmlplus/dom/expatbuilder.py", li
ne 943, in parseString
    return builder.parseString(string)
  File
"/usr/local/lib/python2.2/site-packages/_xmlplus/dom/expatbuilder.py", li
ne 189, in parseString
    parser.Parse(string, True)
xml.parsers.expat.ExpatError: not well-formed (invalid token): line 1,
column 41

-




From aleax at aleax.it  Mon Sep 30 11:19:31 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 30 Sep 2002 15:19:31 GMT
Subject: Is there a way to find the name of the method currently being executed?
References: <mailman.1033009413.19897.python-list@python.org> <k3Ek9.170802$pX1.6081475@news2.tin.it> <mailman.1033140277.13452.python-list@python.org>
Message-ID: <7YZl9.191322$pX1.6844789@news2.tin.it>

Craeg K Strong wrote:

> Thanks to all who responded!
> 
> I should have known to check the python cookbook...
> 
> One question: the sys._getframe() method is new for Python 2.1--
> how likely do you think it is to change in 2.2+
> releases?   The part that worries me is that the
> method starts with an underscore.  That generally means
> "protected/private" or "unpublished" to me.

Yes, the name convention does indicate "internal".

> I just want to know how safe it is to use for
> general business application type programming.

Not very: it's intended to be used for debugging and
development purposes.  If you can cleanly express your
application's intent WITHOUT a function needing to know
its own name, it's by far preferable to do so.  If such
self-knowledge is indispensable to cleanly express your
application's intent, then sys._getframe is as usable as
any other method -- worst case, should it go away one
day, you'll have to install it in your own startup code
(code to emulate sys._getframe for old Python versions is
also available in the Cookbook, and most likely similar
code can be cooked up in the future, as long as there's
_any_ way to get such to-you-indispensable information).


Alex



From martin at v.loewis.de  Mon Sep 30 02:01:19 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 30 Sep 2002 08:01:19 +0200
Subject: sizeof PyTypeObject difference
References: <j5Rl9.29807$IO5.4027049@e3500-atl1.usenetserver.com>
Message-ID: <m34rc8qamo.fsf@mira.informatik.hu-berlin.de>

"Tom Widrick" <lordmacro at hotmail.com> writes:

> I'm trying to make a metatype in C, but there seems to be some difference
> between the size of the type object.
> 
> typedef struct {
>  PyTypeObject type;
> } MyMetaObject;
> 
> 
> and in the tp_basicsize field I place sizeof(MyMetaObject) the value is only
> 196. 

This is a type declararation. In C, you cannot put values into a type
declaration. So I assume you are creating a value (an object) of
struct MyMetaObject, and you put 196 into this value.

If so, you misunderstand the meaning of tp_basicsize: It does not
describe the size of the type object, but the size of instances of the
type.

> This causes a core dump.

That is hard to believe. Declaring the typedef alone can't cause a
core dump. Defining an object of that type, and putting the value of
196 into it can't cause a core dump, either. You must be doing other
things, like using the resulting object in some way, to cause a core
dump.

> What is the cause of this difference? 

The type type is really a variable-sized object. You are seeing the
size of struct etype, which is hidden in typeobject.c.

> And what is the "proper" way to do this?

You can't currently declare a type with additional C fields, since
type objects are variable-sized. One approach is to put additional
things into the type's __dict__. Another approach is to use Christian
Tismer's patches to overcome this limitation; I recommend that you
contact Christian.

Regards,
Martin


From aleax at aleax.it  Mon Sep 23 05:17:48 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 23 Sep 2002 09:17:48 GMT
Subject: Getting started
References: <uojinmcfapjf45@corp.supernews.com> <mailman.1032442884.18747.python-list@python.org> <H3mi9.133635$pX1.4751676@news2.tin.it> <mailman.1032567390.29374.python-list@python.org>
Message-ID: <0%Aj9.129252$ub2.2813083@news1.tin.it>

James J. Besemer wrote:

> 
> Alex Martelli wrote:
> 
>>$a = "2.3";
>>$b = 45;
>>print $a+$b;
>>
>>emitting 47.3 is pretty close to what I consider "weak typing".
>>
> 
> I don't see why you call this "weak typing".  In Perl, the "+" operator
> is defined to work with strings or numbers.  

It's defined to work *INDIFFERENTLY* on numbers, or on strings that
look like numbers, "ignoring" the types of the operands.  Since most
(all?) operations are designed to ignore the operands' types, this
makes those types much weaker than they would be in languages where
the operands' types are respected.


> It may seem more confusing

Not at all: to the typical newbie, it seems LESS confusing to be
allowed to ignore the differences between numbers and strings,
because the significance of the types is so weak.  Like so many
other aspects of Perl, weak typing is *intended* to be a handy
convenience for the user.  It turns out NOT to be, IMNSHO (or
else I'd still be using Perl, no?-), because it ends up as a
collection of trips and pitfalls.  E.g., if $b erroneously turns
out to be "hello world!", $a+$b silently and transparently masks
the error by treating $b just as if its value was 0.  By masking
all kinds of such errors, weak typing makes debugging large
programs a nightmare (in my experience with Perl -- extensive,
though old enough to be mostrly pre-Perl 5 -- and in my experience
with other weakly typed languages such as Basic dialects).


> at first if you expect "+" also to be the string concatenation operator,
> which it is not.  In the above example,
> 
>     print $a . $b;
> 
> will print the string "2.345".

String concatenation has little to do with it.  I have no
special problem against a "stringize and concatenate" operator.
I'm not a "strong-typing evangelist" -- I'm a pragmatist and
my experience tells me that a "stringize and whatever" operator
is not especially error-prone, no more than, say, the rough
Python equivalent "%s%s" % (a, b) would be.


> This all is no different in principle from Python's:
> 
>     a = 2.3
>     b = 45L
>     print a + b

Wrong!  It *is* VERY different in principle to have a hierarchy
of types such that "A is usable where B is required" (often
called A IS-A B: a special, type-safe case of polymorphism --
but such an important special case that many modern languages
HINGE on it, and almost all languages accept some version of
it, possibly limited to numerics), rather than having specific
operators that "treat both operands as A" vs "treat both
operands as B".  The former is NO weakening of typing, in
principle.  The latter is.

I'm somewhat surprised at seeing you make such assertions at
obvious variance with type theory -- I thought you had a solid
CS background (stronger than my own -- I'm just an engineer,
more interested in pragmatics than in theory).  We _do_ share
the fundamentals, I hope?  E.g., my favorite "essential Luca
Cardelli bibliography":

Cardelli, Wegner, "On understanding types, data abstraction, and 
polymorphism", ACM Computing Surveys 1985

Cardelli, "Type Systems", in "Handbook of Computer Science and 
Engineering", CRC Press, 1997.

Cardelli, "Typeful programming", in "Formal Description of Programming 
Concepts", Springer-Verlag, 1989

Abadi, Cardelli, Pierce, Plotkin. "Dynamic typing in a 
statically-typed language", ACM Transactions on PLS, April 1991

Abadi, Cardelli, Pierce, R?my, "Dynamic typing in polymorphic 
languages", Journal of Functional Programming, January 1995

Abadi, Cardelli, Viswanathan, "An interpretation of objects and object 
types", Proceedings POPL 1996


> and getting 47.3.  The "+" operator is defined for floats and also for
> longs.  Since they're all "numbers" it would seem really strange if they
> did not interact naturally like this.  But fundamentally they're

Indeed it would; as far as I know, the CAML dialect of ML (also in
the O'Caml extension, last I looked) is one of the very few
programming languages that doesn't have a polymorphic + operator --
you have to use + for integers only, +. for floating-point.  As I
recall some dialects of Forth had a similar lack of polymorphism
for the *opposite* reason -- typing so weak that once data was on
the stack the system had no idea any more about how it had to be
taken, FP or integer -- CAML (but not Standard ML, if I recall
correctly) ends up in the same place as a kind of reductio ad
absurdum of VERY strong vs weak typing ("the extremes meet":-).

"Promotion" rules are (or should be) simple: whenever (e.g.) a
float is required, an integer is accepted instead and gets
promoted to a float transparently and implicitly (ideally without
loss of information -- in practice, since longs are represented
by unbounded numbers of bits and floats aren't, the "without
loss of information" is only an ideal).  Operator / used to be
a notable exception -- fortunately, that notable exception is
at long last going away.

As everybody but CAML's designers seem to acknowledge, such
polymorphism in no way weakens typing (in a formal sense, as
per the above bibliog, this only holds under the constraint of
no-information-loss).  Perl's treatment of arithmetic operators
cannot be framed this way, nor made consistent with typing
theory: it's _weak_ typing, not just _dynamic_ typing.


> Anyway, why should "2" + 5 == "7" be all that much stranger than "2" * 5
> == "22222"?  If the operation is unambiguous and useful what rule would
> it break?

As I already said, newbies do appear to like the idea of confusing
numbers and strings -- in that sense, such weak typing may be taken
as the reverse of "much stranger"... it's what most beginners seem
to expect (take a tour of duty on the help hotline to get a very
good sample of beginners' misconceptions and problems...:-).  But
experience with Perl (and other weakly typed languages) shows that
such crucial weakening of typing, by hiding mistakes, is a serious
burden to bear in developing and maintaining large programs.

The rule broken is that the implicit-conversions directed graph
must be acyclic -- if type A can be implicitly converted to B,
then B cannot be implicitly converted to A.  It would (I suppose)
be possible to build a consistent (strong) typing system where
all strings can be implicitly converted to numbers, but only
if no numbers where ever implicitly converted to strings (so,
for example, the print statement as it stands in Python today
would have to disappear -- more relevantly, so would the %s
format specifier's ability to apply to any object).  Personally,
I think that would be absurd, but I don't think it would break
the rules (as long as information loss in the conversion could
somehow be contained -- which I think means the conversion rule
would have to be something like "treat as a base-256 number
with the char's ordinal values as digits"... forget "1.0"
being treated as the number 1...!-).

Incidentally, the DAG-rule also explains why it's a good idea
to have a[b] raise an exception if b is floating -- since ints
become floats implicitly, the reverse should not hold.  I'm
not claiming that Python is 100%-perfect is this (as it's not
in information-loss terms, see above), but I think it strikes
a roughly reasonable pragmatic compromise.

The int/long situation is interesting, and shows why the
unification of those two types (currently ongoing) IS most
likely a good idea -- we do want them to offer the same
"interface" to the rest of the program, with the only
diistinction being one of how they're implemented internally;
wanting them totally interchangeable means we should thus
have just one type (with two different implementations --
no problem with that).  I expressed different opinions in
the past, but framing things in terms of type theory helps
me change my mind:-).


  Personally, I think Python would be improved slightly by
> adding this conversion.  It's convenient and there's ample precedent in
> other languages (Awk, Snobol, Icon, Perl, etc.).  But I am sure that
> notion will provoke howls of protest from the True Believers.

The only reason I'm sniggering rather than howling is that I'm
confident that this (expletive deleted) proposal will never be
implemented in Python.


> There are a lot of reasons why Python is better than Perl but strongly
> vs. weakly typed is not one of them.  I don't see a material difference
> in the "strength," per se, of typing between Perl and Python.  They're

Then you are not looking at things in the right way.  Maybe you
do need to refresh your knowledge on type theory.  Python (and
Smalltalk) are much more suitable than Perl (awk, Snobol, etc)
to implement large programs, apart from all other differences,
exactly *because* typing is stronger in Python and Smalltalk
than in those other languages.

> Getting back to the original question, the quintessential example of
> "weak typing" is in old C or Fortran where you can define a function to
> take, say, a string argument but are allowed instead pass in a real or a

In the Fortran language (Fortran IV and Fortran 77 -- the versions
of Fortran of which I had to become a guru: I have no deep knowledge
of earlier or later versions) you are NOT allowed to pass a real to
a function that wants (e.g.) an integer (no strings in Fortran IV).
If you do so, your program's behavior becomes undefined.

No compilers that I know of actually *diagnosed* this programming
mistake (hmmm, perhaps Waterloo Fortran did -- that WOULD be quite
consistent with its didactic orientation -- but it's been too many
years since I last used it, so I don't recall), but that's quite
another issue compared to how the LANGUAGE is defined.  "Old C"
was never precisely specified, so that knowing what was in fact
an error leading to undefined behavior which a given compiler
could not diagnose is harder; taking lint as part and parcel of
the language, as I recall was advised by C's authors and/or main
practitioners, then again we can say that such mismatches were
NOT allowed.


> pointer to a struct and get garbled results or even corrupt the program:
> 
>     printf( "%s\n", 2.5 );        // possible program fault

Variadics are indeed a problem (gcc diagnoses this -- but wouldn't if you
didn't use a constant as the format).

> Another example is all too familiar to MFC programmers:
> 
>     CEdit* edit = (CEdit*)GetDialogItem( ID_FOO );
> 
> GetDialogItem() essentially returns a raw pointer.  The programmer is
> required to explicitly cast the pointer into the proper type, the same
> type as the dialog item identified by the numeric ID.  However, if the
> dialog item itself changes, say, from an Edit box to a List box then the
> cast becomes wrong, some operations on edit may fail catestrophically,
> and -- worst of all -- the compiler cannot detect and diagnose this
> error for you.
>   
> These types of errors are not possible in Python or Perl, which
> illustrates that they're both "strong".

Yes, this is a good example of the dangers of (old-style) casting.
Of course, you could (and probably should) just assign to a plain
CWnd* (no raw pointer involved: it's a pointer to the base class),
then check it's indeed a CEdit* before you proceed, but in practice
programmers DO perform unchecked downcasts and come to grief thereby.


> Curiously, in Python
> 
>     print "%s" % 2.5
> 
> prints the string "2.5" instead of, say, raising an exception.  So
> there's some small precedent for treating numbers as strings in a string
> "context" even in Python.

See above: in a sense, it's exactly BECAUSE we _want_ to have
operations that "transform into a string then ...", that theory
tells us the _reverse_ (making the types-DAG into a graph with
loops) is a disaster.  Incidentally, Java (strongly AND
statically typed) does that, for strings specifically, too
(using the toString method, like Python uses __str__).


> I'll go out on a limb and say that most "dynamic typed" languages are
> also "strong" by necessity.  If the user isn't keeping track of the
> types then the system HAS to.  An exception I can think of would be an
> ancient dialect of Lisp where integers also were machine addresses and
> thus callable without restriction.  E.g., if a function name evaluated
> to a number the interpreter performed a subroutine call to that address
> instead of, say, applying a lambda.  It was the mechanism through which
> system builtins were accessed.  However,  nothing prevented a user from
> "calling" an arbitrary integer and leaping off into never never land.

Yep, that's pretty weak typing indeed -- Forth-ish, I would say.

But you miss the essential distinction: what DRIVES typical ops,
the types of the objects involved, or the operators &c that the
programmer has coded?

In a strongly-typed language, types drive.  In a weakly-typed
language, operators (or the equivalent thereof) drive.  It's not
a black-and-white issue, but it's the key distinction.  In your
"weak Lisp" example, the "call this thing" operator (so to speak:
not _syntactically_ an operator, but that matters not) drives.

Recall your question up above...?

> If the operation is unambiguous and useful what rule would
> it break?

I think you've given a good example of why, even though the definition
WAS unambiguous (if it's an integer, unambiguously this happens)
AND useful (e.g. to call system builtins), this weak Lisp was
not optimally designed for practical use, particularly in the
development and maintenance of large programs.


>>find it very hard to produce substantial, large programs using
>>weakly-typed languages.
>>
> Given my definition of "weakly typed," I agree it's more difficult than
> in strongly typed languages.
> 
> On the other hand, isn't it the case that much of Python is implemented
> in C, which is weakly typed?  So "weakly typed" makes it harder but not
> impossible, at least not for some.

Why, sure.  HUGE programs have indeed been developed even in
assembly language, after all (they may be riddled with so many
bugs to put an ant-hill to shame, maybe, but they're still in
use:-).  C has been used (IS still in use today) to develop
more and larger programs than other languages.  That most of
them could have been developed and could be maintained with 1/10
the effort in Python is another issue.

Human ingenuity (and stubbornness) is strong enough that any
claim that some intellectual-work endeavour is "impossible" (to
a sufficient number of sufficiently-motivated human beings,
with unbounded amounts of cash to keep them going:-) had better
be backed by a very solid mathematical proof...:-).


>> Python is strongly, albeit dynamically, typed,
>>
> FWIW, so is Perl.

Nope: see all of the above.


Alex



From just at xs4all.nl  Thu Sep 12 05:09:27 2002
From: just at xs4all.nl (Just)
Date: Thu, 12 Sep 2002 11:09:27 +0200
Subject: iterators and generators, is this the Python Way???
References: <3D7FF921.2040305@one.net> <just-DB3861.09023412092002@news1.xs4all.nl> <3D804B79.9070003@one.net>
Message-ID: <just-CFB1B3.11092712092002@news1.xs4all.nl>

In article <3D804B79.9070003 at one.net>,
 Michael Schneider <michaels at one.net> wrote:

> Thank you very much, I like this approach.
> 
> I was under the mistaken impression what the
> iterator needed to do.

Hm, I didn't think so: __iter__ needs to return an iterator, and that's 
exactly what you did. It's just that I find it a really nice idiom to 
have __iter__ *be* a generator, it's really nice and compact.

Just


From fgeiger at datec.at  Thu Sep 19 01:55:08 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Thu, 19 Sep 2002 07:55:08 +0200
Subject: Need advice on choosing the technology for intranet based application 
References: <3d88fd1e@news.ivm.net>
Message-ID: <3d8966ab@news.swissonline.ch>

"Nenad Propadovic" <propadovic.nenad at debitel.net> schrieb im Newsbeitrag
news:3d88fd1e at news.ivm.net...

<snip/>
> So I kindly ask for knowledgeable opinions on the following question:
> -which web/application server? I'm considering Zope (seams rather BIG when
> you download & unpack it, and it seams all possible web applications have
> been written for it) and twisted (smaller, seams lower lewel and somewhat
> easier to learn), and I'm wondering if there is something else out there,
> which is both powerfull and flexible and easy to learn and complete. Oh
yes,

Probably you should consider Webware too. It was written by and for people,
who have been frustrated by Zope.

In Webware you get results pretty fast. There is an rel/obj mapper
(MiddleKit), which makes DB handling very easy.

Regards
Franz





From phr-n2002b at NOSPAMnightsong.com  Mon Sep 30 02:36:50 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Sep 2002 23:36:50 -0700
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <E4Bl9.5382$pD4.397@fe05> <an7bvd$burft$1@ID-125932.news.dfncis.de> <an7p7t$cpg$0@216.39.172.122> <mailman.1033338759.6775.python-list@python.org> <7xd6qwqtgz.fsf@ruckus.brouhaha.com> <mailman.1033352259.30842.python-list@python.org> <7xelbcgo56.fsf@ruckus.brouhaha.com> <mailman.1033357298.16341.python-list@python.org> <7x65woks0i.fsf@ruckus.brouhaha.com> <mailman.1033363178.31164.python-list@python.org>
Message-ID: <7xit0oj859.fsf@ruckus.brouhaha.com>

"Chris Gonnerman" <chris.gonnerman at newcenturycomputers.net> writes:
> > With decimal arithmetic, you get $2.97 the first way and
> > $3.00 the second way.
> 
> Excellent example, wrong focus.  The business user (in this
> case, the retail store owner/manager) would define which
> method, and which results, are right.  The second mode is
> the "canonical" way in the manual system, but lately I've
> realized that many (most?) retail POS systems can do it "both"
> ways, handling single items by a variation of method 1 which gives answers
> consistent with method 2.
> 
> I'm glad I didn't have to write and debug that code.

I'm really skeptical that any "business user" can specify a set of
rules that gets all this stuff straight for a complex business
involving cents-off coupons, percentage advertising discounts, etc.
It sounds much more complicated than what we usually call software
implementation, and it sounds like it has to be done differently for
every business situation.  They're almost certainly better off using
some method that gives consistent results at the end.

I worked on one business application where the vendor felt that any
floating point roundoff errors were unacceptable.  He had me implement
exact rational arithmetic (easy in Python using the gmpy package).
All intermediate results were represented as exact integer ratios, so
you were guaranteed that 3 * (1/3) == 1.  This was not the same as
decimal arithmetic.  You would never get $2.97 as an answer for the
banana problem.  But I don't think any customers complained.


From whisper at oz.net  Sat Sep 28 04:17:12 2002
From: whisper at oz.net (David LeBlanc)
Date: Sat, 28 Sep 2002 01:17:12 -0700
Subject: Dirt simple tkinter editor example
Message-ID: <GCEDKONBLEFPPADDJCOEGEOEFDAA.whisper@oz.net>

Can someone point me at one? Ideally, it would support a few text styles
(bold, italic...) and not much else. I'm looking for a working howto
example.

TIA,

David LeBlanc
Seattle, WA USA




From tdelaney at avaya.com  Wed Sep 25 19:50:41 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 26 Sep 2002 09:50:41 +1000
Subject: Decimal arithmatic, was Re: Python GUI app to impress the bos
	s?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A5D9@natasha.auslabs.avaya.com>

> From: danb_83 at yahoo.com [mailto:danb_83 at yahoo.com]
> 
> "Stuart D. Gathman" <stuart at bmsi.com> wrote in message 
> news:<amo0aq$5nf$1 at nntp2-cm.news.eni.net>...
> >...
> > There ought to be severe penalties for idiots that use 
> floating point
> > dollars for financial applications.  If forced to use floating point
> > (e.g. because customer demands BASIC), then keep money amounts in
> > whole pennies (or whatever the smallest currency unit for 
> the country
> > is), and divide by 100 (or whatever) for printing only (or 
> just add the
> > decimal point yourself).
> 
> That's exactly how things were done at my old workplace, except that
> cents were also used for user input.  Most of the time.

It's called Fixed Point arithmetic, and strangely enough, Tim Peters wrote
an excellent implementation ...

Tim Delaney



From ark at research.att.com  Mon Sep  2 10:40:23 2002
From: ark at research.att.com (Andrew Koenig)
Date: Mon, 2 Sep 2002 14:40:23 GMT
Subject: stylistic question -- optional return value
References: <20020828145828.19131.h014.c001.wm@mail.dougfort.net.criticalpath.net> <mailman.1030572688.22121.python-list@python.org> <akkost$cgt$1@newsreader2.netcologne.de> <roy-C1B7F3.09024729082002@reader1.panix.com> <yu99vg5t946h.fsf@europa.research.att.com> <akur66$rp6$0@216.39.172.122>
Message-ID: <yu99vg5oo3oo.fsf@europa.research.att.com>

Bengt> If added information is the exception (;-) rather than the
Bengt> rule, you might want to consider using an exception to return
Bengt> the unusual bag of data.

Unfortunately, it's not.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From loewis at informatik.hu-berlin.de  Sun Sep 15 15:27:04 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 15 Sep 2002 21:27:04 +0200
Subject: Will xml.dom work with HTML docs?
References: <k3Tg9.6083$Lg2.908752@news2.news.adelphia.net> <877khntzvt.fsf@pokey.henrik-motakef.de> <j4y9a3qg3j.fsf@informatik.hu-berlin.de> <871y7vnm9q.fsf@pokey.henrik-motakef.de>
Message-ID: <j41y7vqcjb.fsf@informatik.hu-berlin.de>

Henrik Motakef <henrik.motakef at web.de> writes:

> Does it do so by using an actual DTD, or is HTML-specific knowledge
> hardcoded in it?
> 
> I.e. could this feature be used with other document types?

It's hard-coded (for HTML 4). The support code is generated (as a set
of dictionaries), so you could probably port it to another DTD; I
believe there are constraints to the DTD (i.e. you might need to losen
the content model).

Regards,
Martin


From brueckd at tbye.com  Mon Sep  9 11:02:57 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Mon, 9 Sep 2002 08:02:57 -0700 (PDT)
Subject: Larry Wall's comment on python...
In-Reply-To: <bd751532.0209082142.54236307@posting.google.com>
Message-ID: <Pine.LNX.4.44.0209090749250.10931-100000@ns0.tbye.com>

On 8 Sep 2002, Brian Slesinsky wrote:

> He's repeating the standard newbie complaint.  However, I think the
> fact that so many people have repeated this for so long is
> significant.  Long-term Python programmers tend to forget how weird it
> is to some newcomers to end a function or class just by *stopping*. 
> It feels abrupt and rather fragile, particularly when you don't have a
> firm grasp of the language.  If newbies have complained for this long,
> the complaint's not going away, and if we're serious about making
> Python easy to pick up it might be good to listen.

Yes, the fact that so many newbies complain about it is very significant.  
Perhaps we need to do a better job of presenting the concept. For example,
instead of tossing it out there and then defending against complaints
perhaps it would be better to take the offensive: first make a case for
why the "normal" way to delimit blocks has its drawbacks, and then present
the Python way as a solution to a problem.

But just because people consistently complain about something does not
mean the answer is to bow to their wishes. From an arrogant point of view,
they're newbies, so what do they know? ;-) Seriously though, I enjoy
Python because it _is_ different - and I usually derive benefit from the
ways in which it is different (i.e. - it's not different for arbitrary or
random reasons). It's significant that lots of newbies complain about "no
curly braces", but it's as least as significant that once people use it
for awhile they tend to end up liking it.

> Of course it's unnecessary and should be entirely optional.  Putting
> an "end" after every block would certainly be overly verbose and bad
> style.  But ending major blocks in a way that's syntax-checkable might
> catch the occasional indentation error and/or make error messages

Well, as you know you can always do an "# end" comment which really gives 
much of the same benefit. At the bottom of my Emacs window it tells me 
which block I'm closing (e.g. "Closes block: ...try:"), other editors 
probably do the same. Also, it seems like whenever it's not readily 
apparent which block I'm in, then the _real_ problem is sloppy code on my 
part.

-Dave




From christophe.grimault at sacet.com  Tue Sep  3 06:40:07 2002
From: christophe.grimault at sacet.com (Christophe Grimault)
Date: Tue, 03 Sep 2002 10:40:07 +0000
Subject: Pb with libglade and gtk
Message-ID: <3D749187.F2F1A61E@sacet.com>

Hi all,

I 've a UI created with glade, and when I try to run it [python2 on
redhat7.3] I get :

Traceback (most recent call last):
  File "./run.py", line 3, in ?
    import libglade
  File "/usr/lib/python1.5/site-packages/libglade.py", line 11, in ?
    class GladeXML(_gtk.GtkData):
AttributeError: 'module' object has no attribute 'GtkData'


A fix for this problem ?

Thanks in advance

CG
-------------- next part --------------
A non-text attachment was scrubbed...
Name: christophe.grimault.vcf
Type: text/x-vcard
Size: 271 bytes
Desc: Carte pour Christophe Grimault
URL: <http://mail.python.org/pipermail/python-list/attachments/20020903/f15ce530/attachment.vcf>

From ark at research.att.com  Tue Sep  3 09:20:50 2002
From: ark at research.att.com (Andrew Koenig)
Date: Tue, 3 Sep 2002 13:20:50 GMT
Subject: stylistic question -- optional return value
References: <mailman.1030608587.4698.python-list@python.org> <30292041.0209012344.ab2d4@posting.google.com> <yu99r8gco3m2.fsf@europa.research.att.com> <al0snu$qf7$0@216.39.172.122>
Message-ID: <yu994rd76wgd.fsf@europa.research.att.com>

Bengt> What about defining a class to represent the "data structure,"
Bengt> and passing a reference to an instance of that to your various
Bengt> routines, so that they can populate the structure themselves
Bengt> (via methods/attributes/properties as convenient) instead of
Bengt> returning data for a caller to do the populating with?

That's certainly a possibility, but it's a little heavyweight.

Of the various alternatives I've heard so far, the one I like best is simply
using None to represent the optional information.  That is, I return either
(x, y) or (x, None) depending on whether y exists.

Bengt> BTW, you mentioned recursion in another part of the thread. Is
Bengt> that as in recursive descent? An exception is an interesting
Bengt> alternative way to return a first match, with no need to return
Bengt> anything from unsuccessful visits.

Yes, it is a species of recursive descent.  What I'm doing, though,
is returning generators at each stage.  That way I cover all of the
possibilities.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From peter at engcorp.com  Mon Sep  2 18:52:01 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 02 Sep 2002 18:52:01 -0400
Subject: Why it is so dramatical?
References: <200209021107.25761.b.maryniuk@forbis.lt> <mailman.1030961043.19190.python-list@python.org> <3D73E035.2040300@nospam.attbi.com>
Message-ID: <3d73eb74@news.sentex.net>

djw wrote:
> holger krekel wrote:
> 
>> If you have something like this in a loop it's often
>> *much* better to work with a list and later do a
>>     "".join(stringlist)
>>
>> As you observed yourself this is around 100-1000 times faster. 
>> And consider: if strings were not immutable but modified in-place
>> (like lists) you couldn't use them as dictionary keys.
>> if-it-hurts-don't-do-it-ly, yours Holger
>>
> 
> Stupid question:

It's only a stupid question if you didn't expect this relatively common
answer to such questions:  "feel free to contribute a patch"... :-)

> If ""join() is 100-1000 X faster, why doesn't the Python interperter 
> translate s = s1 + s2 + ... into s = "".join( [ s1, s2, ... ] ) 
> automatically?

In cases like this, you need to know that Python is so dynamic that
the compiler cannot know that the items being added are strings until
runtime.  The "".join() convention does not work well if the items in
the sequence are not already strings.

If you feel compelled to suggest that the compiler could at least do
that for constant strings, see my first point above. :-)

-Peter



From bokr at oz.net  Tue Sep 17 14:56:50 2002
From: bokr at oz.net (Bengt Richter)
Date: 17 Sep 2002 18:56:50 GMT
Subject: raw post data
References: <Wt8h9.9583$Ee4.22928@news-server.bigpond.net.au> <w_fh9.113093$pX1.4054147@news2.tin.it> <gOxh9.13698$Ee4.40115@news-server.bigpond.net.au> <siAh9.118475$pX1.4269024@news2.tin.it>
Message-ID: <am7tti$91b$0@216.39.172.122>

On Tue, 17 Sep 2002 06:52:08 GMT, Alex Martelli <aleax at aleax.it> wrote:

>:B nerdy wrote:
>
>> how could i make it so when i read the stdin it still is there.. cause i
>> use cgi.FormVariables() also
>
>When you instantiate class cgi.FieldStorage (there's no such
>thing as cgi.FormVariables, so I'm just guessing at what it IS
>that you want!) you can pass it an optional first parameter
>that is a file object open for reading -- it uses sys.stdin
>as the default if you don't pass the first parameter.
>
>Thus, the following should work (untested!):
>
>import cgi, sys, cStringIO
>
>copyInput = cStringIO.StringIO(sys.stdin.read())
>fieldStorage = cgi.FieldStorage(copyInput)
>
>
>The raw posted data are now available as the string returned
>by calling copyInput.getvalue(), and are also parsed by cgi to
>obtain fieldStorage.
>
A nit, but it might be good to check that the environment
variable REQUEST_METHOD exists and == 'POST',
and to attempt to read not more than the most you expect to be able to handle,
and to check that what you get with your max read is what it's supposed to be,
e.g., (untested!) (you specify MAX_I_HANDLE) (inside appropriate try/except)

    import os
    ...
    if os.environ.get('REQUEST_METHOD') is None or os.environ.get('REQUEST_METHOD') != 'POST':
        raise ValueError, 'Request method not POST'
    if os.environ.get('CONTENT_LENGTH') is None: raise ValueError, 'Content length unspecified'
    content_length = int(os.environ.get('CONTENT_LENGTH'))
    if content_length == 0: raise ValueError, 'Content length zero'
    elif content_length > MAX_I_HANDLE: raise valueError, 'Content length too great: %s' % content_length 
    copyInput = cStringIO.StringIO(sys.stdin.read(MAX_I_HANDLE))
    if len(copyInput) != content_length:
        raise IOError, 'Data bytes read (%s) not == CONTENT_LENGTH (%s)' % (len(copyInput), content_length)
    ...

Regards,
Bengt Richter


From rjones at ekit-inc.com  Sat Sep  7 01:56:05 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Sat, 7 Sep 2002 15:56:05 +1000
Subject: Sequence or Identity in gadfly
In-Reply-To: <albrbp$rej$1@main.gmane.org>
References: <albrbp$rej$1@main.gmane.org>
Message-ID: <200209071556.05327.rjones@ekit-inc.com>

On Sat, 7 Sep 2002 1:21 pm, Jeff Kowalczyk wrote:
> I don't expect that it does (it would make more sense to do it in your
> python code), but is there a Sequence type in gadfly so that I can make an
> autonumber for ID columns similar this postgres code?:

There's no sequence support built in - you'll have to write a little bit of 
Python code to generate the sequence an hook it into gadfly. This was asked a 
while ago, and Anthony Baxter replied:

http://groups.google.com/groups?selm=mailman.1030335699.8153.python-list%40python.org


    Richard




From borcis at removethis.infomaniak.andthis.ch  Fri Sep 20 11:50:52 2002
From: borcis at removethis.infomaniak.andthis.ch (Boris)
Date: Fri, 20 Sep 2002 17:50:52 +0200
Subject: Python expression (Re: Iraq's offer to readmit arms inspectors)
References: <d8f8142b.0209161913.12cce045@posting.google.com> <vWBh9.26065$Gp1.687517@weber.videotron.net> <3D86FF01.FE0CECC5@geneva-link.ch> <3d87069c.34652637@news.rcn.com> <3D871316.AA1E8FFC@geneva-link.ch> <3d882cd4.9181792@news.rcn.com> <3f74f974.0209180713.27889f74@posting.google.com> <l9ehouci1ga88ub2csdam1qroiik0ogd4r@4ax.com> <3d88c258$0$707$5402220f@news.sunrise.ch> <v4lhouolv1c0ijmihfom94j0lor2lhntli@4ax.com> <3d89be92$0$707$5402220f@news.sunrise.ch> <s6mjous6o5abg3icq12u156v5tijhcnk7h@4ax.com> <3d89ee85$0$712$5402220f@news.sunrise.ch> <B8pi9.149307$5r1.6007886@bin5.nnrp.aus1.giganews.com>
Message-ID: <3d8b4498$0$717$5402220f@news.sunrise.ch>

Terry Reedy <tjreedy at udel.edu> a ?crit dans le message :
B8pi9.149307$5r1.6007886 at bin5.nnrp.aus1.giganews.com...

> Please remove (as I did) comp.lang.python from the list of newsgroups
> for this thread.  It is wildly off-topic here

comp.lang.python was added because the post contained a python expression

filter(lambda W : W not in 'ILLITERATE','BULLSHIT')

with the miraculous property of being runnable and producing the name
of the US president, Mr Bush, when executed; together with being readable
as a metaphor of human behavior illustrated by "Malev" - the person
that the post was responding to. Such dual interpretation of executable
code is witness that Python is closer to Prolog than you would think,
an interesting fact by itself for the python community.

> and should never have been added.

That's a dogmatic opinion that I believe lacking in justification.

if-Iraq-ends-up-Waco-like-what-will-mimic-Oklahoma-City-ly yours,

Boris Borcic
--
assert BOA not in BAOBAB, "Paradise Mismatch Error"




From newt_e at blueyonder.co.uk  Thu Sep 26 17:20:31 2002
From: newt_e at blueyonder.co.uk (Newt)
Date: Thu, 26 Sep 2002 21:20:31 GMT
Subject: toplevel
Message-ID: <zSKk9.10236$uw6.97402053@news-text.cableinet.net>

Hi, silly question time!

In TCL, I would use

set win ".win_goo_skills"
toplevel $win

to create a new window.

How do I do the same using python/Tkinter? The toplevel command doesn't seem
to accept this.

Thanks,

Newt






From bokr at oz.net  Sat Sep  7 15:27:07 2002
From: bokr at oz.net (Bengt Richter)
Date: 7 Sep 2002 19:27:07 GMT
Subject: None or 0
References: <k7ip4-0p3.ln1@junker.stroeder.com>
Message-ID: <aldjub$a8r$0@216.39.172.122>

On Sat, 07 Sep 2002 17:23:00 +0200, =?ISO-8859-1?Q?Michael_Str=F6der?= <michael at stroeder.com> wrote:

>HI!
>
>I'm using the idiom
>
>a = a or b
>
>to get rid of statements like
>
>if a is None:
>   a = b
>
>Now I wonder what happens if both a and b have zero-length values.
>
>Trying in Python 2.2.1 reveals:
>
> >>> repr(0 or None)
>'None'
> >>> repr(None or 0)
>'0'
> >>> repr('' or None)
>'None'
> >>> repr(None or '')
>"''"
> >>>
>
>Is it guaranteed to work like this or should that be avoided?
>

If you want a statement that sets a to a if a is not None and otherwise b,

    a = a or b

is not good, as you can infer from what you show above.

    a = (a, b)[a is None]

should do it, but

    if a is None: a = b

is better as a statement. If you need it in expression context, you could
use the

    (a, b)[a is None]	# alternatively, [a, b][a is None]

or maybe

    def myAorB(a, b):
        if a is None: return b
        return a
    ...
    myAorB(a, b)

For the following combinations, a or b obviously (when they're
distinguishable ;-) always produces b, and (a,b)[a is None] always
produces a except when a is None, in which case it produces b.

 >>> for a in [None, 0, '', [], (), {}]:
 ...     for b in [None, 0, '', [], (), {}]:
 ...         print 'a=%-5s b=%-8s a or b => %-8s  (a,b)[a is None] => %s' % (
 ...             `a`, `b`, `a or b`, `(a, b)[a is None]`
 ... )
 ...
 a=None  b=None     a or b => None      (a,b)[a is None] => None
 a=None  b=0        a or b => 0         (a,b)[a is None] => 0
 a=None  b=''       a or b => ''        (a,b)[a is None] => ''
 a=None  b=[]       a or b => []        (a,b)[a is None] => []
 a=None  b=()       a or b => ()        (a,b)[a is None] => ()
 a=None  b={}       a or b => {}        (a,b)[a is None] => {}
 a=0     b=None     a or b => None      (a,b)[a is None] => 0
 a=0     b=0        a or b => 0         (a,b)[a is None] => 0
 a=0     b=''       a or b => ''        (a,b)[a is None] => 0
 a=0     b=[]       a or b => []        (a,b)[a is None] => 0
 a=0     b=()       a or b => ()        (a,b)[a is None] => 0
 a=0     b={}       a or b => {}        (a,b)[a is None] => 0
 a=''    b=None     a or b => None      (a,b)[a is None] => ''
 a=''    b=0        a or b => 0         (a,b)[a is None] => ''
 a=''    b=''       a or b => ''        (a,b)[a is None] => ''
 a=''    b=[]       a or b => []        (a,b)[a is None] => ''
 a=''    b=()       a or b => ()        (a,b)[a is None] => ''
 a=''    b={}       a or b => {}        (a,b)[a is None] => ''
 a=[]    b=None     a or b => None      (a,b)[a is None] => []
 a=[]    b=0        a or b => 0         (a,b)[a is None] => []
 a=[]    b=''       a or b => ''        (a,b)[a is None] => []
 a=[]    b=[]       a or b => []        (a,b)[a is None] => []
 a=[]    b=()       a or b => ()        (a,b)[a is None] => []
 a=[]    b={}       a or b => {}        (a,b)[a is None] => []
 a=()    b=None     a or b => None      (a,b)[a is None] => ()
 a=()    b=0        a or b => 0         (a,b)[a is None] => ()
 a=()    b=''       a or b => ''        (a,b)[a is None] => ()
 a=()    b=[]       a or b => []        (a,b)[a is None] => ()
 a=()    b=()       a or b => ()        (a,b)[a is None] => ()
 a=()    b={}       a or b => {}        (a,b)[a is None] => ()
 a={}    b=None     a or b => None      (a,b)[a is None] => {}
 a={}    b=0        a or b => 0         (a,b)[a is None] => {}
 a={}    b=''       a or b => ''        (a,b)[a is None] => {}
 a={}    b=[]       a or b => []        (a,b)[a is None] => {}
 a={}    b=()       a or b => ()        (a,b)[a is None] => {}
 a={}    b={}       a or b => {}        (a,b)[a is None] => {}

Regards,
Bengt Richter


From loewis at informatik.hu-berlin.de  Tue Sep 17 12:59:34 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 17 Sep 2002 18:59:34 +0200
Subject: Detecting IPv6 support
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032245506.8580.python-list@python.org> <j4n0qhjbkf.fsf@informatik.hu-berlin.de> <pvsn08lw9k.fsf_-_@agni.research.nokia.com>
Message-ID: <j43cs8ft6x.fsf@informatik.hu-berlin.de>

Pekka Pessi <Pekka.Pessi at nokia.com> writes:

> 	Speaking of IPv6: is the IPv6 enabled on the Pythonlabs Windows
> 	distribution by default (sorry, I don't have a Windows box in my
> 	use at the moment)? Does Python support the older IPv6 stacks
> 	for NT4 and W2K?

No. It requires routines that are not available in the "stock"
winsock2.dll. When Pythonlabs Python 2.2 was released, it also
required an installation of the preview IPv6 release, which could not
be redistributed freely.

At the moment, you need CVS Python, and VC.NET, to get IPv6 suport on
Windows. This then works both with the released XP, and the MS
research IPv6 stack.

> 	What is the recommended method of finding out if Python and
> 	the underlying OS support IPv6? I can, for instance, create an
> 	IPv6 socket with a Python version that does *not* support
> 	IPv6.

What system is this? If you can create an IPv6 socket, why does it not
support IPv6 (i.e. how do you know it does not support IPv6)? If you
do .listen(), .getsockname() on that socket, what does it tell you?

Regards,
Martin


From shalehperry at attbi.com  Thu Sep  5 15:44:17 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 5 Sep 2002 12:44:17 -0700
Subject: Psycos and Pyros don't seem to play well together
In-Reply-To: <72F73B808A78D511A73600034771D9FF717EB7@dc_exchange2.howrey.com>
References: <72F73B808A78D511A73600034771D9FF717EB7@dc_exchange2.howrey.com>
Message-ID: <200209051244.17210.shalehperry@attbi.com>

On Thursday 05 September 2002 11:22, PoulsenL at capanalysis.com wrote:
> I am working on a distributed application using Pyro that I thought would
> benefit from Psyco.  All is well until I add "from psyco.classes import *"
> and then it gives me some Pyro based error messages (be happy to send them
> along if requested) when I launch the client.  I am using the "quickstart"
> method to get my objects up and running on Pyro quickly.  Anyone try to use
> these two modules together before?  What were your results?
>
> Thanks,
>
> Loren

and this is why 'from foo import *' is a bad idea.  You should just do 'import 
psyco' because you are probably seeing a name conflict.



From aleax at aleax.it  Wed Sep 25 10:51:21 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 25 Sep 2002 14:51:21 GMT
Subject: problem with dictionaries within lists
References: <pan.2002.09.25.05.21.07.477231.167@ihatespam.shaw.ca> <S4fk9.163219$pX1.5840444@news2.tin.it> <pan.2002.09.25.14.28.41.435139.181@ihatespam.shaw.ca>
Message-ID: <J3kk9.142193$ub2.3103834@news1.tin.it>

Grant Hallam wrote:

> The problem with blanking out the dictionary with dict{} is that it wipes
> the keys out.  As the loop goes through it assigns the matching second
> half of the list to a matching key.  If you insert a print statement on
> the dict you can see the values in the dict are changing.  Thus while it
> is the same dict the key values within the dict are different.  Thats why
> I'm somewhat confused as to why its the same dic value.  As for the tabs,

It's the same dict object because you didn't copy it.  You just appended
many references to it to the list.  So, just use:

>> webdatabaselist.append(dict.copy())

appending *copies* of the dict to the list, and you should be OK.

BTW, avoid naming your own variables like builtins -- list, dict,
file, int, str, &c -- you'll come to grief eventually when you do
that and then some other piece of your code tries to use the
builtin in the normal way.  It's always possible to find a more
significant name, anyway, so avoiding the types' names is no burden.
(This has nothing to do with your specific problem -- it's just
general advice).


Alex



From timr at probo.com  Wed Sep 18 02:22:47 2002
From: timr at probo.com (Tim Roberts)
Date: Tue, 17 Sep 2002 23:22:47 -0700
Subject: hexacodes in python?
References: <mailman.1032292881.7684.python-list@python.org>
Message-ID: <9n6gou8haak3o8njn8h38g55s6bn5gdapc@4ax.com>

jubafre at brturbo.com wrote:

>what is the hexacode for '?' in python?
>where can i take the hexacodes?	

This is somewhat operating-system dependent, but it is not really
Python-dependent.  In Windows, "lower case C with cedilla" is hex E7,
decimal 231.  In Unicode, it is U+00E7.

I used the Windows 'charmap' application to look this up.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From j_r_ashley at MAKGifts.com  Wed Sep  4 13:45:57 2002
From: j_r_ashley at MAKGifts.com (James Ashley)
Date: Wed, 04 Sep 2002 17:45:57 GMT
Subject: maximum number of threads?
References: <3d75b761$0$26216$afc38c87@news.optusnet.com.au>
Message-ID: <slrnance4n.1s0.j_r_ashley@minerva.jimrthy>

In article <3d75b761$0$26216$afc38c87 at news.optusnet.com.au>, Steven wrote:
> Basically its a voice server, or rather a voice to text server, so
> information produced might be sporadic, or fairly continuous. 
Up until about a year ago, I worked for a company which did 
"speech recognition applications."  I shuddered when I saw this--my
initial assumption was that you're trying to do both "dictation-style"
and "small, formal, grammar-based" ASR at the same time, or some such.
It scared me <G>.

But, then, to actually be on topic:

> Is there a maximum number of threads, and any opinions on whether the GIL
> and context switching in the interpreter is liable to make it unworkable (or
> a case of 'build and see')?

I think it really depends on your expected server load.  If you're going
to start out with just a few clients, it might be best to just start out
with python threads.  I think you might be better served starting out
using jython, though.  That depends, of course, on a lot of other
factors (like any issues getting JNI going with the actual speech
recognition engine...I'm guessing you're doing some embedding).

When I left that company, we were moving into a situation with hundreds
of clients using "one server."  At that point, we had to create a
distributed network cluster and use that for the backend.  The speech
recognition part wasn't much of a problem...it had been designed with
that in mind from the ground up.  Ironically enough, our biggest problem
(or, rather, the piece that took the most work) was the database
backend (it turns out that Access could only handle about a dozen
clients).

And that's far more than enough.  I presume one of the threading gurus
around here is much more qualified than I to give you actual advice
about threading.

Regards,
James


From robin at jessikat.fsnet.co.uk  Wed Sep 18 14:56:06 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 18 Sep 2002 19:56:06 +0100
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com>
 <Xns928CED692541cliechtigmxnet@62.2.16.82>
 <C05DA40F31B09329.0165201CAF094801.13AF9332EB64347A@lp.airnews.net>
 <s4Qh9.479726$q53.16320265@twister.austin.rr.com>
 <vATh9.666$oY5.33977741@newssvr14.news.prodigy.com>
Message-ID: <oY+rxnAGxMi9Ew5+@jessikat.demon.co.uk>

In article <vATh9.666$oY5.33977741 at newssvr14.news.prodigy.com>, Dennis
Reinhardt <DennisR at dair.com> writes
.......
>
>Would he be impressed by a user interface that powers 2.1 BILLION web
>documents? (ref: alltheweb.com)
>
>I am about 2 weeks away from releasing HTUIL, an integrated client for
>Windows, hosting Python, Perl, and/or native PLD2 displaying in a standard
>web browser.  See www.dair.com/hellopy.exe for a "hello" program written in
>Python and distributed as a standalone EXE on Windows.
>
>--
>
>Dennis Reinhardt

Well hoping Dennis is genuine and has better control over virii than my
outdated anti-virals I tried running this 360k hello world.

Now I'm confused, I have K-Meleon 0.6 installed as my standard, but
hellopy.exe tries to use Netscape 4.78 (also installed on my machine)
and fails to display anything properly (I see the source in the browser
window).

Probably one of those pesky registry settings over which only Billy
Goats has any control.
-- 
Robin Becker


From oren-py-l at hishome.net  Sun Sep 22 02:47:00 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Sun, 22 Sep 2002 02:47:00 -0400
Subject: "time" module under AIX?
In-Reply-To: <yibelbm7fkj.fsf@coxeter.math.toronto.edu>
References: <yibhegi7l40.fsf@coxeter.math.toronto.edu> <yibelbm7fkj.fsf@coxeter.math.toronto.edu>
Message-ID: <20020922064700.GA4436@hishome.net>

On Sun, Sep 22, 2002 at 01:34:36AM -0400, marco wrote:
> marco <marco at reimeika.ca> writes:
> 
> > I've just compiled python 2.2.1 under AIX and all seemed
> > to go well. However, I'm having some trouble importing
> > a few modules such as "time" and "math" e.g.
> 
> OK, it seems it was a permission problem. I compile/install
> stuff as a non-privileged user and then change ownership.
> However, it turns out that "make install" doesn't use the
> "install" command consistently and sometimes just copies
> stuff (without setting any permissions).

Please submit a bug report at http://sourceforge.net/bugs/?group_id=5470

If you also submit a patch it will be great!

	Oren




From robin at jessikat.fsnet.co.uk  Wed Sep  4 14:39:16 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 4 Sep 2002 19:39:16 +0100
Subject: jython memory usage
References: <hVm5ydAY2hd9Eww$@jessikat.demon.co.uk>
 <3d76366a$1_5@news.bluewin.ch> <rnLbuKAv5jd9Ewzz@jessikat.demon.co.uk>
 <3d76494d$1_5@news.bluewin.ch>
Message-ID: <YR13DVAUNld9EwXH@jessikat.demon.co.uk>

In article <3d76494d$1_5 at news.bluewin.ch>, Samuele Pedroni
<pedronis at bluewin.ch> writes
.....
>> >
>> >
>> well actually we're running with the default jython.bat which doesn't
>> seem to specify a memory, 1Gig is probably the default or maybe I ran
>> out of page space. The python process never gets over 90Mb grrrhhhh.
>
>then maybe I'm puzzled,
>
>have you checked process memory usage or is just that you get
>an "out of memory" and are estimating based on your total memory.
>
>Because the normal jython.bat uses the JVM default which is
>to scream "out of memory" after 64Mb of object memory has
>been allocated, see -Xmx here
>
>http://java.sun.com/j2se/1.4/docs/tooldocs/windows/java.html#options
>
>regards.
......
we're watching it on the task monitor/performance. No other process
running, this particular thing just eats up the world.
-- 
Robin Becker


From djc at object-craft.com.au  Fri Sep 20 22:57:07 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 21 Sep 2002 12:57:07 +1000
Subject: Python Sybase slow compared to Perl?
References: <Xns928FA7E6BA09Dcmay4yahoocom@165.112.130.8>
Message-ID: <m3hegknj7g.fsf@ferret.object-craft.com.au>

> I have always written my database scripts in Perl.  I am very
> interested in Python, and would like to start using it for all my
> scripting.  I am not trying to start a flame war, I'd just like to
> understand why a particular script I wrote runs so much faster under
> Perl.  The script in question takes roughly 3 times as long to run
> under Python.  It is very possible that there is a faster way to do
> this under Python.

[snip]

> Here is the Perl code snippet:

[snip]

>    while ( @data = $sth->fetchrow_array() ) {

[snip]

> and here is the Python code snippet:

[snip]

>    while 1:
>        row = c.fetchone()

snip]

> No matter what table how many records there are, the Python script
> always takes longer to run (usually by a facter of 3 or more).

I suspect that some of the difference is due to the fact that you are
fetching an array of rows using Perl but only a single row at a time
in Python.

Having said that, the current implementation of the Python module does
not do array binding for cursors.  The Sybase.py module is built on
top of an extension module called sybasect which does support array
binding.

To find out if that is the problem (until I get around to fixing the
cursor to do array binding) you could try doing this:

   # output the actual data
   num_recs = 0
   if os.path.isfile( "%s.gz" % dumpfile ):
       os.remove( "%s.gz" % dumpfile )
   file = os.popen( "gzip -c > %s.gz" % dumpfile, "w" )
   result_sets = db.execute( "SELECT %s FROM %s" % ( select, table ) )
   for row in result_sets[0]:
    
       line = []
       for i in range( len( row ) ):
           line.append( "%-*s" % ( lengths[i], row[i] ) )

       print >> file, "".join( line )
       num_recs += 1

   file.close()
   c.close()

Note that the c.fetchone() method retrieves rows one at a time from
the server so does not consume much memory when there are a huge
number of rows returned.  The db.execute() array binds a buffer for 16
rows so retrieves the entire result 16 rows at a time before
returning.  It will use a large amount of memory if there are a large
number of rows returned.

> I am still a Python newbie, so I may be doing something wrong.  Has anyone 
> else experienced similar results?  Is this due to a poorly optimized 
> Sybase module?  I really want to use Python, and I don't mind giving up a 
> little speed, but this is more than I can accept.
> 
> I ran the Python script using the profiler and the output is below.  Any 
> help would be much appreciated.  Thanks.

> profile output:
> 
>  Ordered by: standard name
> 
>  ncalls  tottime  percall  cumtime  percall filename:lineno(function)
>       1    0.000    0.000 1230.540 1230.540 <string>:1(?)
>       4    0.000    0.000    0.000    0.000 Sybase.py:114(_get_diag)
>       2    0.000    0.000    0.000    0.000 Sybase.py:141(_ct_errors)
>       3    0.000    0.000    0.000    0.000 Sybase.py:148(_row_bind)
>  900345   96.240    0.000   96.240    0.000 Sybase.py:166(_extract_row)
>  900348  177.670    0.000  273.910    0.000 Sybase.py:180(_fetch_rows)
>       3    0.000    0.000    0.010    0.003 Sybase.py:229(__init__)
>       3    0.000    0.000    0.000    0.000 Sybase.py:246(__del__)
>  900361   83.230    0.000  267.090    0.000 Sybase.py:258(_lock)
>  900361   86.720    0.000  269.000    0.000 Sybase.py:262(_unlock)
>       3    0.000    0.000    0.010    0.003 Sybase.py:312(close)
>       3    0.000    0.000    0.000    0.000 Sybase.py:328(execute)
>  900348  178.280    0.000  988.260    0.001 Sybase.py:365(fetchone)

[snip]

The other thing which may be causing some of the slowness is the
locking.  The module allows threads to share connections which
requires a fair amount of lock maintenance.

- Dave

-- 
http://www.object-craft.com.au


From andymac at bullseye.apana.org.au  Fri Sep  6 00:25:31 2002
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Fri, 6 Sep 2002 14:25:31 +1000 (EST)
Subject: Python "compiler" is  too slow for processing large data files???
In-Reply-To: <3d6d097d$0$321$e2e8da3@nntp.cts.com>
Message-ID: <20020906141327.S86922-100000@bullseye.apana.org.au>

[posted and mailed]

On Wed, 28 Aug 2002, Ron Horn wrote:

{...}

> Simple example - I can import or exec this file to load my data (my real app
> has int, float, and string data):
> ------ try5a3.py --------
> list1 = [
>     (323, 870, 46, ),
>     (810, 336, 271, ),
>     (572, 55, 596, ),
>     (337, 256, 629, ),
>     (31, 702, 16, ),
> ]
> print len(list1)
> ---------------------------
>
> Anyway, as my data files went from just a few lines, up to about 8000 lines
> (with 10 values in each line for total of about 450KB of text), the time to
> 'exec' the file became too slow (e.g. 15 seconds) and used too much memory
> (e.g. 50MB) (using ms-windows, python 2.2.1).  It is the "compile" phase,
> because if I re-run, and there is *.pyc file available, the import goes very
> fast (no compilation required).

Hmmm...  Your problem looks familiar.

A change to Python's parser was checked in to CVS by Tim Peters on both
the head (2.3 to be) and 2.2 maintenance branches, which works around
malloc()/realloc()/free() issues on several platforms in the face of very
long expressions - see test_longexp.py in Python's regression test suite.

The symptoms include gross memory consumption and/or very long runtimes
for test_longexp.py.

The fix will be available in 2.2.2 as well as 2.3 when released.

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andymac at bullseye.apana.org.au  (pref) | Snail: PO Box 370
        andymac at pcug.org.au             (alt) |        Belconnen  ACT  2616
Web:    http://www.andymac.org/               |        Australia




From loewis at informatik.hu-berlin.de  Wed Sep 18 02:50:14 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Sep 2002 08:50:14 +0200
Subject: How to make a C extension module backwards compatible?
References: <mailman.1032319522.8608.python-list@python.org> <m3sn07hpjk.fsf@mira.informatik.hu-berlin.de> <mailman.1032325763.493.python-list@python.org>
Message-ID: <j4znufbxll.fsf@informatik.hu-berlin.de>

Andrew McNamara <andrewm at object-craft.com.au> writes:

> That's a good question - this is more of a learning exercise than a
> project with a real need. I guess the most useful attribute would be
> sub-classing.

I see. In that case, there is no way around ifdefs - the type object
has grown extra fields, which the compiler will complain about if you
are using an older version of Python.

Regards,
Martin


From pieroul at attglobal.net  Mon Sep 30 19:10:17 2002
From: pieroul at attglobal.net (Pierre Rouleau)
Date: Mon, 30 Sep 2002 19:10:17 -0400
Subject: Is there a docutils / reStructuredText system for C++ source code
 available?
Message-ID: <3D98D9D9.8000009@attglobal.net>

I have been using docutil and reStructuredText for Python code and it 
works great. Pydoc does a superb job of helping in setting up an 
internal web site that provides documentation for our Python code, let 
alone the ability to run doctest test and show auto testing of python 
documentation on line!

That works great for our Python code.  Now i want to do the same thing 
for our C++ code.  Since Python has been partly implemented in C / C++ 
i tough there might already exist some tools for parsing C++ and some 
comments (given some formatting conventions) and generate the same kind 
of output the html tool of docutil does.

Does any one know any thing that resembles that?

Thanks
-- 
         Pierre Rouleau




From Oschler at earthlink.net  Wed Sep 11 19:55:47 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Wed, 11 Sep 2002 23:55:47 GMT
Subject: SpamBayes PIK file?
Message-ID: <7KQf9.1760$Lg2.232748@news2.news.adelphia.net>

I want to look at the SpamBayes project so I downloaded the zip file from
source forge.  Inside there's just a PIK file.  How do I read a PIK file?

thx





From mennosimons at gmx.net  Tue Sep 17 05:52:47 2002
From: mennosimons at gmx.net (Ulli Stein)
Date: Tue, 17 Sep 2002 11:52:47 +0200
Subject: faster way for adding many strings
Message-ID: <am6tgt$lqm$01$1@news.t-online.com>

Hi,

What is the faster way: appending many strings to a list and then joining 
them, or writing to 
StringIO and then getvalue()?

willi 


From sismex01 at hebmex.com  Wed Sep 25 12:40:22 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Wed, 25 Sep 2002 11:40:22 -0500
Subject: Larry Wall's comment on python...
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233025D48F5@mail.hebmex.com>

> 
> Alan Daniels wrote:
> 
> >Sometimes I wonder if Guido would have saved a lot of debate 
> > by declaring, long ago, that Python code would never allow
> > tabs in it, period. Which may be a bit extreme, I'll admit. :-)

Actually, it's the only universally-workable solution
to this "problem".

> Ironically, Guido originally ruled (in the style guide, IIRC) 
> that thou shalt use tabs and tabs are 8 spaces, "as god
> intentended them to be."

Yeesh! Why in God's name would he say such a dumb thing?
I mean, even ancient mechanical typewriters have movable
tab settings. Rigid, 8-char tabs is just so... neolithical.

> 
> Unfortunately, he later changed position to the present, 
> unsatisfying, compromise position.

And so, we all lose.

> -- 
> James J. Besemer		503-280-0838 voice
> 2727 NE Skidmore St.		503-280-0375 fax
> Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
> 				http://cascade-sys.com	
>

-gus










Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From jafo at tummy.com  Wed Sep 18 13:19:00 2002
From: jafo at tummy.com (Sean Reifschneider)
Date: Wed, 18 Sep 2002 11:19:00 -0600
Subject: New RPMs at python.org
Message-ID: <20020918171900.GB13272@tummy.com>

I've just released a new .src.rpm and binary RPMs for Red Hat 7.3.
Included in the new release is a Makefile.pre.in (in the -devel package)
for use with building Zope, and inclusion of /usr/bin/python2.2 to better
operate with the source and Red Hat releases.  The lack of it was a
hold-over from previous 2.x RPMs.

Find them at:

   http://python.org/2.2.1/rpms.html

Sean
-- 
 What no spouse of a programmer can ever understand is that a programmer is
 working when he's staring out the window.
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com, ltd. - Linux Consulting since 1995.  Qmail, Python, SysAdmin



From dhein at acm.org  Mon Sep  9 09:01:40 2002
From: dhein at acm.org (Coot)
Date: 9 Sep 2002 06:01:40 -0700
Subject: Running IDLE from within a Tkinter application
Message-ID: <a3a96207.0209090501.4f62f22e@posting.google.com>

Is there any way to start an IDLE "Python Shell" window from within an
executing Tkinter application _AND_ have the shell see the global
variables of the main application???

I can launch IDLE from within the application, but I cannot figure out
how to do so in such a way that it sees the global variables of the
application.


From michael at stroeder.com  Tue Sep 24 07:35:25 2002
From: michael at stroeder.com (=?ISO-8859-1?Q?Michael_Str=F6der?=)
Date: Tue, 24 Sep 2002 13:35:25 +0200
Subject: Sessionhandling
References: <c5b3549a.0209230650.34a30fca@posting.google.com> <f72dac0d.0209231211.5614b398@posting.google.com> <amp3h2$1ot7$1@ariadne.rz.tu-clausthal.de> <bvo56-tf1.ln1@junker.stroeder.com> <ampi2g$28ci$1@ariadne.rz.tu-clausthal.de>
Message-ID: <t8v56-mc4.ln1@junker.stroeder.com>

Lutz Schroeer wrote:
> Michael Str?der <michael at stroeder.com> wrote:
> : Most users do *not* configure their browsers anyway. Therefore it's 
> : almost useless to use HTTP_ACCEPT_LANGUAGE or similar env vars.
> 
> Well, at least you can try.

As you can see in PyWebLib I did. I'm also using it in web2ldap to 
determine the preferred language for HTML templates. Guess what? Most 
german users see the english HTML output...

> My solution to the orignal poster's problem
> would be to detect the browser's language configuration and additionally
> offer the possibility to change the language and store it in a (maybe
> persistent) cookie.

Note that someone suggested to use HTTP_ACCEPT_LANGUAGE for the orignal 
poster's problem to avoid session handling at all.

Ciao, Michael.



From sandskyfly at hotmail.com  Sat Sep 14 04:02:17 2002
From: sandskyfly at hotmail.com (Sandy Norton)
Date: 14 Sep 2002 01:02:17 -0700
Subject: Argh! Python 2.2 + threading + sockets = trouble
References: <R_ug9.753$Ee4.1406@news-server.bigpond.net.au>
Message-ID: <b03e80d.0209140002.44c6b816@posting.google.com>

Freddie <freddie at madcowdisease.org> wrote in message news:<R_ug9.753$Ee4.1406 at news-server.bigpond.net.au>...
> Hi,
> 
> I'm a fairly new Python programmer, but I've written a decent sized 
> project (threaded IRC offer bot), and I wrote a threaded IPv6 capable 
> IRC bouncer yesterday. Of course, it doesn't work quite as I expected.
> 
> MadCowOffer (offerbot) has rarely works properly with Python 2.2. It 
> will sometimes manage to start all the threads, then it's fine, but it 
> will usually just stop working after the IRC connection thread is started.

A more precise description of your problem will often help others help
you; more specifically, stating that an app 'just stops working' is
not a great problem description.

Localize the problem and someone may be able to point you in the right
direction.

cia,

Sandy


From iqbal at colorado.edu  Wed Sep  4 14:44:02 2002
From: iqbal at colorado.edu (Naveed Iqbal)
Date: 4 Sep 2002 11:44:02 -0700
Subject: How do i print to a printer using python?
Message-ID: <6afb9d2c.0209041044.43c5f5f1@posting.google.com>

I have to print to a standard printer. I take the data from the
keyboard and print it using the printer. Please help me....my job
depends on this!!!!
thanks
naveed


From sandorlevi at yahoo.com  Fri Sep 27 16:11:58 2002
From: sandorlevi at yahoo.com (Levente Sandor)
Date: 27 Sep 2002 13:11:58 -0700
Subject: Problem with ClientCookie
References: <3d946d0f$0$10364@echo-01.iinet.net.au>
Message-ID: <8b5e42a6.0209271211.4f0c6d9f@posting.google.com>

Probably the error is caused by the "secure" word in the set-cookie
headers, because it doesn't contain "=", so it cannot be split by
re.split(r"\s*=\s*", param, 1).
I never used the ClientCookie package, but you may try to replace line
627 in the source with the following one:
if '=' in param: k, v = re.split(r"\s*=\s*", param, 1)
(be sure to retain the original indentation of the line)
----
levi

"Rob Hall" <bloke at ii.net> wrote in message news:<3d946d0f$0$10364 at echo-01.iinet.net.au>...
> I am trying to use ClientCookie to access a HTTPS website under Win98.  I
> have installed the W32 SSL wrappers.
> 
> The code I have is similar to as follows (have removed authorisation
> detail):
> 
> import urllib
> import ClientCookie
> 
> if __name__ == '__main__':
>     urllib.URLopener.version = 'Mozilla/4.0 (compatible; MSIE 5.5; Windows
> NT 5.0)'
>     print 'Logging on...'
>     request =
> urllib2.Request('https://www.quicktrade.com.au/cgi-bin/login?USERNAME=myUser
> Name&PASSWORD=myPassword')
>     result = ClientCookie.urlopen(request)
> 
> I have set HTTP_DEBUG = 1 in _ClientCookie.py
> 
> I get the following response...
> 
> 
> Logging on...
> Page requested...
> send: 'GET /cgi-bin/login?USERNAME=myUserName&PASSWORD=myPassword
> HTTP/1.0\r\n'
> send: 'Host: www.quicktrade.com.au\r\n'
> send: 'User-agent: Python-urllib/2.0a1\r\n'
> send: '\r\n'
> reply: 'HTTP/1.1 200 OK\r\n'
> header: Date: Fri, 27 Sep 2002 14:21:31 GMT
> 
> header: Server: Stronghold/2.4.2 Apache/1.3.6 C2NetEU/2410 (Unix)
> 
> header: set-cookie:
> key=Kpfz5sbkkPrwsM7eng5ihtfuKq6sAlSdwTXhn/jUiEOSOCYJAC5r/Vlq+5uS9wqkpHJzOWHi
> m4v+^Va7SldrYP9Vqea2vDKtZWnOiP7Ff5ckrrM61DDM+tIGO44jUyKPt2yof5Uz7lvUu0ErVqJk
> DWAYX^7x66TxqZ2esB823pOue6zNxcmhY1SiS7pPTd5w3X/5ncX3njMZs_^; Path=/; secure
> header: set-cookie: acct=myUserName STGBNET; Path=/; secure
> header: set-cookie: level=1; Path=/; Domain=.quicktrade.com.au; secure
> header: set-cookie: ID=1802009989; Path=/; Domain=.quicktrade.com.au; secure
> header: set-cookie: w1=None; Path=/; secure
> header: P3P: CP="NON DSP ADMa OUR IND UNI FIN"
> header: Connection: close
> header: Content-Type: text/html
> 
> Traceback (most recent call last):
>   File
> "C:\WINDOWS\Desktop\backup\uni\project\development\htmlProcessor\htmlProcess
> or.py", line 22, in ?
>     result = ClientCookie.urlopen(request)
>   File "C:\PYTHON22\lib\site-packages\ClientCookie\_ClientCookie.py", line
> 1387, in urlopen
>     return _opener.open(url, data)
>   File "C:\PYTHON22\lib\urllib2.py", line 322, in open
>     '_open', req)
>   File "C:\PYTHON22\lib\urllib2.py", line 301, in _call_chain
>     result = func(*args)
>   File "C:\PYTHON22\lib\site-packages\ClientCookie\_ClientCookie.py", line
> 1334, in https_open
>     return self.do_open(httplib.HTTPS, req)
>   File "C:\PYTHON22\lib\site-packages\ClientCookie\_ClientCookie.py", line
> 1319, in do_open
>     self.c.extract_cookies(result, req, redirect=redirect)
>   File "C:\PYTHON22\lib\site-packages\ClientCookie\_ClientCookie.py", line
> 627, in extract_cookies
>     k, v = re.split(r"\s*=\s*", param, 1)
> ValueError: unpack list of wrong size
> 
> I can't find anything to help me debug this.  Is the error because the web
> server is returning more than one cookie?  or is it just returning a bad
> format?
> 
> Any help is greatly appreciated.
> 
> Rob


From sismex01 at hebmex.com  Wed Sep 25 10:31:43 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Wed, 25 Sep 2002 09:31:43 -0500
Subject: Newbie References Question
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233025D4490@mail.hebmex.com>

> 
> Hi,
> 
> let say I have:
> i1 = 1
> i2 = 2
> and
> p = [i1,i2]
> 
> How can I have:
> p=[*i1,*i2]
> rather than
> p=[**i1,**i2] as it is curently ?
> 
> ( Sorry for this nasty hybrid syntax )
> 
> Textualy, how if I want that p[0] refere to the 'i1' 
> container, and not to
> the refererence it contains ?
> 
> What I want is that if now I set:
> i1 = 4
> that automaticaly:
> p -> [4,2]
> and not keeping [1,2]
> 
> Thanks in advance.
> 
> --
> guy rabiller
> 3d animator / td
> grabiller at 3dvf.net
> http://grabiller.3dvf.net
> 

Repeat after me: "there is no spoon"

Really though; you're chasing a ghost; the "container" you're
looking for is (in the case of CPython) an element in a dictionary,
any definition in a namespace is a definition in the
dictionary which implements said namespace.

Also, when you do a "i1 = 4", you're actually deleting the
previous binding for the name "i1", and rebinding it to
the object "int(4)".

Why not tell us what you're trying to do, and maybe someone
can come with a scheme to let you do it. :-)

Good luck!

-gus




Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From see_reply_address at something.invalid  Sun Sep 22 20:46:52 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Mon, 23 Sep 2002 12:46:52 +1200
Subject: Destructor never called ???
References: <mailman.1032549652.29064.python-list@python.org> <ami6v5$phc$1@panix1.panix.com>
Message-ID: <3D8E647C.8010707@something.invalid>

Aahz wrote:

> I'm of the opinion that I understand roughly half of what you're talking
> about and if I count as half a person, you can just round up.


Except in situations where the law requires us to
round you down. :-)

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From loewis at informatik.hu-berlin.de  Wed Sep  4 12:50:40 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 04 Sep 2002 18:50:40 +0200
Subject: Tkinter report_callback_exception?    'module' object has no attribute 'print_exception' ?
References: <2621b014.0209040715.1bf9fce3@posting.google.com>
Message-ID: <j4sn0pd7hb.fsf@informatik.hu-berlin.de>

schull at digitalgoods.com (Jon Schull) writes:

> As I interpret this message, the environment is telling me that
> traceback has no "print_exception."   But of course, it does.

That is indeed surprising. I'd check whether traceback really is what
you think it is, at that point:

print traceback.__filename__
print dir(traceback)

HTH,
Martin


From peeryc at hotmail.com  Fri Sep 27 10:29:02 2002
From: peeryc at hotmail.com (Christopher Peery)
Date: 27 Sep 2002 07:29:02 -0700
Subject: Python Process that can survive
Message-ID: <571739c3.0209270629.10575f47@posting.google.com>

Alright... I love python but I think I may have found a definite
 problem. I can't seem to run a python process, kick it to the
 back-ground, and then kill the terminal that started it. With most
 other things, the python process would keep running in the
 back-ground regardless. For me, the process always dies with the
 terminal.

 So my question: is there a way to separate a python script from the
 underlying terminal so that this will not happen.

 I'm hoping for some scheme that's supported by the language. I've
 found a few hacks on-line where the python process is daemonized
 but this requires shutting down all the standard io streams.
 There's got to be a better way especially since it can be done in
Perl very easily with no special tricks.

By the way, I'm doing this on a linux box.

Chris


From bokr at oz.net  Mon Sep  2 19:37:34 2002
From: bokr at oz.net (Bengt Richter)
Date: 2 Sep 2002 23:37:34 GMT
Subject: stylistic question -- optional return value
References: <mailman.1030608587.4698.python-list@python.org> <30292041.0209012344.ab2d4@posting.google.com> <yu99r8gco3m2.fsf@europa.research.att.com>
Message-ID: <al0snu$qf7$0@216.39.172.122>

On Mon, 2 Sep 2002 14:41:57 GMT, Andrew Koenig <ark at research.att.com> wrote:

>Sebastian> Why not create an additional optional *argument* (say a list, or a
>Sebastian> function), to which you optionally provide your debug information?
>
>It's not debug information.  I used that term only as an analogy.
>
>Anyway, I don't want to separate the optional information from the
>rest of it because everything needs to go into the same data structure
>eventually.
>
What about defining a class to represent the "data structure," and passing
a reference to an instance of that to your various routines, so that they
can populate the structure themselves (via methods/attributes/properties
as convenient) instead of returning data for a caller to do the populating with?

BTW, you mentioned recursion in another part of the thread. Is that as in recursive
descent? An exception is an interesting alternative way to return a first match, with
no need to return anything from unsuccessful visits.

Regards,
Bengt Richter


From sismex01 at hebmex.com  Wed Sep 25 10:00:48 2002
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Wed, 25 Sep 2002 09:00:48 -0500
Subject: Decimal arithmatic, was Re: Python GUI app to impress the bos
	s?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233025D43B3@mail.hebmex.com>

> > 
> > There ought to be severe penalties for idiots that use 
> > floating point dollars for financial applications.
> > If forced to use floating point (e.g. because customer
> > demands BASIC), then keep money amounts in whole pennies
> > (or whatever the smallest currency unit for the country
> > is), and divide by 100 (or whatever) for printing only (or 
> > just add the decimal point yourself).
> > 
> 
> In many circumstances this is exactly what you can't do.
> The minute you need to calculate percentages you will get in to
> the things like .0001254 of a cent, no matter what is you
> smallest unit of currency.
> 
> I worked for a insurance company (using Business Basic) many
> years ago on MAI and Prime gear, and we did everything as floats
> with 14 places after the decimal point,and only rounded when
> a human needed to see a number, and then used standard accounting
> practices for rounding,
> 
> Rgds
> 
> Tim
> 

OK, so don't use whole pennies (or smallest currency) as your
base, use 10,000ths of a penny, but still use integer arithmetic
for your calculations.

Floating point has too many rounding errors to be trustworthy;
or, looking at it from another angle, floating-point calculations
cannot be 100% accurate because of the way it's implemented.

-gustavo




Advertencia: 
La informacion contenida en este mensaje es confidencial y restringida y
esta destinada unicamente para el uso de la persona arriba indicada, Esta
comunicacion representa la opinion personal del remitente y no refleja
necesariamente la opinion de la Compa?ia. Se le notifica que esta
estrictamente prohibida cualquier difusion, distribucion o copia de este
mensaje. Si ha recibido esta comunicacion o copia de este mensaje por error,
o si hay problemas en la transmision, favor de comunicarse con el remitente.


Todo el correo electr?nico enviado para o desde esta direcci?n ser?
procesado por el sistema de correo corporativo de HEB. Tal correo
electr?nico esta sujeto a ser almacenado y puede ser revisado por alguien
ajeno al recipiente autorizado con el prop?sito de monitorear que se cumplan
las normas de seguridad de la empresa.



From syver-en+usenet at online.no  Fri Sep  6 13:32:25 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Fri, 06 Sep 2002 17:32:25 GMT
Subject: Windows NT login script
References: <3D781715.5070505@nowhere.com>
Message-ID: <uofbbuir2.fsf@online.no>

c <noone at nowhere.com> writes:

> I am a newbie to Python and I'm wondering how difficult it would be to
> 
> write a Windows NT/2000 login script using Python?
> 
> Can anyone think of a good way to do it?

I am not really into this, but couldn't you start a python script from
a windows batch file (.bat or .cmd)

Like this:

bla bla batch stuff
net use bla bla
python myScript.py withAnArgument orTwo
bla bla batch stuff

-- 

Vennlig hilsen 

Syver Enstad


From fgibbons at hms.harvard.edu  Fri Sep 20 13:56:21 2002
From: fgibbons at hms.harvard.edu (Frank Gibbons)
Date: Fri, 20 Sep 2002 13:56:21 -0400
Subject: saving state in Python/CGI
Message-ID: <5.0.0.25.2.20020920134912.00ad3230@hms.harvard.edu>

Hi,

I've been using Python (and Jython) for about 5 months now. So far, so 
great. I'm building a CGI app that has about 5 different stages, and I need 
to maintain state between them. Specifically, it's a scientific tool that 
requires the user to upload a (possibly quite large) datafile, processes 
it, does some computation, then hands it back (formatted) with results.

I need access to the data on each page, yet don't want to pass it as a 
hidden field, because of its potential size. In Perl, I could just use the 
save() method from CGI.pm to save the form on the server side, pass a 
session key in a hidden field, then retrieve the data using the key. For 
the life of me, I can't find an equivalent way in Python. cgiFormStorage 
looks like a dictionary, but it's immutable, so you can't retrieve data 
from disk and add it to the form. cgi.py appears to have no analog to 
CGI.pm's save() method, at least not that I've been able to find. I've 
tried pickling the form, but that doesn't work either (it saves something 
to disk alright, but not the large data retrieved from a multi-part form 
file upload).

Perhaps I've missed something. Maybe that's just not how you do things in 
Python. Can anyone help?

Thanks,

-Frank Gibbons

PhD, Computational Biologist,
Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA.
Tel: 617-432-3555       Fax: 
617-432-3557       http://llama.med.harvard.edu/~fgibbons




From sholden at holdenweb.com  Thu Sep 12 11:57:00 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 12 Sep 2002 11:57:00 -0400
Subject: class variable won't icrement!
References: <3d8054ef$0$234$4d4ebb8e@news.nl.uu.net> <alpner$bm$1@wanadoo.fr> <alq9hf$bn$1@panix1.panix.com>
Message-ID: <PQ2g9.191834$l_4.58187@atlpnn01.usenetserver.com>

"Aahz" <aahz at pythoncraft.com> wrote in message
news:alq9hf$bn$1 at panix1.panix.com...
> In article <alpner$bm$1 at wanadoo.fr>,
> Eric Brunel  <eric.brunel at pragmadev.com> wrote:
> >
> >I always found that inheritance of class attributes to instance
attributes
> >was really misleading, and I even think it would be better if it were
> >simply removed from Python. IMHO, it would far more conform to the
> >"explicit is better than implicit" rule that is one of the basis of
Python.
>
> If I think of it as "scope resolution" rather than "inheritance", I find
> that it makes more sense.  It works exactly the same way as hiding a
> global with a local.

Well, not quite exactly the same way. If a local hides a global then there's
no way to access the global without explicit use of a namespace.

A class variable is available by instance-qualified name up until the time
the instance variable with the same name is created, which is why statements
like

    self.var += 1

can be so confusing. Especially when you compare operations that give
different results for mutable and immutable objects.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From richard.papworth at bbsrc.ac.uk  Tue Sep  3 05:07:29 2002
From: richard.papworth at bbsrc.ac.uk (Richard Papworth)
Date: 3 Sep 2002 02:07:29 -0700
Subject: Core dump on Tru-64 when using 2.2
Message-ID: <695dd93e.0209030107.7ba70082@posting.google.com>

Does anyone know the source of the following error message?

exception system: exiting due to multiple internal errors:
        exception dispatch or unwind stuck in infinite loop
        exception dispatch or unwind stuck in infinite loop
exception system: exiting due to multiple internal errors:
        exception dispatch or unwind stuck in infinite loop
        exception dispatch or unwind stuck in infinite loop
Abort process (core dumped)


Build information:
Python 2.2 (#2, Feb 20 2002, 11:44:40) [GCC 2.95 19990728 (release)]
on osf1V4

I've only experienced this problem over the past couple of days.
Firstly when attempting to use minidom to parse a string, then when
using the help system under the interactive interpreter.

I'm puzzled 'cos I'm sure these are activities I've carried out
without a hitch over the past few months. So why the problem now? Any
suggestions?

Thanks in advance

Richard


From deltapigz at telocity.com  Tue Sep 10 13:18:21 2002
From: deltapigz at telocity.com (Adonis)
Date: Tue, 10 Sep 2002 13:18:21 -0400
Subject: test (ignore)
Message-ID: <3d7e2957$1_2@nopics.sjc>

just a test, please ignore.

Adonis




From mmaddox at hcsmail.com  Fri Sep  6 23:28:20 2002
From: mmaddox at hcsmail.com (mmaddox at hcsmail.com)
Date: Fri, 6 Sep 2002 23:28:20 -0400 (EDT)
Subject: Larry Wall's comment on python...
Message-ID: <200209070328.g873SKVm012455@mail.hcsmail.com>

You're right. The PLACEMENT of blocks in my style is arbitrary. My
intention was to indicate the STYLE of block delimiting was arbitrary
(Nehemiah Jordan: when a task is not seen in meaningful context it is
experienced as being arbitrary). This is not totally in line with my
intentions either, as all programming language rules could be deemed
arbitrary by someone unwilling to accept the context in which they exist.

Now, I LOVE Python. I also love C++. Smalltalk and Lisp are interesting.
Perl is useful to someone somewhere. I said some things that are not as
precise as I would have liked, and I don't want to explore the semantics
of this mailing list and each individual post.	Python is cool. Let's talk
about Python.

-Michael

peter at engcorp.com wrote:
> mmaddox at hcsmail.com wrote:
> > Although I grant the Python structure and the column-based
> > structures are not completely analogous, they BOTH burden the
programmer
> > with arbitrary rules to govern the physical layout of the code.
Although
> > this layout may be considered more elegant from a readability
standpoint,
> > it's the enforcement of the layout that is irritating to the more
> > Libertarian of us.
> 
> I think you have the wrong meaning for "arbitrary".  From the
> sounds of it, you have exactly reversed which is arbitrary (your
> own "libertarian" coding style) and which is "governed by fixed
> rules" (Python layout).
> 
> > Having read Wall's comments on Slashdot, I think he is really just
mildly
> > complaining about ridiculously long page-widths in his editor, not
really
> > slashing Python seriously. He's just giving some props to his own baby
at
> > the expense of another. (He had to say something, right?) You must
agree
> > that Python can lead you to horizontal scrolling in your editor - the
bane
> > of user interface design.
> 
> I agree with the comments about Larry "having" to say something like
this.
> 
> I definitely don't understand how Python, more or less than any other
> language, leads to more horizontal scrolling.  Pretty much all
> Python code I've ever seen (though I haven't seen yours yet :-) is
> roughly 80 characters max, and mine own has been about that since I
> started writing C code 16 years ago.
> 
> Do you mean there is something special about Python not delimiting
> blocks in the traditional ways which somehow leads you to write
> much wider lines of code?
> 
> -Peter
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 
> ?

--
Michael F. Maddox
Senior Software Engineer
Hayes
Tallahassee, FL, USA




From donn at drizzle.com  Sat Sep 21 02:06:43 2002
From: donn at drizzle.com (Donn Cave)
Date: Sat, 21 Sep 2002 06:06:43 -0000
Subject: Python threading?
References: <f9Pi9.10275$Lg2.2191294@news2.news.adelphia.net> <mailman.1032571709.25544.python-list@python.org>
Message-ID: <1032588401.625298@yasure>

Quoth Trent Mick <trentm at ActiveState.com>:
| [Robert Oschler wrote]
|> How can I tell if the version of Python I am running is thread-enabled?
|
| Run Lib/test/test_thread.py with your Python build. If you are on any
| common platform then Python, unless specifically crippled, is
| thread-enabled.

I think you know that doesn't include NetBSD, and I believe it doesn't
include MacOS either.  I'd call both of those common platforms, though
you can draw that line anywhere you want.

|> Also, what is a reasonable number of threads to expect to be able to run
|> before context switching overhead becomes a problem (I'm using a PIII 500
|> Mhz with 512MB ram if that helps).
|
| I have no experience putting that many threads to use.

Of course it would depend, too, on what "a problem" means, and on what
operating system the application is hosted on.  And it might be worth
mentioning that the Python parts of the application will not execute
in parallel, anyway, thanks to a global interpreter lock that can be
held by only one thread at a time.

	Donn Cave, donn at drizzle.com


From jubafre at brturbo.com  Sun Sep  8 13:56:26 2002
From: jubafre at brturbo.com (jubafre at brturbo.com)
Date: Sun, 8 Sep 2002 14:56:26 -0300 (GMT-03:00)
Subject: transform a list of lists in a lit of strtings???
Message-ID: <-1588066870.1031507786228.JavaMail.nobody@webmail2.brturbo.com>

I have a list of lists like this:

labels=[['D1'], ['D2'], ['D3']]

and i want to transform ia a list of strings

labels2==[ 'D1' ,  'D2' ,  'D3' ] <<<<<<<----------- how i can get it????

	
Juliano Freitas
www.gebrasil.hpg.com.br 

From pyth at devel.trillke.net  Thu Sep 19 09:40:00 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 19 Sep 2002 15:40:00 +0200
Subject: Getting started
In-Reply-To: <uojinmcfapjf45@corp.supernews.com>; from lbrannma@cablespeed.com on Thu, Sep 19, 2002 at 06:03:25AM -0700
References: <uojinmcfapjf45@corp.supernews.com>
Message-ID: <20020919154000.B30315@prim.han.de>

Lance wrote:
> Hi All,
> 
> Last night I attended a Weak typing, C++ Templates, and Python talk by Bruce
> Eckel. It was great, my introduction to Python.
> 
> I'm sold on Python and weak typing. 

I am not. But i much prefer dynamic over static typing :-)
What is 'weak typing' supposed to mean, anyway?  I have heard that a couple
of times from java-people because they tend to think that python
is not strongly typed.

> Would you be able to suggest an indispensable learning Python book?

David Beazley wrote "Python essential reference" which is a very good
book for programmers. It doesn't cover the most recent developments
but it still is one of the best IMO.

    holger



From python-list at hendel.net  Wed Sep 25 03:06:55 2002
From: python-list at hendel.net (Manuel Hendel)
Date: Wed, 25 Sep 2002 09:06:55 +0200
Subject: problems defining an empty list
Message-ID: <20020925070654.GB57911@partagas.as.de.cw.net>

This is what I'm trying to do:

    # empty list for lines per domain
    linesperdomain = []
    linesperdomain = [domain, domainlines]
    print linesperdomain

domainlines is also a list.

Error:

linesperdomain = []
                  ^
IndentationError: unindent does not match any outer indentation level


I've done it the same way in the same script several times before,
what's wrong here.

I don't understand this error.

Can someone help.

Thanks,
    Manuel

-- 
This American system of ours, call it Americanism, call it capitalism, call it 
what you will, gives each and every one of us a great opportunity if we only 
seize it with both hands and make the most of it. 
-Al Capone 



From s_gherman at yahoo.com  Tue Sep 10 15:22:31 2002
From: s_gherman at yahoo.com (Sorin Gherman)
Date: 10 Sep 2002 12:22:31 -0700
Subject: web crawler help?
References: <JARd9.6927$yt3.3340577@newssrv26.news.prodigy.com>
Message-ID: <ac259413.0209101122.54cc774c@posting.google.com>

"koko" <kokohh at hotmail.com> wrote in message news:<JARd9.6927$yt3.3340577 at newssrv26.news.prodigy.com>...
> is there any sample for basic web crawler, that ask for a starting url and
> log the url and extract the hyperlinks?
> thx

    It's a very simple one in Mark Pilgrim's "Dive into Python" book,
whose text is freely available at: http://diveintopython.org/
    Check the "HTML processing" chapter. It contains a urllister.py 9
lines program, followed by a 7 lines usage example which does just
that: given a URL for a HTML file, it lists the hyperlinks inside it.

/sorin gherman


From max at alcyone.com  Mon Sep  2 14:59:12 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 02 Sep 2002 11:59:12 -0700
Subject: How to parse over a Python expr?
References: <8sjvmugochmuvo4fnfaelq4m59plj6lg0r@4ax.com> <3D6FD585.111E32AE@alcyone.com> <s7tvmukn21cpi99ecksvr2up7u4aurm130@4ax.com> <3D7002C4.22DD2AB3@alcyone.com> <hn07nucub4elvt0dkauff7ee2jh8n8l2pa@4ax.com>
Message-ID: <3D73B500.D1304D13@alcyone.com>

Stefan Franke wrote:

> Are you sure? I think a TQS with a single quote inside would be
> sufficient to miss the right TQS end:
> 
> '''That's all'''
> 
> These a 7 single quotes. Parsed as single-quote strings, the last one
> opens a string until EOF.

Yep, you're right (provided the triple-quoted string is the same kind
that appears within it).  I realized that after I posted.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ There is nothing so subject to the inconstancy of fortune as war.
\__/ Miguel de Cervantes
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From bloke at ii.net  Tue Sep 17 21:19:36 2002
From: bloke at ii.net (Rob Hall)
Date: Wed, 18 Sep 2002 09:19:36 +0800
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <siOh9.451392$me6.54442@sccrnsc01>
Message-ID: <3d87d495$0$28699@echo-01.iinet.net.au>

> Actually both may be a solution. I've recently had good success at
> developing an with  a VB (not .net) front end and doing most of the
> work in Python. Use each for what it's best at. Based on my experience
with
> project, I'd say that Python take about 25% of the code of VB to
accomplish
> the same objective. Plus Python objects tend to do what is expected so
that
> you can often guess at the syntax and be right whereas this is seldom true
> with
> VB objects.


Interesting...  Can you point me to any links that may give me some pointers
in doing the same?

Rob




From Doc at goodweb.de  Wed Sep  4 11:18:03 2002
From: Doc at goodweb.de (Thomas Korb)
Date: 04 Sep 2002 17:18:03 +0200
Subject: 4Suite / McMillan Installer / py2exe
Message-ID: <doc.m3admxkclw.fsf@goodweb.de>

Hi! I am using the XSLT-Processor from 4Suite (0.12.0a2), i.e.

  from Ft.Xml.Xslt.Processor import Processor
  from Ft.Xml.InputSource import DefaultFactory
  from Ft.Xml import ReaderException

On Linux, I can use McMillan's Installer to create a distribution;
on Windows, it fails (problems with a Math.py-module in the 4Suite
package). And py2exe complains about missing modules.

I can overcome the py2exe-problem by importing everything from
Ft.Xml.Xslt, i.e.

  import Ft.Xml.Xslt.ApplyImportsElement
  import Ft.Xml.Xslt.ApplyTemplatesElement
  etc. etc. (a lot of etc. :-)

(Too many module to include them on the commmand line with py2exe's
'-p' option; more than 1024 characters!)

After that, py2exe still complains about missing modules (some of
which even do not exist on the system; I think at least some of them
are Mac-specific!), but it works (as far as I could check up to now).

But now the program takes quite a time to start (many imports), and
I actually do not really trust this solution. Moreover, I would have 
to check and test all the above imports with each new version of 4Suite.

Does someone know how to use 4Suite with the McMillan Installer or
py2exe on Windows? Any help or hints would be highly appreciated!

</Doc>

-- 
-----------------  Dr. Thomas Korb  -  G O O D W E B . D E s i g n  ---
-----------------  DOMAIN: www.goodweb.de / E-MAIL: Doc at goodweb.de  ---


From =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?=  Thu Sep 19 08:08:25 2002
From: =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?= (=?iso-8859-5?Q?=B4=D0=DC=F8=D0=DD_=B3=2E?=)
Date: Thu, 19 Sep 2002 14:08:25 +0200
Subject: RegEx editors
References: <amaq0d$4aton$1@ID-144898.news.dfncis.de>
Message-ID: <3d89be32$1@news.mt.net.mk>

> Does anybody know of any good graphical regex editors out there?

As of KDE 3.0 its a standard option in (most) every find dialog.
It is also part of QT 3 I think.



-- 
??????

hundred-and-one symptoms of being an internet addict:
21. Your dog has its own home page.



From erikprice at mac.com  Wed Sep  4 11:42:05 2002
From: erikprice at mac.com (Erik Price)
Date: Wed, 4 Sep 2002 11:42:05 -0400
Subject: quick question about generator.next()
Message-ID: <DD431E96-C01C-11D6-BB50-00039351FE6A@mac.com>

A quick question about generators:

If I have a generator object, the first value produced by the generator 
object does not actually get produced until the first time I call 
generator_object.next() , correct?

In other words, simply instantiating the generator object doesn't 
"activate" the generator function, rather it just creates it -- only 
when the next() method is called does the generator's internal logic 
"activate", right?

(Where are generator objects getting their next() method from anyway... 
is it implicitly inherited behind the scenes when I use the "yield" 
keyword to turn a function into a generator?)


Thank you,

Erik





--
Erik Price

email: erikprice at mac.com
jabber: erikprice at jabber.org




From magnus at thinkware.se  Fri Sep 20 09:34:29 2002
From: magnus at thinkware.se (=?ISO-8859-1?Q?Magnus_Lyck=E5?=)
Date: Fri, 20 Sep 2002 15:34:29 +0200
Subject: isFloat: Without Exception-Handling
References: <ama6nt$n6i$06$1@news.t-online.com>
Message-ID: <3D8B23E5.3040005@thinkware.se>

Thomas Guettler wrote:
> Is there a way to write the following
> method without using exceptions?
> 
> def isFloat(string):
>     is_float=1
>     try:
>         float(string)
>     except:
>         is_float=0
>     return is_float

Something like this perhaps:

import re

def isFloat(S):
     floatRE = r"^[-+]?(\d+\.?\d*|\.\d+)([eE][-+]?\d+)?$"
     return re.match(floatRE, str(S)) is not None

I'm not sure it's faster though.


-- 
Magnus Lycka, Thinkware AB
Alvans vag 99, SE-907 50 UMEA, SWEDEN
phone: int+46 70 582 80 65, fax: int+46 70 612 80 65
http://www.thinkware.se/  mailto:magnus at thinkware.se



From david.abrahams at rcn.com  Mon Sep 30 16:22:55 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Mon, 30 Sep 2002 16:22:55 -0400
Subject: SWIG - C++ - Inheritance
References: <d31m9.3742$Qh1.491539@news20.bellglobal.com>
Message-ID: <anae03$qnd$1@bob.news.rcn.net>

"news1.sympatico.ca" <mtremblay at golemlabs.com> wrote in message
news:d31m9.3742$Qh1.491539 at news20.bellglobal.com...
> Hello,
>
> The problem comes when I come to create a new class in python that
inherits
> from CThread.  Here is the python code sample
> =============== test.py =================
> import my_module
>
> class MyThread(my_module.CThread):
>     def __init__(self):
>         pass
>     def process(self):
>         print "hey"
>
> a = MyThread()
> a.run()
> ============================================
> my_module is exposed to Python using SWIG.  When I try to execute this, I
> get the following error :
>
>     #############################################
>     Traceback (most recent call last):
>       File "U:\Alexandra\test\pythontool\martin.py", line 14, in ?
>         a.run()
>       File "U:\Alexandra\test\pythontool\hector.py", line 649, in run
>         def run(*args): return apply(_hector.CThread_run,args)
>     TypeError: Type error. Expected _p_CThread
>     #############################################
>
> It appears that the swig module does not recognize the a variable as a
> CThread.  I guess there is something I am doing wrong.  It does not even
> call the process function so the problem does not appear to be with the
pure
> virtual function.  Anybody can help ? It would be really appreciated.

I don't know much about SWIG, but if it's anything like Boost.Python the
problem is that you've shielded the base class __init__ function. Something
has to create the C++ CThread object, and it would make sense if
CThread.__init__ were responsible for that job. Just remove the __init__
function from MyThread, or have it call CThread.__init__(self) and you
should be OK.

--
-----------------------------------------------------------
           David Abrahams * Boost Consulting
dave at boost-consulting.com * http://www.boost-consulting.com





From just at xs4all.nl  Mon Sep  9 18:46:54 2002
From: just at xs4all.nl (Just)
Date: Tue, 10 Sep 2002 00:46:54 +0200
Subject: Windows/DOS: double clicking a .py file
References: <mailman.1031608392.11528.python-list@python.org>
Message-ID: <just-7B849D.00465410092002@news1.xs4all.nl>

[Just]
> > impossible to actually  see  the exception... Is there a way 
> > to keep the dos prompt window in this case?

[Bjorn Pettersen]
> Try something like the following... We've had much success with
> automatically emailing the traceback back to the developer using the
> cgitb module -- you can never trust users to give you good error reports
> <wink> besides the cgitb module provides much more context than the
> standard exceptions... Let me know if you'd like the code...
> 
> try:
>     your code here
> except:
>     import msvcrt, traceback
>     traceback.print exc()
>     print 'Hit a key to exit'
>     while not msvcrt.kbhit():
>         pass

Ah, that looks like a good trick, I will try that.

Thanks for your code offer, but in most cases I will have to ask for the 
input data anyway, so a simple traceback is good enough for now.

Just


From hbl at st-andrews.ac.uk  Tue Sep 17 14:46:21 2002
From: hbl at st-andrews.ac.uk (Hamish Lawson)
Date: Tue, 17 Sep 2002 19:46:21 +0100
Subject: Will "Python in a Nutshell" be too short?
Message-ID: <5.1.1.6.0.20020917181205.035d0a58@spey.st-andrews.ac.uk>

O'Reilly doesn't yet list the forthcoming "Python in a Nutshell" by Alex 
Martelli, but according to Amazon.com's page for the book, it will have 400 
pages. Is that indeed the plan? I know that O'Reilly aims to keep fat out 
of its books, and that Alex is striving to be comprehensive yet concise, 
but I worry that the book's proposed length will simply be too short to 
cover the material I'm hoping will be included.

I regularly consult David Beazley's "Python Essential Reference", but the 
usefulness to me of this otherwise excellent book is compromised by the 
number of (platform-independent and not obsolescent) standard modules and 
packages that had to be left out of the book's 380 pages (htmllib, 
telnetlib, urllib2, CongigParser, unittest, tkinter, xml, etc.).

Many of us were hoping that "Python in a Nutshell" would not only be a 
comprehensive reference of pretty much all of the modules in the standard 
library that aren't obsolescent or platform-specific, but would also cover 
some of the most popular third-party libraries - e.g. DB-API, PIL, 
mxDateTime, Numeric, win32all, wxPython, mod_python, ReportLab.

My congratulations to Alex and O'Reilly if they reckon they can somehow 
manage to fit this all into just 400 pages; but if this won't be possible, 
as I worry must be the case, then may I make a plea for the book's length 
to be reconsidered. It would be good to have a one-stop Python reference 
book that would be a suitable companion to the task-oriented "Python 
Cookbook" - I'm hoping "Python in a Nutshell" can still be that book.

By comparison, O'Reilly's "Perl in a Nutshell" is 800 pages and has the 
following chapters: Introduction to Perl; Installing Perl; The Perl 
Executable; The Perl Language; Function Reference; Debugging; Packages, 
Modules, and Objects; Standard Modules; CGI Overview; The CGI.pm Module; 
Web Server Programming with mod_perl; Databases and Perl; XML and Perl; 
SOAP; Sockets; Email Connectivity; Usenet News; FTP; Lightweight Directory 
Access with Net::LDAP; The LWP Library; Perl/Tk; Win32 Modules and 
Extensions; OLE Automation; ODBC Extension for Win32.


Hamish Lawson




From gregoryk at futures.wharton.upenn.edu  Wed Sep 18 18:21:01 2002
From: gregoryk at futures.wharton.upenn.edu (GREGORY KNESER)
Date: 18 Sep 2002 22:21:01 GMT
Subject: Calling function from command line
References: <amao3t$mug$1@netnews.upenn.edu>
Message-ID: <amau8d$sq0$1@netnews.upenn.edu>

GREGORY KNESER (gregoryk at futures.wharton.upenn.edu) wrote:

Thanks to those who followed up.  

I also got referred to an article at 
http://www.linuxjournal.com/article.php?sid=3946

involving:

import sys
sys.argv[1]


that was helpful.

Thanks again,
Greg


From gua81 at XXXyahoo.com  Thu Sep 26 07:15:03 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Thu, 26 Sep 2002 21:15:03 +1000
Subject: Getting last char in string.
References: <mailman.1033027776.19152.python-list@python.org> <amumk5$p52$1@lust.ihug.co.nz> <amuna6$erv$1@news1.wdf.sap-ag.de>
Message-ID: <amuq6t$sv9$1@lust.ihug.co.nz>

"Daniel Dittmar" <daniel.dittmar at sap.com> wrote in message
news:amuna6$erv$1 at news1.wdf.sap-ag.de...
> >>> s = ''
> >>> s [-1]
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
> IndexError: string index out of range
> >>> s [-1:]
> ''
> s [-1] returns a single element, s [-1:] returns a sequence. But in the
> current Python definition (implementation?), single characters are strings
> of length 1.
>
> Daniel

if s[-1] returns a single element, why woulnd't it return " ?
Single character are strings of length 1(as u said) and " is single
character right?




From christopher.saunter at durham.ac.uk  Mon Sep 23 12:46:17 2002
From: christopher.saunter at durham.ac.uk (Christopher Saunter)
Date: Mon, 23 Sep 2002 16:46:17 +0000 (UTC)
Subject: How to call a procedure in exe file
References: <mailman.1032767369.29784.python-list@python.org> <4rEj9.225917$z91.9596261@bin3.nnrp.aus1.giganews.com>
Message-ID: <amnggp$mcu$1@sirius.dur.ac.uk>

Terry Reedy (tjreedy at udel.edu) wrote:
<snip>
: IF you can start download with keys instead of just mouse, and if
: program takes keypresses through generic standard input functions
: (like C getchar()) rather than specific getkey function, start it at
: the other end of a pipe.
<snip>
: IF you can find (or write or modify open-source code to make) a
: mouse-click simulator program thats responds to dynamic input via
: stdin (rather than only responding to prewritten script read as
: startup), use that.

A rather neater and less trouble prone way of scripting things (when you
are reduced to sendkeys etc....) is Girder, and application designed to
interface IR remotes to PCs.  There is decent support for trapping and
replaying windows messages, and it can be commanded from network sockets
as well as IR etc.  So you can use it as a sort of network controlled
version of Win31's macro recorder (whatever happend to that?...)

---

cds



From loewis at informatik.hu-berlin.de  Mon Sep 16 04:46:55 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 16 Sep 2002 10:46:55 +0200
Subject: [xml] convert funny chars to char entites?
References: <vNah9.10101$Ee4.23987@news-server.bigpond.net.au> <m3elbuze67.fsf@mira.informatik.hu-berlin.de> <bKgh9.10807$Ee4.29949@news-server.bigpond.net.au>
Message-ID: <j43csanwxs.fsf@informatik.hu-berlin.de>

":B nerdy" <thoa0025 at mail.usyd.edu.au> writes:

> Traceback (most recent call last):
>   File "./batch-maid.py", line 85, in ?
>     generate_browse_options()
>   File "./batch-maid.py", line 70, in generate_browse_options
>     f.write(options_composer)
> UnicodeError: ASCII encoding error: ordinal not in range(128)
> 
> whats going on?

You should start with telling us what f is, and what options_composer
is. How are we supposed to know?

> how can i fix it?

I assume f is a file object, and options_composer is a Unicode
object. In that case, you need to determine the encoding you want to
use, and encode the Unicode object, e.g. with

  f.write(options_composer.encode("utf-8"))

Regards,
Martin



From gerhard.haering at opus-gmbh.net  Wed Sep 18 11:42:11 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 18 Sep 2002 15:42:11 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <Xns928CED692541cliechtigmxnet@62.2.16.82> <6_Ph9.479711$q53.16315681@twister.austin.rr.com> <am9jj9$c2m$1@wanadoo.fr> <3d8868b1$0$28683@echo-01.iinet.net.au> <1Y0i9.508708$UU1.84704@sccrnsc03>
Message-ID: <slrnaoh7cr.2e4.gerhard.haering@haering.opus-gmbh.net>

"bap" wrote:
> Does wx work with ActiveState Python?

Works for me (wxPython 2.3.2.1, Python 2.2.1 from ActiveState, Windows 2000
SP3).

-- Gerhard


From Oschler at earthlink.net  Wed Sep 11 19:38:02 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Wed, 11 Sep 2002 23:38:02 GMT
Subject: SuSE 7.3 Yast2-able Python 2.2.1?
References: <IjPf9.1729$Lg2.215939@news2.news.adelphia.net> <m3k7lsgnb2.fsf@mira.informatik.hu-berlin.de>
Message-ID: <utQf9.1755$Lg2.228666@news2.news.adelphia.net>

"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m3k7lsgnb2.fsf at mira.informatik.hu-berlin.de...
> "Robert Oschler" <Oschler at earthlink.net> writes:
>
>
> If you can't satisfy the dependencies, you can try installing building
> the source RPM.
>

Martin,

I'm waiting for 8.3 :).  Regarding the source RPM's, I tried installing it
with kpackage and with rpm and got the same dependency problems.  I guess
I'll have to use the vanilla tarball.

thx





From anthony at computronix.com  Fri Sep 13 12:55:21 2002
From: anthony at computronix.com (Anthony Tuininga)
Date: 13 Sep 2002 10:55:21 -0600
Subject: calling Oracle functions
In-Reply-To: <3d8210a3@news.mt.net.mk>
References: <3d81f811$1@news.mt.net.mk>
	<mailman.1031929598.17983.python-list@python.org>  <3d8210a3@news.mt.net.mk>
Message-ID: <1031936122.472.95.camel@chl0151.edmonton.computronix.com>

On Fri, 2002-09-13 at 10:22, ?????? ?. wrote:
> 
> > Oracle differentiates between functions and procedures so you must
> > identify that you are calling a function, not a procedure. The DB API
> > does not allow for this, so you must use an anonymous PL/SQL block and
> > some non DB API code. The way to do this is as follows:
> 
> ok. I also tried that, I mean a PL/SQL block... but
> 
> > Replace the data types as appropriate.
> 
> > v_Vars = v_Cursor.setinputsizes(p_Result = cx_Oracle.NUMBER)
> 
> didn't know about this :)) thanks.

You're welcome.

> ...
> > If anyone has suggestions about how to improve this, fire away... :-)
> 
> yes, put it in the documentation. 
> Actually is there any documentation about cx_Oracle, past the README
> file?

Bits and pieces in a variety of places. The assumption (to date) has
been that the DB API is sufficient for almost everything and that you
really shouldn't use extensions.... :-) However, over the couple of
years that I have been using the DB API I have discovered a fair number
of deficiencies with respect to Oracle, in particular when it comes to
performance and quirks of Oracle. It hasn't been a problem for me since
I wrote cx_Oracle, but it has been my desire to consolidate the DB API
document with descriptions of the extensions used by cx_Oracle -- I hope
to do something about this in the next month or so.

> Now a simpler question:
> 	is it possible to specify the data source name directly not through
> 	the alias defined in tnsnames.ora. In perl DBD I can specify a
> 	connection string like cs = "dbi:Oracle:host=1.10.100.200;sid=TEST"?

cx_Oracle is passing the string directly to Oracle, so whatever it
accepts, cx_Oracle will accept. Specifically, you can take the entire
string that is in your tnsnames.ora file and pass it directly as in

v_TNS =
(DESCRIPTION=(ADDRESS_LIST=(ADDRESS=(PROTOCOL=TCP)(HOST=opus)(PORT=1521)))(CONNECT_DATA=(SID=DEV)))

v_Connection = cx_Oracle.connect("user", "pw", v_TNS)

and that works just fine. Does that answer your question?

-- 
Anthony Tuininga
anthony at computronix.com
 
Computronix
Distinctive Software. Real People.
Suite 200, 10216 - 124 Street NW
Edmonton, AB, Canada  T5N 4A3
Phone:	(780) 454-3700
Fax:	(780) 454-3838
http://www.computronix.com




From sholden at holdenweb.com  Wed Sep 11 21:39:42 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 11 Sep 2002 21:39:42 -0400
Subject: [FEEDBACK] Is this script efficient...is there a better way?
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <L9Rf9.168314$l_4.127784@atlpnn01.usenetserver.com> <4ZRf9.1782$Lg2.248733@news2.news.adelphia.net>
Message-ID: <3hSf9.168947$l_4.98525@atlpnn01.usenetserver.com>

"Bob X" <bobx at linuxmail.org> wrote in message
news:4ZRf9.1782$Lg2.248733 at news2.news.adelphia.net...
[ ... ]
> Thank you very much...as a follow up:
>
> #1 Can I use the same sys.exit("message here") on the IO errors as well?

Sure. I assumed you would generalize the first example to the rest ...

> #2 For the "print looks like a debugging statement"...how would it be
> done different...in the context of my script and not a total re-write?

I simply thought that since you were writing the output to the outFile
anyway it seems a bit redundant to be printing it. Plus the readlines()
includes the line ending, so you'd get double-spaced output fom the print
statements anyway.

> #3 I will test your shorter one as I have stayed away from RE till now.

Well, obviously there may be problems if the keywords contains any
characters special to regular expressions. Nut my example whould wqork for
testing purposes.

> #4 On your shorter one should xreadlines still be done for memory sake?

Yes. This is a general improvement.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From sam at localhost.localdomain  Wed Sep 11 17:32:10 2002
From: sam at localhost.localdomain (newbie)
Date: Wed, 11 Sep 2002 22:32:10 +0100
Subject: I must be an idiot, please pity me!
Message-ID: <VAOf9.2572$Pa7.155732@newsfep1-gui.server.ntli.net>

Hello all. I'm getting very desperate now, so I have to ask for some help.
I'm working on a small project and have slipped into what must be
namespace hell.

I have a few controlling CGI scripts that use functions from other modules
to build parts of a page. Each function in these other modules contains
the SQL processing and HTML building code grouped by related data. The
idea being I can reuse the functions in the other scripts. All works fine,
providing each modules connects to the database as a stand alone job, or I
pass a database connection object thing to each and every function. I
suspect that both of these are wrong, due to the ugliness or nasty system
overhead of the datbase connections.

What I thought I should be able to do is:

CGI1.py:

import mod1, mod2

connect to database

create part of page using mod1.functiona 
create part of page using mod1.functionb 
create part of page using mod2.functionz

close database


But it doesn't seem to work like that. I've trawled though google and deja
for days (plus a couple of new rider books), and no one seems to hit this
problem. I must have a HUGE misunderstanding somewhere. I know mod1 and
mod2 cannot see CGI.py namespace, but there must be a way of sharing the
database connection, mustn't there?

Help or a kick up the bottom would be really appreciated here!

Sammy.


From dwelch91 at nospam.attbi.com  Sat Sep 21 01:51:05 2002
From: dwelch91 at nospam.attbi.com (dwelch)
Date: Sat, 21 Sep 2002 05:51:05 GMT
Subject: PyQt setup on Linux Mandrake 8.2
Message-ID: <3D8B4898.20808@nospam.attbi.com>

I would like to setup PyQt 3.4 on my Mandrake 8.2 box. I am having 
difficulty, and I am worried that if I do the wrong thing, KDE will go 
belly-up on me since it is Qt based.

First, question is: why isn't QTDIR defined by default? Is QTDIR a 
runtime setting or a development setting?

So I set QTDIR to /usr/lib/qt2 (that seems to be where the libqt2 RPMs 
files are). Next question is: Since I have libqt2 already on my system, 
how do I add the Qt development pieces without screwing up Qt/KDE? Is 
/usr/lib/qt2 even the correct location?

When I run SIP 3.4 build.py, it complains about tmake not being 
available. Is tmake included with Qt2 development files? SIP doesn't 
seem to have any docs, so I'm not sure how to proceed.

Also, my RPM manager says that I have libqt2-devel (2.3.1-29mdk) 
installed. No tmake listed in the files, though. And, PyQt-devel 
(2.5-1mdk) is also listed as installed, but there aren't any .py files 
included, only .sip files (I found this curious).

Installing software on Linux makes me feel like such an idiot!

Any help or pointers would be much appreciated.

Don




From ken at hotmail.com  Fri Sep 20 20:41:43 2002
From: ken at hotmail.com (Ken)
Date: Sat, 21 Sep 2002 10:41:43 +1000
Subject: Python question on cgi
Message-ID: <amgf6b$5ob93$1@ID-49758.news.dfncis.de>

Hi all, how do I pass parameters to a backend C++ program and how do I know
when does the C++ program ends so that I can read the output file produced
by the C++ program and print to the HTML interface?

Thanks




From info at voidhobo.de  Tue Sep 17 14:04:15 2002
From: info at voidhobo.de (John D. Boy)
Date: 17 Sep 2002 18:04:15 GMT
Subject: CGI question
Message-ID: <slrnaoerk7.mgo.jboy@otaku.freeshell.org>

Hello there!

I am working on a simple CGI script to use as a framework for a website
of mine. Basically, I want it to work as follows: When somebody accesses
the site, a CGI (called e.g. index.py) is run. It takes an html-template
and fills it with content from another file, called e.g. index.dat, and
returns the completed html-document. When I want to present another
page, I want the url to look like this: http://my.domain.name/?page. In
that case, index.py would fill the same template with the content of a
file called e.g. page.dat. 

Now, my problem is to find out what follows behind the question mark in
my CGI script. Unless the key has a value
(http://my.domain.name/?key=value), it is not in the cgi.FieldStorage
dictionary. Is it simply impossible, or am I not using the right
function calls? 

Any help is appreciated.
Thanks,
John



-- 
John D. Boy                               "Information Superhighway" is 
email      <mailto:info at voidhobo.de>       just an anagram for "I'm on 
hypertext  <http://jboy.chaosnet.org>       a huge wispy rhino fart."


From tjreedy at udel.edu  Wed Sep 25 00:45:38 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 25 Sep 2002 04:45:38 GMT
Subject: Getting started
References: <mailman.1032917851.23864.python-list@python.org>
Message-ID: <Sbbk9.262614$z91.11379176@bin3.nnrp.aus1.giganews.com>

"Adam Taylor" <ataylor at ucsd.edu> wrote in message
news:mailman.1032917851.23864.python-list at python.org...
> Based on this definition, it seems to me that Python _is_ untyped,
> since the type of a variable can change arbitrarily in mid-program.

Python types objects and their values rather than names and the fixed
chunks of memory that names usually refer to in name-typed languages.
Given a definition of type that sees typeness as inhering only in
names rather than objects, then yes, Python is 'untyped'.  But how
useful is such a viewpoint?  It obscures a real difference between
data models.

If a language has nested scopes and allows the type of a variable to
be different in different scopes,  then the type associated with a
particular name can change arbitrarily in mid-program (at scope
changes) even if the name is strongly and statically typed within each
scope.  One can say that the name represents different variables in
the different contexts (scopes), but the net result is similar to what
Python does.  Ie, one can view the rebinding of a name to an object of
different type as implicitly defining a new scope with respect to that
name.

Terry J. Reedy




From phr-n2002b at NOSPAMnightsong.com  Fri Sep 20 18:42:53 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 20 Sep 2002 15:42:53 -0700
Subject: e vs exp()?
References: <3a838998.0209191933.1f22e977@posting.google.com> <amflhl$3eb$1@peabody.colorado.edu> <FQMi9.171268$AR1.7178378@bin2.nnrp.aus1.giganews.com>
Message-ID: <7xy99wqo42.fsf@ruckus.brouhaha.com>

"Terry Reedy" <tjreedy at udel.edu> writes:
> At least two of us expected x**y == pow(x,y) == exp(y*log(x)) (a
> standard method, and the only one I could find in Abrahamowitz and
> Stegun, Handbook of Mathematical Functions) but there must be a more
> direct method being used at least in some C libraries now.

I would expect that the C library just uses the machine instructions
for exp and pow, if they are available.  All x86 processors with an
FPU have these instructions.  It's actually possible that an AMD
Athlon could give slightly different results than an Intel Pentium.



From bdesth at nospam.free.fr  Sun Sep  8 21:54:14 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Sun, 08 Sep 2002 21:54:14 -0400
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <3D7958DB.50903@nospam.free.fr> <ald3g5$1p8lpp$1@ID-77524.news.dfncis.de>
Message-ID: <3D7BFF46.6020306@nospam.free.fr>

Thorsten Kampe wrote:
> * laotseu
> 
>>Rod Stephenson wrote:
>>
>>>Slashdot has a list of questions posed to Larry Wall (perl). When
>>>asked his thoughts on other scripting languages, he makes the
>>>following observation about python
>>>
>>>"Python is cool to look at small bits of, but I think the "outline"
>>>syntax breaks down with larger chunks of code. I'm with Aristotle on
>>>the structure of discourse--a story should have a beginning, and
>>>middle, and an end. So should blocks"
>>>[...]
>>
>>def fun(args):
>>#BEGIN
>>     if ceci or cela:
>>     #BEGIN
>>         do_this(with_that)
>>         [snip a lot of code]
>>     #END
>>#END
>>
>>So what the problem with blocks ?
> 
> 
> Nope. Bad, bad readability, sorry.
> 
> #||<--- LOOK HERE, LARRY, THIS IS WHERE THE FUNCTION _*BEGINS*_!
> def fun(args):
>     #||<--- LOOK HERE, LARRY, THIS IS WHERE THE IF-CLAUSE _*BEGINS*_!
>     if ceci or cela:
>         do_this(with_that)
>         [snip a lot of code]
>     #||<--- LOOK HERE, LARRY, THIS IS WHERE THE IF-CLAUSE _*ENDS*_!
> #||<--- LOOK HERE, LARRY, THIS IS WHERE THE FUNCTION _*ENDS*_!
> 
> 
> Thorsten

Ho my !

Isn't it a little bit too much ?-)

laotseu



From dummy1 at net-es.dk  Sun Sep  8 16:22:00 2002
From: dummy1 at net-es.dk (Per Jensen)
Date: Sun, 08 Sep 2002 22:22:00 +0200
Subject: [ANN] gallery2.py now useful
Message-ID: <pan.2002.09.08.20.22.00.533492.1999@net-es.dk>

gallery2.py makes static galleries from directories with pictures. It is 
not the final version, but beta 10 is very useful for me.

New features in version 2 are 'sourcedir' and 'destdir' options along with
extraction of EXIF timestamp.

'Date', 'Where', 'Who' and 'What' properties can be written in a xml file
and included on the html pages.

Feel free to check it out on http://www.net-es.dk/~pj/python

/Per


From peter at engcorp.com  Fri Sep 13 21:33:17 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 13 Sep 2002 21:33:17 -0400
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <3D7958DB.50903@nospam.free.fr> <mailman.1031964141.9990.python-list@python.org>
Message-ID: <3d8291cc@news.sentex.net>

Ken Seehof wrote:
> Definitely not.  Too wide by far.  You need to reduce the indent to one 
> space.
> 
> #||<--- LOOK HERE, LARRY, THIS IS WHERE THE FUNCTION _*BEGINS*_!
> def fun(args):
>  #||<--- LOOK HERE, LARRY, THIS IS WHERE THE IF-CLAUSE _*BEGINS*_!
>  if ceci or cela:
>   do_this(with_that)
>   [snip a lot of code]
>  #||<--- LOOK HERE, LARRY, THIS IS WHERE THE IF-CLAUSE _*ENDS*_!
> #||<--- LOOK HERE, LARRY, THIS IS WHERE THE FUNCTION _*ENDS*_!

Oh, I didn't recognize this until now.  It's COBOL!  Everything
with a leading # is executable code, right?  ;-)

-Peter



From david.abrahams at rcn.com  Fri Sep 27 18:59:15 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Fri, 27 Sep 2002 18:59:15 -0400
Subject: Is Boost Python version 2 ready?
References: <MvZk9.12124$Lg2.3584877@news2.news.adelphia.net>
Message-ID: <an2pt9$g5h$1@bob.news.rcn.net>


"Robert Oschler" <Oschler at earthlink.net> wrote in message
news:MvZk9.12124$Lg2.3584877 at news2.news.adelphia.net...

> Is Boost Python version 2 ready for primetime yet?

I'd say so, except for docs. We're not making any interface-breaking changes
at this point, so anything you write will continue to work for some time to
come. I'm finishing the last feature before release as we speak, and
everything's been passing all tests on a wide range of platforms for quite a
while. Docs will be finished over the next week or so.

If you want a reasonably stable version, you might try checking out the
RC_1_29_0 branch of Boost.

Also, the C++-sig is the best place to ask further questions
(http://www.python.org/sigs/c++-sig/)

--
-----------------------------------------------------------
           David Abrahams * Boost Consulting
dave at boost-consulting.com * http://www.boost-consulting.com




From python-list at hendel.net  Wed Sep 11 17:54:23 2002
From: python-list at hendel.net (Manuel Hendel)
Date: Wed, 11 Sep 2002 23:54:23 +0200
Subject: problem with lists
Message-ID: <20020911215423.GA30455@habana.easygolucky.de>

I want to create a list of domains, every domain only once.

So long I did the following, but it didn't work:

#!/usr/bin/env python
#
import sys, string, re

inputfile = open(sys.argv[-1], "r")

lines = inputfile.readlines()
lines = lines[3:]

for line in lines:
        line = string.strip(line)[1:-1]
        fields = string.split(line, "|")

        domains = []

        if not domains.index(fields[3:4]):
            domains.append(fields[3:4])

print domains

This brings the following error message:

:!./pop3create.py pop3data
[No write since last change]
Traceback (most recent call last):
File "./pop3create.py", line 22, in ?
if domains.index(fields[3:4]):
ValueError: list.index(x): x not in list

Can someone explain this to me?

Thanks,
    Manuel

-- 
Immature love says, "I love you because I need you." Mature love says, "I need 
you because I love you." 
-Erich Fromm 



From cnetzer at mail.arc.nasa.gov  Sat Sep 28 00:23:47 2002
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Fri, 27 Sep 2002 21:23:47 -0700
Subject: Help please with menubars
In-Reply-To: <3d9516b3$1_7@goliath.newsgroups.com>
References: <3d9516b3$1_7@goliath.newsgroups.com>
Message-ID: <200209280423.VAA20949@mail.arc.nasa.gov>

On Friday 27 September 2002 20:54, mongo57a at comcast.net wrote:
> Specifically: I have
> an "exit" item on the menubar and if the user clicks on "exit", the
> window should be closed. What I don't want is a pulldown with "exit"
> as a button option.

I'm assuming you are talking about Tkinter here:

If so, your 'menubar' is a Tk menu, with cascade buttons in it.  It is 
also possible to attach a "command" button to it:

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
import Tkinter
Tk = Tkinter

def output_func():
    print "T.I.S.M. is a fine Aussie band."
    return

root = Tk.Tk()  #This opens a Toplevel()

menubar = Tk.Menu( root )
menubar.add_command( label='T.I.S.M.', command=output_func )

root.configure( menu=menubar )   #Attach menubar to the Toplevel

root.mainloop()
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

And if that doesn't work for you, then use the 'bind' command on the 
button or widget that you want to trigger the exit, and look for 
"<ButtonPress-1>" events.  Attach that to an exit function.

Other window toolkits should have analogous methods of doing this.

P.S. In my version, I just reuse the root Toplevel; you can also attach 
a menubar at creation time.

-- 

Chad Netzer
cnetzer at mail.arc.nasa.gov



From ajs at ix.netcom.com  Mon Sep 16 20:20:18 2002
From: ajs at ix.netcom.com (Arthur)
Date: Mon, 16 Sep 2002 20:20:18 -0400
Subject: Not sure how to phrase this question
Message-ID: <000b01c25de0$010816c0$9865fea9@arthur>

I wrote -
>I'll certainly take a simple solution to my problem - though I think some
elucidation of the behavior in my original post might >be of some interest
to folks other than just myself, in any case.

Is the simple solution that the factory should be a function, not a class?

Art





From sholden at holdenweb.com  Sun Sep  8 14:52:12 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 8 Sep 2002 14:52:12 -0400
Subject: Larry Wall's comment on python...
References: <mailman.1031349792.2374.python-list@python.org>
Message-ID: <j1Ne9.42624$l_4.38685@atlpnn01.usenetserver.com>

<mmaddox at hcsmail.com> wrote ...
[ ... ]
> Having read Wall's comments on Slashdot, I think he is really just mildly
> complaining about ridiculously long page-widths in his editor, not really
> slashing Python seriously. He's just giving some props to his own baby at
> the expense of another. (He had to say something, right?) You must agree
> that Python can lead you to horizontal scrolling in your editor - the bane
> of user interface design.
>
Well, using a four-space indent helps a lot -- tabs are passe nowadays.
Further, lines can easily be broken and continued, which helps to make long
lines more manageable. lastly, of course, if your code has nine levels of
indentation it might well be time to refactor it for improved readbility.

> Yes, I am new here. Hi, everyone.

Hi, and welcome!

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                        pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From matt at mondoinfo.com  Wed Sep 25 14:58:54 2002
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Wed, 25 Sep 2002 18:58:54 GMT
Subject: combining mailbox and email modules
References: <NPck9.581867$UU1.104088@sccrnsc03>
Message-ID: <slrnap41re.1mh.matt@happy-hour.t-island.com>

On Wed, 25 Sep 2002 06:36:29 GMT, Bob Horvath <nospam at horvath.com>
wrote:

> I am having trouble figuring out how to get the mailbox and email
> modules to play together so that I can use email module functions on
> every mail in the mailbox.  Is there a way to feed a factory from
> the email module to mailbox module?

Yes, there is:

>>> import mailbox
>>> import email
>>> i=mailbox.Maildir("Mail/testmail",email.message_from_file)
>>> i.next()
<email.Message.Message instance at 0x812410c>
>>> i.next()
<email.Message.Message instance at 0x8124224>

Regards,
Matt


From huxley1 at westnet.com.au  Tue Sep 24 07:15:28 2002
From: huxley1 at westnet.com.au (Jeremy Rew)
Date: Tue, 24 Sep 2002 19:15:28 +0800
Subject: Copying Elements using PyXML
Message-ID: <5.1.1.6.0.20020924190151.00b96068@localhost>

I have 2 XML documents. I would like to copy the entire structure of one 
document into a child element of the other, but i cannot copy the root node 
across:
I get the following error from Zope:
Error Type: HierarchyRequestErr
Error Value: Node manipulation results in invalid parent/child relationship.

the structures are as follows:

<person>
         <name/>
         <address/>
         <phone/>
</person>

and would like to copy this into the following format:

<xformimport>
         <docID col="blah" id="blahblah"/>
         <document>
                 <person>
                         <name/>
                         <address/>
                         <phone/>
                 </person>
         </document>
</xformimport>

I have written all the code to build 'xformimport','docID' and 'document', 
but just need to copy 'person' across.
I figure i may be missing an easy solution here, so any help/ideas will be 
appreciated.
I am using PyXML 0.8.1 and python 2.1.3.

thanks
Jeremy




From brueckd at tbye.com  Wed Sep 18 12:31:41 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 18 Sep 2002 09:31:41 -0700 (PDT)
Subject: Earing money with Python software (was: More on Protecting Source
 Code)
In-Reply-To: <nkem5-fei.ln1@junker.stroeder.com>
Message-ID: <Pine.LNX.4.44.0209180912110.32336-100000@ns0.tbye.com>

On Wed, 18 Sep 2002, Michael Str?der wrote:

> brueckd at tbye.com wrote:
> > free-for-personal-use-but-companies-must-pay approach when possible.
> 
> The problem with getting some bucks from a company is the amount of time 
> spent during decision taking and ordering process. For you to earn 300 
> bucks for your software the company spends another 500+ bucks for 
> working time until the money gets out.

Well, it really varies by company of course. Most places I've worked only 
require that much bureaucracy for larger purchases. For small stuff (under 
a few hundred dollars) it usually amounts to me informing a superior that 
tool X is needed and worthwhile. After that I just buy it myself and turn 
in a simple expense report.

> IMHO that's why it's so hard to 
> sell small packages of rather cheap software to big companies.

That's why you don't sell it for $300. Seriously: if it's a small and
cheap package then $300 (I'm assuming US$300) is often too high for small
businesses and many individuals anyway, and for large companies it might
be expensive enough to trigger more purchasing overhead (at my current job
that price could go either way). So... you sell your product with a single
seat license for, say, $100 and then offer a site license for e.g. $500
as well as a renewable support package for $1000 a year or whatever. You 
also give 'em discounts when you release a new version next year.

-Dave




From bob at eleaston.com  Fri Sep  6 20:50:45 2002
From: bob at eleaston.com (Bob Easton)
Date: Fri, 6 Sep 2002 20:50:45 -0400
Subject: Interrupting Python
References: <mailman.1031345649.9840.python-list@python.org> <Ef9e9.7309$rU2.218110@news4.srv.hcvlny.cv.net> <albel2$7te$0@216.39.172.122>
Message-ID: <R3ce9.9745$rU2.313350@news4.srv.hcvlny.cv.net>

OS = Win2000
Python = 2.2
Python execution environment = PythonWin

Using catchCtlC.py in a command line environment works as you expect. When I
hit Ctrl-Break, Windows displays "^C" and then a command prompt.  If I hit
Ctrl-C, the exception handler catches it and displays the results as
specified in catchCtlC/py

However, my preferred execution environment is PythonWin, not a command
prompt. There, neither Ctrl-break, not Ctrl-C have any effect.  That could
easily be a bug with PythonWin.

--
Bob Easton


"Bengt Richter" <bokr at oz.net> wrote in message
news:albel2$7te$0 at 216.39.172.122...
> On Fri, 6 Sep 2002 17:38:55 -0400, "Bob Easton" <bob at eleaston.com> wrote:
>
> >Thanks to all.  My system is Win2K. Python simply ignores Ctrl-c and
> >Ctrl-break. I think I'll fall back to using file existence as a simple
> If this were true as generally as you put it, we should be looking for a
bug.
>
> Would you put the following listed catchCtlC.py file in a convenient
directory,
> and then go there with **a plain "DOS" command window**, and type the
command
>
>     catchCtlC.py 1000
>
> It should complete in a short time. Then try it again, hitting Ctrl-C
before
> it finishes. If you have a fast machine and need more time, add another
zero
> to the 1000. Let us know if no interrupt is reported for either Ctrl-C or
Ctrl-Break.
>
> Then tell us what's different about your program.
>
> ---< catchCtlC.py >------------------
> i=x=0
> def foo(n):
>    global i,x
>    for i in xrange(n):
>        x = i**100
>
> if __name__ == '__main__':
>     import sys
>     try:
>         foo(int(sys.argv[1]))  # call your thing here instead
>     except KeyboardInterrupt:
>         print 'Interrupted by Ctrl-C'
>     else:
>         print 'Exited w/o interrupt'
>     print 'FHOI, the last i**100 x value: %s**100 = %s' % (i, x)
> ---< catchCtlC.py >------------------
>
> >semaphore.  That will be good enough and cheap enough for this
application.
> >
>
> >--
> >Bob Easton
> >
> >
> >"David LeBlanc" <whisper at oz.net> wrote in message
> >news:mailman.1031345649.9840.python-list at python.org...
> >> > I have a script that can run, accessing network resources, for
> >> > several days.
> >> > Since the script does not normally need keyed input, exception
> >processing
> >> > does not raise the keyboard exception until after the program
> >> > ends normally.
> >> > I would like to be able to interrupt if from the keyboard, but have
not
> >> > learned the trick.  How can I do this?
> >> >
> >> > --
> >> > Bob Easton
> >>
> >> I have had the experience of Python progs not responding to ctrl-c or
> >ctrl-d
> >> on Windows, but I found that ctrl-break works every time.
> >>
> >> Dave LeBlanc
> >>
> >>
> Also, please post responses below what you are responding to ;-)
>
> Regards,
> Bengt Richter




From hst at empolis.co.uk  Fri Sep 13 12:47:11 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Fri, 13 Sep 2002 17:47:11 +0100
Subject: Help with regular expression using findall and .*?
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB04F417@hendrix.empolisuk.com>

czrpb wrote
> 
> Harvey:
> 
> Great thanks!! And thanks for sticking to my question's 
> requirements. <wink!>
> 
> Ok, this is what we thought around here. But what I do not 
> understand is why any backtracking data is being kept? The 
> '?' in '.*?' means it is non-greedy right? When would 
> backtracking ever occur using '.*?'? What am I missing?
> 
> <<q
> 
> 
> On Fri, 13 Sep 2002, Harvey Thomas wrote:
> 
> > czrpb wrote
> > > 
> > > Could anyone help out with rewriting (still using regular 
> expressions)
> > > the following so that it does not cause an exception:
> > > 
> > > import re
> > > 
> > > s1=('macro\n'+'a'*200+'\norcam\n')*10
> > > s2=('macro\n'+'a'*20000+'\norcam\n')*10
> > > 
> > > p=re.compile(r'macro.*?orcam',re.DOTALL)
> > > 
> > > for x in re.findall(p,s1):
> > >     print x
> > > 
> > > for x in re.findall(p,s2):
> > >     print x
> > > 
> > > thanks!! Quentin Crain
> > > 
> > 
> > You need to be very careful about using .*? as the engine 
> "only" allows 10,000 backtracks
> > 
> > Try this
> > 
> > p = re.compile('macro(?:[^o]+|o(?!rcam))*orcam')
> > for x in p.findall(s2):
> >     print x
> > 
> > HTH
> > 
> > Harvey
> > 

Don't think I, or even Mr. Friedl can explain this concisely. If you are interested in knowing about the mechanics of REs, the best reference book is the 2nd edition of "Mastering Regular Expressions" by Jeffrey Friedl ISBN 0-596-00289-0.

A good tip though is to try and use negated character classes whenever possible rather than use "."

Harvey

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.



From erno-news at erno.iki.fi  Tue Sep 10 07:52:52 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 10 Sep 2002 14:52:52 +0300
Subject: UNIX processes
References: <mailman.1031654172.15956.python-list@python.org>
Message-ID: <kuhegycb8r.fsf@lasipalatsi.fi>

In article <mailman.1031654172.15956.python-list at python.org>, "Bo
M. Maryniuck" <b.maryniuk at forbis.lt> writes:

| Hi, all.
| Somebody know how to list all the processes in RIGHT way on Linux/UNIX?
| I need to find the zombies, stopped processes etc.

use the ps program (ps -e works on linux). there is no other
standard api for it.

  -- erno



From steven_shaw at adc.com  Wed Sep 11 19:54:00 2002
From: steven_shaw at adc.com (Steven Shaw)
Date: 11 Sep 2002 16:54:00 -0700
Subject: dynamism
References: <66715c02.0209091923.265fe77e@posting.google.com> <Xns92855D48384F9duncanrcpcouk@127.0.0.1> <3d7e209a_1@news.iprimus.com.au> <mailman.1031680455.27705.python-list@python.org>
Message-ID: <66715c02.0209111554.48aeedc8@posting.google.com>

Michael Chermside <mcherm at destiny.com> wrote in message news:<mailman.1031680455.27705.python-list at python.org>...
> Steven Shaw wrote:
> >>An object of a builtin type, or with __slots__ defined in the class and
> >>all base classes does not have a __dict__ attribute.
> > 
> > 
> > Does this mean that the attributes are stored in a more efficient way?
> > 
> > Is there a way to implement "record-based" object layout rather than
> > "associative-array based" object layout?
> 
> In short, YES.

Excellent.

> 
> See http://python.org/doc/2.2.1/whatsnew/sect-rellinks.html . 
> Unfortunately, http://www.python.org/peps/pep-0253.html doesn't give 
> full details since they haven't been written up yet.

Thanks for the links.


From bslesins at best.com  Mon Sep  9 01:42:46 2002
From: bslesins at best.com (Brian Slesinsky)
Date: 8 Sep 2002 22:42:46 -0700
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com>
Message-ID: <bd751532.0209082142.54236307@posting.google.com>

Rod Stephenson <viking_kiwi at yahoo.com> wrote in message news:<wkr8g7rn4c.fsf at yahoo.com>...
> Slashdot has a list of questions posed to Larry Wall (perl). When
> asked his thoughts on other scripting languages, he makes the
> following observation about python
> 
> "Python is cool to look at small bits of, but I think the "outline"
> syntax breaks down with larger chunks of code. I'm with Aristotle on
> the structure of discourse--a story should have a beginning, and
> middle, and an end. So should blocks"
> 
> I'm not quite sure what he's trying to get at here - I guess that for
> a long heavily indented chunk of code, you could lose track of the
> overall structure, but I don't write code this way.
> 
> Any comments?

He's repeating the standard newbie complaint.  However, I think the
fact that so many people have repeated this for so long is
significant.  Long-term Python programmers tend to forget how weird it
is to some newcomers to end a function or class just by *stopping*. 
It feels abrupt and rather fragile, particularly when you don't have a
firm grasp of the language.  If newbies have complained for this long,
the complaint's not going away, and if we're serious about making
Python easy to pick up it might be good to listen.

I've long thought that Python would be improved by an optional "end"
statement.  Example:

class Foo:

    def bar(self, x):
        # ...

    # (imagine 20 more short methods definitions here)

end Foo
# or "end class Foo", or "end class", or, just "end"

Of course it's unnecessary and should be entirely optional.  Putting
an "end" after every block would certainly be overly verbose and bad
style.  But ending major blocks in a way that's syntax-checkable might
catch the occasional indentation error and/or make error messages
clearer.  The main drawback is another reserved word, but it's one
that's reserved in several other programming languages.

I also think there would be a certain satisfaction in having a proper
ending, even for folks who are entirely comfortable with Python's
indentation.  It's like "Q.E.D." at the end of a proof, or "The End"
at the end of children's book, or signing a message even though the
"From" header line makes it entirely redundant.

- Brian Slesinsky


From hst at empolis.co.uk  Mon Sep  2 07:06:08 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Mon, 2 Sep 2002 12:06:08 +0100
Subject: Making regex suck less
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220CF9@hendrix.empolisuk.com>

John La Rooy wrote
> 
> Carl Banks wrote:
> > Gerhard H?ring wrote:
> > 
> >>>which means the real time is not spent in the compile() 
> function, but
> >>>in the match or find function. So basically, couldn't one 
> come up with
> >>>a *human readable* syntax for re, and compile that instead?
> >>
> >>That's equally powerful? Most probably not.
> > 
> > 
> > Why not?  It won't be as fast, but it should be able to do 
> anything a
> > regexp can do, and would be much more versatile.
> > 
> >
> I think the main problem is that *human readable* doesn't map really 
> well onto regular expressions.
> 
> What would the equivalent of r"(.)(.)(.)\3\2\1"
> This means a "palindrome of 6 characters"
> But it is unlikely that the human readable processor would understand 
> that (isn't it??)
> 
> It would be more likely to look like this (I haven't put too much 
> thought into this)
> "anything,anything,anything,same_as_3rd,same_as_2nd,same_as_1st"
> or would you like to suggest something else?
> 
> palindrome_6 = re.compile(r"(.)(.)(.)\3\2\1")
> palindrome_6 = 
> re.compile("anything,anything,anything,same_as_3rd,same_as_2nd
> ,same_as_1st")
> 
> Sure there are some cases where the re is loaded with meta 
> characters...
> 
> hmmm
> OK is this about writing maintainable code or people not wanting to 
> learn all the ins and outs of re's?
> 
> John

I used to use OmniMark a lot when it was free. With OmniMark's
equivalent of REs, the palindrome would be

any => char1 any => char2 any => char3  char3 char2 char1

I selected a non-trivial OmniMark RE from old code at random and came up with

('<!DOCTYPE' white-space+ [any except white-space]+ white-space+ "PUBLIC" white-space+ '"'
    upto-inc('"')) => a.whole (white-space+ "SYSTEM"? white-space* '"'
    upto-inc('"'))?

The (untested) Python RE equivalent is something like

"""(?P<a.whole><!DOCTYPE\s+[^\n]+\s+"PUBLIC"\s+"[^"]+")(?:\s+(?:"SYSTEM")?\s+"[^"]*")?"""

or, more readably if compiled with the re.VERBOSE flag

"""
(?P<a.whole>
   <!DOCTYPE
   \s+[^\n]+
   \s+"PUBLIC"
   \s+"[^"]+")
(?:\s+(?:"SYSTEM")?
\s+"[^"]*")?
"""


Which is the easiest to understand?

I'm used to REs, so I don't find the verbose Python RE too difficult to read. When I was learning
OmniMark, however, it was nice to be able to use "digit" and "letter" rather than
"\d" and [a-zA-Z] as creating non-trivial effective search patterns is never easy.


_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.



From jcoppens at usa.net  Tue Sep  3 20:48:39 2002
From: jcoppens at usa.net (John)
Date: 3 Sep 2002 17:48:39 -0700
Subject: Strange library problem
Message-ID: <c2ba8c6.0209031648.4984487e@posting.google.com>

Hi all.

I installed quite a few packages trying to upgrade to Gnome 2.0 and
got a
problem in another area. Tried to run eRoaster again, and the the
following
message:
-------
Traceback (most recent call last):
  File "/usr/local/bin/eroaster", line 9, in ?
    from main import Application
  File "/usr/local/lib/eroaster/main.py", line 49, in ?
    from gnome.ui import GnomeApp, GnomeFileEntry, GnomeErrorDialog,
GnomeAbout, GnomeOkDialog
  File "/opt/gnome/lib/python2.2/site-packages/gtk-1.2/gnome/ui.py",
line 3, in ?
    import _gnomeui
ImportError: /opt/gnome/lib/libgnome.so.32: undefined symbol:
libiconv_open
-------
I checked libgnome, and didn't find any reference to libiconv in it
(using ldd).
So, I wonder why is libgnome needing libiconv_open at all?

Sorry - this is maybe a gnome problem, but I didn't find any ref at
all to this
on the 'net, and this may have something to with the python config?

John

BTW, I separated the 'import _gnomeui, gnome' from ui.py in two lines
to
detect which was the one failing.


From e_fax_t at hotmail._ZAPME_.com  Thu Sep 26 16:36:19 2002
From: e_fax_t at hotmail._ZAPME_.com (Justin)
Date: Thu, 26 Sep 2002 22:36:19 +0200
Subject: readlines() and "binary" files
References: <3d90cacf$0$30457$afc38c87@sisyphus.news.be.easynet.net> <9r8k9.46422$V7.11170409@twister.socal.rr.com> <3d91de2d$0$30457$afc38c87@sisyphus.news.be.easynet.net> <9yqk9.143972$ub2.3161265@news1.tin.it>
Message-ID: <3d936f05$0$30456$afc38c87@sisyphus.news.be.easynet.net>

> Pity you ruled out defining your own iterator --

I had ruled it out by laziness, but since you were kind enough to
provide an implementation, I'll definitely use it! Thanks a lot,
it works nicely. Gee I can't wait for that nutshell to crack !-}

Regards,

Justin.




From unendliche at hanmail.net  Fri Sep 13 09:44:19 2002
From: unendliche at hanmail.net (Seo Sanghyeon)
Date: 13 Sep 2002 06:44:19 -0700
Subject: Python evaluator via Javascript
Message-ID: <45e6545c.0209130544.1664d78f@posting.google.com>

Python FAQ Entry
4.15. Is it possible to write obfuscated one-liners in Python?
(http://www.python.org/cgi-bin/faqw.py?req=show&file=faq04.015.htp)

As we know from the FAQ, Python is a nice language for writing
obfuscated one-liners. (wink)

Now you can test your one-liners on the web -- using 'Python
evaluator via Javascript'. Since it is for one-liners, you should
type a expression. No statement allowed.

Try the following: (taken from the FAQ)

# without newline
filter(None,map(lambda y:y*reduce(lambda x,y:x*y!=0,
map(lambda x,y=y:y%x,range(2,int(pow(y,0.5)+1))),1),range(2,1000)))

Mysteriously, it works. I couldn't believe it.

Copy the following 116 lines (2825 bytes) and execute. It took about
3 hours for me to write. Perhaps I have too much free time in my hands.

Tested on Win 98, Python 2.2.1, MSIE 6.0.

---- Python evaluator via Javascript ----

def prepare():
    import os
    if not os.path.exists('httpd'):
        os.mkdir('httpd')
    if not os.path.exists('httpd/cgi-bin'):
        os.mkdir('httpd/cgi-bin')
    fp = open('httpd/eval.htm', 'w')
    fp.write(dhtml)
    fp.close()
    fp = open('httpd/cgi-bin/eval.py', 'w')
    fp.write(pycgi)
    fp.close()

def server():
    import os
    os.chdir('httpd')
    from BaseHTTPServer import HTTPServer
    from CGIHTTPServer import CGIHTTPRequestHandler
    serv = HTTPServer(('', 1234), CGIHTTPRequestHandler)
    serv.serve_forever()

def main():
    import thread
    thread.start_new_thread(server, ())
    import time
    time.sleep(10)
    import webbrowser
    webbrowser.open('http://localhost:1234/eval.htm')

def idle():
    while 1:
        pass

dhtml = '''
<html>
<head>
<title>Python Evaluator via Javascript</title>
<script>
function run() {
  line = document.getElementById('expr').value;
  document.getElementById('console').value += '>>> '+line+'\\n';
  document.getElementById('expr').value = '';
  eline = escape(line).replace('+', '%2B');
  addr = 'cgi-bin/eval.py?expr='+eline;
  document.getElementById('cgi').src = addr;
}
function cls() {
  document.getElementById('console').value = '';
}
</script>
</head>
<body>
<div style="font-family: Lucida Console; position: absolute;
 left: 10px; top: 10px; width: 500px; height: 300px;">
 <textarea id="console" style="width: 500px; height: 300px;"
  disabled></textarea>
</div>
<div style="font-family: Lucida Console; position: absolute;
 left: 10px; top: 310px; width: 40px; height: 30px;">
 &gt;&gt;&gt;
</div>
<div style="position: absolute;
 left: 50px; top: 310px; width: 340px; height: 30px;">
 <input type="text" id="expr" style="width: 340px;">
</div>
<div style="position: absolute;
 left: 390px; top: 310px; width: 60px; height: 30px;">
 <input type="button" style="width: 60px;" value="Eval"
  onClick="run();">
</div>
<div style="position: absolute;
 left: 450px; top: 310px; width: 60px; height: 30px;">
 <input type="button" style="width: 60px;" value="Clear"
  onClick="cls();">
</div>
<iframe id="cgi" width="0" height="0">
</body>
</html>
'''

pycgi = '''
import cgi
form = cgi.FieldStorage()

from math import *

try:
    expr = form['expr'].value
    val = str(eval(expr))
except:
    val = ''

print 'Content-Type: text/html'
print
print \'\'\'
<html>
<head>
<title>Python Evaluator via Javascript</title>
<script language="JavaScript">
function run() {
  ret = document.getElementById('val').value;
  parent.document.getElementById('console').value += ret+'\\\\n';
}
</script>
</head>
<body onLoad="run();">
<input type="hidden" id="val" value="%s">
</body>
</html>
\'\'\' % val
'''

prepare()
main()
idle()

---- Python evaluator via Javascript ----

Quiz: Why line 102 has four backslashes?

-- Seo Sanghyeon


From agenkin at cdf.toronto.edu  Wed Sep  4 17:23:01 2002
From: agenkin at cdf.toronto.edu (Arcady Genkin)
Date: Wed, 4 Sep 2002 21:23:01 GMT
Subject: MD4 and DES for unicode strings (samba passwd file
 manipulation)
References: <uoit1mxubi.fsf@bashful.cdf.toronto.edu> <8pid9.5889$e5.1033526@newsb.telia.net>
Message-ID: <uod6rttpoq.fsf@bashful.cdf.toronto.edu>

martin at v.loewis.de (Martin v. Loewis) writes:

"Fredrik Lundh" <fredrik at pythonware.com> writes:

> if you want a string of 16-bit little-endian unicode characters,
> make sure you have one:
>
>     m = MD4.new()
>     m.update(password.encode("utf-16-le"))
>     m.hexdigest().upper()
>
> (I don't have the MD4 library, so I cannot verify this.  But if
> it works, it works for both 8-bit strings with ASCII text, and
> Unicode strings)

Yep, this works.  Thank you both very much!
-- 
Arcady Genkin : CDF sysadmin : http://cdf.toronto.edu/~agenkin/contact.html

guilt is the cause of more disauders
than history's most obscene marorders                       (E.E. Cummings)


From mhammond at skippinet.com.au  Wed Sep 18 08:55:28 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 18 Sep 2002 12:55:28 GMT
Subject: COM object registration
References: <afph9.9468$yt3.4669302@newssrv26.news.prodigy.com> <T_Rh9.17227$Ee4.51953@news-server.bigpond.net.au> <uCTh9.10343$yt3.4866032@newssrv26.news.prodigy.com>
Message-ID: <4J_h9.19439$Ee4.59690@news-server.bigpond.net.au>

dsavitsk wrote:
>>>>a = win32com.client.Dispatch('ecp.MODULE_REGISTER2')
>>>
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
>   File "C:\Python\win32com\client\__init__.py", line 92, in Dispatch
>     dispatch, userName =
> dynamic._GetGoodDispatchAndUserName(dispatch,userName,clsctx)
>   File "C:\Python\win32com\client\dynamic.py", line 81, in
> _GetGoodDispatchAndUserName
>     return (_GetGoodDispatch(IDispatch, clsctx), userName)
>   File "C:\Python\win32com\client\dynamic.py", line 72, in _GetGoodDispatch
>     IDispatch = pythoncom.CoCreateInstance(IDispatch, None, clsctx,
> pythoncom.IID_IDispatch)
> com_error: (-2147467259, 'Unspecified error', None, None)

It is likely that an exception is being encountered that we can get at 
pretty easy.

Add a named param "debug=1" to all UseCommandLine calls.  Then, on both 
machines, start Pythonwin and select the debug trace collector.  Rerun 
your tests and see if any additional interesting exceptions are reported.

Mark.



From jeff at sasmor.com  Sun Sep 15 10:04:39 2002
From: jeff at sasmor.com (Jeff Sasmor)
Date: Sun, 15 Sep 2002 14:04:39 GMT
Subject: Python11
References: <xwMg9.1433$Dy4.464@nwrddc03.gnilink.net> <m3admk47pf.fsf@mira.informatik.hu-berlin.de>
Message-ID: <Xr0h9.2585$Dy4.200@nwrddc03.gnilink.net>

I suppose in some sense it would produce
some greater exposure for Python; esp. given
the larger marketing exposure provided by
O'Reilly than would be possible for a
largely volunteer organization that had seemed
to be the direction for Python 11. One hopes that
Python11 won't be lost in the much larger OSCON..

--
#--------------------------------
Jeff Sasmor
jeff at sasmor.com
"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m3admk47pf.fsf at mira.informatik.hu-berlin.de...
> "Jeff Sasmor" <jeff at sasmor.com> writes:
>
> > Is there any definitive information on the next Python conference,
> > Python 11?  I've read the conferences-discuss archive on
> > Python.org (http://mail.python.org/pipermail/conferences-discuss/)
> > and the discussion seems to be tapering off.  IIRC, by last year
> > at this time there was already a call for papers.
>
> Python11 will be colocated with the O'Reilly Open Source Convention
> (OSCON), which will take place July 7-11, 2003 in Portland, USA.
>
> I believe O'Reilly will organize the specific schedule, including
> setting dates for the call for papers, submission deadlines, etc.
> Given that this is still ten months away, I don't see any urgency.
>
> HTH,
> Martin




From phr-n2002b at NOSPAMnightsong.com  Wed Sep  4 17:07:22 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 04 Sep 2002 14:07:22 -0700
Subject: encryption (passwords)
References: <3d75d115$0$221$4d4ebb8e@news.nl.uu.net> <7xd6ru2gfi.fsf@ruckus.brouhaha.com> <3d75f0f0$0$230$4d4ebb8e@news.nl.uu.net>
Message-ID: <7xsn0pzcol.fsf@ruckus.brouhaha.com>

"Gumuz" <gumuz at looze.net> writes:
> although this is just a play-learn-project and security is not really an
> issue, i'd like to include encryption in it as well to get experience in
> this, too.

If you want to learn programming cryptography, the book you want is
"Applied Cryptography" by Bruce Schneier.  You might like to implement
everything yourself rather than using a library.

> Actually, I am trying to create a sort of simple instant messenger
> server+client.

I think this should use Diffie-Hellman key exchange to choose the
decryption keys.  DH is a way for two people to agree on a key based
on random parameters they each choose.  The traditional version goes:

   Alice and Bob agree beforehand on public parameters g and P,
   where P is a big prime number (like 300 digits) with certain
   properties and g is a generator of Z//p.  If you don't understand
   that, don't worry too much--there are standard values of g and P
   that you can use.

   To start talking to each other, Alice chooses a secret random number x,
   and Bob chooses a secret random number y.  They do not reveal these
   numbers to anyone.  Alice instead computes the number X = g**x mod P
   and sends X to Bob.  Bob computes Y = g**y mod P and sends Y to Alice.
   Computing these modular exponentials is trivial in python because
   of Python's built-in long integers and its 3-argument pow function.
   You can just say Y = pow(g,y,P).

   Since Alice receives Y from Bob and knows x already, she can compute
   K = Y**x mod P.  Notice Y**x mod P == (g**y)**x mod P = g**(yx) mod P.
   Bob likewise can compute K = X**y mod P = (g**x)**y mod P = g**(xy) mod P.
   Since multiplication is commutative, xy==yx so both have found the same K.
   The coolness here is the computation required knowing at least one of
   the secrets, x or y.  An eavesdropper knowing neither secret doesn't
   have any easy way to find K.

   At the end of the conversation, Alice and Bob should both erase their
   secret values x, y, and K from computer memory.  That means the
   conversation can never be recovered by a third party, even by forcing
   Alice and Bob to reveal their passwords, turn over their computers, etc.
   The keys are gone forever, like burning a document.

Note you will also have to protect against "man in the middle"
attacks, and deal with some other subtleties, to make DH secure--the
description above is just to sketch the process.  See Applied
Cryptography for more info.


From sabren at manifestation.com  Thu Sep 26 00:50:53 2002
From: sabren at manifestation.com (Michal Wallace)
Date: Thu, 26 Sep 2002 00:50:53 -0400 (EDT)
Subject: revenge of the python puzzle
Message-ID: <Pine.LNX.4.44.0209260039440.18996-100000@hydrogen.sabren.com>


Okay, so you guys think you're so smart. Try this one!  No
contest this time, I've learned my lesson.  (I suppose if
someone really wants some free hosting bad enough to solve
this puzzle, you could talk me into it, but I'm just posting
this for revenge... Um, I mean "fun".)

If you decode it, see if you can guess what it does before
you run it. Please don't spoil the output for anyone else
who might want to try it... Hint: the earlier thread might
have a clue. :)



wbjig_ucpxmy, kxzp_xzli_h_fmyi \
    ='ilhg.zuuiyf(ghsp(izwxmyi))',\
"zkvjg wmfi hs sm ajwx vjy"
hv 'h za sazbp': 'hp djsp ahqxp bjy'
"hp's sm ajwx kmbsi pxzy 'xiggm kmbgf'"
"kxc, pxisi yitp ghyis zgamsp gmmo ghoi uibg!"

'kxhgi(<>){wxmau;s/uibg/ucpxmy;ubhyp}';mbzyqis='vbjhp';
ezf="'iyhpgzlm bjmc oyhbf'=qsa;qsa gzemgqtqsa pyhbu";
iaupc="";qgmezg izwxmyi;ilhg=[];

'ejp z siwbip xibi hs hy cmjb bizwx'
zuugis=mbzyqis, "izwxmyi=izwx"
fiv fm(pzso="fm kxzp h zso"):
     qgmezg izwxmyi; itiw pzso
"zyf djsp hy wzsi cmj'bi ymp uibugitif,"
kmbsi,ymk = (ezf.sughp('t'),"!")
vmb izwx hy kmbsi:
    itiw zuugis[1]
    hv izwx hy kmbsi:
        fm(wbjig_ucpxmy)

kmbsi = ilhg; "ilhg".sughp().bilibsi()
[izwx.bilibsi() vmb izwx hy kmbsi]
"zyf ymk ki'li bizwxif pxi vhyzg libsi";
azu(fm, azu(iaupc.dmhy, kmbsi))



Cheers,

- Michal   http://www.sabren.net/   sabren at manifestation.com 
------------------------------------------------------------
Switch to Cornerhost!             http://www.cornerhost.com/
 Low Priced, Reliable Blog Hosting, With a Human Touch. :)
------------------------------------------------------------




From mark at freelance-developer.com  Wed Sep 18 19:14:26 2002
From: mark at freelance-developer.com (Mark Nenadov)
Date: Wed, 18 Sep 2002 19:14:26 -0400
Subject: Need advice on choosing the technology for intranet based application
References: <3d88fd1e@news.ivm.net>
Message-ID: <20020918.191423.706043324.1365@x1-6-00-d0-09-7e-c7-3e.wido3.on.cogeco.ca>

In article <3d88fd1e at news.ivm.net>, "Nenad Propadovic"
<propadovic.nenad at debitel.net> wrote:


> Hello pythonistas,
> I'm about to write an application related to injury prevention in big
> industrial companies. It's supposed to be intranet-based (because the
> admins are reluctant to install it on x (where x>>1) clients). So: the
> client is a web-browser. I still don't know whether I will get a server
> on my own, or I'll have to fit into an unknown web server environment,
> both options being realistic.

Ah, yes. Isn't it fun not knowing where an application will reside? :)

> By now it seems the logic of the application will be simple. It's going
> to be a relational database with several tables, a modest number of
> users with write priviledges, some more with read priviledges only.

Sound quite simple.

> Performance is not a big issue.
> The GUI has a part I consider complex: I have to present some statistics
> in form of charts or graphs.

Maybee I should recant my last statement! :) The web can be a nightmare for
complex user interfaces.

> web/application server? I'm considering Zope (seams rather BIG when you
> download & unpack it, and it seams all possible web applications have
> been written for it) and twisted (smaller, seams lower lewel and
> somewhat easier to learn), and I'm wondering if there is something else
> out there, which is both powerfull and flexible and easy to learn and

Zope would be a good choice. Personally, though, I would stick to a
combination of Apache and Python/PHP/Perl/etc for a project like this.
Your project will be fairly small (although the GUI and chart generation
may be somewhat complex). It can't be too huge if you plan to get it done
by November in your spare time! :)  I think Python (with Apache) would be
a good choice.

One thing, you didn't mention was what languages you have used (you 
mentioned what you wrote, but not what you wrote it in). If you mentioned
what languages you have worked with in the past could, that could help 
people give you better suggestions.

> -I assume the simplest way to present a graph/chart kind of graphic is
> to create a jpg/giff/tiff on the server side, and just show it on the
> client? Thanks to everybody taking the time to read the message, and
> even more to those who answer it :-)

It really depends. You may want to generate them as images.  Or maybee
PDF's? Would PDF be a viable option for you? ReportLAB 
(www.reportlab.com) has a very good PDF generating library for Python. I've
worked with the library and it is amazing.

~Mark (http://www.freelance-developer.com)


From Oschler at earthlink.net  Sat Sep 28 22:13:00 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sun, 29 Sep 2002 02:13:00 GMT
Subject: Fast-track to installing/implementing a SWIP module?
Message-ID: <Mktl9.13603$Lg2.3951792@news2.news.adelphia.net>

I am a Python newbie so bear with me here.  I just downloaded the PyProlog
package, which allows Python to access SWI-Prolog through a C-code SWIP
module.  I just need the basics such as, proper directory to install files
and instructions for making sure the SWIP module can find SWI-Prolog, and
for making sure the package will run properly in Python. Is there a URL or
doc that can get me started?  I would assume that I have to compile the 'C'
source for the SWIP module, and the resultant object file (.so?) in a proper
directory?  These the kinds of questions I need to get answered.  I read the
3 README's in the PyProlog package but nothing seemed to suffice.

thx





From wlfraed at ix.netcom.com  Sun Sep 29 23:17:02 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Sun, 29 Sep 2002 20:17:02 -0700
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <mailman.1032962673.23086.python-list@python.org> <an7bvd$burft$1@ID-125932.news.dfncis.de> <an7p7t$cpg$0@216.39.172.122> <mailman.1033338759.6775.python-list@python.org>
Message-ID: <enf8na.86m.ln@ix.netcom.com>

Chris Gonnerman fed this fish to the penguins on Sunday 29 September 
2002 03:33 pm:


> You ask how it is a different problem.  Simple.  A human,
> training in decimal math in grade school, knows that
> 
>     1/3 = 0.3333333...
> 
> and would *expect* it to round to 0.33 (still in two decimal

        So, in 2 decimal place arithmetic

        a = 1.00 / 3.00
        b = a * 3.00

gives
        b = 0.99

which is NOT equal to 1... But this is okay because it's done is 
"decimal arithmetic"? <G>

        The telling phrase is "training in DECIMAL math"... Most computers 
don't work in decimal math, so the user /should/ be ready to learn the 
ins&outs of binary floating point...

        For every example of "what is wrong with binary floating point" 
someone else can find a similar case when using decimal math. Even VB's 
"currency" type (which looks to be a scaled 64-bit integer) carries 4 
decimal places -- which allows for "proper" results when displaying 
final results in two decimal places with rounding.

        1/3 => 0.3333   internally
        3*0.3333 => 0.9999
        1.00 for display after rounding.


--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From duncan at NOSPAMrcp.co.uk  Thu Sep 12 10:57:51 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Thu, 12 Sep 2002 14:57:51 +0000 (UTC)
Subject: Is there a portable way to copy files in Python?
References: <3D80A78B.5030206@impathnetworks.com>
Message-ID: <Xns9287A1A862C3Aduncanrcpcouk@127.0.0.1>

Pierre Rouleau <prouleau at impathnetworks.com> wrote in 
news:3D80A78B.5030206 at impathnetworks.com:

> I could write code that identifies the operating system (the os.path 
> way) and use os.system("cp " + source_filename + " " + destination) for 
> Unix-type and os.system("copy " + source_filename + " " + destination) 
> but i would prefer to use a simple call.
> 
> Is this available in the Python library?  If not, why is it not part of 
> the library?

Try the shutil module.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From bokr at oz.net  Tue Sep  3 19:45:17 2002
From: bokr at oz.net (Bengt Richter)
Date: 3 Sep 2002 23:45:17 GMT
Subject: trailing newline in string.splitfields()
References: <Gv3d9.287721$me6.36853@sccrnsc01>
Message-ID: <al3hid$ipp$0@216.39.172.122>

On Tue, 03 Sep 2002 14:17:10 GMT, Micah Mayo <astrophels at yahoo.com> wrote:

>I was splitting up some fields in a textfile using string.split, which 
>works fine, but that it usee ' ' as the delimiter. I've found that some 
>of the things I want to split include a space or two, so I've gone to 
>using string.splitfields, which also works fine, but for one thing.
AFAICT string.splitfields is now an alias for string.split. The docs say
"""
splitfields(s[, sep[, maxsplit]]) 
      This function behaves identically to split(). (In the past, split() was only used
      with one argument, while splitfields() was only used with two arguments.) 
"""

>Unlike string.split() it also grabs the newlne('\n') at the end of each 
>line.
>
>The way I've worked around this is as such:
>
>var1,var2,var3 = string.splitfields(stringtosplit,',')
>list.append(var1)
>list.append(var2)
>list.append(var3[:-1])
 ^^^^--you really ought not to use builtin names for variables ;-)

If you are not otherwise using the three var's, you could write

    myList.extend(stringtosplit.strip().split(','))

if you don't care about leading or trailing whitespace. Otherwise, to duplicate
what you are doing, you could write

    myList.extend(stringtosplit[:-1].split(','))

e.g.,

 >>> myList = list('abc')
 >>> myList
 ['a', 'b', 'c']
 >>> stringtosplit = '  <-2sp..1sp-> ,no_spaces,inside spaces and end.\n'
 >>> myList.extend(stringtosplit[:-1].split(','))
 >>> myList
 ['a', 'b', 'c', '  <-2sp..1sp-> ', 'no_spaces', 'inside spaces and end.']

>
>So, I slice of the character, easy enough. And I'm generally satisfied 
>with that - but I was wondering if there may be, for the cause of more 
>beautiful code, a way to tell string.splitfields() to do it automatically.
>
Regular expressions will let you do things like splitting on comma with
optional white space on either side of the comma, but not touching white
space elsewhere. Or ignore delimiters inside of quotations, etc. You might
not need those options every time, but sometimes you do.

Regards,
Bengt Richter


From peter at engcorp.com  Fri Sep  6 23:02:09 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 06 Sep 2002 23:02:09 -0400
Subject: Larry Wall's comment on python...
References: <mailman.1031349792.2374.python-list@python.org>
Message-ID: <3d796c18@news.sentex.net>

mmaddox at hcsmail.com wrote:
> Although I grant the Python structure and the column-based
> structures are not completely analogous, they BOTH burden the programmer
> with arbitrary rules to govern the physical layout of the code. Although
> this layout may be considered more elegant from a readability standpoint,
> it's the enforcement of the layout that is irritating to the more
> Libertarian of us.

I think you have the wrong meaning for "arbitrary".  From the
sounds of it, you have exactly reversed which is arbitrary (your
own "libertarian" coding style) and which is "governed by fixed
rules" (Python layout).

> Having read Wall's comments on Slashdot, I think he is really just mildly
> complaining about ridiculously long page-widths in his editor, not really
> slashing Python seriously. He's just giving some props to his own baby at
> the expense of another. (He had to say something, right?) You must agree
> that Python can lead you to horizontal scrolling in your editor - the bane
> of user interface design.

I agree with the comments about Larry "having" to say something like this.

I definitely don't understand how Python, more or less than any other
language, leads to more horizontal scrolling.  Pretty much all
Python code I've ever seen (though I haven't seen yours yet :-) is
roughly 80 characters max, and mine own has been about that since I
started writing C code 16 years ago.

Do you mean there is something special about Python not delimiting
blocks in the traditional ways which somehow leads you to write
much wider lines of code?

-Peter



From DennisR at dair.com  Wed Sep 18 10:41:40 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Wed, 18 Sep 2002 14:41:40 GMT
Subject: Python comment stripper
References: <16Nh9.248$Wc6.17087796@newssvr13.news.prodigy.com> <3d8807b1$1@news.sentex.net>
Message-ID: <Eg0i9.561$Y33.34967690@newssvr13.news.prodigy.com>

> This doesn't seem like a good reason to me.  Why not just use .pyc files
> all the time and be done with it?

I am looking to package Python programs as EXE files for upload/download
over the internet.  The "good reason" is faster upload/download over dialup
lines, not hard disk space.  Consider the situation at developer and
recipient systems:

On developer machine: .py files are copied to a staging directory and
comments stripped in the staging directory.  The original .py files are not
changed and so maintenance is not an issue.  The eventual output of process
is a single EXE compressed file, uploaded to internet..

On recipient machine: compressed EXE file is downloaded via internet.  When
run, the EXE contents are automatically unpacked to a temporary directory,
The temporary directory is deleted at conclusion of running the EXE.
Recipients will do nothing with the comments, especially since they are
bound inside the EXE.

See www.dair.com/hellopy.exe for an example (Windows platform only, sorry)
which has imported os.py.  The example is 361 Kb.  It could be smaller still
if the comments are stripped.  This matters more and more if library files
are included since the uncompressed size of the lib directory is about 1.9
Mb.

--

Dennis Reinhardt

DennisR at dair.com
http://www.dair.com




From jepler at unpythonic.net  Tue Sep 24 18:21:26 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 24 Sep 2002 17:21:26 -0500
Subject: Strange values in Accept-Language header?
In-Reply-To: <f72dac0d.0209231303.12349f2e@posting.google.com>
References: <f72dac0d.0209231303.12349f2e@posting.google.com>
Message-ID: <20020924172114.A938@unpythonic.net>

On Mon, Sep 23, 2002 at 02:03:02PM -0700, Alan Kennedy wrote:
> Dear all,
> 
> I know that this is not quite the correct place to post this question,
> but I'm not sure where else to go with it (suggestions? nice ones :-).
> And I know that there are plenty of web developers/masters here, and
> they might be interested.....
> 
> On looking at my server logs for the Accept-Language headers (to
> answer another post in this group), I came across some strange values
> that are definitely not ISO language codes. For example,
> 
> Accept-Language: en-us,x-ns1rDGeT4e2FpA,x-ns2f67971hgDw1
> Accept-Language: x-ns13V8cN8S1Xz9,x-ns2T329gfxKa7d
> 
> I have no idea what these could mean (I've changed the actual values
> BTW, on the remote off-chance that somebody's highly personal
> information was coded in there)

Searching on google turns up two hits in quoted mail message headers,
both with Mozilla 4.7x.
    http://www.google.com/search?q=accept-language+x-ns1

I wonder whether really ancient mozilla cvs might have some clues.  I
groveled around a bit in bonsai but didn't come up with anything.

Jeff



From robin at jessikat.fsnet.co.uk  Mon Sep 23 03:56:09 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Mon, 23 Sep 2002 08:56:09 +0100
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com>
 <mailman.1031377930.4225.python-list@python.org>
 <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com>
 <3D8E6353.7030303@something.invalid>
Message-ID: <2IYsVKAZksj9EwSH@jessikat.fsnet.co.uk>

In article <3D8E6353.7030303 at something.invalid>, Greg Ewing <see_reply_a
ddress at something.invalid> writes
......
>If you use spaces only, you're also choosing how other
>people will see it in their editors (unless they're
>willing to use a tool to re-indent it).
>
>Using *tabs* only is the only way to get truly
>preference-independent indentation.
>
here here, but I'm always wrong
-- 
Robin Becker


From wyojustin at hotmail.com  Thu Sep 26 20:52:00 2002
From: wyojustin at hotmail.com (Justin Shaw)
Date: Fri, 27 Sep 2002 00:52:00 GMT
Subject: This is driving me insane!
References: <mailman.1033084697.25598.python-list@python.org>
Message-ID: <QYNk9.312259$z91.13399844@bin3.nnrp.aus1.giganews.com>

Lists are mutable.  When you append to self.errorlog with self.grouptotals,
you have two references to the same list.  One in self.errorlog[0] and one
in self.grouptotals.  Therefore when you increment self.grouptotals in
grouptotalcounter() self.errorlog gets incremented too.  You can either use
tuples in self.errorlog or you can COPY self.grouptotals into self.errorlog
if you want independent references.


class Test:
    def __init__(self):
        self.errorlog = []
        pass

    def run(self):
        self.grouptotals = []
        for i in range(2):
            self.grouptotals.append(0)
        print self.grouptotals            # [0, 0]
        self.errorlog.append(self.grouptotals[:]) #  Note use of [:] to
shallow-copy list
        print self.grouptotals
        self.grouptotalcounter()
        return self.errorlog

    def grouptotalcounter(self):
        for i in range(2):
            self.grouptotals[i] += 1

"Derek Basch" <dbasch at yahoo.com> wrote in message
news:mailman.1033084697.25598.python-list at python.org...
> Hello,
> I created a very simple class that totals some data
> but I am not getting the output from it that I would
> expect at all.
> Here is the class:
> http://www.geocities.com/dbasch/insane.txt
>
> I would expect to get:
> [0, 0]
> [0, 0]
> [[0, 0]]
>
> but I am getting:
> [0, 0]
> [0, 0]
> [[1, 1]]
>
> I dont understand how the 'grouptotals' list is
> getting the incremented values prior to the
> 'grouptotalcounter' function being called.
> furthermore, I can place a 'print grouptotals' right
> after the errorlog append and it isnt the incremented
> values! Please tell me this isnt a normal thing!
> Thanks,
> Derek Basch
>
>
> __________________________________________________
> Do you Yahoo!?
> New DSL Internet Access from SBC & Yahoo!
> http://sbc.yahoo.com
>




From gmcm at hypernet.com  Thu Sep 12 08:30:51 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 12 Sep 2002 12:30:51 GMT
Subject: Mimicing intepreter import behaviour in embedded application
References: <uu1l0bfco.fsf@online.no>
Message-ID: <Xns928756AFCC67Bgmcmhypernetcom@199.171.54.214>

Syver Enstad wrote:

> 
> I have embedded python in a C++ application. Now I want python code
> running in this application to reach modules in the same way as code
> running in the standard interpreter. So I set up a test comparing the
> sys.path variable under python and my application with the intpreter
> embedded. 
> 
> Most of the two sys.path's are equal except:
> Regular python:                   Embedded:
>  ('d:\\devtools\\python22\\DLLs', 'D:\\devtools\\Python22\\Lib'),
>  ('d:\\devtools\\python22\\lib', 'D:\\devtools\\Python22\\DLLs'),
>  ('d:\\devtools\\python22\\lib\\lib-tk',
>   'D:\\devtools\\Python22\\Lib\\lib-tk'),
> This line is caused by the differing locations of the .exe files
>  ('d:\\devtools\\python22',
>  'h:\\mydocuments\\Kode\\C++\\Applications\\pyexe'), 
>  ('d:\\devtools\\python22\\lib\\site-packages\\Numeric', 
>   'D:\\devtools\\Python22'),
> Here I miss the .pth file that I have for PIL.
>  ('d:\\devtools\\python22\\lib\\site-packages\\PIL',
>   'D:\\devtools\\Python22\\lib\\site-packages\\Numeric')]
> 
> Is there a way to more closely mimic "the real interpreter"?

Usually when you embed you need to provide your own getpath.c
so you can set things up as you want them (& most embeddres
want the opposite - they don't want their app screwed up if
the user has a different Python version installed).

As for .pth files, they are processed in site.py. There's a
C flag for that. You may also need to have your own site.py
in your exe's directory.

-- Gordon
http://www.mcmillan-inc.com/


From dougfort at dougfort.net  Thu Sep 26 20:17:06 2002
From: dougfort at dougfort.net (dougfort)
Date: Thu, 26 Sep 2002 17:17:06 -0700 (PDT)
Subject: ANN: Tim Peters' FixedPoint on SourceForge
Message-ID: <20020926171707.18832.h014.c001.wm@mail.dougfort.net.criticalpath.net>

This project makes Tim Peters' FixedPoint math class available to the community
as a SourceForge project. 

Check out http://fixedpoint.sourceforge.net

Especially follow the link to the HTML rendering of the LatTex documentation. Joe
Griffin did the documentation, and I think it's really good.

There has been a vigorous discussion on this list about Python in monetary
applications. Here's your chance to follow up. We solicit bug reports, feature
requests, and general discussion.

What would it take to get you to use FixedPoint in a serious application....
 benchmarks?
 celebrity endorsements?
 sample code?
 other?
 
Please let us know. Use the discussion forum on SourceForge, or post it right
here.

Doug Fort <dougfort at dougfort.net>
http://www.dougfort.net



From rich.browne at hp.com  Wed Sep 11 12:26:20 2002
From: rich.browne at hp.com (Rich Browne)
Date: 11 Sep 2002 09:26:20 -0700
Subject: Building a python executable archive
Message-ID: <6c20157a.0209110826.3e98a77@posting.google.com>

I'm trying to build python as an archive bound executable.  

This sounds dumb, Python is inherently dynamic, so why am I doing
this?
Because the python executable will have to execute without any dynamic
dependencies on shared libs outside the build environment that I plan
to install it in.

I have succeeded in statically linking most of the shared objects in
lib-dynload, by modifing Modules/Setup.dist.  I have edited the
Makefile so that it will not build any shared objects, changed some
linker/compiler options to link archive and specified dynload_stub.o
and the dynamic loader.  I also undefined the preprocessor directives
HAVE_DYNAMIC_LOADER and DLOPEN.  This works, but I'm looking for an
easier(automated)way to do this.  Is there a configure option that
will allow me to do this more easily?  And is there anything that I'm
forgetting?  It seems to work fine.  All the formerly shared objects
are in the list of builtin_modules and I've used some of them, but I
haven't thoroughly tested this.

Also, what is the difference between the *.pyo and *.pyc files.


From loewis at informatik.hu-berlin.de  Mon Sep 16 03:25:05 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 16 Sep 2002 09:25:05 +0200
Subject: Pickling slotted instances
References: <mailman.1031805375.9030.python-list@python.org> <j4fzwf349u.fsf@informatik.hu-berlin.de> <mailman.1032144560.28257.python-list@python.org> <m3it16ze98.fsf@mira.informatik.hu-berlin.de> <mailman.1032158478.7315.python-list@python.org>
Message-ID: <j44rcqpfam.fsf@informatik.hu-berlin.de>

Andrew McNamara <andrewm at object-craft.com.au> writes:

> >1. How to determine all slots? It has been proposed that __slots__
> >   should return a tuple that includes base slots as well;
> >   alternatively, iterate over base types yourself. Alternatively,
> >   return slots in __dict__.
> 
> How is this handled in an instance now? Does an instance have slots
> for all it's direct slots as well as the slots defined by it's base
> classes? If so, it should probably be serialised in exactly this manner?

I don't understand the question. By "instance", do you mean an object
of type InstanceType, or something else? Instances don't have slots.

Or do you mean the attributes of an instance? The attributes of an
instance are all kept in the instance's __dict__, regardless of the
class in which they are assigned.

Or, do you mean, by "instance", an object of a certain type which has
slots? If so, inst.attr is the same as 

getattr(inst.__class__, 'attr').__get__(inst)

(if not inst.__dict__.has_key('attr')). So 'attr' will be searched in
the class, using the method resolution order. This happens to return 
a member descriptor, namely <member 'attr' of `inst.__class` objects>.


> >   has not assigned the slot, requesting the slot value raises an
> >   AttributeError.
> 
> It would seem simple enough to include a "not set" token in the serial
> stream.

You need to specify how exactly that token is represented.

> I'm sort of surprised that slots were added without having all the
> bases covered, but I'm presuming that they are really a manifestation
> of the internal machinery that was added for class/type unification,
> and have been exposed in the language itself so people can see if it's
> a useful feature.

"The bases", for me, would put "proper syntax" before support for
automatic pickling, e.g. via

class X(object):
  def foo              # defines a slot foo.
  def bar[transient]   # defines a transient slot.

New-style classes are really the same as traditional types. In the
past, you could not pickle arbitrary types unless you have special
machinery (such as copy_reg manipulation); so it is not surprising
that you can't pickle certain cases of new-style classes
automatically. Notice that you *can* pickle them - just define the
__getstate__.

Regards,
Martin


From ebbingeo at logica.com  Wed Sep 11 05:28:52 2002
From: ebbingeo at logica.com (Onno Ebbinge)
Date: 11 Sep 2002 02:28:52 -0700
Subject: EDI parsing
Message-ID: <746127ef.0209110128.53665a23@posting.google.com>

I'm writing an application (in Python) that is part of a bigger 
infrastructure for B2B messaging. One communication channel is an EDI 
interface to the B2B messaging system.

I need to parse a (subset) of the UN EDIFACT (aka EDI) standard 
(homepage: http://www.unece.org/trade/untdid/welcome.htm). After doing 
the initial reading of the standard it looks like it is quite hard to 
make a parser for this standard (and ?constructor').

My questions:

1) Is there a python EDI module? I can't find any :(
2) Is there a (python) EDI to XML converter?
3) Is there anything that I can use to easily interface with EDI?
   (preferably in lib and DLL form)

If the above is not available or fails... 

[You have to know that I've never written a parser in Python before. 
 The last (big) parser I wrote was in C with a few years back the help 
 of lex and yacc if memory serves me right.]

What is the best approach to writing an EDI parser in Python?

Thanks for your input,

Onno


From Oschler at earthlink.net  Wed Sep 25 17:14:08 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Wed, 25 Sep 2002 21:14:08 GMT
Subject: Python threading?
References: <f9Pi9.10275$Lg2.2191294@news2.news.adelphia.net> <amogmi$r1l$1@panix1.panix.com> <T5Rj9.13058$6g7.39977@news-server.bigpond.net.au> <MyUj9.134183$ub2.2937335@news1.tin.it>
Message-ID: <AGpk9.11786$Lg2.3232428@news2.news.adelphia.net>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:MyUj9.134183$ub2.2937335 at news1.tin.it...
> Mark Hammond wrote:
>         ...
<snip>

> Python may not yet support event-driven programming quite as smoothly
> and seamlessly as Ousterhout's Tcl, but, thanks to such developments
> as twisted, I think we're drawing progressively closer.  With a solid
> enough underlying framework to dispatch events for you, I think event
> driven programming (perhaps with a few simple specialized threads off
> to one side, interfacing to some slow external resource, and getting
> work requests off a Queue, posting results back to a Queue...) can
> be made conceptually simpler as well as more scalable than threading.
> After all, in event-driven / async programming, ONE thing is happening
> at a time -- you only "switch" at known points, and between two such
> points you need not worry about what is or isn't atomic, locking, &c.
> It's also easier to debug, I think...
>
>

Having written _many_ state machines in my life, which async programming
boils down to in one form or another, I feel it depends on the task.  When
you have an async event driven model, the code that services a
"transaction-package" ends up spreading functionality that services the
"transaction-package" application's code landscape.  This can frequently be
a headache.  As an example of a "transaction-package", take all the steps
involved in servicing an incoming TCP connection, with custom authentication
of a user, and perhaps a header and data object transacttion or two,
finishing with a proper ending and cleanup of the transaction, and with
subsequent notification of all necessary modules and data structures of the
results of the transaction.  For dessert add a centralized clean error
handler that can log errors propertly and gracefully recover the app
regardless of where in the distributed event flow the error took place.

Contrast this with a simple procedure call where a linear series of function
calls are made, each with appropriate error handling and status adjustment,
in a pleasant centrally located manner.

There is an upcoming massively multiplayer game called EVE which is using
Stackless Python which makes this point quite well.  Game AI suffers from
the same design considerations as mentioned above.  Here's the link if
you're interested (I'm not associated with them in any way, just considering
taking a look at Stackless).  Look for question "8.5 How is the game logic
implemented in EVE?":

http://www.eve-online.com/faq/faq_08.asp

thx





From johandewitte at hotmail.com  Tue Sep  3 08:08:51 2002
From: johandewitte at hotmail.com (Johan de Witte)
Date: 3 Sep 2002 05:08:51 -0700
Subject: simpleparse example
References: <3d62a770.10000379@newszilla.xs4all.nl>
Message-ID: <cfb4599.0209030408.2c4497af@posting.google.com>

baronvm2 at yahoo.com wrote in message news:<3d62a770.10000379 at newszilla.xs4all.nl>...
> Hi,
> 
> I'm trying to use SimpleParse to process semistructured data. I have
> defined the language and can produce a taglist. 
> 
> How can I use the dispatchprocessor to process the taglist? Does
> anybody have an example.
> 
> Thanx,
> Hans


I got tons of code I used to parse weakly structured HTML pages, and
get rid of unwanted data. But in fact it is not limited to HTML, but
can be used for almost any kind of (more or less) structured data. If
you are interested you can have all of it, if you are not going to use
it for commercial purposes.

Johan


From tmohr at s.netic.de  Sun Sep 15 11:57:45 2002
From: tmohr at s.netic.de (Torsten Mohr)
Date: Sun, 15 Sep 2002 17:57:45 +0200
Subject: Tkinter in PyRun_SimpleString(char* s)
References: <am12be$4j0$1@schleim.qwe.de> <_OWg9.109539$pX1.3874894@news2.tin.it>
Message-ID: <am2alp$3v3$1@schleim.qwe.de>

Hi,

thanks for that hint.

> As you can see, the problem is with sys.argv, which you have
> not set.  You can set it up from the Python code you execute,
> before you import Tkinter, or with C-API call PySys_SetArgv.

I've now added that call but still when i try to exec the script:


Fatal Python error: PyThreadState_Get: no current thread


Is there any rule what needs to be initialized/called before
executing a python script?  The simplest example works fine,
but i seems that that's not all that needs to be done when
i try to import Tkinter.  I didn't find anything in the docs,
maybe i missed it?


Best regards,
Torsten.


# python code
from Tkinter import *

class Application(Frame):
        def say_hi(self):
                print "hi there, everyone!"

        def createWidgets(self):
                self.QUIT = Button(self)
                self.QUIT["text"] = "QUIT"
                self.QUIT["fg"]   = "red"
                self.QUIT["command"] =  self.quit

                self.QUIT.pack({"side": "left"})

                self.hi_there = Button(self)
                self.hi_there["text"] = "Hello",
                self.hi_there["command"] = self.say_hi

                self.hi_there.pack({"side": "left"})


        def __init__(self, master=None):
                Frame.__init__(self, master)
                self.pack()
                self.createWidgets()

app = Application()
app.mainloop()


/* C code */
        Py_SetProgramName(argv[0]);
        PySys_SetArgv(argc, argv);

        Py_Initialize();
        PyRun_SimpleString(script);

        Py_Finalize();



From DGustafson at attbi.com  Thu Sep 19 18:19:23 2002
From: DGustafson at attbi.com (Dave Gustafson)
Date: Thu, 19 Sep 2002 22:19:23 GMT
Subject: Creating a typelibrary for calling Python COM server from c#
References: <yvoi9.516983$UU1.86221@sccrnsc03> <uit11stin.fsf@online.no>
Message-ID: <L3si9.518827$UU1.86524@sccrnsc03>

After trying to figure this out, I'm starting to feel either really stupid,
or lazy, or just burned out...
Does anybody out there have any working examples of how to do this?

Thanks,
Dave


"Syver Enstad" <syver-en+usenet at online.no> wrote in message
news:uit11stin.fsf at online.no...
> "Dave Gustafson" <DGustafson at attbi.com> writes:
>
> > There was a thread recently that suggested creating a typelibrary in
> > order
> >
> > to access a Python based COM server from C# (or .NET).  Can anyone
> > please
> >
> > suggest an easy way to create such a library???
>
> Create an .idl file. MSDN has resources for the syntax of this file.
>
> Run the midl compiler on the .idl file (midl.exe), this will generate
> a typelibrary (.tlb) file and c source and header files.
>
> If you don't have Visual Studio, I believe that you can get the midl
> compiler by downloading the Microsft Platform SDK (also on msdn).
>
> --
>
> Vennlig hilsen
>
> Syver Enstad




From Peer.Griebel at entire.de  Thu Sep 19 03:14:39 2002
From: Peer.Griebel at entire.de (Griebel, Peer)
Date: Thu, 19 Sep 2002 09:14:39 +0200
Subject: Copy with __slots__
Message-ID: <673E8D215B3BD311A9C90090279BBB8402829E7D@mail.ebssoftware.de>

> It's of course not a matter of new vs old classes -- new classes
> would let you use the dictionary-update method just as well... as
> long as you didn't use __slots__.

Ok. I was a little unprecise.
 
> Regarding performance -- *measure* things, man!  Intuition about
> what runs slowly, what rapidly, is notoriously misleading.  E.g.:
> 
> import time, copy
> 
> class noslots(object):
>     pass
> 
> class withslots(object):
>     __slots__ = 's1', 's2'
>     def __copy__(self):
>         result = self.__class__.__new__(self.__class__)
>         result.s1 = self.s1
>         result.s2 = self.s2
>         return result

...

> [alex at lancelot ~]$ python -O slorno.py
> 1.04 noslots
> 0.29 withslots
> 
> "much more slowly", hmm...?  Looks over 3 times faster to *ME*...

Sorry. But even timing shows that slots are slower. My class has exaclty 10
attributes (some more that your example) which have to be copied. Copying
each slot explicitly takes 2.28 sec (100000 iterations).

Using a class without slots and following method:
    def __copy__(self):
        c = _EmptyClass()
        c.__class__ = self.__class__
        c.__dict__.update(self.__dict__)
        return c
copying takes only 1.4 sec. So there is some price to pay to use slots. (To
be precise: In my code I directly call __copy__ but do not use copy.py -
because it's much faster. And my method isn't called __copy__ but clone.)

BUT: I just timed my whole program. Although copying takes more time the
program as a whole runs faster! The runtime drops from 16.3 sec to 14.5 sec.
So using slots seems to speed up other accesses to attributes significantly.

> > 2. The approach is error prone. When simply updating a dict 
> I'm not going
> > to miss a newly added attribute. But using your method I 
> have to remember
> > the method whenever I add an attribute.
> 
> Yes, that IS part of the price you pay for getting the optimization
> that __slots__ affords: you do have to know exactly what you're doing.
> Thus, __slots__ should be used when the program is working, in the
> optimization phase, for classes such that saving a __dict__ per
> instance IS an important optimization, only (classes of which MANY
> instances exist, basically).

You are right. Saving a dict made me switch to use slots. And now I learnt
that I also can save time using slots.

So my advice now is: Use slots! The program runs faster, uses less memory.
And explicitly coding a __copy__ function which copies each attribute isn't
a bad thing per se. It makes the things more explicit.


One more thing: Does anybody know if copy.py in CVS correctly handles
objects with slots? I think it should be modified...

Peer



From jepler at unpythonic.net  Sun Sep 29 10:47:39 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Sun, 29 Sep 2002 09:47:39 -0500
Subject: Help please with menubars
In-Reply-To: <3d9516b3$1_7@goliath.newsgroups.com>
References: <3d9516b3$1_7@goliath.newsgroups.com>
Message-ID: <20020929094729.A938@unpythonic.net>

On Sat, Sep 28, 2002 at 03:54:52AM +0000, mongo57a at comcast.net wrote:
> Specifically: I have an "exit" item on the
> menubar and if the user clicks on "exit", the window should be closed. What
> I don't want is a pulldown with "exit" as a button option. Anybody know how
> to do this? Thanks.

I have never read a single guide to User Interface Design that said this
was a good idea.

First, all modern GUIs have a standard decoration for "Close Window",
as well as a standard for File->Exit (plus Ctrl-Q/Command-Q shortcut).
Why add another, non-standard way to exit the application?  (A similar
argument exists for not placing "Exit" on a toolbar:  Toolbars are for
frequently-accessed commands, but Exit is invoked only once per session)

Second, having a destructive command invoked simply by clicking on the
menubar violates the user's expectation that an action will not be taken
until she chooses an item from inside a menu.  In a pulldown menu there is
a visual cue that an item is a cascade (the right-pointing triangle) but
there is no such standard visual cue in the toplevel menu.  Instead, they
are all assumed to act as cascades.

Third, if you're really insistent on doing this, code like the following
"works" for me:
    >>> import Tkinter
    >>> x = Tkinter.Tk()
    >>> m = Tkinter.Menu(x)
    >>> x.configure(menu=m)
    >>> def p(): print "clicked" 
    ... 
    >>> m.add_command(label="Blah", underline=0, command=p)
    >>> # I click "Blah" or press Alt-B
    clicked

While the current crap^H^Hop of "skinnable" applications seem to indicate
otherwise, there's really no value in making your program arbitrarily
different from the standard programs that run on your platform.  I hope
you'll reconsider.  Eventual users of your software will thank you.

Jeff



From gerrit.muller at philips.com  Mon Sep 23 06:22:52 2002
From: gerrit.muller at philips.com (Gerrit Muller)
Date: Mon, 23 Sep 2002 12:22:52 +0200
Subject: How to get memory usage on Windows?
References: <amilhk$bi8$1@reader1.tiscali.nl> <mailman.1032715528.24198.python-list@python.org>
Message-ID: <3D8EEB7C.C1B2F01F@philips.com>


Emile van Sebille wrote:
> 
> GerritM:
> [snip]
> > how much memory is used per phase and/or per function. Is there any
> function
> > returning this information. I noted that on Unix the resource module
> gives
> > access to this kind of information, but I could not find any Windows
> > alternative.
> >
> 
> You could try pslist from the pstools package at www.sysinternals.com to
> get an idea of total memory in use by process.
> 
> --
> 
> Emile van Sebille
> emile at fenx.com

I tried pslist from www.sysinternals.com at a windows2000 machine, which
works!
According to their documentation it will not work on windows98.

Sysinternals provide an attractive set of utilities anyway.

Thanks for the pointer, Gerrit
-- 
--------------------------------------------------------------------
Gerrit Muller                Philips Research Laboratories Eindhoven
Building:   WDC 2 - 007              Prof. Holstlaan 4
Phone: +31 40 27 45242               5656 AA  Eindhoven
Fax:   +31 40 27 45033               The Netherlands
mailto:gerrit.muller at philips.com
http://www.extra.research.philips.com/natlab/sysarch/index.html


From jk3380 at naida.org  Sun Sep 22 13:30:57 2002
From: jk3380 at naida.org (Pino Gargiulo)
Date: Sun, 22 Sep 2002 19:30:57 +0200
Subject: Smalltalk vs. Python and COM was Re: Byte Magazine and Smalltalk
References: <MXFh9.23885$jF4.2037258@twister.southeast.rr.com> <3d87a686@tobjects.> <rd8fougqbgdivimn7vuest3ugoj57lvrr6@4ax.com> <Xns928CA588EF17Dtrippyparcplacecom@64.164.98.7> <3d87dd7c$1@tobjects.> <g9Vh9.313655$kp.1041470@rwcrnsc52.ops.asp.att.net> <Xns928D58DF09B6trippyparcplacecom@64.164.98.6> <20020918.115128.1605894428.1527@cybermesa.com> <amkpk0$or2$1@fe2.cs.interbusiness.it> <amkqfs$ldg$1@reader1.tiscali.nl>
Message-ID: <amkup0$qtb$1@fe2.cs.interbusiness.it>

"GerritM" <gmuller at worldonline.nl> wrote in message
news:amkqfs$ldg$1 at reader1.tiscali.nl...
> Long long ago I worked in a large scale Objective-C project. Objective-C
> inherited many of its good characteristics from Smalltalk. Now I am using
> Python quite a lot. It has again a lot of the good (dynamic)
> characteristics, with on top an excellent set of predefined modules. One
of
> them are the win32 modules by Mark Hammond, documented on line of course,
> but also described extensively in "Python programming on Win32" by Mark
> Hammond and Andy Robinson, ISBN 1-56592-621-8. I am using the ActiveState
> distribution, which is a complete distribution including this win32 stuff.
>
> In short highly recommended!

Gerrit

Thanks for your advice and prompt reply!

If I understand you correctly you are suggesting that Python is the way to
go
for COM m$ API rather than Smalltalk .. again Id like to hear suggestions
on which product. My dabbling experience is just with VWNC

TNX again

Pino






From marklists at mceahern.com  Sun Sep 15 09:16:52 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sun, 15 Sep 2002 08:16:52 -0500
Subject: Stackless Python, eventual merge?
In-Reply-To: <gNQg9.6056$Lg2.882744@news2.news.adelphia.net>
Message-ID: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com>

[Robert Oschler]
> [...] Do you think that it [stackless] eventually
> will be merged with the main Python code base?

Note which category PEP 219 is in:

  http://www.python.org/peps/

// m
-




From roy at panix.com  Mon Sep  9 08:07:06 2002
From: roy at panix.com (Roy Smith)
Date: Mon, 09 Sep 2002 08:07:06 -0400
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <bd751532.0209082142.54236307@posting.google.com> <3d7c8d8a@news.sentex.net>
Message-ID: <roy-8A7122.08070609092002@reader1.panix.com>

Peter Hansen <peter at engcorp.com> wrote:
> There's some truth in that.  I often end my files with a line that
> reads "# EOF", which helps me not try to scroll down to find more
> code

I would think your editor beeping (or whatever your particular editor 
does when you try to scroll past the end of file) would do the trick.  
But, it could be worse.  We had a guy here who insisted on putting stuff 
like this at the ends of his files:


#####################################
#                                   #
#  ###############################  #
#  # End of <filename>           #  #
#  ###############################  #
#                                   #
#####################################


From max at alcyone.com  Fri Sep  6 18:45:33 2002
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 06 Sep 2002 15:45:33 -0700
Subject: Beginners Question - use of variables
References: <mailman.1031345109.5380.python-list@python.org>
Message-ID: <3D79300D.CD4E96FC@alcyone.com>

Martin Klaffenboeck wrote:

> in Perl I used to save such a thing in a multidimensional array like
> 
> array(
>       array(string, string, int, timestamp),
>       array(string, string, int, timestamp),
>       array(string, string, int, timestamp))
> 
> So how do I save that in python and how do I get fast to one of this
> strings or int or timestamp... ?

A list of 4-tuples, or just a list of lists, comes to mind.  Difference
between Python and Perl is that with Python you don't have to worry
about awkward manipulation of references to those subarrays.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ There is nothing so subject to the inconstancy of fortune as war.
\__/ Miguel de Cervantes
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From contato at upclick.com.br  Fri Sep 27 12:08:45 2002
From: contato at upclick.com.br (UpClick - Mais clicks em seu site)
Date: Fri, 27 Sep 2002 13:08:45 -0300
Subject: Seu site com mais acessos
Message-ID: <10301112-22002952716845230@upclick.com.br>

Se a sua empresa possui um site na Internet, precisa
conhecer o UP Click. 

De forma simples, objetiva e econ?mica, o UP Click 
aumenta as visitas do seu site. E voc? n?o precisa
ser nenhum expert em Internet para isso. Basta ter
a consci?ncia que o sucesso de um site depende 
da quantidade de pessoas que o visitam... 

Confira: http://www.upclick.com.br 







----------------------
E-mail promocional enviado as empresas associadas
da Amcham. Nossa empresa n?o possui nenhum
vinculo comercial com a Amcham. 






From BPettersen at NAREX.com  Tue Sep 17 03:32:45 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Tue, 17 Sep 2002 01:32:45 -0600
Subject: file mode attributes...
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E201920125335F@admin56.narex.com>

> From: Martin v. L?wis [mailto:loewis at informatik.hu-berlin.de] 
> 
> "Bjorn Pettersen" <BPettersen at narex.com> writes:
> 
> > Thanks. Do you think there would be any interest in adding 
> > these kinds of utility functions to e.g. the os.path module?
> 
> I doubt that. Your implementation is valid only on Windows, 
> on Unix, you need to determine whether you are interested in 
> user, group, or other bits, and you also have not only read 
> and write, but also execute. The current API is as portable 
> and convenient is it can get.
[snip]

Ok. Thanks for considering it.

> > win32all isn't standard on Windows <wink>?
> 
> No; that's a huge code base, and nobody has proposed to 
> incorporate it into Python proper.

Notice the <wink>? Seriously though, I can't imagine any Python/Windows shop that don't have win32all installed by default (hats off to Mark Hammond for excellent work). I know I made it a requirement when I had sysadmin install Python on all of our machines <grin>.

-- bjorn



From elf at drizzle.com  Wed Sep 18 13:49:39 2002
From: elf at drizzle.com (Elf Sternberg)
Date: Wed, 18 Sep 2002 17:49:39 -0000
Subject: CGI question
References: <slrnaoerk7.mgo.jboy@otaku.freeshell.org>
Message-ID: <1032371377.675310@yasure>

In article <slrnaoerk7.mgo.jboy at otaku.freeshell.org> 
    "John D. Boy" <info at voidhobo.de> writes:

>I am working on a simple CGI script to use as a framework for a website
>of mine. Basically, I want it to work as follows: When somebody accesses
>the site, a CGI (called e.g. index.py) is run. It takes an html-template
>and fills it with content from another file, called e.g. index.dat, and
>returns the completed html-document. When I want to present another
>page, I want the url to look like this: http://my.domain.name/?page. In
>that case, index.py would fill the same template with the content of a
>file called e.g. page.dat. 
>
>Now, my problem is to find out what follows behind the question mark in
>my CGI script. Unless the key has a value
>(http://my.domain.name/?key=value), it is not in the cgi.FieldStorage
>dictionary. Is it simply impossible, or am I not using the right
>function calls? 

        I wrote something similar recently.  You can find it at:

http://www.drizzle.com/~elf/code/narrator/

        It's a bit different from what you want.  I wanted to write a
CGI script that would manage a "book", meaning front matter, table of
contents, and chapters; I wanted it to be able to distinguish between
browsers that use style sheets and those that do not, and supply
different HTML to meet different needs.  And I wanted to obscure the
fact that it's a CGI.  URLs needed to look like:

        http://www.foo.com/stories
        http://www.foo.com/stories/index.html
        http://www.foo.com/stories/chapter1.html
        http://www.foo.com/stories/chapter2.html
        etc...

        The trick is that the element "stories" in the above URL is the
CGI script: everything after that is passed to the program by the
environment variable "PATH_INFO".  By using this line:

         pathinfo = os.environ.get( 'PATH_INFO', '' )

        ... and then using a lookup table, I can identify what the user
asked for and respond to it.  I can then supply the four necessary
pieces (splash page, index page, chapters page, and 404 page).  The big
advantage to this is that robots, spiders and other indexing engines
will troll the story archives and index them because they _look like_
they're plain ol' HTML.  The other advantage is that the story entries
themselves can be compressed using GZIP, saving me on disk space, which
my ISP charges for.

        If you're using Apache, the settings needed to get this working
are trivial.  In your .htaccess file, put:

DefaultType application/x-httpd-cgi
XBitHack on
ErrorDocument 404 /~elf/error.html

        Apache will attempt to run everything that does not have an
extension and has the executable bit set.  

        Hope this helps.

                Elf

--
Elf M. Sternberg
http://www.drizzle.com/~elf/

No bastard ever won a war by dying for his country.  He won it by making
the other poor dumb bastard die for his country.  - Gen. George S. Patton.


From duncan at NOSPAMrcp.co.uk  Tue Sep 10 04:16:12 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 10 Sep 2002 08:16:12 +0000 (UTC)
Subject: dynamism
References: <66715c02.0209091923.265fe77e@posting.google.com>
Message-ID: <Xns92855D48384F9duncanrcpcouk@127.0.0.1>

steven_shaw at adc.com (Steven Shaw) wrote in
news:66715c02.0209091923.265fe77e at posting.google.com: 

> 
> I thought you could just extend an object's attributes at run-time
> like this: 
> 
>>>> p = object()
>>>> p.x = 1
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> AttributeError: 'object' object has no attribute 'x'

New style classes (which are those with 'object' or other builtin classes 
as a base class) do not all allow you to add arbitrary attributes:

If you have an object which is a builtin type (object, int, dict, ...) then 
you cannot add attributes. If you have an object which is a subclass of a 
builtin type, then you can add attributes unless the subclass defined a 
member __slots__.

The __slots__ attribute in a class, if defined, lists all of the attribute 
names settable in instances of that class. If no __slots__ attribute is 
directly defined in a class then you may add arbitrary attributes (this 
applies even if __slots__ is set in a base class).

An object of a builtin type, or with __slots__ defined in the class and 
all base classes does not have a __dict__ attribute. 

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From max at alcyone.com  Fri Sep 20 04:28:22 2002
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 20 Sep 2002 01:28:22 -0700
Subject: newbie seeks advise.
References: <77c84db2.0209192137.51167ebb@posting.google.com> <mailman.1032501210.16162.python-list@python.org>
Message-ID: <3D8ADC26.8F3CF3BD@alcyone.com>

Gerhard H?ring wrote:

> If you want to program a max() function yourself, you can convert the
> tuple to a list using list(data), then sorting it, then finding the
> largest element yourself.

Or just scan the list, keeping track of the largest element you've seen
so far.  When you're done, the largest one you've seen is the largest in
the list.  This has better performance than sorting.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ He who has a _why_ to live can bear with almost any _how_.
\__/ Friedrich Nietzsche
    REALpolitik / http://www.realpolitik.com/
 Get your own customized newsfeed online in realtime ... for free!


From sross at chat.carleton.ca  Wed Sep 25 09:52:33 2002
From: sross at chat.carleton.ca (Sean Ross)
Date: 25 Sep 2002 13:52:33 GMT
Subject: Removing all occurences of a character from a string.
References: <amr8nh$rd8$1@driftwood.ccs.carleton.ca> <DL9k9.201596$216.7001230@bin4.nnrp.aus1.giganews.com> <amrav1$4rk$1@driftwood.ccs.carleton.ca> <mailman.1032928353.13514.python-list@python.org>
Message-ID: <amsf31$g97$1@driftwood.ccs.carleton.ca>

Yes. That's exactly what I was looking for. Thank you very much.

Mike C. Fletcher (mcfletch at rogers.com) wrote:
> Under later Python versions (such as 2.2.1), you can do the following:

> attr = [[item.replace("'","") for item in element] for element in attr]

> Which is rather elegant IMO.  It also has the side-effect of _not_ 
> rewriting the data in-place in your original list, instead just 
> replacing that list when we are finished creating the new list, which 
> may or may not be desirable for you.

> HTH,
> Mike

> PS: Always nice to hear from another Canadian Pythonista :)

> Sean Ross wrote:
> > Thank you Justin.
> > 
> > How about this, then:
> > 
> > I have a list of lists of these strings and I want to apply replace to each
> > string in the list of lists.
> > 
> > So, say:
> > 
> > attr = [ ["'gill-attachment'", "'a'", "'d'", "'f'",
> > "'n'"],["'gill-spacing'", "'c'", "'d'", "'w'"], ["'gill-size'", "'b'",
> > "'n'"]]
> > 
> > I can do the following:
> > for i in xrange(0, len(attr)):
> >     for j in xrange(0,len(attr[i])):
> >          attr[i][j] = attr[i][j].replace("'","")
> > 
> > to get what I want; which is:
> > [ ['gill-attachment', 'a', 'd', 'f', 'n'],['gill-spacing', 'c', 'd', 'w'],
> > ['gill-size', 'b', 'n']]
> > 
> > Is there a nicer way to do this?
> > Sean
> ...




From jb at cascade-sys.com  Wed Sep 25 23:52:40 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Wed, 25 Sep 2002 20:52:40 -0700
Subject: Getting started
References: <uojinmcfapjf45@corp.supernews.com> <mailman.1032442884.18747.python-list@python.org> <H3mi9.133635$pX1.4751676@news2.tin.it> <mailman.1032567390.29374.python-list@python.org> <AFik9.51323$b%3.25248@fe03> <mailman.1032979596.18965.python-list@python.org> <3D920950.323B25B3@alcyone.com>
Message-ID: <3D928488.3060906@cascade-sys.com>

Erik Max Francis wrote:

>The problem here is that there is no official "formal definition."  
>

That's a real problem, as a definition for the term was the OP's 
original question.

>Some
>use "weak typing" to mean any type system where the typing can be foiled
>by casts; 
>

The majority of Google references and the one Cardelli reference I was 
able to locate all support this view.  Some of the Google references are 
from University course curriculum, so it's what's being taught to CS 
students.  It's the most authoritative and meaningful definition anyone 
has brought to the table thus far.

>some use it to mean any system where operations, not the
>objects themselves, determine the types of the objects.  
>

By my count, only 2 or 3 people (including yourself), all on this list, 
use this interpretation.  From Google, nobody gave this definition.  The 
only definitions I found contrary to the first alternative was that 
"weak typing" was synonymous with "dynamic typing", which I think 
everybody here rejects.

Furthermore, with elaborate review of the abstractions involved and 
details of the Python implementation, I showed that the above statement 
is really meaningless.  On the one hand, it IS how Python intrinsics are 
implemented internally (operations literally inspect object types, not 
the reverse).  But OTOH it is NOT the abstraction we tend to think of 
when we use the language (objects define operations).  It's one of two 
ways of looking at a language but it's not a way to distinguish one 
language from another.  You have objects and you have operations which 
are associated somehow.  In actuality, neither the object or the 
operation "determines" or makes any decision about anything.  It's the 
runtime environment that glues it all together.  

With user defined types (different implementation than system 
intrinsics), we know there's a dictionary of function names pointed to 
by each object.  But "who" does this lookup?  Who applies the function 
to self (and to self only by convention)?  There's no code in the Object 
to do this.  Nor is there any code in the member function to do the 
lookup.  So, it's not really fair to say that either one "determines" 
what to do, is it?  Else, in which one does the apply() occur?  

In fact, in Python, the answer is none of the above.  It is the 
LOAD_ATTR primitive operator (".") that examines its two arguments 
(faulting if they're not specifically an object reference and a string) 
and then does the dictionary lookup which fetches the resulting 
attribute and pushes it onto the stack.  Then (if the result is a 
function) it's is the CALL_FUNCTION (or a related) operator which looks 
at its arguments on the top of stack and figures out how to proceed 
(primitive function, a method or what?).  So in the end, even Python's 
OOP features are based on (multiple) explicit "operators that have to 
examine the types of its arguments in order to know what operation to 
perform."  But this is in contradiction to how we all "know" objects 
work, so in reality the distinction is meaningless.

>(I favor the
>latter definition, since it carries more information about the
>language.)  
>

It doesn't really tell you anything.  

Else, in 25 words or less, why does Perl's "+" operator supposedly work 
the one way and Python's "*" operator work the other way?

>This is simply a case of differing terminology.
>

In this thread I've been twice accused of, Humpty Dumpty like, using 
terminology of my own device in lieu of some mythical, more 
authoritative source.  And yet many explicit 3rd party references 
(including one of Alex's own) contradicts his interpretation in favor of 
the one I have thus substantiated.  

I dunno -- I tracked down and presented all the references I could find. 
 If the definition I offer from that perspective is not sufficiently 
WORTHY for this group then clearly I have wasted my time.  

You can lead a pig to pearls....

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From ianb at colorstudy.com  Mon Sep 16 15:30:38 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 16 Sep 2002 14:30:38 -0500
Subject: Stackless Python, eventual merge?
In-Reply-To: <j4znuio0kr.fsf@informatik.hu-berlin.de>
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com>
	<mailman.1032107006.6411.python-list@python.org>
	<am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> 
	<j4znuio0kr.fsf@informatik.hu-berlin.de>
Message-ID: <1032204639.2425.84.camel@dsl-65-184-205-5.telocity.com>

On Mon, 2002-09-16 at 02:28, Martin v. L?wis wrote:
> Michael Schneider <michaels at one.net> writes:
> > Could someone please explain to a "confused" python user. Why the
> > current version of stackless python would not be distributed like
> > threads???
> 
> Very easy: because nobody makes it so. That, in turn, is because
> nobody works on making it so. 
> 
> Christian (rightfully) points out that this is not his obligation:
> this is free software, he is not obliged to do anything - unless you
> pay him to do something. Nobody else is obliged to incorporate
> Stackless into Standard Python either. If you want this to happen,
> contribute.

I don't think this represents the current state of Stackless --
Christian has tried to get it into the standard Python, and it is not
for lack of desire or effort that that did not happen.  I think it could
be considered a difference of opinion between him and Guido -- in part,
Stackless introduces more complexity which Guido did not want to
maintain, and in part because Stackless is not portable to forms of
Python (not to Jython in particular, and certainly any .NET Python --
probably a Parrot Python as well).

So the issue is primarily technical/political, and not a matter of
resources.  

  Ian





From d_blade8 at hotmail.com  Thu Sep 26 17:11:56 2002
From: d_blade8 at hotmail.com (Lemniscate)
Date: 26 Sep 2002 14:11:56 -0700
Subject: Silly question about ungzipping a file
Message-ID: <7396d2b2.0209261311.1fdd2d3a@posting.google.com>

Hi everyone,

This may be a ridiculously easy question, but I've kind of hit a wall
and I was wondering if I am just missing something.  I want to
automate the retrieval and unzipping of a *.gz file.  The issue is
that the file, when it is unzipped, usually has a size somewhere in
the range of 128MB.  Occassionally, I get memory errors when I try to
run it.  Here is a quick idea of what I am doing...

>>> import gzip
>>> myFile = gzip.open('LL_tmpl.gz')
>>> file('output.txt', 'w').write(myFile.read())
>>> 

Now, is there any way to do this so that less memory is used?  I mean,
if I wanted to do some processing on the resulting output file, I
would use xreadlines or something like that to keep memory consumption
to a minimum.  Is there something roughly equivalent that I am not
noticing in the gzip documentation.  Let me also say that I have tried
the following as well:

>>> myFile = gzip.open('LL_tmpl.gz')
>>> fout = file('output.txt', 'w')
>>> while myFile.readline():
... 	fout.write(myFile.readline())
... 	fout.write('\n')
... 	
>>> fout.close()
>>> myFile = gzip.open('LL_tmpl.gz')
>>> fout = file('output2.txt', 'w')
>>> while myFile.readline():
... 	fout.writelines(myFile.readline())
... 	
>>> fout.close()


These do solve my memory problem, but there are other issues.  First
of all, my CPU gets pegged and it takes FOREVER (okay, not forever,
but about ...let me test real quick... at least 4-5 times as long, and
my computer is pretty much useless during that time (side note:  can
you tell I am working on a woefully underpowered machine?)).  Is there
something in-between that anybody can think of?  The other, and much
more immediate, issue is puzzling to me.  It seems that the resulting
files from the code are only about 65MB (64.7 to be exact) versus
129MB.  I'm sure I'm just missing something simple, but why is that? 
Thanks for your time.

Lem


From sholden at holdenweb.com  Wed Sep 25 15:41:37 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 25 Sep 2002 15:41:37 -0400
Subject: filter(lambda W : W not in 'ILLITERATE','BULLSHIT') (Re : python expression)
References: <d8f8142b.0209161913.12cce045@posting.google.com> <vWBh9.26065$Gp1.687517@weber.videotron.net> <3D86FF01.FE0CECC5@geneva-link.ch> <3d87069c.34652637@news.rcn.com> <3D871316.AA1E8FFC@geneva-link.ch> <3d882cd4.9181792@news.rcn.com> <3f74f974.0209180713.27889f74@posting.google.com> <l9ehouci1ga88ub2csdam1qroiik0ogd4r@4ax.com> <3d88c258$0$707$5402220f@news.sunrise.ch> <v4lhouolv1c0ijmihfom94j0lor2lhntli@4ax.com> <3d89be92$0$707$5402220f@news.sunrise.ch> <s6mjous6o5abg3icq12u156v5tijhcnk7h@4ax.com> <3d89ee85$0$712$5402220f@news.sunrise.ch> <B8pi9.149307$5r1.6007886@bin5.nnrp.aus1.giganews.com> <3d8b4498$0$717$5402220f@news.sunrise.ch> <WSLi9.165658$z91.6788612@bin3.nnrp.aus1.giganews.com> <3d8c6982$0$719$5402220f@news.sunrise.ch>
Message-ID: <Edok9.1832$4K3.1065@fe02>

"Boris" <borcis at removethis.infomaniak.andthis.ch> wrote ...
>
> Terry Reedy <tjreedy at udel.edu> a ?crit ...
> >
> > "Boris"
> > >
> > > filter(lambda W : W not in 'ILLITERATE','BULLSHIT')
> >
> > This somewhat-well-known and certainly well-worn programmer joke
>
> Well, *I* invented this formula myself soon after Mr Bush revealed to
> the world that "USA" is written in "CRUSADE" just like "US" is in
> "BUSH". Adding : "Either you are with US, or you are against US".
>  [...other matters not dealt with here...]

Wouldn't

>>> filter(lambda W : W not in "Illiterate", "Bullshit")
'Bush'

seem a little more, erm, literate?

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From DennisR at dair.com  Sun Sep 29 22:23:38 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Mon, 30 Sep 2002 02:23:38 GMT
Subject: Python's import: why doesn't it work?
References: <9ac02e81.0209290839.4f4d6475@posting.google.com> <874rc8d9c9.fsf@nospam.eml.cc> <mailman.1033319977.7828.python-list@python.org> <3d976fbc$0$23666$91cee783@newsreader02.highway.telekom.at> <8428na.7vj.ln@ix.netcom.com> <3d979184$0$20332$79c14f64@nan-newsreader-03.noos.net>
Message-ID: <KAOl9.1635$vX6.52299899@newssvr21.news.prodigy.com>

> >         python -v script.py 2>output.txt
> yes it works on MS-DOS too.

Stderr redirect works as shown on Windows 2000 (which is not DOS!), but does
not work for Windows 98 and earlier.

--

Dennis Reinhardt

http://www.dair.com




From bh at intevation.de  Mon Sep 16 13:55:01 2002
From: bh at intevation.de (Bernhard Herzog)
Date: 16 Sep 2002 19:55:01 +0200
Subject: Mutable Stings
References: <mailman.1032181643.21256.python-list@python.org> <m3admidiys.fsf@mira.informatik.hu-berlin.de>
Message-ID: <6qznuhx1je.fsf@thetis.intevation.de>

martin at v.loewis.de (Martin v. Loewis) writes:

> "Griebel, Peer" <Peer.Griebel at entire.de> writes:
> 
> > So my question: Is this correct - or did I overlook something?
> 
> You are not overlooking anything: Your analysis is correct, the code
> is wrong. Feel free to submit a patch to SF.

Not necessary. I submitted a bugreport when this came up the last time
and it has been fixed in CVS since then.

   Bernhard

-- 
Intevation GmbH                                 http://intevation.de/
Sketch                                 http://sketch.sourceforge.net/
MapIt!                                           http://www.mapit.de/


From matt at mondoinfo.com  Thu Sep 26 13:58:52 2002
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Thu, 26 Sep 2002 17:58:52 GMT
Subject: combining mailbox and email modules
References: <NPck9.581867$UU1.104088@sccrnsc03> <slrnap41re.1mh.matt@happy-hour.t-island.com> <LAxk9.79207$gA4.27843@sccrnsc02>
Message-ID: <slrnap6ims.1g6.matt@happy-hour.t-island.com>

On Thu, 26 Sep 2002 06:14:03 GMT, Bob Horvath <nospam at horvath.com>
wrote:

> OK, I tried that.  Looks like the email package is being picky about 
> things where I can't see a problem.
> 
> 
> I keep getting the following error:
> 
> email.Errors.HeaderParseError: Not a header, not a continuation
> 
> Which I assume means a line that doesn't start with a space or tab 
> that also does not contain a ':'.
> 
> The problem is that it looks good to me.

I'd run your program in such a way that you can do a pdb.pm() after
the exception is raised and see what's in the variable line. I'd bet
that the message got some extra newlines in it somewhere along the way
but it's hard to tell after it has been wrapped for news posting.

Regards,
Matt


From tundra at tundraware.com  Wed Sep  4 20:00:07 2002
From: tundra at tundraware.com (Tim Daneliuk)
Date: 05 Sep 2002 00:00:07 GMT
Subject: [ANN]  'hb' 1.88 Budget Management/Python Tutorial Released
Message-ID: <3D769D65.50708@tundraware.com>

'hb' 1.88 Budget Management/Python Tutorial is released at:

       http://www.tundraware.com/Software/hb/


'hb' is a modest Budget Management system written in Python.  It is
appropriate for home budget management and small business budgets.

'hb' is primarily intended to be a helpful tutorial for people
learning Python by examining a complete program.  The program
illustrates interaction with the user, printing on a line printer,
disk I/O, and uses all the major Python data structures and many of
the language idioms.  It comes with extensive documentation and the
code is heavily commented.

The FreeBSD port has also been submitted.

------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com



From wweexxsseessssaa at telusplanet.net  Mon Sep  2 15:51:44 2002
From: wweexxsseessssaa at telusplanet.net (John Hall)
Date: Mon, 02 Sep 2002 19:51:44 GMT
Subject: Making regex suck less
References: <3d725881.345921@news.t-online.de> <un50ccotaak86b@corp.supernews.com> <al0e96$oms$0@216.39.172.122>
Message-ID: <7sf7nuo1eaa7eaq447pjqjd9klv8ct4di5@4ax.com>

On 2 Sep 2002 19:30:46 GMT, bokr at oz.net (Bengt Richter) wrote:

>I agree about new syntax, but I wouldn't mind having a re.help(regexp) function
>for interactive use that would just explain in 'English' what the regexp expression
>stands for. It would be a nice easy double check on whether I wrote what I meant,
>and helpful for understanding someone else's magic. It shouldn't be that hard to do.
>

Many years ago, in my PL/I & IBM Mainframe days, I wrote a gizmo to
check the result of some tricky JCL (Job Control Language) parameters.
(Required because if it did not do what the user intended, it could
fail in an overnight run and the user wouldn't know until morning)

Not only did it take an expression and show the result, but it gave a
decision-by-decision commentary on why the result was what it was, so
provided a live tutorial by way of a simulator. It was moderately
difficult to write, IIRC, but most difficult was checking that the
simulator was in accord with reality.

However I don't understand regexs well enough to attempt this yet.

-- 
John W Hall <wweexxsseessssaa at telusplanet.net>
Calgary, Alberta, Canada.
"Helping People Prosper in the Information Age"


From usenet at thinkspot.net  Sun Sep  8 00:33:09 2002
From: usenet at thinkspot.net (Sheila King)
Date: Sat, 07 Sep 2002 21:33:09 -0700
Subject: Possible bug in Email and Multifile Modules ?
References: <mailman.1031430611.6805.python-list@python.org>
Message-ID: <aldral.n0.1@kserver.org>

On Sat, 7 Sep 2002 16:29:15 -0400, "Steven D. Arnold"
<stevena at neosynapse.net> wrote in comp.lang.python in article
<mailman.1031430611.6805.python-list at python.org>:

> In my application, I'm trying to process a message based on its
> header, but I can't even get the header information, because instead
> of trying to intelligently handle a broken MIME message, an exception
> is raised.  Since I have no alternative but to find a way to
> automatically handle even broken messages, I either have to patch the
> email module to do what I want or reproduce much of the same logic in
> my own module.  I'd like to patch the email module and submit the
> patch for inclusion in the standard library.  What is the process for
> this?  Any other suggestions or comments would be appreciated as well.

Steve,

The developers of the email module have in fact been working on just such a
thing for the last several months. My understanding is that it is much
better than the email module that was released with Python 2.2.

Look here:
http://sourceforge.net/projects/mimelib/

Hope this helps,

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From printers at sendme.cz  Fri Sep 20 04:24:54 2002
From: printers at sendme.cz (A)
Date: Fri, 20 Sep 2002 10:24:54 +0200
Subject: Is there a widget or program  for a database browsing
Message-ID: <3D8AF776.22712.D2CB4CD@localhost>

Hi,
In my program users will search  a database and I need to display choosen
record(s ) in a nice graphical way. Is there a module( program), that I could use together 
with Python, that  could easily  show the database record(s) and take care of all thinks 
during browsing? Thanks for help.
 Ladislav






From ak at silmarill.org  Tue Sep 17 21:09:37 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Wed, 18 Sep 2002 01:09:37 GMT
Subject: Writing linux term color codes to a file
Message-ID: <e3gfouc157qqe6oj2i17g51edv8in2s26g@4ax.com>

Hello,

I ran into a bit of a problem here.. I'm making a small app that
visually creates bash prompts (just for fun of it), and at one point
it has to write the prompt to .profile, optionally with linux terminal
color codes, like \[044;\]m or something like that (you know what I'm
talking about). They don't come out right *in* the file. I tested the
simple case where I'd have a few such codes in a string in python
prompt and I'd look at it and it looks fine, but as soon as I write it
out to a file and look at the file, it's "cooked".

I hope you know what I mean, if anyone wants I can post direct pastes.

Looks like write() does some sort of processing on the text; is there
any way to turn that off?

Thanks a lot,

 - Andrei


From vincent_a_primavera at netzero.net  Thu Sep 26 07:31:48 2002
From: vincent_a_primavera at netzero.net (Vincent A. Primavera)
Date: 26 Sep 2002 11:31:48 +0000
Subject: comp.lang.python...
Message-ID: <1033039908.10951.7.camel@del-vap-bos.ralphpill.com>

Hello,
	Stupid question...  What information do I need to set up
comp.lang.python with my news client(Pan)?

	Thank you,

	Vincent A. Primavera.







From martin at v.loewis.de  Wed Sep 18 17:27:49 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 18 Sep 2002 23:27:49 +0200
Subject: Python as shared library on Unix
References: <mailman.1032382224.9592.python-list@python.org>
Message-ID: <m3ptvbj8dm.fsf@mira.informatik.hu-berlin.de>

"Rolf Kalbermatter" <r.kalbermatter at hccnet.nl> writes:

> Under Windows I have linked my app dynamically to the pythonXX.dll using
> LoadLibrary/GetProcAddress and that works very nicely. Under Linux I see
> no option to build a python.so.2.2.1 or similar to link my application
> to with the dl* functions.

I recommend that you not use a shared library. Instead, just link
libpython2.2.a to your application.

> My question is if I have overlooked some option in ./configure or
> somewhere else or if this is not possible out of the box? 

It's not possible out of the box. Python 2.3 will support that.

> If it is not available as a standard, what are the changes to get it
> working by modifying the ./configure and or Makefile scripts?

You can backport the changes from 2.3 if you want to.

> I can of course always link the Python core into my application which
> by the way is a shared library itself which is loaded dynamically
> into the main application but coming from the Windows world, this
> seems kind of unelegant to me.

I think linking the static library is more elegant. It reduces the
external dependencies, and thus the number of problem sources, in your
application. It also increases performance.

Regards,
Martin



From markus.vonehr at ipm.fhg.de  Tue Sep  3 07:39:49 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Tue, 03 Sep 2002 13:39:49 +0200
Subject: hackicon, tkicon
Message-ID: <3D749F85.962D8D39@ipm.fhg.de>

Hi,

is there already a hackicon or tkicon module for
python 2.2 ?
The modules are useful for changing the window icon on win32.

Thanks for any hints,

Markus


From lordmacro at hotmail.com  Mon Sep 30 01:29:51 2002
From: lordmacro at hotmail.com (Tom Widrick)
Date: Mon, 30 Sep 2002 00:29:51 -0500
Subject: sizeof PyTypeObject difference
Message-ID: <j5Rl9.29807$IO5.4027049@e3500-atl1.usenetserver.com>

I'm trying to make a metatype in C, but there seems to be some difference
between the size of the type object.

typedef struct {
 PyTypeObject type;
} MyMetaObject;


and in the tp_basicsize field I place sizeof(MyMetaObject) the value is only
196. It is the same if I put sizeof(PyTypeObject).

This causes a core dump.
However,

>>> type.__basicsize__
444

If I put 444, it works fine.

What is the cause of this difference? And what is the "proper" way to do
this?
444 + sizeof(all of my added fields of my meta type)?
Thanks






From me at aaronsw.com  Wed Sep  4 23:12:50 2002
From: me at aaronsw.com (Aaron Swartz)
Date: 4 Sep 2002 20:12:50 -0700
Subject: Graham's spam filter
References: <mailman.1029992314.27908.python-list@python.org> <3D647B4B.28D9BF04@alcyone.com> <mailman.1030050253.11330.python-list@python.org> <3D655AE6.A097AA25@alcyone.com> <ak44j1$1f2ft4$2@ID-125932.news.dfncis.de>
Message-ID: <c2ef5940.0209041912.16e102fb@posting.google.com>

I've been using bogofilter[1], Eric Raymond's Graham-derived spam
filter which threw away base64-encoded data and 90% of all spam that
got past the filter was base64-encoded. Therefore, I think that base64
content really needs to be decoded. I wrote a base64-decoding filter
in Python for it and the problem has gone away.

Aaron Swartz [http://www.aaronsw.com/]


From michaels at one.net  Mon Sep 16 07:49:41 2002
From: michaels at one.net (Michael Schneider)
Date: Mon, 16 Sep 2002 07:49:41 -0400
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de>
Message-ID: <3D85C555.10009@one.net>

Thanks Martin,

That is a good news.  So there is no current technical reason,
just a resource issue (not this this is not an issue).

We are comming into beta at work, so I am pretty booked now.

In a couple of weeks, I will send email Christian, and see if I can help
this in any way.

I did see that Christian had a paypal link setup.  This is a great idea for
open source code.  If you are too booked to constribute, you can send
a little bit of money ( or alot of money).



Thanks Martin,
Mike

Martin v. L?wis wrote:

>Michael Schneider <michaels at one.net> writes:
>
>>Could someone please explain to a "confused" python user. Why the
>>current version of stackless python would not be distributed like
>>threads???
>>
>
>Very easy: because nobody makes it so. That, in turn, is because
>nobody works on making it so. 
>
>Christian (rightfully) points out that this is not his obligation:
>this is free software, he is not obliged to do anything - unless you
>pay him to do something. Nobody else is obliged to incorporate
>Stackless into Standard Python either. If you want this to happen,
>contribute.
>
>Regards,
>Martin
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020916/956d9eec/attachment.html>

From tjreedy at udel.edu  Fri Sep 27 15:56:52 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 27 Sep 2002 19:56:52 GMT
Subject: comp.lang.python...
References: <mailman.1033054480.24506.python-list@python.org> <an1ip7$rgs$00$1@news.t-online.com> <mailman.1033131395.24431.python-list@python.org> <slrnap910v.5gd.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <8K2l9.327516$z91.14251915@bin3.nnrp.aus1.giganews.com>

"> However, there's a mail-to-news-gateway (and reverse) available at
> http://gmane.org/, which NNTP access for many Python mailing lists.
> The data you'll need are:
>
> NNTP server: news.gmane.org
> group name: gmane.comp.python.general
>
> No username or password required for reading. If you post, you'll
need
> to answer a one-time confirmation message for every group.
>
> I'd still prefer to use comp.lang.python diretly. But this is quite
> cool for all the SIG groups, as you can read them without actually
> subscribing.

I have been using gmane.comp.python.devel to follow the development
list and it is much nicer than downloading archive, which I did
before.

TJR




From buzzard at urubu.freeserve.co.uk  Tue Sep 10 14:55:18 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Tue, 10 Sep 2002 19:55:18 +0100
Subject: Graph Layout Algorithms
References: <alksio$d2l$1@newsg3.svr.pol.co.uk> <mailman.1031666838.31065.python-list@python.org>
Message-ID: <allf7h$n4j$1@newsg1.svr.pol.co.uk>

"Mike C. Fletcher" <mcfletch at rogers.com> wrote in message
news:mailman.1031666838.31065.python-list at python.org...
> Boa Constructor has some code in their UML view that I contributed to do
> this using topological sorts.  The topological sorting algos (one by me,
> one by Tim) are available seperately here (as toposort.py):
>
> http://members.rogers.com/mcfletch/programming/
>
> if you don't want the GPL contamination of using Boa's copy.
>
> Basically, you do the topo-sort, then lay out each generation
> above/below next/previous generations.  For really nice-looking graphs,
> you'd want to do some post-processing to make children show up
> underneath parents as much as possible (i.e. minimise number of crossing
> arcs in the graph), and maybe do some work with spacing so that each
> generation takes up approx. the same amount of space.
>
> HTH,
> Mike
>

Given me something to think about.  Cheers Mike.

Duncan




From gua81 at XXXyahoo.com  Fri Sep 27 04:40:13 2002
From: gua81 at XXXyahoo.com (CheapSkate)
Date: Fri, 27 Sep 2002 18:40:13 +1000
Subject: Making directory in cgi!
Message-ID: <an15gg$s3h$1@lust.ihug.co.nz>

HI,
let say my createdir.cgi is in
/bug/ger/createdir.cgi
 and doing ls in /bug/ger/ will list out
.
..
(file)createdir.cgi
(dir)homer

how do I create directory inside homer?
like homer/simpsons/
???

what are the library that I need to import?

Thanks......
(I know this is a simple Q, so please excuse me... im a newbiee)




From Oschler at earthlink.net  Tue Sep 10 13:48:22 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Tue, 10 Sep 2002 17:48:22 GMT
Subject: Trouble upgrading to 2.2.1
References: <aXof9.29813$w51.7932884@news2.news.adelphia.net> <slrnans81k.jc.gerhard.haering@lilith.my-fqdn.de> <_qpf9.29862$w51.7937790@news2.news.adelphia.net> <slrnans9pd.jc.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <Gfqf9.29954$w51.7943981@news2.news.adelphia.net>

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote in message
news:slrnans9pd.jc.gerhard.haering at lilith.my-fqdn.de...
Ditto for libssl.
>
> HTH,
>
> -- Gerhard

Gerhard,

I managed to get things going by using rpm with the flags "-ivh --nodeps".
I'm beginning to wonder if OpenSLL was installed in a way that it is not
registered in the rpm database.  Python2 runs fine so it is able to find the
symlinks that I created.  Perhaps it's a SuSE Linux/yast2 thing.

thx





From mcfletch at rogers.com  Sat Sep 21 16:24:36 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Sat, 21 Sep 2002 16:24:36 -0400
Subject: Python 2.2.1 compatible with Tcl/Tk 8.4?
References: <yu9965x05vkr.fsf@europa.research.att.com> <j4ptv89zwk.fsf@informatik.hu-berlin.de> <3D8CBFD1.4010808@aon.at> <m3k7lfruq2.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3D8CD584.7070602@rogers.com>

Associated query:
	Anyone compiled TOGL for TCL/Tk 8.4?  I don't see anything on the TCL 
newsgroup or the TOGL discussion lists, and I have one error report that 
it doesn't work on a Linux machine.  TOGL appears to use the tkint file 
from TK, so I'm guessing it's minimally version-dependent, though it 
should, if I understand correctly, build with 8.0.5+.

Anyway, if anyone has built it, and has diffs, I'd like to hear about it 
so I can look into making PyOpenGL compatible with 8.4.

Thanks all,
Mike

Martin v. Loewis wrote:
> Gregor Lingl <glingl at aon.at> writes:
> 
> 
>>1. When working with Python 2.2.1, are there any advantages of
>>using Tcl/Tk 8.4  over the version which comes with Python2.2.1 for Window?
> 
> 
> I don't know; I guess you should refer to the Tcl release announcement
> for details.
> 
> 
>>2. Which steps would I have to perform, to change to the new version?
> 
> 
> You need to recompile _tkinter.pyd. See PCbuild/readme.txt for
> details.
> 
> Regards,
> Martin

-- 
_______________________________________
   Mike C. Fletcher
   Designer, VR Plumber, Coder
   http://members.rogers.com/mcfletch/





From tjreedy at udel.edu  Sat Sep  7 17:48:17 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 07 Sep 2002 21:48:17 GMT
Subject: Smoothing a discrete set of data
References: <3csmdso5.fsf@morpheus.demon.co.uk>
Message-ID: <Buue9.492218$m91.19280936@bin5.nnrp.aus1.giganews.com>

"Paul Moore" <gustav at morpheus.demon.co.uk> wrote in message
news:3csmdso5.fsf at morpheus.demon.co.uk...
> I have a set of data, basically a histogram.

Actually, the two examples you give are *not* histograms.  A histogram
is a discrete frequency graph: how many numbers fit in each bit.  Note
that binning numbers like this them tosses away any order and other
covariate info.  Unimodal histograms are usually smoothed by moment
matching: mean, standard deviation, and possibly more.

>The data is pretty  variable, and I'd like to "smooth" it to find
trends.

This is too vague to be useful.  Most everything in statistics and
data analysis involves "smoothing" of some sort.

>  Actually, this comes up a *lot* for me - some examples:

Of course: see above

> sampled IO rates on a machine
> - I want to look for trends (is IO higher overnight or during the
day,

If you take measurements (samples) every hour, for instance, you have
a time series.  There are many books on this subject alone.  If you
reduce the time variable to a binary day/night variable, then you
would apply methods for two groups of data.  See below.

> etc) or fuel consumption for my car (do I use less fuel since I had
> the service).

This is a standard question with standard methods.  If you ignore
order and other covariates and group measurements as before and after,
a t-test or signed rank test would be appropriate.  If you add a
factor like city versus country driving between gas stops, then you
need analysis of variance.  On the other hand, you could graph milage
between service stops and see if there is a (downward) trend.

> Normally, what I do with things like this is draw a graph, and try
to
> spot the trends "by eyeball". But it feels to me like I should be
able
> to write something which smooths the data out. I just don't see how

As a statistician, I am a fan of eyeballing raw data (with appropriate
caveats about testing what you think you see) in addition to numerical
analysis.  However, each statistical procedure is aimed as answering a
question (and many are based on some assumption about the data).  So
you need to better formulate what you want to know.

Terry J. Reedy





From mertz at gnosis.cx  Wed Sep  4 12:13:04 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Wed, 04 Sep 2002 12:13:04 -0400
Subject: simple xml pickle
References: <mailman.1031104505.31870.python-list@python.org>
Message-ID: <QEjd9kKkXEbH092yn@gnosis.cx>                

Michael Stenner <mstenner at phy.duke.edu> wrote previously:
|  the "old" xml_pickle.py, which appears to not work at all with
|  python2.2, but is otherwise appealing....
|  the "new" xml.pickle as it appears in the Gnosis tools.  This is
|  less appealing because it's very large and distributed over several
|  files...

I think there are some good reasons why the "new" gnosis.xml.pickle is
organized into a package namespace.  And there are some new capabilities
and modularity that go along with that.

However, the "old" xml_pickle (use the latest such one, labelled 0.51)
can be fixed up for Python 2.2 by adjusting two uses of dir().  That
built-in function changed meaning in Python 2.2--or at least it changed
behavior.  Just replace occurrences of 'dir(obj)' with
'obj.__dict__.keys()', and I'm pretty sure all will be well.

Yours, David (the xml_pickle guy)...

--
 mertz@   _/_/_/_/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY:_/_/_/_/ v i
gnosis  _/_/                    Postmodern Enterprises         _/_/  s r
.cx    _/_/  MAKERS OF CHAOS....                              _/_/   i u
      _/_/_/_/_/ LOOK FOR IT IN A NEIGHBORHOOD NEAR YOU_/_/_/_/_/    g s




From fredp at mygale.org.nospam  Thu Sep 19 07:47:04 2002
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 19 Sep 2002 11:47:04 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <96904b50.0209181149.429e8533@posting.google.com>
Message-ID: <Xns928E8CC3E7680PaCmAnRDLM@161.48.128.20>

lpgray at uop.com (Lee Gray) said :

> Thanks for all the suggestions, folks!
> 
> I've collected lots of links to a variety of very nice screenshots,
> and I've been playing with the wxWindows demo and Boa Constructor. Boa
> was a bit disappointing, not as an app, but because of its current
> alpha status - but it's got great potential and I'll definitely look
> forward to future releases. I'll check out PythonCard too.

This just in : the Python Daily URL (http://www.pythonware.com/daily/) 
mentions this collection of classes and tools for UML modeling and 
programming. I don't understand it all, but the sceenshot of the graphical 
model editor sure is a killer :-)

http://www.abo.fi/~iporres/Projects/fog0000000023.html

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


From arosado at softhome.net  Mon Sep  2 22:00:03 2002
From: arosado at softhome.net (Andres Rosado)
Date: Mon, 02 Sep 2002 22:00:03 -0400
Subject: Python to surpass C performance by 2030
Message-ID: <5.1.0.14.0.20020902220001.00beaa40@mail.softhome.net>

At 08:12 AM 8/27/2002 -0400, you wrote:
>learn python today, and in 28 years you will have a job writing
>the speed-critical parts of c programs in python :)

I don't know how to take this, is either as a joke or for real. If it's for 
real, you are doing the same kind of approximation Mark Twain did about the 
lenght of the Mississippi River.


-----------------------------------
Andres Rosado
Email: andresr at despammed.com
ICQ: 66750646
Homepage: http://andres980.tripod.com/

A word to the wise is enough.
                 -- Miguel de Cervantes




From ajs at ix.netcom.com  Mon Sep 16 20:08:55 2002
From: ajs at ix.netcom.com (Arthur)
Date: Mon, 16 Sep 2002 20:08:55 -0400
Subject: Not sure how to phrase this question
Message-ID: <000501c25dde$79a8be60$9865fea9@arthur>

Arthur:
> Would hate to have to resort to truely understanding
> this stuff to get to where I am hoping to get.
>

Emille:

>It would help to know where that is... ;-)

Something that I guess can be called a simple class factory.  I have a
routine that will analyze *args sufficiently to provide information as to
which class to be called from the ClassFactory to which *args is
passed.

In the end I want to create an instance of the called/returned class, not of
the ClassFactory class.

Ran into the behavior of my original post while playing around with
alternative ways to go.

I'll certainly take a simple solution to my problem - though I think some
elucidation of the behavior in my original post might be of some interest to
folks other than just myself, in any case.

I have posted some other questions/comments as to what I am trying to
accomplish, mostly on tutor.  Seem to be getting alot of blank stares.
Suffice it to say that what I think I am doing is designing a user friendly
interface for scripting, geared to the famous non-programmer.

In short, I want to avoid a lot of named arguments, want to have the user
interface with "argument smart" classes (in the manner of how I think method
overloading generally works), going so far as wanting positional arguments
that are in effect order insensitive.  Give a Intersection class
a plane and a line in any order and it will know what you are trying to get
at. No Intersection(plane=planeinstance, line=lineinstance). No error on
Intersection(planeinstance,lineinstance) because the class has be set up to
receive only (lineinstance, planeinstance).

Like most things with Python, its all quite doable. In fact its pretty much
done in a way that I am satisfied is modular and maintainable.

But for this one last area of confusion.

Sorry you asked?

Art













From loewis at informatik.hu-berlin.de  Fri Sep 13 09:27:03 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 13 Sep 2002 15:27:03 +0200
Subject: Calling Python from C# - please help
References: <6de16ac9.0209130508.5d1b56c6@posting.google.com>
Message-ID: <j4elbyt3yw.fsf@informatik.hu-berlin.de>

adila001 at yahoo.com (AdrianC) writes:

> I am using Python 2.1 and I am tring to use my Python COM component.
> From VB6 everything is OK, but in .NET i don't know how to set the
> reference to my COM.

I've never used the COM bridge in .NET, but I believe you need to
create a type library first.

Regards,
Martin


From martin at v.loewis.de  Mon Sep  2 01:41:09 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 02 Sep 2002 07:41:09 +0200
Subject: saving cyrillic text files
References: <mailman.1030031434.27667.python-list@python.org>
Message-ID: <m3elcd7xu2.fsf@mira.informatik.hu-berlin.de>

"Denis S. Otkidach" <ods at fep.ru> writes:

> If you have created python script using GUI text editor (that
> uses cp1251) and opened resulted file with console editor (that
> uses cp866) or vice versa, than you get garbled text.  Try using
> the same editor for both things.

That would explain the observed behaviour. Could the original poster
kindly report whether this is indeed the problem?

Thanks,
Martin

P.S. I really hate OEMCP, and that they use it in the console.


From percy at thunderbolt.ee.itb.ac.id  Fri Sep 13 06:52:51 2002
From: percy at thunderbolt.ee.itb.ac.id (Percy Tambunan)
Date: Fri, 13 Sep 2002 17:52:51 +0700 (JAVT)
Subject: asking to much?
Message-ID: <Pine.BSF.4.44.0209131737390.16557-100000@thunderbolt.ee.itb.ac.id>

hi guys
I want to rewrite the mysql-navigator program using PyQt, it is originally
written (by shi bok jang) using Qt designer actually. I found that it is
quite hard and time-wasting ( that's what i think ).

If any of you have done it, i hope , please ... can i have the source code?
( in PyQt of course ).
please help me guys... i will be so thankful!
bye...





From lordmacro at hotmail.com  Fri Sep 20 06:14:53 2002
From: lordmacro at hotmail.com (Tom Widrick)
Date: Fri, 20 Sep 2002 05:14:53 -0500
Subject: implementing reduce protocol in C
References: <3D8A7DA1.7090506@hotmail.com> <62Ai9.136323$pX1.4852975@news2.tin.it>
Message-ID: <3D8AF51D.5040305@hotmail.com>

Alex Martelli wrote:
> This is stealing references to the five objects you pass as items.
> I'm not sure what effects this might have, but doesn't look good.
> Perhaps it's a contributing factor to strange behavior (though it
> would cause anomalies at pickling time, not when unpickling...).
> 
> Looks like a leak (no decref of args) but that's not the specific
> problem's that's biting you, I think.
> 

Plain as the blank stare on my face, I'm a CPython newbie :)

static PyObject *
proto_reduce(self)
	HyProtoObject *self;
{
	PyObject *args;
	PyObject *result;

	args = PyTuple_New(5);
	if(args == NULL)
		return NULL;

	PyTuple_SET_ITEM(args, 0, self->po_name);
	Py_INCREF(self->po_name);
	PyTuple_SET_ITEM(args, 1, self->po_parents);
	Py_INCREF(self->po_parents);
	PyTuple_SET_ITEM(args, 2, self->po_children);
	Py_INCREF(self->po_children);
	PyTuple_SET_ITEM(args, 3, self->po_doc);
	Py_INCREF(self->po_doc);
	PyTuple_SET_ITEM(args, 4, self->po_dict);
	Py_INCREF(self->po_dict);

	result = Py_BuildValue("(OOO)", self->ob_type, Py_None, args);
	Py_DECREF(args);
	return result;
};

Any better?

> Do you expose this as the __basicnew__ static method of your type?
> I'm not sure how you do so, as the definition of staticmethod
> lives in funcobject.c, not exposed in the .h, I think.  Or, ditto
> for classmethod.  Haven't tried exposing a __basicnew__ method
> from a C-coded type, yet.  If I had to, I think I'd try to run
> from C the Python-level exposed staticmethod built-in -- but since
> as I said I didn't try that yet, I may be missing something.
> Alex

I have stuck it into my methodlist as follows:

static PyMethodDef proto_methodlist[] = {
	{"chparents",	 (PyCFunction)proto_chparents, METH_VARARGS,
	  "change the parents of the object"},
	{"__basicnew__", (PyCFunction)proto_basicnew, METH_NOARGS,
	 "helper for pickle"},
	{"__reduce__",   (PyCFunction)proto_reduce, METH_NOARGS,
	 "helper for pickle"},
	{"__setstate__", (PyCFunction)proto_setstate, METH_VARARGS,
	 "helper for pickle"},
	{0}
};

I see. I did not know that basicnew had to be a staticmethod on my type.
That may have something to do with the problem, and I do not know how
to make a staticmethod on a C type either. In the doc, it says no 
arguments at all, and I assume the only way to do that is to have it a 
staticmethod? I'll scour the source to see if I can find a way to do
that.

How was this done prior to staticmethods? Was there no support for the 
basicnew way of doing it?

Thanks,
Tom




From loewis at informatik.hu-berlin.de  Thu Sep  5 06:11:45 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 05 Sep 2002 12:11:45 +0200
Subject: Why read-only nested scopes?
References: <te5cnu8mlfcr2gfoupii51gs9v2mpnquo6@4ax.com> <j4wuq1d7mh.fsf@informatik.hu-berlin.de> <r2icnusrtibq6ctbn8fl1subugc37p068v@4ax.com> <j4d6rt61nd.fsf@informatik.hu-berlin.de> <m7ncnugbkm0qghf66qcub6e37b64fesdkq@4ax.com>
Message-ID: <j4it1k4ufy.fsf@informatik.hu-berlin.de>

Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:

> >x = 1
> >
> >def foo():
> >  x = 2
> >  def bar():
> >    x = 3
> >    def baz():
> >      global x
> >      x = 4
> >
> >If you invoke baz(), which of the variables is changed? How do you
> >denote that you want to change the others?
> 
> I think that crawling up the scope hierarchy and stopping at the first x
> that one finds (in the above case, the one defined in bar) is a sensible
> solution. 

This solution is not backwards-compatible. global always refers to
globals(), programs may rely on that.

Regards,
Martin


From skip at pobox.com  Tue Sep  3 22:58:04 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 3 Sep 2002 21:58:04 -0500
Subject: Grumble about too strict an attitude about backward compatibility...
Message-ID: <15733.30396.670049.67330@12-248-11-90.client.attbi.com>

I don't want to wake any sleeping dogs, however, as I'm sitting here
watching a bunch of Fortran function names fly by as SciPy builds, I'm
reminded of the fairly recent threads about backward compatibility,
Python-in-a-tie, etc.   Here I am compiling with a Fortran 90/95 compiler
(Sun's Forte thing-a-ma-bob) and see function names like

        lpni:
        klvna:
        chgubi:
        cyzo:
        klvnb:
        rmn2so:
        csphik:

spew forth.  So, while it's great that this same large library compiles and
runs on compilers back to at least Fortran 77 (and probably earlier),
programmers are still stuck with the same cryptic function and data names
they had to deal with 30+ years ago, all in the name of backward
compatibility.

What's the Python connection?  Other than a reminder not to get to slavish
about backward compatibility, I note that these same function names will
then go on to pollute the Python namespace because all this Fortran code is
automatically wrapped using f2py.

-- 
Skip Montanaro
skip at pobox.com



From martin.klaffenboeck at gmx.at  Fri Sep  6 11:14:44 2002
From: martin.klaffenboeck at gmx.at (Martin Klaffenboeck)
Date: Fri, 06 Sep 2002 17:14:44 +0200
Subject: pygtk
References: <pan.2002.09.06.13.30.48.241597.8751@ns.kleinerdrache.org> <mailman.1031320328.24656.python-list@python.org>
Message-ID: <pan.2002.09.06.15.14.43.914664.8751@ns.kleinerdrache.org>

On Fri, 06 Sep 2002 16:51:41 +0200, Michael Stenner wrote:

> On Fri, Sep 06, 2002 at 03:30:48PM +0200, Martin Klaffenboeck wrote:

>> (I think I have questions in both.)
> 
> Perhaps you should subscribe to both lists then.

Ok, so I'll do that.
 
Martin

-- 
     Virtuelle Jugendarbeit auf http://create.kleinerdrache.org
            ...  bald kommt ein neues Layout  ...


From hwlgw at hotmail.com  Mon Sep 30 08:56:19 2002
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 30 Sep 2002 05:56:19 -0700
Subject: XMLRPC calls don't work in threads, or rather, they don't play well with others ! (help)
References: <3D97D190.9030003@bcm.tmc.edu>
Message-ID: <cb035744.0209300456.70c45108@posting.google.com>

[Steve Ludtke]
> Both of the server.()  calls prevent other threads from running

Um, do you really think they prevent other *threads* from running?  Or
would it be the case that there can be only one server call being
handled at any time at a *specific url:portnumber*.


'''
Multithreading will rot your teeth.
        -- Michael Swaine
'''


From chris.gonnerman at newcenturycomputers.net  Fri Sep 20 08:05:28 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Fri, 20 Sep 2002 07:05:28 -0500
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <bd751532.0209082142.54236307@posting.google.com> <allrrs$6hm$2@solaris.cc.vt.edu> <ambpgq$dre$1@nntp9.atl.mindspring.net> <006501c25fd5$ecc80740$0101010a@local> <3D8AAC3C.9020300@cascade-sys.com>
Message-ID: <003001c2609e$04d0fea0$0101010a@local>

----- Original Message ----- 
From: "James J. Besemer" <jb at cascade-sys.com>


> Chris Gonnerman wrote:
> >
> >No smiley this time... I really hate typing curly after 
> >curly.  It makes me wonder that anyone *likes* the freakin' 
> >things so much that they named a LANGUAGE after them.
>
> They are a hell of an improvement over "begin" and "end".

I disagree slightly... I don't have to shift for the literal
"begin" and "end" statements, and I don't have to reach either.
I'm left-handed; my right pinky is probably my weakest finger.
So, while I don't really *like* "begin" and "end" as in 
Pascal, I rarely curse while typing them.

At least the : is right under that finger.
 
> But I agree, leaving them out altogether is even better.

Absolutely.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From news at agapow.net  Sat Sep 28 07:19:56 2002
From: news at agapow.net (Paul-Michael Agapow)
Date: Sat, 28 Sep 2002 12:19:56 +0100
Subject: [Mac] Editor on Macintosh?
Message-ID: <1fj7za2.1erbqps1uvzpj8N%news@agapow.net>

Sadly, it seems like the excellent editor Pepper is going away. (See
<http://www.hekkelman.com>.) I found it the best editor for Python on
the Macintosh, having never been really happy with the alternatives. 

So what are other MacPython using? There are some Java-based editors,
but I'm not persuaded that that they'll perform adequately. There's are
numerous X11 based options, but who wants to swap into X everytime they
need to edit a file? The once excellent editor Alpha seems to have
disappeared permanently into testing for its next release.

Ideas? Anyone? Bueller?

-- 
Paul-Michael Agapow (news at agapow.net)



From duduca007 at yahoo.com  Tue Sep  3 16:33:59 2002
From: duduca007 at yahoo.com (Carlos Moreira)
Date: Tue, 3 Sep 2002 13:33:59 -0700 (PDT)
Subject: basic cgi question
In-Reply-To: <67f6bbe5.0209031225.56b1b7bd@posting.google.com>
Message-ID: <20020903203359.81846.qmail@web11405.mail.yahoo.com>

--- Jeffrey Keil <keilj_33 at yahoo.com> wrote:
> Greetings:
> 
> I've used Perl in the past for cgi programs. I
> thought I'd play around
> a little with Python. I've tried a few experiments.
> 
> I have programs that prints out to the browser.
> Unfortunately, it
> prints out everything, including the "import"
> statements, "print",
> etc. I'm not quite sure what I'm doing wrong.

Send us your source code.

	[]'s

Cadu Moreira
KaduSoft President
--------------------------
[......]

__________________________________________________
Do You Yahoo!?
Yahoo! Finance - Get real-time stock quotes
http://finance.yahoo.com



From oren-py-l at hishome.net  Tue Sep 17 02:24:12 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Tue, 17 Sep 2002 02:24:12 -0400
Subject: More on Protecting Source Code
In-Reply-To: <GCEDKONBLEFPPADDJCOEAEDDFAAA.whisper@oz.net>
References: <GCEDKONBLEFPPADDJCOEAEDDFAAA.whisper@oz.net>
Message-ID: <20020917062412.GA85457@hishome.net>

On Mon, Sep 16, 2002 at 08:26:06PM -0700, David LeBlanc wrote:
> I have read what I consider to be the narrow minded and short sighted
> views wrt to protecting IP in the form of Python source code, and I
> would like to make a few points.

Yes, there are narrow-minded people on both sides of the intellectual
property debate.

> * Not everyone believes that all or most software ought to be free. I 
> for one, don't.

I don't think that software 'ought' to be this or 'ought' to be that. In
other words, I believe in freedom.

> * Owners of property have the right to control it. If you disagree and 
> own a nice car or house, I'll be right over :->

I don't think anyone would disagree with that. The disagreement seems to 
be whether information is property or not. By its nature it isn't but it
can be made more property-like by laws that restrict people's freedom to
use information. 

What I find funny is what people consider to be their 'rights'.

You have the right not to use Python if it doesn't meet your requirements 
for obscuring your source code. You also have the right to modify Python 
to suit your needs because the copyright owners of Python have chosen to 
give you a license to do it. And you also have the right to hire someone 
to do it for you.

The state has given you the 'right' to use the power of its court and 
police to forcibly restrict people that use information you have created 
in violation of copy-restriction laws.

But that is not enough for you. You also want the 'right' to complain about
something that was given to you freely and expect that righteous rants 
about your view of intellectual property laws will somehow compel people 
to modify it to your specification without compensation.

	Oren




From inigoserna at terra.es  Wed Sep  4 13:48:14 2002
From: inigoserna at terra.es (=?ISO-8859-1?Q?I=F1igo?= Serna)
Date: 04 Sep 2002 19:48:14 +0200
Subject: ANN: lfm 0.9
Message-ID: <1031161776.1238.0.camel@lethe.infernus.org>

Hello, 

After 6 months since last release, I bring you a new version (0.9) of lfm.
Read more and download it in next link:

http://www.terra.es/personal7/inigoserna/lfm

lfm is a 'midnight commander'-clone written in Python and licensed under
GNU Public License. It should work on any UNIX with a modern (n)curses
library and Python curses module compiled in. See README file or visit
the web page for more information. 


Changes since version 0.8: 

Version 0.9 ("...and the day arrived") - 2002/09/05:

  + ZIP files can be un/compressed now. ZIP vfs works too
  + added 'rebuild vfs' and 'rebuild_vfs' question / option. Configurable.
    There is no need to wait until vfs file is rebuilt
  + Applied some good patches from Bartosz Oler:
    1. colors customization: Now you can customize the colors lfm use in
       the configuration file, 'colors' section. Each color is defined by
       a string with its name. It looks like this:
	      element foreground_color background_color
    2. Allow preferences values to contain colons in the configuration file.
    3. Lack of a bookmark's definition shouldn't be an error.
  + pyview: new features:
      - read from stdin
      - go to / set bookmarks
      - open shell
  + pyview doesn't show blank screen a the end of the file, now it shows last
    line
  + lfm works now with "from __future__ import division", prepared for Python 3.0
  + set python2 as default interpreter for setup.py, lfm, lfm.py, pyview and
    pyview.py
  + my email address has changed
  + Bugs fixes. See ChangeLog for complete list
      - lfm:
          * create temporary files with mask 0066 and directories with perms
            0700, so only owner can read/write them
          * avoid zombie processes when using 'fork' (now fork twice or threads)
          * show correct filename when an error occurs while uncompressing file
          * don't compress '..' directory
          * when lfm exits after checking command line options, make 'lfm' shell
            script don't show path error
          * when un/compressing files cursorbar must remain in the same file
          * when, at start, lfm can't enter into a dir due to directory permissions
          * crash if len(line) == width of cursorbar window
		  * don't append '*' to historic in entries
		  * fix cursor bar position after sorting
      - pyview:
          * when user hasn't permissions to read file, exit gracefully instead
            of crashing
          * last char in file is not showed


As always, comments and suggestions are welcome. 

Best regards,
I?igo Serna
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 196 bytes
Desc: Esta parte del mensaje esta firmada digitalmente
URL: <http://mail.python.org/pipermail/python-list/attachments/20020904/c7d5bef3/attachment.sig>

From jacek.generowicz at cern.ch  Tue Sep 24 05:20:34 2002
From: jacek.generowicz at cern.ch (Jacek Generowicz)
Date: 24 Sep 2002 11:20:34 +0200
Subject: instance creation
References: <265e380f.0209240052.4c2988f3@posting.google.com>
Message-ID: <tyffzvzwxp9.fsf@pcitapi22.cern.ch>

paulmg at digitalbrain.com (Paul MG) writes:

>     def load(self, id):
>       i = open("appointment"+str(id))
>       self = pickle.load(i)

> - Assigning to 'self' in the load method seems to really not work!

This associates the object returned by "pickle.load(i)" with the local
variable "self", and in the process removes any previous association
between "self" and an object; in your case, the association
established when the "load" method was called with "self" as an
argument. The only way to affect the object referred to by "self"
through "self" is by assigning to its attributes (or invoking member
functions which do this) ... along these lines:

tmp = pickle.load(i)
self.name = tmp.name
self.details = tmp.details

Though I suspect you could find neater ways by using a different
approach altogether.



From whisper at oz.net  Mon Sep 23 13:40:23 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 23 Sep 2002 10:40:23 -0700
Subject: Sessionhandling
In-Reply-To: <c5b3549a.0209230650.34a30fca@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEOEFIFCAA.whisper@oz.net>

http://www-hkn.eecs.berkeley.edu/~dyoo/python/ has an example of using
cookies with python.

I found it by entering "python cookies" (without the quotes) in yahoo. A
bunch of other links came up too.

HTH

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of beachboy
> Sent: Monday, September 23, 2002 7:50
> To: python-list at python.org
> Subject: Sessionhandling
>
>
> Hi!
> Please can anyone help me.
> My problem:
> I'm trying to program a web application with python. This tool has to
> provide diffrent languages. The user has to choose the language at the
> beginning. So i have to save the information which language the user
> chooses. I think this is a perfect example for using sessions.
> The only problem is I've got no idea how to realise this in python.
> Please help me as soon as you can because I have to solve this problem
> very quickly.
>
> hoping to hear from you soon
> beachboy
> --
> http://mail.python.org/mailman/listinfo/python-list




From peter at engcorp.com  Thu Sep 26 08:34:23 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 26 Sep 2002 08:34:23 -0400
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <%8ci9.138477$5r1.5541622@bin5.nnrp.aus1.giganews.com> <amltd3$3k$1@glue.ucr.edu> <3D911FE1.1050801@something.invalid> <mailman.1032976178.17709.python-list@python.org>
Message-ID: <3D92FECF.273F335B@engcorp.com>

Mike Rovner wrote:
> 
> "Greg Ewing" <see_reply_address at something.invalid> wrote in message
> news:3D911FE1.1050801 at something.invalid...
> > thp at cs.ucr.edu wrote:
> >
> > Personally, I rather like
> >
> >     a if b else c
> >
> > but Guido doesn't like that either, unfortunately.
> >
> > Perhaps I'll slip it into Pyrex when no-one's looking...:-)
> 
> It's not as readable as one think ;)

Agreed!  If nothing else, it fails the test of Least Surprise.
You have to stop and think "does the conditional test *really*
lie in the middle of this expression?".

> I, personally, read this as 'take a if b is true else take c' which
> corresponds to 'b? a: c' in C.

Surely that's exactly what he meant?

> So, please, don't create ambiguity from nowhere.

I'm not so sure it's ambiguous, as surprising.  Would anyone
really interpret that differently, after they stopped to think?

-Peter


From jb at cascade-sys.com  Sun Sep 15 03:46:55 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sun, 15 Sep 2002 00:46:55 -0700
Subject: threads or queue for this task
References: <jij5ouglsri5gbbff6ingi6cfjohisaahe@4ax.com> <GmCg9.2354$Ee4.6690@news-server.bigpond.net.au> <cqp7ou8d0jjcdtrqetu9v9ls9gst2cp4qi@4ax.com>
Message-ID: <3D843AEF.3090806@cascade-sys.com>

robin at execulink.com wrote:

>Er, quite. For some reason I had not thought of simply having the
>worker threads look for work. I was thinking more in terms of a Queue,
>where the workers are fed work. But this is likely not even needed. 
>
>In my case I would have them loop indefinitely. Something simple like:
>
>nap = 1
>while 1:
>    try:
>        LookForWork()
>		time.sleep(nap)
>	except MyInterrupt:
>		break
>	except:
>		raise
>
>Having individual threads sleeping will not adversely effect
>performance I imagine. 
>
In general I would say it's not a problem.  So long as they're sleeping, 
they're not making any significant demands on the system.  

On the other hand, you probably DO want to eliminate explicit sleeps 
like this if you can.  E.g., the sleeps force a delay if data arrives to 
be processed during the sleep.  Or even if data arrives while processing 
the current command -- instead of immediately going on to the next 
command, the thread doesn't even look for the next command until after 
it takes  a manditory nap, which in most cases you don't want..

If you are clever, proper design of LookForWork() can automatically 
suspend the threads when they have nothing to do.  E.g., that's one of 
the great features of using a Queue -- threads are put to sleep if they 
read an empty queue and they are woken up automatically when something 
it put into the queue.  Similarly, if your thread is reading a Socket() 
-- it'll sleep until data is available.  Then you don't need the 
explicit sleeps.

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	







From gvdbogae at vub.ac.be  Fri Sep 27 16:40:59 2002
From: gvdbogae at vub.ac.be (Geert-Jan Van den Bogaerde)
Date: 27 Sep 2002 22:40:59 +0200
Subject: Python Process that can survive
In-Reply-To: <571739c3.0209271111.6514c32f@posting.google.com>
References: <mailman.1033137937.24205.python-list@python.org> 
	<571739c3.0209271111.6514c32f@posting.google.com>
Message-ID: <1033159259.1337.5.camel@gandalf>

On Fri, 2002-09-27 at 21:11, Christopher Peery wrote:
[snipped]
> Yeah I found this hack on the web and I can't believe that this is
> only way to do this in python. You're basically stripping the process
> of all it's stream and sending it on its way. This is really messy.

This is the standard way of daemonizing a process under UNIX, copied
(and translated to Python) from W. Richard Stevens' Advanced Programming
in the UNIX Environment. I don't see what's so messy about it.

Why do you care if the standard input, output and error streams are
closed? Since you're closing the terminal they're writing to and reading
from, they're useless anyway. You can reopen them after daemonizing to
redirect them to a file or some other stream, if you want.

Geert-Jan Van den Bogaerde




From shalehperry at attbi.com  Thu Sep  5 20:11:42 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 5 Sep 2002 17:11:42 -0700
Subject: try blocks in Python vs. C++
In-Reply-To: <roy-13BC67.19534805092002@reader1.panix.com>
References: <roy-13BC67.19534805092002@reader1.panix.com>
Message-ID: <200209051711.42404.shalehperry@attbi.com>

On Thursday 05 September 2002 16:53, Roy Smith wrote:
> I'm reading Stroustroup's C++ book, and came upon a piece of advice to
> avoid using try blocks.  I use try blocks a lot in Python and gather
> it's common practice among pythonistas.
>
> Is there something fundamentally different about try blocks in Python
> and C++ which makes them more attractive to use in one vs. the other?
> On the surface, they appear to be virtually identical in just about
> every respect.  Is it simply a cultural difference?  Or just the
> author's own personal style?

If I recall, Stroustrup's advice is to let higher level code deal with errors 
and not to avoid try's entirely.  The python equivalent would be:

def func():
    ...
    ...
    i_may_except
    ...
    ...

try:
    func()
except ....:
    ...

if you write code like:

def func():
    try:
      do this
    except ...:
      ...

    try:
      do that
    except ...:
      ...

it is no better than the old style of testing each and every function call's 
return value.  You spend more time reading and writing error handling code 
than anything else.



From eugene1977 at hotmail.com  Sun Sep 15 05:48:49 2002
From: eugene1977 at hotmail.com (eugene kim)
Date: Sun, 15 Sep 2002 09:48:49 +0000
Subject: python + java for web development?
References: <am0nea$mhd$1@newsreader.mailgate.org> <mailman.1032054377.2006.python-list@python.org>
Message-ID: <am0rlq$nk6$1@newsreader.mailgate.org>

> Of course, CPython is very well suited for web development, too. I'd
> even prefer it, since you don't have to touch Java, then ;-)
> 
> -- Gerhard
> 

i assume cpython is c integration with java in any form
well, my friends will be writing java..
so i need my program can interact with java programs..
assume cpython is not suitable for this..

i'll check jython.. thank you


From adila001 at yahoo.com  Fri Sep 13 09:08:10 2002
From: adila001 at yahoo.com (AdrianC)
Date: 13 Sep 2002 06:08:10 -0700
Subject: Calling Python from C# - please help
Message-ID: <6de16ac9.0209130508.5d1b56c6@posting.google.com>

I am using Python 2.1 and I am tring to use my Python COM component.
>From VB6 everything is OK, but in .NET i don't know how to set the
reference to my COM.
Thanks!


From thomas.heller at ion-tof.com  Fri Sep 27 13:55:16 2002
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: 27 Sep 2002 19:55:16 +0200
Subject: watching mutables?
References: <mailman.1033137396.14482.python-list@python.org> <an211j$oi$1@news.hccnet.nl> <Xns9296C58C9ADFCcliechtigmxnet@62.2.16.82>
Message-ID: <k7l7qpvf.fsf@ion-tof.com>

Chris Liechti <cliechti at gmx.net> writes:

> anton at vredegoor.doge.nl (Anton Vredegoor) wrote in 
> news:an211j$oi$1 at news.hccnet.nl:
> 
> > On Fri, 27 Sep 2002 09:34:55 -0500, "Mark McEahern"
> > <marklists at mceahern.com> wrote:
> > 
> >>l = Watched()
> >>l.append(1)
> > 
> > Thanks for reacting to my question. However, I am looking for
> > something that calls a defineable function if *any* mutable variable
> > in the current scope changes. Variables should be 'marked' to install
> > this function for them. For example 'watch(m,f)' indicates that
> > function f is called if mutable variable m is changed. I do not want
> > function f to be a method of object m since I want object m to be
> > *any* mutable python object and furthermore I do not want to change
> > the orginal object. The call of function f would be triggered by *any*
> > change of object m. This would make debugging (and some other things I
> > am tinkering with) a lot easier. Is it possible?
> 
> see EventObj:
> http://oomadness.tuxfamily.org/p-pyobjtools.php
> 
> chris
> -- 
> Chris <cliechti at gmx.net>
This is what I get when clicking on the download link:

<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN">
<html><head>
<title>404 Not Found</title>
</head><body>
<h1>Not Found</h1>
<p>The requested URL /downloads/PyObjTools-0.1.tar.gz was not found on this server.</p>
<p>Additionally, a 404 Not Found
error was encountered while trying to use an ErrorDocument to handle the request.</p>
</body></html>

Thomas


From brueckd at tbye.com  Fri Sep 27 01:06:46 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Thu, 26 Sep 2002 22:06:46 -0700 (PDT)
Subject: Question about select()
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F06233026397F8@mail.hebmex.com>
Message-ID: <Pine.LNX.4.44.0209262203290.24123-100000@ns0.tbye.com>

On Thu, 26 Sep 2002 sismex01 at hebmex.com wrote:

> Does select() work on bound sockets, to detect
> incoming connections?
> 
> I'd like to be able to have a timeout when accepting
> connections from a server socket.

Yes. When the listening socket is readable there is a new socket waiting 
to accept:

>>> from socket import *
>>> import select
>>> s = socket(AF_INET, SOCK_STREAM)
>>> s.bind(('127.0.0.1',5555))
>>> s.listen(1)
>>> select.select([s],[],[],1.0)
([], [], [])
>>> select.select([s],[],[],1.0)
([<socket object, fd=3, family=2, type=1, protocol=0>], [], [])

-Dave




From cliechti at gmx.net  Tue Sep 10 19:10:00 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 11 Sep 2002 01:10:00 +0200
Subject: Maybe a dumb question
References: <3D7E6B0E.4060709@caramail.com>
Message-ID: <Xns9286BDDE1F08cliechtigmxnet@62.2.16.82>

Fabien HENON <fabien.henon at caramail.com> wrote in 
news:3D7E6B0E.4060709 at caramail.com:

> I have written an editor for Povray using Python, Tkinter and Pmw
> 
> It is available at :
> 
> http://pyvon.sourceforge.net
> 
> The software includes one main file and 6 others which are for the 
> language customization. People can choose between French, English, 
> German,...
> All the translations for these languages are stored in files called 
> menu_francais.py, menu_english.py, menu_deutsch.py ....when the editor 
> is started.
> 
> I would like to clean the directory a bit and put all these 'languages 
> files' into a sub-directory called Lang.
> How do I tell python to look into a particular sub-directory when 
> importing a file.
> 
> I tried the os.path.join syntax, but to no avail.
> 
> Below is the bit of code that imports the language file from the current 
>   working directory :
> 
> 
> try :
>    homedir = os.path.expanduser('~')
>    fd = open(homedir+"/.pyvonrc", 'r')
>    for line in fd.readlines():
>       if line.startswith("Language :"):
>      line=line[11:]          
>      if   line.startswith("francais"):from menu_francais import *
>          elif line.startswith("english"):from menu_english import *
>          elif line.startswith("deutsch"):from menu_deutsch import *
>          ...............

do you have considered the ConfigParser module? it would make it easy to 
parse a config file...

and for your problem, if you avoid "from x import *" , which isn't 
recomended for broader usage anyway, you could use something like that:


config = ConfigParser.ConfigParser()
config.read([os.path.join(homedir, "/.pyvonrc")]) 
try:
    	lang = __import__(config.get("Environment", "Language"))
except ImportError:
    	print "Wanrning: could not load language file, using default"
    	lang = __import__("english.py")


which would require an entry like that in the config file:
--------
[Environment]
Language=deutsch
--------

you could also look at other ways to localize your program.
i think "locale" and "i18n" should help in a google search.

chris



-- 
Chris <cliechti at gmx.net>



From jingguoming at yahoo.com.cn  Wed Sep 11 22:29:00 2002
From: jingguoming at yahoo.com.cn (jgm2163)
Date: 11 Sep 2002 19:29:00 -0700
Subject: how can i make a executable program with a python2's script
References: <dd083388.0209060003.1188000c@posting.google.com> <3D7924BF.805@motorola.com> <mailman.1031351050.14744.python-list@python.org>
Message-ID: <dd083388.0209111829.671b2a1b@posting.google.com>

Michael Stenner <mstenner at phy.duke.edu> wrote in message news:<mailman.1031351050.14744.python-list at python.org>...
> On Fri, Sep 06, 2002 at 04:57:19PM -0500, Stephen Boulet wrote:
> > You can put
> > 
> >     #/usr/bin/env python
> > 
> > at the top of the file and do a chmod +x file. Is this the answer to the 
> > right question?
> 
> Although since you're trying to use python 2.2.1 on a RHL 7.2 box, you
> may need to make that
> 
>  #!/usr/bin/env python2
> 
> Otherwise, you'll get python 1.5.2, which can lead to confusing errors.
> 
> 					-Michael
> > jgm2163 wrote:
> > > os     :RedHat 7.2
> > > Python :Python2.2.1
> > > 
> > > now,i had writen a python's script (extend with .py)
> > > how can i make a executable program on RedHat 7.2?
> > 
> > 
> > -- 
> > http://mail.python.org/mailman/listinfo/python-list

sorry,i want to make a release program on the computer that donot
install python2.1 (because that python1.5 was installed on the redhat
7.2 defaultly).


From whisper at oz.net  Tue Sep 17 06:13:19 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 17 Sep 2002 03:13:19 -0700
Subject: More on Protecting Source Code
In-Reply-To: <am6qv1$7q8$1@newstoo.ericsson.se>
Message-ID: <GCEDKONBLEFPPADDJCOECEEJFAAA.whisper@oz.net>

EXCUSE ME, but I'm not criticizing anyone! I am suggesting that Python could
be better. I've also been honest enough to say that I don't know what or how
that could be. Python didn't get where it is because people impuned the
motives who had things to say about it.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Frithiof Andreas
> Jensen
> Sent: Tuesday, September 17, 2002 2:07
> To: python-list at python.org
> Subject: Re: More on Protecting Source Code
>
>
>
> "David LeBlanc" <whisper at oz.net> wrote in message
> news:mailman.1032232999.22661.python-list at python.org...
>
> > * Why invest a substantial amount of time and money developing in a
> language
> > that makes it trivial to gain access to the work product?
> >
>
> Why indeed - why do you bother?
>
> Why is existing Copyright law not sufficient for you?
>
> > beyond those that
> > are fairly closely licensed and/or have substantial parts of the app
> written
> > in C/C++.
>
> Do that then - there is nothing intrinsic in the Python licensing that
> restricts what you can do with it; I seem to recall that you can
> pretty much
> do whatever you bloody like, including wrapping Python in your own product
> and selling it!
>
> > A better solution is to make a .pyc file approximately as hard
> as a binary
> > .exe file to decompile - however that could be done.
>
> And according to the generous license awarded to you by the very people
> whoms work you are critisizing, *you* can indeed do that; you are in fact
> *encouraged* to make you own enhancements to Python and release them.
>
> Simple Darwinism will show whether others feels the same way.
>
> One could, in theory, package the application with a custom build of the
> Python interpreter so that each interpreter uses it's own unique set of
> bytecodes. Performance will likely suck and you will have a real
> issue with
> upgrades, extension modules and maintenance of the many unique
> applications
> shipped - but again this is your choice to make.
>
> Personally, I do not care - I like Python exactly the way it is, including
> the "openess", since that is an essential part of simplifying the
> development task. If I was really, really worried about some
> "IP-leakage" I
> would stick the core algorithm in an extension library - that
> would not stop
> someone from just ripping the software though.
>
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list




From see_reply_address at something.invalid  Tue Sep 24 22:06:19 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Wed, 25 Sep 2002 14:06:19 +1200
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <mailman.1031377930.4225.python-list@python.org> <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com> <3D8E6353.7030303@something.invalid> <dvbtouciidaufjneja3s4t89ds0r7jd76k@4ax.com>
Message-ID: <3D911A1B.1000109@something.invalid>

Courageous wrote:

> But makes impossible certain constructs that involving lining
> indentation level up with specific items of text in surrounding
> lines


Actually, it doesn't. You use tabs to get up to the
current statement indentation level, and then spaces
after that for the purpose of lining-up.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From cliechti at gmx.net  Tue Sep 24 15:03:18 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 24 Sep 2002 21:03:18 +0200
Subject: Nokia 3310 SMS Application, done:connecting to serial port(linux/win), done:reading some hex like \x00 . . ., problem:write/send command! noresponse
References: <52facf1d.0209190533.3048e62c@posting.google.com>
Message-ID: <Xns9293D63A140D6cliechtigmxnet@62.2.16.82>

ptbjrii at hotmail.com (Polerio Babao Jr.II) wrote in 
news:52facf1d.0209190533.3048e62c at posting.google.com:
> I wanted to create an SMS application using Nokia 3310 with
> DAU-9P(data cable), I am now able to get some data from the cell
> phone. I'm using Pyserial and Siomodule. I use pyserial in my linux
> system. Out of curiosity I've even tested the pyserial on my mouse to
> prove that I'm really reading data from the external device connected
> on my com2 or /dev/ttyS1 on linux. From this you would perhaps believe
> that I've done the first phase of my plan.
... 
> When I connect to the port for the first time I get the ser.read()
> output like this \x00\x0c I think. But that's two response is what I
> get on my first connect. After that when I envoke ser.read() again, I 
> get nothing.

had those too, that comes from the port opening

> The problem is, this still gives no response, when I type
> ser.write('ATCGMI') or any other command, I get no response. Please
> correct me if I'm wrong for using the Nokia At command on 3310, I
> doubt that this At command applies only to some device of nokia
> products like 6110 series, 71.. etc.

1st of all, i don't think that your phone understand AT commands at all...
my 3210 doesn't. instead there are two other protocolls: fbus and mbus

both protocols are binary. and to be able to talk to the phone i had to set 
RTS to low! complely counter intuitive...

i have played around with fbus. i can send the sources to anyone 
interested. i was able to read out the phone book and encoded (7bit 
compressed or so) SMS, but it was some months ago. the sources are not 
cleaned up but the low level part is usable.

> I also tried writing to Mr. Frank Stajano, but it's already 2 weeks
> and still I get no response. Perhaps his busy or doesn't want to talk
> about this matter. Frank Stajano had already created an application
> like this for his research and I think he even asks for help in this
> group. His output an SMS python scripts is also not available.

i'd be interested too, if you hear something.

> The Thing that I really want to know is what kind of command will I
> use to get some response from the nokia 3310? Is it also a hex
> commands in which I will envoke \x00,.. etc commands. Or is it a flat
> ASCII file also. I hope you could share some idea which regards to
> this. Any idea regarding this would be of great help to me, I hope you
> could give me one. Thanks

search for gnokii and you can have my sources for a start. the protocols 
are not published, but the people at that project did some reverse 
engineering.

sorry for the late response, but i completely missed that thread, just saw 
it in the weekly summary.

chris

-- 
Chris <cliechti at gmx.net>



From henrik.motakef at web.de  Sun Sep 15 06:45:05 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 15 Sep 2002 12:45:05 +0200
Subject: regular expression help [port]
References: <qPYg9.7377$Ee4.17838@news-server.bigpond.net.au>
Message-ID: <87wupnsf9q.fsf@pokey.henrik-motakef.de>

":B nerdy" <thoa0025 at mail.usyd.edu.au> writes:

> Greetings, this is a regular expression for an email address
> "^[a-z0-9_]+@[a-z0-9\-]{2,}\.[a-z0-9\-\.]{2,}$"

No, it's not. The format of email addresses is a lot more complicated
than this. Just for a start, you don't allow a dot before the @, so
my address wouldn't match.

J. Friedl, author of "Mastering Regular Expressions", wrote a perl
script that generates a regular expression to check email addresses
for syntactical validity, see
<http://examples.oreilly.com/regex/readme.html>. Is is huge and
ugly. Additionaly, it is almost useless, because a regex cannot tell
you if an email address really exists, which is most often what you
want to know.

The only sane way to make sure that an email address really exists is
to send an email to it and to wait for a reply.

> i wish to valid a variable against this string. would someone be so kind to
> supply a few lines of code to do this?

>>> import re
>>> pattern = re.compile("^[a-z0-9_]+@[a-z0-9\-]{2,}\.[a-z0-9\-\.]{2,}$")
>>> pattern.match('":B nerdy" <thoa0025 at mail.usyd.edu.au>')
>>> pattern.match('thoa0025 at mail.usyd.edu.au')
<_sre.SRE_Match object at 0x8679f40>

Regards
Henrik


From tjreedy at udel.edu  Thu Sep 12 08:57:26 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 12 Sep 2002 12:57:26 GMT
Subject: Overloading the ">>" operator in Python 2.2.1
References: <fhYf9.1418$b5.128061@news02.tsnz.net>
Message-ID: <Wa0g9.597233$2p2.24664063@bin4.nnrp.aus1.giganews.com>

"Andy Elvey" <andy.elvey at paradise.net.nz> wrote in message
news:fhYf9.1418$b5.128061 at news02.tsnz.net...
> Hi all .
>    I'm using Python 2.2.1 under Win98.  I'm doing some experimenting
with
> parsing, and as part of that, I'm trying to overload the right-shift
> operator (>>) so that I can use it as follows (testing against a
sequence
> that I read in) -
>   a >> b   would mean  "a is followed by b" .
>    So,  you could have a function like seqtest("string_to_test" ,
> "foo">>"bar" ) , and that function would return true or false,
depending on
> whether foo was followed by bar.
>  (  I know it's possible to use regexes for this,  but being able to
create
> a sequence-testing operator like this would be useful .)
>     Many thanks in advance ......

You did not explicitly ask a question, but here's a possible answer
anyway:

>>> class wierd(str):
...   def __rshift__(self,other):
...     print self, 'followed by', other
...
>>> foo = wierd('foo')
>>> bar = wierd('bar')
>>> foo>>bar
foo followed by bar

You cannot add attributes to built-in types.  You can only subclass
them (in 2.2+).

Terry J. Reedy






From brian at sweetapp.com  Thu Sep 26 13:51:33 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Thu, 26 Sep 2002 10:51:33 -0700
Subject: unicode woes
In-Reply-To: <amuhrk$4cn$05$1@news.t-online.com>
Message-ID: <00c501c26585$5a328030$df7e4e18@brianspiv1700>

Ulli Stein wrote:

> Some weeks ago the management decided to deploy unicode so that we can
> handle every charset uniformly. The problem which showed at first: we
> realized that we cannot change the encoding in site.py afterwards,
i.e. we
> have to specify the encoding all the way.

Why do you want to change the encoding of site.py at all?

> It got worse and worse because the unicode encoding spread like a
virus
> through all the source code.

That's exactly what should happen. You should minimize your use of
strings in a Unicode-aware application. 

> Furthermore, in Python you have no way to detect, which encoding a
special
> unicode string atually is.

I have no idea what this means.


> We ended up having our own u() function. 

This function is insane. Why do you think that you need it?

Maybe you could talk a bit more about what your application is supposed
to do. I've written large Unicode-aware applications in Python and it
wasn't that difficult (at least the Python aspect wasn't).

Cheers,
Brian




From SBrunning at trisystems.co.uk  Tue Sep 17 09:57:13 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 17 Sep 2002 14:57:13 +0100
Subject: Good book choice?
Message-ID: <31575A892FF6D1118F5800600846864DCBD7F6@intrepid>

> From:	Erik Price [SMTP:erikprice at mac.com]
> Plus, you can't tell me that the inclusion of a CD and the 
> envelope/publishing used to include the CD doesn't add a few units of 
> currency to the book's cost.
 
Certainly does in the UK - books with cover mounted CDs attract VAT (sales
tax), whereas books without do not.

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




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



From gerhard.haering at gmx.de  Fri Sep 13 23:40:28 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 14 Sep 2002 03:40:28 GMT
Subject: Part Time Help Needed
References: <uo5bbmrtuccc2d@corp.supernews.com>
Message-ID: <slrnao5btb.8ni.gerhard.haering@lilith.my-fqdn.de>

Demi Vitalis wrote in comp.lang.python:
> Part Time Work at Home: PHP Programmers
> [...]

X-Complaints-To: abuse at supernews.com

Done.

-- Gerhard


From martin at v.loewis.de  Mon Sep  2 01:16:39 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 02 Sep 2002 07:16:39 +0200
Subject: Converting Unicode to integer
References: <f003aff3.0208270106.116fe6f9@posting.google.com> <3D6B68D6.EBB57001@engcorp.com> <akfvki$eot$1@bagan.srce.hr> <akg6a8$ahe$1@mailgate2.lexis-nexis.com> <akgd1s$lmf$1@bagan.srce.hr> <0vQa9.122140$Aw4.5151890@bin2.nnrp.aus1.giganews.com> <akhq5k$c85$1@bagan.srce.hr>
Message-ID: <m3ofbh7yyw.fsf@mira.informatik.hu-berlin.de>

"Nikola Plejic" <Nikola.Plejic at pu.CARNet.hr> writes:

> Thanks, but this still doesn't work - I get a result of 55 instead of 10...
> :(
> 
> Any ideas, anyone?

Can you please obtain and report the output of the following statements?

def handler(**kw):
    a = txt.text
    b = txt2.text
    print type(a),repr(a),type(b),repr(b)
    a = int(a)
    b = int(b)
    print type(a),repr(a),type(b),repr(b)
    c = a + b
    print type(c),repr(c)
    c = str(c)
    print type(c),repr(c)
    lbl.text = c

Regards,
Martin


From gerhard.haering at gmx.de  Tue Sep 24 13:18:27 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 24 Sep 2002 17:18:27 GMT
Subject: PicklingError with md5.md5() object Von: Ulli Stein <mennosimons@gmx.net> Datum: Dienstag, 24. September 2002 17:03:09 Gruppen: comp.lang.python Folgenachricht an: mennosimons@gmx.net Keine Referenzen
References: <amq1td$eli$07$1@news.t-online.com>
Message-ID: <slrnap17j2.j7.gerhard.haering@lilith.my-fqdn.de>

Ulli Stein wrote in comp.lang.python:
> Hi,
> 
> while trying to pickle a md5 instance I get the following error:
> 
>>>> import cPickle, md5
>>>> f=open("/tmp/t", "w")
>>>> p = cPickle.Pickler(f)
>>>> m=md5.md5()
>>>> p.dump(m)
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> cPickle.UnpickleableError: Cannot pickle <type 'md5.md5'> objects
> 
> 
> What should I do?

If you really need this, you'll probably have to patch md5 (and for
additional bonus points, sha and hmac, too) to support pickling. I
think, however, that this will be far from a trivial task, as you'll
have to dig through the internals of the MD5_CTX struct.

Do you really need to pickle md5 objects? I currently can only imagine
use cases where pickling full digests (from the .digest() or
.hexdigest() methods) makes sense. Do you really need to pickle
half-computed digests?

-- Gerhard


From cs26 at tecco.at  Fri Sep 13 04:52:49 2002
From: cs26 at tecco.at (Chris)
Date: 13 Sep 2002 01:52:49 -0700
Subject: concurrent programming
Message-ID: <fe31142f.0209130052.33174fcc@posting.google.com>

Hi!

I have a problem in my project I startet a few weeks ago. The
requirement is to program a Task Scheduler who gets a Python script
with the tasks spezified. So I have to write some extensions in C++
which provide the functionality. So far so good, now comes the
problem. It should be possible to write something like:
<code>
PARBEGIN
    if ....
        exeute some shell scripts on distributed computers
        some method calls from the extension module, e.g. write in a
database
    do something else
    while ...
    ...
PAREND
</code>

This means that the code bracketed in PARBEGIN PAREND should be
executed parallel. The contrary would be SEQBEGIN SEQEND for
sequential execution.

The Task Scheduler is for testing some programs, so the scripts should
be easy to write and not too complicated. My idea would be to handle
this with a call of a function from the extension module which gets a
function pointer to a method with the code inside PARBEGIN PAREND as a
parameter, so that I can do a callback from the C++ module e.g.:
<code>
def f():
    if ....
        exeute some shell scripts on distributed computers
        some method calls from the extension module, e.g. write in a
database
    do something else
    while ...
    ...

exec_par(f())
</code>

But how can this be executed concurrent? Is there another possibility
in Python?

thx


From sholden at holdenweb.com  Thu Sep 26 21:23:33 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 26 Sep 2002 21:23:33 -0400
Subject: Newbia alert .. working my way through Easy Python Montana Edu page, need help..
References: <MPG.17f540bc75626c66989683@news.worldonline.nl> <8b8gma.dj3.ln@ix.netcom.com>
Message-ID: <xnOk9.1$x_6.0@fe03>

"Dennis Lee Bieber" <wlfraed at ix.netcom.com> was reckless enough to opine...
> Highwaykind fed this fish to the penguins on Friday 20 September 2002
> 06:27 am:
>
> > I'm asked ( on the website : www.honors.montana.edu\_jjc\easytut
> > \node10.html ) to write a small programm to calculate rectangle,
> > square and circle thingies( forgot the English word .. ) and I came up
> > with this code :
> >
The word, of course, is "areas".

>         <snip>
>
> > But the thing keeps telling me width is not defined, also I can;t use
> > the ELIF command .. THink I could do with some help :)
> >
I saw Alex's post, and thought all his advice very sound. Nice of you to
reply, by the way, telling him it was all working. Even the martellibot
likes to know it's appreciated :-).

>
>         Where to begin... I see way too many errors to intersperse
corrections.
>
> 1)      Python indentation controls block structure!

While true, this is not by ityself helpful. Alex's fairly exact pointer
about the indentation of the print statement, plus conjecture that this
might solve the elif problem apparently did the trick.

> 2)      You never assign values to those names it is complaining about --
> rather you are inputting to "a", "b", etc. but never passing /those/ to
> the functions.
>
>         Compare:
>
> import math     #get the long definition of PI
>
> def area_rect(width, height):
>         return width * height
>
> def area_square(side):          # squares are special rectangles
>         return area_rect(side, side)
>
> def area_circle(radius)
>         return math.pi * (radius * radius)      # r*r is likely faster
than r**2
>
Your comment might be important in the context of a program that had to
compute the area of three gazillion circles, but is definitely premature
optimization as far as the original poster is concerned.
>
> option = "prompt"       #filler for first pass
> while option != "q":
>         if option == "r":
>                 w = input("Width")
>                 h = input("Height")
>                 print "Area of Rectangle is ", area_rect(w, h)
>
>         elif option == "s":
>                 s = input("Side")
>                 print "Area of Square is ", area_square(s)
>
>         elif option == "c":
>                 r = input("Radius")
>                 print "Area of Circle is ", area_circle(r)
>
>         else:
>                 print "Options:"
>                 print " 'r'     Calculate Area of Rectangle"
>                 print " 's'     Calculate Area of Square"
>                 print " 'c'     Calculate Area of Circle"
>                 print " 'q'     Quit program"
>                 print " <other> Display this help"
>
>         option = raw_input("Option")
>
This is a nice command loop structure, although it would work just as well
if the initial statement were replaced by

    option = None

That probably only makes a difference to certain people, so feel free to
ignore it -- the existing code will certainly not work any worse...

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From martin at v.loewis.de  Tue Sep  3 16:18:52 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 03 Sep 2002 22:18:52 +0200
Subject: https in Python?
References: <ac259413.0209031134.3778c6ac@posting.google.com>
Message-ID: <m37ki2su6r.fsf@mira.informatik.hu-berlin.de>

s_gherman at yahoo.com (Sorin Gherman) writes:

> Is it possible to handle https using Python's libraries? 

Sure. Just use httplib.HTTPS. This requires SSL support in the socket
module, which is only available on Unix, and only if you have OpenSSL.

> If so, what's involved with using Python to read my hotmail
> messages, for instance?

Dunno, since I don't know how you read your hotmail messages.

Regards,
Martin


From g0w1s0 at yahoo.ca  Mon Sep  2 09:01:13 2002
From: g0w1s0 at yahoo.ca (Cesar)
Date: Mon, 02 Sep 2002 09:01:13 -0400
Subject: curses: AttributeError: initscr
References: <un5dhs4f7qlte8@corp.supernews.com> <krkuka.va3.ln@ix.netcom.com>
Message-ID: <un6o8oqmgbol71@corp.supernews.com>

Dennis Lee Bieber wrote:

>> 
>> 
>> I'm running python 1.5/2.2.1 under redhat linux 7.2.
>> 
>         So which is it when you get the error?

both

> 
>         I could be wrong (I've not kept track of when certain changes took
> effect) but does 1.5 understand library directories? On this system
> (Mandrake) curses is a directory with an __init__.py file.
> 
>         What does (from python) dir(curses) return?
> 
>>> dir(curses)
['__builtins__', '__doc__', '__file__', '__name__', 'sys']

the file curses.py was in the same directory where my script was.

this file contained only some testing instructions.

when I removed the curses.py file, all remains ok.

thanks


From op73418 at mail.telepac.pt  Fri Sep  6 13:18:39 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Fri, 06 Sep 2002 18:18:39 +0100
Subject: Difference between static method and a function as a class attribute?
References: <pqchnu40hs1dgqc4bct2shi4lbpl1oe80i@4ax.com> <j4znuvb3nm.fsf@informatik.hu-berlin.de>
Message-ID: <nlohnusn3fi2b2ksltv1kcgpl6h4340sgv@4ax.com>

On 06 Sep 2002 16:20:45 +0200, loewis at informatik.hu-berlin.de (Martin v.
L?wis) wrote:

>Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:
>
>> What are the differences, if any, between 1 and 2?
>
>In 1, test.funtest will be an unbound method; in 2, it will be a
>function.

Haaaa (a light goes on in an otherwise dark and empty skull) so the
function is implicitely mutated into an unbound method! Wasn't aware of
this, thought that Python did nothing at all and that is why I asked -
Always learning new things.

>
>> And if there are any differences do they really matter, that is, does
>> anybody has any test cases where 1 and 2 are not interchangeable?
>
>Sure:
>
>def m():
>    print "called ::m"
>class t1(object):
>    method = m
>
>class t2(object):
>    def m():
>        print "called t2::m"
>
>    method = staticmethod(m)
>
>t2.method()
>t1.method()
>
>gives
>
>called t2::m
>Traceback (most recent call last):
>  File "a.py", line 13, in ?
>    t1.method()
>TypeError: unbound method m() must be called with t1 instance as first argument (got nothing instead)
>
>Regards,
>Martin

Thanks and all the best,
Gon?alo Rodrigues



From tim.one at comcast.net  Wed Sep  4 11:44:17 2002
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 04 Sep 2002 11:44:17 -0400
Subject: quick question about generator.next()
In-Reply-To: <DD431E96-C01C-11D6-BB50-00039351FE6A@mac.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEAGBCAB.tim.one@comcast.net>

[Erik Price]
> A quick question about generators:
>
> If I have a generator object, the first value produced by the generator
> object does not actually get produced until the first time I call
> generator_object.next() , correct?

Correct, although in the most frequent case you don't need to call .next()
explicitly:

    for object in some_generator_function(...):

Calls to next() are generated by "for" statements for iterable objects of
all kinds (generators, sequences, dicts, file objects, ... anything with an
__iter__ method).

> In other words, simply instantiating the generator object doesn't
> "activate" the generator function, rather it just creates it -- only
> when the next() method is called does the generator's internal logic
> "activate", right?

Partly; calling the generator function *does* do some real work, including
evaluating the actual arguments, and setting up the function's formal
arguments as initialized local variables.  The *body* of the generator
function is not entered at this time, though.

> (Where are generator objects getting their next() method from anyway...
> is it implicitly inherited behind the scenes when I use the "yield"
> keyword to turn a function into a generator?)

The PEP explains this.  A generator function returns a generator iterator,
and it's the latter that has a next() method.  A compile-time effect of
"yield" is to light a "hey, this function is a generator-function" flag bit
in the compiled code object.




From sholden at holdenweb.com  Sun Sep  8 15:26:10 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 8 Sep 2002 15:26:10 -0400
Subject: serve_forever() and non-blocking connection
References: <ale100$bj4$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <9xNe9.43496$l_4.8372@atlpnn01.usenetserver.com>

"Colin Thefleau" <quatrelle at gmx.ch> wrote in message
news:ale100$bj4$1 at nets3.rz.RWTH-Aachen.DE...
> Hi
>
> I have a problem programing a small proxy with non-blocking sockets. I
> guess asyncore would be better, but I have no idea how to implement it
> and the application is runing now (just too slow).
> So I would like to use threads. First here a small example of what I use
> for the connection (inside  a function):
>
> http_proxy = HTTPServer(configuration.listen, RequestHandler)
> http_proxy.serve_forever()
>
> Now this serve_forever() process one request at a time.
> Ok this is working but too slow. I tried to to use handle_request()
> instead of serve_forever() in a while loop and put some
>
> thread.start_new_thread(http_proxy.handle_request, ())
>
> in there. But this looks too dirty and apparently does not help much.
>
> So here my newbee question: what would be the (better) right approach to
> be able so serve many connections at the same time? Knowing that each
> process can be quite CPU hungry.
>
> Any advice is greatly aprecieted


Try using the ThreadingMixIn, something like this

    class ThreadingHTTPServer(SocketServer.ThreadingMixIn,
BaseHTTPServer.HTTPServer)
    ...
    http_proxy = ThreadingHTTPServer(configuration.listen, RequestHandler)
    http_proxy.serve_forever()

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                        pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From fgeiger at datec.at  Wed Sep  4 14:30:23 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Wed, 4 Sep 2002 20:30:23 +0200
Subject: [wxPython] ListCtrl: How can I change the column header text at runtime?
Message-ID: <al5jek$eah$1@rex.ip-plus.net>

Hi all,

with an InsertColumn(colnbr, "This is the header's text") I can add a column
with a header text "This is the header's text". How can I change the text at
runtime, say, after a click on an other control? I had no luck with
GetItem()/SetItem(): I can get at the items of the list elements only but
not at the headers.

AS Python 2.1.3, wxPython 2.3.2.1.

Any hints?

Best regards
Franz GEIGER





From k.weinert at gmx.net  Tue Sep 17 12:27:21 2002
From: k.weinert at gmx.net (Karsten Weinert)
Date: 17 Sep 2002 09:27:21 -0700
Subject: Q: handling exceptions
Message-ID: <fa576652.0209170827.4ba29201@posting.google.com>

Hello,
being new to python, I have a question about handling exceptions. I
want to catch an exception which is raised from a module not written
by me, and display a messagebox with the message of the exception
rather than halting the program. The problem is: how can I access the
message string of the exception?
Here is an example:

import mx.ODBC.Windows
import win32api, win32con

conn=mx.ODBC.Windows.Connect("...")
curs=conn.cursor()
try:
  curs.execute("INSERT INTO test VALUES ('abc',1)"
except mx.ODBC.Windows.IntegrityError:
  # how can I set exceptionstr now?
  win32api.MessageBox(0, "Integrity Error", exceptionstr, win32con.MB
OK)
  pass

Any help appreciated,
Karsten.


From oliver.smith at removethisbit.baesystems.com  Mon Sep  9 07:40:12 2002
From: oliver.smith at removethisbit.baesystems.com (Olly Smith)
Date: Mon, 9 Sep 2002 12:40:12 +0100
Subject: Returning lists from python COM servers
References: <3d7731f5$1@pull.gecm.com> <3d7738f5$1@pull.gecm.com> <qFce9.1228$_l1.2030@news-server.bigpond.net.au>
Message-ID: <3d7c87ff@pull.gecm.com>

"Mark Hammond" <mhammond at skippinet.com.au> wrote in message
news:qFce9.1228$_l1.2030 at news-server.bigpond.net.au...
> Olly Smith wrote:
> > uhmmmmm, strangely enough, changing it to
> >
> > ==
> > return [tags,vals]
> > ==
> >
> > on the server side fixed it!
> >
> > Now the question is: WHY!?
>
> 'cos some of the pythoncom code sucks.  Returning a tuple can trick the
> framework into thinking you are actually trying to return the *real*
> values, as well as the underlying CSCODE in the COM call.  Returning a
> list fixes this.  I need to look into this a little deeper, but fear b/w
> compatibility will screw me.
>

OK :)

I'll remember to only chuck lists across the COM boundary.  Just out of
interest, are there any other distributed object architectures that are
better supported by Python? (corba, soap, ......)

Thanks,

Olly




From duncan at NOSPAMrcp.co.uk  Tue Sep 17 05:12:06 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 17 Sep 2002 09:12:06 +0000 (UTC)
Subject: don't understand error message
References: <mailman.1032251060.4906.python-list@python.org>
Message-ID: <Xns928C66E3994D7duncanrcpcouk@127.0.0.1>

Manuel Hendel <manuel at hendel.net> wrote in 
news:mailman.1032251060.4906.python-list at python.org:

> I do the following,
> 
> domains = []
> 
> for line in lines:
>     line = string.strip(line)[1:-1]
>     fields = string.split(line, "|")
>     if fields[3] not in domains:
>         domains.append(fields[3])
> 
> error:
> 
> Traceback (most recent call last):
>   File "./pop3create.py", line 16, in ?
>     if fields[3] not in domains:
> IndexError: list index out of range
> 

It simply means that at least one line of your file doesn't have 4 fields 
after you have finished your processing. If the line only has two "|" 
characters in it, then you will only get three fields which you can access 
with indices 0 to 2. Trying to use an index out of range for a list will 
produce this error.

Try sticking a "print fields" line just before the if statement, the last 
line printed will show you which line didn't match.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From fredrik at pythonware.com  Wed Sep  4 03:14:11 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 04 Sep 2002 07:14:11 GMT
Subject: Gratitude due to Red Hat? More work for the PBF?
References: <Iyn89.11545$aC5.4453@atlpnn01.usenetserver.com> <mailman.1030142061.12228.python-list@python.org> <xuyptw3knfx.fsf@halden.devel.redhat.com>
Message-ID: <7pid9.5888$e5.1033416@newsb.telia.net>

Trond Eivind Glomsr?d wrote:

> A program originally using python 1.5 and newt would, in almost all cases, work
> just as well with python 2.1 if the newt module was available.

unless it depends on what repr returns for longs, uses malformed \x escapes,
passes multiple arguments to list append or the socket functions, depends on
certain undocumented regular expression features, and so on.

or did "A program" really mean "our programs, which we've tested on 2.1" ?

</F>




From peter at engcorp.com  Mon Sep  2 11:53:48 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 02 Sep 2002 11:53:48 -0400
Subject: file processing question
References: <akvg3g$5bn$04$1@news.t-online.com>
Message-ID: <3D73898C.B0AC0D61@engcorp.com>

Axel Grune wrote:
> 
> Hi, Im trying to read the data of a file with following method
> 
> file = open( inputfile, "r")
> data = file.read()
> file.close()
> 
> but 'data' is always smaller then the file size, e.g. I have a file
> which is 25677824 bytes but file.read() reads only 4295 (file.tell() =
> 25677824).
> Im using python 2.2 under win2k.

Jere gave you the solution, but not the precise reason.  Under
Windows, the Ctrl-Z character (ASCII 26) is treated as End Of File
and causes the read to terminate if you open the file in "text"
mode instead of "binary" mode.  It also causes CR/LF combinations
(ASCII 13 and 10, respectively) to be combined into a single
"newline" character which is actually LF again.  If your file
is text but contains Ctrl-Z, you'll have to use binary to read
it in, but if you still want the line-ending conversion to be
performed you'll have to do it yourself.

-Peter


From loewis at informatik.hu-berlin.de  Fri Sep 13 05:19:13 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 13 Sep 2002 11:19:13 +0200
Subject: Question about Tutorial-Unicode
References: <Xns9288683A9F87Agrzegorzrumatowski@130.133.1.4>
Message-ID: <j44rcuuu0e.fsf@informatik.hu-berlin.de>

Grzegorz Rumatowski <grzegorz.rumatowskiNOSPAM at chello.at> writes:

> I can't find any explanation. Mayby anyone of you?

That's a bug in the tutorial. You shouldn't use non-ASCII characters
in a Unicode literal in Python 2.2; this will only work in Python 2.3.

In the meantime, use u"\xe4\xf6\xfc" for that string, or one of the
other notations (\u00e4 or \N{LATIN CAPITAL LETTER A WITH DIAERESIS}).

The example works sometimes, but only in a terminal (not in IDLE), and
only if the terminal's encoding is Latin-1.

Regards,
Martin



From DennisR at dair.com  Wed Sep 18 17:57:43 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Wed, 18 Sep 2002 21:57:43 GMT
Subject: Python GUI app to impress the boss?
References: <96904b50.0209171305.6d65f574@posting.google.com> <Xns928CED692541cliechtigmxnet@62.2.16.82> <C05DA40F31B09329.0165201CAF094801.13AF9332EB64347A@lp.airnews.net> <s4Qh9.479726$q53.16320265@twister.austin.rr.com> <vATh9.666$oY5.33977741@newssvr14.news.prodigy.com> <oY+rxnAGxMi9Ew5+@jessikat.demon.co.uk>
Message-ID: <rF6i9.686$xY4.42252550@newssvr13.news.prodigy.com>

> Well hoping Dennis is genuine and has better control over virii than my
> outdated anti-virals I tried running this 360k hello world.
I think I am genuine.  The base code compiles entirely from source in the
PLD2 compiler language I developed (www.pld2.com); I think I have adequate
control.

> Now I'm confused, I have K-Meleon 0.6 installed as my standard, but
> hellopy.exe tries to use Netscape 4.78 (also installed on my machine)
> and fails to display anything properly (I see the source in the browser
> window).
Netscape instead of K-Meleon is a "registry thing".  The code is looking for
the association of ".htm" and running whatever the associated program is.  I
suspect if you have a local file called eggs.htm, that Netscape comes up
when you double click the file.  Perhaps K-Meleon comes up if you click
spam.html.  I dunno but the program is looking no further than the "open"
verb of .htm.

I have duplicated your experience with both Netscape and Opera on Win98.  I
have neither a fix nor a facile explanation at this time.  I have repeatedly
run both browsers on other CGI code in both Perl and native PLD2.  I suspect
a recent change here is at fault.  I am working on fixing this now and will
let you know.

Thanks very much for the report.
--

Dennis Reinhardt

DennisR at dair.com
http://www.dair.com




From nospam at here  Thu Sep  5 14:04:12 2002
From: nospam at here (Ricardo)
Date: Thu, 5 Sep 2002 19:04:12 +0100
Subject: How to get os.popen subprocess PID?
References: <3D775167.6060307@pp.htv.fi> <1031248567.61579.0@doris.uk.clara.net>
Message-ID: <1031248831.62067.0@doris.uk.clara.net>

> Instead of using os.popen() you could import the popen2 module and use the
> Popen3 or Popen4 classes.
>
> Objects of these classes contain a 'pid' attribute.
>

Sorry, forgot to mention these classes are available on Unix only ...





From hst at empolis.co.uk  Tue Sep 10 06:13:47 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Tue, 10 Sep 2002 11:13:47 +0100
Subject: Pickle problem
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220D0E@hendrix.empolisuk.com>

I'm trying to pickle a complicated data structure which includes a dictionary whose entries are created by (simplified):

a[x] = re.compile(y).match

The pickle fails, however, with the error:

PicklingError: Can't pickle <built-in method match of _sre.SRE_Pattern object at
 0x00838700>: it's not found as __main__.match

I'd be grateful if someone could tell me if it is possible to pickle the data structure and what I'm doing wrongly.

Thanks

Harvey

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.



From shalehperry at attbi.com  Mon Sep 23 13:21:34 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Mon, 23 Sep 2002 10:21:34 -0700
Subject: parsing tabs in a string
In-Reply-To: <006201c26324$4148ae90$025e10ac@calibredd.com>
References: <006201c26324$4148ae90$025e10ac@calibredd.com>
Message-ID: <200209231021.34368.shalehperry@attbi.com>

On Monday 23 September 2002 10:11, Andrew Alzner wrote:
> Hello,
>
> I'm getting to know Python and have a quick question for something I don't
> understand.
>
> I'm trying to parse information extracted from a spreadsheet which is tab
> separated. Here is a simplified version of what I have:
>
> list=['"103-01a-17"\t2\t33\t256\t227\n']
> y=list[0]
> print y.find("\t")
> print y.find("\t",1)
>
> I thought the second .find should return the second occurence of "\t". To
> get the index of the second tab I have to type:
>
> print y.find("\t",13)
>
> This doesn't make sense to me. Can someone help me please.
>

>>> s = 'I      am      a       string'
>>> s.find('\t')
1
>>> s.find('\t',2)
4
>>> s[4]
'\t'
>>> s[3]
'm'

the second number tells find() where to start looking.  It uses the same 
notation as slicing.

>>> s.find('\t', 2, 4)
-1
>>> s.find('\t', 2, 5)
4

so for your code just catch the number find() returns, add one and continue.



From jbublitzNO at SPAMnwinternet.com  Thu Sep 26 00:52:46 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Thu, 26 Sep 2002 04:52:46 GMT
Subject: pyqt 3.4 -- libqtcmodule.so undefined symbol
References: <m265wtubxp.fsf@boiko_linux.demogr.mpg.de>
Message-ID: <3D929290.3090900@SPAMnwinternet.com>

Serge Boiko wrote:
> I've just compiled pyqt v.3.4 on Mandrake Linux 9.0 (rc3) and any pyqt
> applications I tried crashes with the message:

> Traceback (most recent call last):
>   File "/usr/lib/python2.2/site-packages/eric/eric.py", line 6, in ?
>     from qt import QApplication, QTranslator, QTextCodec
>   File "/usr/lib/python2.2/site-packages/qt.py", line 39, in ?
>     import libqtc
> ImportError: /usr/lib/python2.2/site-packages/libqtcmodule.so: 
> undefined symbol: _ZNK9QSGIStyle9classNameEv

> I have python 2.2.1 and gcc version 3.2-1mdk. Compilation ran
> smoothly.
> What's wrong?


Your version of the Qt libs wasn't compiled
with QSGIStyle (and some other similar stuff)
included. See README.Linux for the work around
(you shouldn't need to rebuild Qt, although
that is another possible fix)

If all else fails, try searching the archive
at:

http://mats.gmd.de/mailman/listinfo/pykde

or subscribe/post to that mailing list.

You can also keep an eye on

http://sourceforge.net/projects/pykde

for Mandrake RPMs (for PyQt as well as PyKDE),
but I just checked and those aren't available
yet.

Jim



From bobx at linuxmail.org  Wed Sep 11 21:09:59 2002
From: bobx at linuxmail.org (Bob X)
Date: Thu, 12 Sep 2002 01:09:59 GMT
Subject: [FEEDBACK] Is this script efficient...is there a better way?
References: <VaOf9.1723$Lg2.201506@news2.news.adelphia.net> <mailman.1031781254.1680.python-list@python.org>
Message-ID: <HPRf9.1780$Lg2.247316@news2.news.adelphia.net>

> 1) readlines() loads the entire file into a list so if you have a 30+ mb file 
> you just ate 30+mb of memory.  Try using xreadlines() instead, it reads the 
> file line by line and is much more memory friendly.
Very cool...I had missed that!

> 
> 2) do you expect to find more than one keyword in a particular line?  If not 
> you could save some iterations by stopping the inner line.find() loop as soon 
> as one item is found.
I could but it doesn't matter after the first on a line. How would I 
stop it?

> As a final comment you need to be aware that the more keywords you look for 
> the slower this will be.  However there is not a way to get around that, it 
> is just something to keep in mind.
> 
Yeah...the keywords vary...I have used as little as 1 to about 150...



From see_reply_address at something.invalid  Thu Sep 26 18:54:55 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Fri, 27 Sep 2002 10:54:55 +1200
Subject: conditional expressions
References: <33b491f.0209181820.17b6ea70@posting.google.com> <%8ci9.138477$5r1.5541622@bin5.nnrp.aus1.giganews.com> <amltd3$3k$1@glue.ucr.edu> <3D911FE1.1050801@something.invalid> <mailman.1032976178.17709.python-list@python.org> <3D92FECF.273F335B@engcorp.com>
Message-ID: <3D93903F.5090601@something.invalid>

Peter Hansen wrote:

> If nothing else, it fails the test of Least Surprise.
> You have to stop and think "does the conditional test *really*
> lie in the middle of this expression?".


I think you're misapplying Least Surprise here.
If you actually read it, then it means just what
it says. If it meant something *other* than what
it says, that would be a violation of Least Surprise.

>>I, personally, read this as 'take a if b is true else take c' which
>>corresponds to 'b? a: c' in C.
>>
> Surely that's exactly what he meant?


Yes! You just have to train yourself to believe
your eyes. :-)

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From xozip at hotmail.com  Wed Sep 18 00:04:50 2002
From: xozip at hotmail.com (sensei zeki)
Date: Wed, 18 Sep 2002 04:04:50 GMT
Subject: Parsing Logs...Please Advise
Message-ID: <CXSh9.504829$UU1.83831@sccrnsc03>

I am looking for a good tool to parse system logs.  While I am familiar with
the concepts of several languages, I have never acutally used them directly.

I know phyton is a good text tool, but is it the one I should use?  or
should I use awk, perl, or sed?

Thanks for any advice...

An example log looks like this...
-----------------------------------
Begin
-----------------------------------
Fri Jun 21 14:38:56 2002  Entered ROTATION
Fri Jun 21 14:40:37 2002  New customer has inserted $5.00. (bill)
Fri Jun 21 14:42:30 2002   studio:  animating (#1)...
Fri Jun 21 14:42:30 2002    animation: 1
Fri Jun 21 14:44:07 2002    hwpcl:  dithering...
Fri Jun 21 14:44:09 2002    hwpcl:  downloading...
Fri Jun 21 14:44:13 2002    hwpcl:  done.
Fri Jun 21 14:44:34 2002  Customer has bought 1 print(s) of Pic 0.
Fri Jun 21 14:44:35 2002  Entered ROTATION
Fri Jun 21 15:19:48 2002  New customer has inserted $5.00. (bill)
Fri Jun 21 15:22:34 2002   studio:  animating (#5 warped)...
Fri Jun 21 15:22:34 2002    animation: 15
Fri Jun 21 15:24:01 2002    hwpcl:  dithering...
Fri Jun 21 15:24:03 2002    hwpcl:  downloading...
Fri Jun 21 15:24:07 2002    hwpcl:  done.
Fri Jun 21 15:24:28 2002  Customer has bought 1 print(s) of Pic 0.
Fri Jun 21 15:24:29 2002  Entered ROTATION
Fri Jun 21 16:55:52 2002  New customer has inserted $5.00. (bill)
Fri Jun 21 16:59:04 2002   studio:  animating (#1)...
Fri Jun 21 16:59:04 2002    animation: 1
Fri Jun 21 17:00:27 2002    hwpcl:  dithering...
Fri Jun 21 17:00:29 2002    hwpcl:  downloading...
Fri Jun 21 17:00:33 2002    hwpcl:  done.
Fri Jun 21 17:00:55 2002  Customer has bought 1 print(s) of Pic 0.
Fri Jun 21 17:00:56 2002  Entered ROTATION
-------------------------------------------------------
END
-------------------------------------------------------




From rajarshi at presidency.com  Mon Sep  9 16:31:08 2002
From: rajarshi at presidency.com (Rajarshi Guha)
Date: Mon, 09 Sep 2002 16:31:08 -0400
Subject: a re question
References: <pan.2002.09.09.16.10.38.482129.20940@presidency.com>
Message-ID: <pan.2002.09.09.16.31.05.192526.26649@presidency.com>

On Mon, 09 Sep 2002 16:10:39 -0400, Rajarshi Guha wrote:

> Hi,
>   I have a file with lines of the format:
> 
> 001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23 001 Xya FGh
> 143557789   7 100 09/05/2002 20:23:23
> 
> I am trying to extract the 9 digit field and the single digit field
> immediatley after that.
> 
> When I use Visual Regexp to try out the regexp
> 
> (\d{9,} {3,}\d)
> 
> it highlights the 2 fields exactly.
> 
> But when I use the following Python code I get None:
> 
>>> s='001 Abc D Efg 123456789   7 100 09/05/2002 20:23:23' p =
>>> re.compile(r'(\d{9,} {3,}\d)')
>>> print p.match(s)
>>> None

OK - I got it working using p.split() rather than p.match()
Why does'nt p.match() the line but p.split() is able to split it using
the same regex?


From kaci_tizi_ouzou2000 at yahoo.ca  Tue Sep 10 08:51:17 2002
From: kaci_tizi_ouzou2000 at yahoo.ca (Kaci Tizi Ouzou)
Date: 10 Sep 2002 05:51:17 -0700
Subject: Emiting user programatically
Message-ID: <68599ccc.0209100451.28db2e2a@posting.google.com>

Hi all,

I have a GUI wich launches another process (C code) and
I want to be able to stop/kill the process on a user request.
For this, I would like to emit a user(asynchronous) signal
that will be caught by a signal handler within my C
code. My problem is how can I emit such a signal from 
Python.

Thanks all.

Kaci


From sholden at holdenweb.com  Fri Sep 13 08:08:05 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 13 Sep 2002 08:08:05 -0400
Subject: Newbie Looking for direction
References: <uo23o95eb8tg87@corp.supernews.com>
Message-ID: <0Akg9.3474$WV.3470@fe03>

"mowestusa" <justnotworking at yahoo.com> wrote ...
> I have never programmed in a language before.
> I have developed complex databases with Access (Office 95 version) just
with
> the visual tools.


Languages are the only way. Graphical manipulations such as report
generation in Access are all very well, but you hit the limits sooner or
later.

> I would like to learn a computer language or database as a hobby that I
> would use to keep track of things in my real job.
> I have worked through "Non-programmer's Tutorial for Python".
> I'm reading "Programming Python (First Edition, got cheap)" right now.
>
"Programming Python" is a good book, but its focus is perhaps a little
narrow. There are many online tutorials you can add to your book with, so
you shouldn't *need* to buy anything else.

> Most of the programs that I would like to develop would be relatively
simple
> databases.

Nothing wrong with that. You can create them visually if that's easier. You
just need to learn how to use DB-API modules. For Access I'd recommend
mxODBC, which doesn't have to be licensed for personal use.

> I have Access 95 and Paradox 10 (with runtime).  I'm wondering if I'm
going
> down the right path in learning Python or if I would be better served by
> just learning to use Paradox 10 well.  The one problem I have with Paradox
> or Access is that they seem so slow. Even if you have a simple database it
> takes a long time to start up and run.  I would also like to make things
> that I could share with colleagues who have no interest in programming,
but
> could use some of the database tools I would develop for our line of work.
>
You might also want to cover Gordon MacMillan's installer, to know how to
distribute stand-alone Python applications. If you're going to be creating
databases on the fly then you might have some interesting challenges, but
nothing insurmountable.

> So far it seems like I'm a long way away from being able to be productive
> with Python, but that maybe because my end goal is too complex.
>
By aiming for the stars you do increase your chances of actually getting off
the ground. I would make sure that you plan a series of steps, though, so
you get to check off some accomplishment each day. Otherwise a project can
seem too daunting sometimes.

regards, and good luck.
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From aahz at pythoncraft.com  Wed Sep  4 17:42:16 2002
From: aahz at pythoncraft.com (Aahz)
Date: 4 Sep 2002 17:42:16 -0400
Subject: maximum number of threads?
References: <3d75b761$0$26216$afc38c87@news.optusnet.com.au>
Message-ID: <al5uno$7ff$1@panix1.panix.com>

In article <3d75b761$0$26216$afc38c87 at news.optusnet.com.au>,
Steven <sadams123 at optushome.com.au> wrote:
>
>Is there a maximum number of threads that can be running 'simultaneously' in
>Python 2.2 on Win2k/XP?
>
>I'm choosing between a select or threads based design, threads would be
>easier, but wasn't sure whether I'll exceed some maximum number of threads.
>
>Basically its a voice server, or rather a voice to text server, so
>information produced might be sporadic, or fairly continuous. An arbitrary
>number of subscribers subscribe to the system and get voice information sent
>out as it comes in (and if it matches the language the subscriber wants),
>threads make it simple since each subscription could be separate.
>
>Is there a maximum number of threads, and any opinions on whether the GIL
>and context switching in the interpreter is liable to make it unworkable (or
>a case of 'build and see')?

I'd want the actual voice-to-text processing to be in some kind of
library that releases the GIL.  Other than that, a modern computer
should easily support 20-50 threads, possibly as many as 200.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From rsrchstr at msn.com  Sun Sep 29 19:42:21 2002
From: rsrchstr at msn.com (mike henley)
Date: Mon, 30 Sep 2002 00:42:21 +0100
Subject: idea for a much needed application
References: <576c1752.0209291326.56dcdf31@posting.google.com> <JTKl9.1293$_X3.109133@newsfep1-win.server.ntli.net>
Message-ID: <an8355$bs5bh$1@ID-157863.news.dfncis.de>

"Richard Reynolds" <richiereynolds at GETRIDOFTHISntlworld.com> wrote in
message news:JTKl9.1293$_X3.109133 at newsfep1-win.server.ntli.net...
> check out www.zoneedit.com

Thanks, they only provide dynamic DNS; what i wish for is something to
easily combine dynamic DNS with instant messaging presence, so a personal
webserver can have "presence".




From chris.gonnerman at newcenturycomputers.net  Wed Sep 11 08:48:49 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Wed, 11 Sep 2002 07:48:49 -0500
Subject: Windows/DOS: double clicking a .py file
References: <just-A936EC.17062909092002@news1.xs4all.nl>
Message-ID: <001701c25991$97a38040$0101010a@local>

----- Original Message ----- 
From: "Just" <just at xs4all.nl>


> Sorry if this is a Windows FAQ, I'm a newbie to Windows, but not to 
> Python.
> 
> If I double-click a Python script (or use the start command in the dos 
> prompt) a new dos prompt window is opened and my program is run. The 
> window goes away when the program is done. However, the window _also_ 
> goes away when the program raises an exception, making it virtually 
> impossible to actually _see_ the exception... Is there a way to keep the 
> dos prompt window in this case?

import sys, traceback

try:
    do_something_here()

except:
    # show the error
    traceback.print_exc(file=sys.stderr)
    # save the error
    try:
        import time
        fp = open("my_error_log.txt", "a")
        fp.write("\n%s\n" % time.ctime(time.time())
        traceback.print_exc(file=fp)
        fp.close()
    except:
        pass
    # email the error
    try:
        import smtplib, StringIO
        fp = StringIO.StringIO()
        traceback.print_exc(file=fp)
        conn = smtplib.SMTP("my.email.server.com")
        conn.sendmail("me at my.email.server.com", 
            [ "me at my.email.server.com" ], fp.getvalue())
        conn.quit()
        fp.close()
    except:
        pass
    # hold the screen
    sys.stderr.write("\nPress ENTER to Continue: ")
    sys.stdin.readline()

A couple of notes:

I import traceback at the very top, to prevent any possibility
that it can't be imported later.

I don't import smtplib or time at the top (unless I need them
anyway) since I won't (hopefully) need them all the time.

You mustn't screw with sys.stdin or sys.stderr if you expect 
this to work (yes, you could use msvcrt.getch() on Windows,
which is where you said you are, but I like cross-platform
code).

Code above is untested, but I use similar code in many of my
programs; any error is most likely a typo.

Standard disclaimers apply.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net




From calves at sobral.org  Sun Sep  8 09:11:37 2002
From: calves at sobral.org (calves at sobral.org)
Date: Sun, 08 Sep 2002 13:11:37 GMT
Subject: =?iso-8859-1?Q?Justi=E7a?= Eleitoral apreende mais urnas falsificadas no DF (fwd)
Message-ID: <20020908131137.6914.qmail@linux.sobral.org>

Srs.,
Precisamos de sua ajuda pra impedir que nosso governo corrupto aqui no 
Brasil fraudem as elei??es do pr?ximo dia 06/10/02 pra se perpetuar no 
poder. Divulgem e levem essa mensagem at? os organismos internacionais que 
possam obrigar o governo brasileiro a realizar as elei??es pelo modo 
tradicional (voto escrito). 

Justi?a Eleitoral apreende mais urnas falsificadas no DF (Urnas Eletr?nicas) 


05/09/2002, Linha Aberta 

A fiscaliza??o da propaganda eleitoral do Tribunal Regional Eleitoral (TRE) 
do Distrito Federal apreendeu hoje mais duas urnas eletr?nicas "viciadas", e 
j? fala em 30 pe?as apreendidas, al?m de 80 carca?as para montagem. Segundo 
o presidente do Diret?rio Regional, Wilmar Lacerda, as den?ncias t?m sido 
feitas por militantes petistas ao Diret?rio Regional do PT-DF, que mobilizou 
a fiscaliza??o do TRE. Os equipamentos foram violados e tinham sido 
programados para computar votos para o presidenci?vel Jos? Serra (PSDB) e 
candidatos ligados ao atual governo do PMDB. Segundo o coordenador da 
campanha petista de Geraldo Magela ao governo do Distrito Federal, Eduardo 
Daer, as urnas recusam voto no 13, ou qualquer outro candidato que n?o os 
programados. 

"Este ? um procedimento escancarado do governo Roriz que demonstra n?o ter 
qualquer respeito ? legisla??o, promovendo um desmando total durante o 
processo eleitoral", indignou-se. Segundo ele, as pr?ticas irregulares n?o 
se restringem ? falsifica??o de urnas, mas tamb?m ? chantagem e compra de 
votos por representantes do governo em reuni?es comunit?rias. "As 
ilegalidades s?o tantas, que as pessoas se escandalizam e denunciam, 
promovendo uma mobiliza??o ainda maior da milit?ncia petista", relatou Daer. 

Milit?ncia horrorizada 

"Essas urnas produzidas irregularmente estavam sendo usadas na Brazl?ndia e 
em Sobradinho (cidades sat?lite pr?ximas a Bras?lia) para ensinar as pessoas 
da periferia a votar em candidatos da Coliga??o Serra-Roriz", disse Lacerda, 
declarando que tem informa??o de muitas outras urnas nestas condi??es. 

Daer, por sua vez, relata que fiscais do Ibama (Instituto Brasileiro do Meio 
Ambiente e dos Recursos Naturais Renov?veis) encontraram as 80 carca?as e 
componentes para montagem de urnas no Parque Nacional de Bras?lia. Segundo o 
petista, eles levaram o material ? 3? Delegacia de Pol?cia no Cruzeiro, onde 
o delegado de plant?o tirou as tarjas de identifica??o de "simulador de urna 
eletr?nica", alegando que tudo n?o passava de material de inform?tica. O 
delegado disse aos fiscais para "jogar fora" o equipamento. Eles 
encaminharam, ent?o, o material ? Superintend?ncia da Pol?cia Federal. 
"Nossos recursos s?o poucos, mas a milit?ncia est? escandalizada e reagindo 
?s den?ncias de abuso econ?mico que se proliferam", disse Daer. 

Mal-estar governista 

Segundo divulgou a Justi?a Eleitoral regional, as urnas falsificadas est?o 
programadas para computar o voto presidencial para Jos? Serra (PSDB) e de 
governador para Joaquim Roriz (PDMB-DF), candidato ? reelei??o. Para o 
Senado, seriam computados votos para Paulo Octavio (PFL) e Jofran Frejat 
(PPB). Tadeu Filipelli (PMDB) receberia votos para C?mara dos Deputados e 
Edimar Pireneus (PMDB), para a C?mara Legislativa. Regularmente, os nomes 
constantes das urnas para testes s?o fict?cios. "Os proporcionais s?o todos 
ligados ao governador", disse Lacerda. 

O dirigente do PT informou que o programa eleitoral do PT na TV e r?dio est? 
denunciando o esc?ndalo. A falsifica??o de urnas tamb?m est? sendo 
amplamente divulgada pela imprensa local, causando mal-estar ao processo 
eleitoral do DF. A coliga??o que defende a candidatura de Geraldo Magela ao 
governo do Distrito Federal est? acompanhando o processo. "O TRE e a PF 
abriram inqu?rito sobre o caso e est?o investigando as responsabilidades", 
disse Lacerda. A pena prevista para esse tipo de crime no C?digo Eleitoral ? 
de pris?o de at? tr?s anos, al?m do pagamento de multa. 

Linha Aberta 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020908/c8e0cf30/attachment.html>

From h.baumgartl at chello.NOSPAM.nl  Thu Sep 26 17:51:06 2002
From: h.baumgartl at chello.NOSPAM.nl (Henry Baumgartl)
Date: Thu, 26 Sep 2002 21:51:06 GMT
Subject: lost in Python / Tkinter
Message-ID: <ejLk9.82555$724.7422408@amsnews02.chello.com>

Hello all,

Still trying hard to get to grips with OO concepts in Python2.2.1

i'm testing my efforts on an application with ui's in Tkinter
the idea is to have an application module from where subsequent modules are
de- and activated.

getting the root window, container module (application menu's etc.), and the
first selection module to work is no problem. (Each module is basically a
class based on Frame)

the problem starts when the selection ends and i need to get back to the
container instance to build, for instance, a text editor based on the
selection made.

the selection instance i am in doesn't know root or the container instance,
so i can't get the new module build in the container frame.

Strangely enough, when i use

import from module Class
Class()

the frame and everything in it is built without complaints underneath the
container frame in the root window.

Could anyone please point me in the right direction to deal with my problem.

Best regards,
Henry

Please remove NOSPAM. from my address when answering directly to me.




From gregoryk at futures.wharton.upenn.edu  Wed Sep 18 16:36:13 2002
From: gregoryk at futures.wharton.upenn.edu (GREGORY KNESER)
Date: 18 Sep 2002 20:36:13 GMT
Subject: Calling function from command line
Message-ID: <amao3t$mug$1@netnews.upenn.edu>

Hi,

I have a file func.py which includes a working function definition with some 
inputs and some defaults for those inputs.

I want to type

c:\python22\python func.py func1('input1','input2')

or some such thing and run the function with the inputs I fed it rather 
than the defaults.  Is such a call possible?

I've tried searching the web and this group with little success on this 
topic.

Thanks much,
Greg


From anthony at computronix.com  Fri Sep 13 11:18:37 2002
From: anthony at computronix.com (Anthony Tuininga)
Date: 13 Sep 2002 09:18:37 -0600
Subject: calling Oracle functions
In-Reply-To: <3d81f811$1@news.mt.net.mk>
References: <3d81f811$1@news.mt.net.mk>
Message-ID: <1031930318.472.83.camel@chl0151.edmonton.computronix.com>

Oracle differentiates between functions and procedures so you must
identify that you are calling a function, not a procedure. The DB API
does not allow for this, so you must use an anonymous PL/SQL block and
some non DB API code. The way to do this is as follows:

Replace the data types as appropriate.

v_Vars = v_Cursor.setinputsizes(p_Result = cx_Oracle.NUMBER)
v_Cursor.execute("""
    begin
      :p_Result := sch.pkg.myfunction(:p_Param1, :p_Param2, :p_Param3);
    end;""",
    p_Param1 = 1,
    p_Param2 = "Some string",
    p_Param3 = "A different string")
print "Result:", v_Vars["p_Result"].getvalue()

If anyone has suggestions about how to improve this, fire away... :-)

On Fri, 2002-09-13 at 08:37, ?????? ?. wrote:
> Does anyone know, how can I call Oracle stored functions from Python.
> I'm using cx_Oracle.
> 
> The function takes three parameters and returns one result.
> 
> I tried:
> 
> 	params = ['param1', 'param2', 'param3']
> 	v_Cursor.callproc("SCH.PKG.MyFunction", params)
> 
> and I got the error:
> 
> 	cx_Oracle.DatabaseError: ORA-06550: line 1, column 8:
> 	PLS-00221: 'MyFunction' is not a procedure or is undefined
> 
> But I can execute the function from sqlplus
> 
> 	declare result number;
>     
> 	BEGIN
> 	result := SCH.PKG.MyFunction('param1', 'param2', 'param3');
> 	END;
> 		  
> Any hints?
> 
> 
> -- 
> ??????
> 
>             Yes, I've heard of "decaf." What's your point?
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
-- 
Anthony Tuininga
anthony at computronix.com
 
Computronix
Distinctive Software. Real People.
Suite 200, 10216 - 124 Street NW
Edmonton, AB, Canada  T5N 4A3
Phone:	(780) 454-3700
Fax:	(780) 454-3838
http://www.computronix.com




From aglyportREMOvethispart at nospam.yahoo.com  Wed Sep  4 22:07:18 2002
From: aglyportREMOvethispart at nospam.yahoo.com (Anton Graph)
Date: Wed, 04 Sep 2002 19:07:18 -0700
Subject: Memory fault - core dumped   while compiling python 2.2.1 on
  SCO
References: <3D757EFA.9050401@nospam.yahoo.com> <yu99sn0qqw1r.fsf@europa.research.att.com> <m3bs7etem9.fsf@mira.informatik.hu-berlin.de> <3D765A05.6030004@nospam.yahoo.com> <j48z2h61j9.fsf@informatik.hu-berlin.de> <3D768753.8030302@nospam.yahoo.com> <m3n0qx8j3w.fsf@mira.informatik.hu-berlin.de> <3D76A139.1030304@nospam.yahoo.com> <al69pn$sg$0@216.39.172.122>
Message-ID: <3D76BC56.7020503@nospam.yahoo.com>

Bengt Richter wrote:
> On Wed, 04 Sep 2002 17:11:37 -0700, Anton Graph <aglyportREMOvethispart at nospam.yahoo.com> wrote:
> 
> 
>>Martin v. Loewis wrote:
>>
>>>Anton Graph <aglyportREMOvethispart at nospam.yahoo.com> writes:
>>>
>>>- What is the precise version of the operating system you are using?
>>>
>>
>>SCO_SV bongo 3.2 5.0.5 i386
>>
>>
>>>- I'm troubled by the fact that in both configurations (with an
>>>  without gcc), the compiler is always referred-to as "cc". Are you
>>>  sure you are not using gcc?
>>>
>>
>>I'm positive it's SCO cc:
>>
>>Wed Sep  4 16:17:04 PDT 2002 bongo:/u/src/agl/txb/txt $ gcc -v
>>Reading specs from /usr/local/lib/gcc-lib/i386-pc-sco3.2v5.0.5/2.95.2/specs
>>gcc version 2.95.2 19991024 (release)
>>Wed Sep  4 16:17:07 PDT 2002 bongo:/u/src/agl/txb/txt $ /usr/local/bin/cc
>>bash: /usr/local/bin/cc: No such file or directory
>>Wed Sep  4 16:17:14 PDT 2002 bongo:/u/src/agl/txb/txt $ /usr/bin/cc
>>Usage: cc [ options ] files ...
>>Wed Sep  4 16:17:18 PDT 2002 bongo:/u/src/agl/txb/txt $ /bin/cc
>>Usage: cc [ options ] files ...
>>
> 
> How about
>     cc -v
> ?
> 
> what does
> 
>     which cc
> 
> give you? And
> 
>     which cc | xargs -ld
> ?



$ which cc
/bin/cc
$ /bin/cc -v
Usage: cc [ options ] files ...
$ which cc | xargs -ld
xargs: d is not a valid number of lines, must be a positive integer



From dcswest at flashmail.com  Thu Sep  5 21:20:18 2002
From: dcswest at flashmail.com (Dennis B.)
Date: 05 Sep 2002 18:20:18 -0700
Subject: Privacy and Inheritance
In-Reply-To: <3D776CC2.CA928E38@fluent.com>
References: <mailman.1031112136.4594.python-list@python.org> 
	<3D776CC2.CA928E38@fluent.com>
Message-ID: <1031275229.240.42.camel@localhost>

Hello Mark;

And thank you for that clarification!  Wasn't even aware of a convention
as such and it certainly answers my question!!!

Leads me to wonder why we wouldn't enforce this in Python too with the
interpreter and without workarounds, rather than strictly by having well
behaved programmers, but it does seem good how much Python encourages
behavior as such...

Thank you again,


On Thu, 2002-09-05 at 07:40, Mark Moales wrote:
> Dennis,
> 
> I think the point of the double underscores is to create "private"
> attributes similar to C++'s or Java's private keyword.  In other words,
> only the class that defines the attribute can access it.  Subclasses and
> others can't.  We use a single underscore here to denote "protected"
> attributes, or attributes that are accessible by subclasses.  However,
> the only way to enforce this in Python is by having well behaved
> programmers ;-)  Here's an example:
> 
> class BaseClass:
>     def __init__(self):
>         self.publicVar = 'Foo'
>         self._protectedVar = 'Bar'
>         self.__privateVar = 'Spam'
> 
> class SubClass(BaseClass):
>     def aMethod(self):
>         print self.publicVar
>         print self._protectedVar
>         # Can't do this because it's private
>         # print self.__privateVar
> 
> Mark
> 
> "Dennis B." wrote:
> > 
> > Greetings;
> > 
> > Have enough confidence in Python to realize this is probably just my
> > misunderstanding, but wondering why Python mangles private, double
> > underscore attributes with the name of the base class even if they're
> > inherited and the instance is that of the sub class.
> > 
> > Like the following doesn't even seem to work:
> > 
> > class StandardNormal:
> > 
> >     def letEventZScore(self, eventZScore):
> >         self.__eventZScore = float(eventZScore)
> > 
> > class DerivativeNormal(StandardNormal):
> > 
> >     def getCentral(self):
> >         if self.__eventZScore < 0:
> > 
> > Unless I change that last line to something like:
> > 
> >         if self._StandardNormal__eventZScore < 0:
> > 
> > Maybe just missing the point all together, but wondering if anyone'll
> > point out anything a little more elegant and/or intuitive.
> > 
> > Thank you and have a great day,





From aahz at pythoncraft.com  Thu Sep 19 10:37:02 2002
From: aahz at pythoncraft.com (Aahz)
Date: 19 Sep 2002 10:37:02 -0400
Subject: monetary applications [was: Python GUI app to impress the bos s?]
References: <31575A892FF6D1118F5800600846864DCBD817@intrepid> <mailman.1032362843.10509.python-list@python.org> <23891c90.0209190014.26c5bb9@posting.google.com>
Message-ID: <amcnee$c21$1@panix1.panix.com>

In article <23891c90.0209190014.26c5bb9 at posting.google.com>,
Paul Boddie <paul at boddie.net> wrote:
>
>In that case, what do you think of this module? ;-)
>
>  http://www.boddie.org.uk/python/downloads/ScaledDecimal.py
>
>I'll gladly release this into the public domain, but some commentary
>would be nice - I fear that I may not have implemented it as well as
>the "numeric gods" might like.

Can you clearly explain the differences between your module and
FixedPoint.py?  If not, the numeric gods probably won't accept it.  ;-)
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From gerhard.haering at gmx.de  Sat Sep 21 10:37:52 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sat, 21 Sep 2002 16:37:52 +0200
Subject: pyrhon and oracle
In-Reply-To: <B__i9.505088$q53.17034835@twister.austin.rr.com>
References: <5b8834c2.0209201412.50e77861@posting.google.com> <pan.2002.09.20.23.29.54.67293.2172@verizon.net> <B__i9.505088$q53.17034835@twister.austin.rr.com>
Message-ID: <20020921143752.GA4615@lilith.ghaering.test>

* Lee Gray <lee100 at sport.rr.com> [2002-09-21 14:03 +0000]:
> Satheesh Babu wrote:
> > cx_Oracle might be the *easiest* to use. I use DCOracle while working
> > with Zope, but for standalone development, cx_Oracle is tops.
> > 
> > Download the driver from
> > http://www.computronix.com/utilities.shtml
> > 
> > go to Python DB-API specs
> > http://www.python.org/topics/database/DatabaseAPI-2.0.html
> > 
> > Here's an example using DB-API
> > http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/81189
> > 
> > vsb
> > http://vsbabu.org/
> > 
> 
> I'm just looking into Python, so these articles don't mean much to me yet... 
> Does all this mean that Python doesn't talk to an Oracle DB "out of the 
> box"?

Yes.

> For Windows, can Python use ADO?

Yes.

There are binaries for cx_Oracle, DCOracle2 and DCOracle, though, which
you can download and install in a matter of minutes.

If you have installed an Oracle 8 client or higher, you can use
cx_Oracle or DCOracle2. For Oracle 7 clients, you need DCOracle.

There are also modules for ADO (alpha) and ODBC, but I'd (and I do,
personally) go with one of the native Oracle modules.

-- Gerhard



From magnus at thinkware.se  Mon Sep 16 09:17:31 2002
From: magnus at thinkware.se (=?ISO-8859-1?Q?Magnus_Lyck=E5?=)
Date: Mon, 16 Sep 2002 15:17:31 +0200
Subject: Converting Python app to C++ completely
References: <aklm6c$r1r$1@news.swman.net.uk>
Message-ID: <3D85D9EB.9020102@thinkware.se>

Alan James Salmoni wrote:
> I have written an application (a package for statistical analysis - 
> SalStat) using Python and wxPython, not just because I wanted to, but 
> also to learn the language. I found both Python & wxPython to be superb 
> for the task.

Not surprising ;)

> Now, what I want to do is write the thing in C++. The reason? heh, 
> because I want to ;) Well, actually I want to learn C++ too, and I would 
> like to have the performance of a compiled program for larger data sets 
> (just for fun and comparison more than anything).

<wink>
"Premature optimization is the root of all evil!" - Don Knuth
</wink>

A common approach is to factor out the performance critical
parts (once you have actually measured) and rewrite them in
C or C++. The profiler is a good tool for this measuring.

Writing extension modules is one solution, I'm sure you are
aware of the efforts of SciPy/weave and Pyrex etc as well.
But that's beside the point if the purpose is to learn C++.

For learning C++, I can heartily recommend Scott Meyers' books,
at least if you know some C. See
http://www.thinkware.se/cgi-bin/thinki.cgi/Effective_20C_2b_2b
www.amazon.com/exec/obidos/search-handle-url/index=books&field-author=Meyers%2C%20Scott 

> I know very little about C++, but have a few questions about converting 
> from Python code. I am asking here because Pythonistas are more likely 
> to know C++ than vice versa (and CLP is one of the friendliest groups 
> around - creep, creep).
> 
> 1) How easy is it to convert existing Python code to C++? I read that 
> each line of C++ code is roughly equivilent to 5-8 lines of C++ code, 
> but is the conversion a simple task or is it troublesome?

I programmed a lot in both languages, but I haven't really
done that. While I have glanced at your program, I haven't
really looked through the source. I guess the only true
answer is that it depends a lot on how your code is written.
It's certainly possible to write C++ and Python programs in a
very similar structure.

I imagine it's not so difficult to map wxPython to wxWindows,
and with most logic being arithmetic operations on floats or
integers, I imagine it won't be so bad. I guess you don't
depend so much on the Python Standard Libraries.

Considerig this, I think your C++ program will be less than
five times larger than the equivalent Python program. Maybe
factor three is more likely.

> 2) Are there any pitfalls to beware of - I understand that dealing with 
> memory and pointers might be a problem, but like I said earlier, I just 
> don't really know?

Certainly. C++ is full of pitfalls. I mean, C is well known for
all it's pitfalls, and while C++ fixes a few of those, it adds
a lot more complexity to the language.

Meyers' books are concerned with these issues.

> 3) Does C++ have equivilent functional constructs like reduce and map? 
> These helped to make my code much cleaner, tighter and faster, and I 
> would miss them dearly.

Sure. They are part of the Standard Library (previously know as STL)
of the new, Standard C++. Unfortunately it seems to me that learning
the details of the Standard Library is a comparable undertaking as
learning the details of the Python standard library. And the C++
Standard Library will only give you the functional equivalent of
the core Python language without any libraries at all (well, except
for maybe math, array, sys and struct). In other words, the bulk
of the C++ Standard Library only remedies the lack of richness in
the C++ syntax. It's nothing like the "batteries included" approach
of Python.

I'm probably exaggerating a bit, but the standard containers and
algorithms are really awkward in comparision with the straight
forward approach of Python.

> btw - the code I am writing doesn't rely upon any platform specific 
> stuff at all and is quite portable in Python - I will be aiming for the 
> same thing in Python and assume it will be possible.

I guess you mean that you aim for the same thing in C++.
I'm afraid Standard C++ is not yet so standard, so that might
not be the easiest thing...



From jester_xp at yahoo.com  Wed Sep 25 11:29:26 2002
From: jester_xp at yahoo.com (Mark Jose)
Date: Wed, 25 Sep 2002 08:29:26 -0700 (PDT)
Subject: parsing from a position found with a regexp
Message-ID: <20020925152926.3102.qmail@web13113.mail.yahoo.com>

Hi,

I would like to parse only a portion of a particular
file. The portion I am interested in starts after the
last occurrence of a particular string in the file.
Ideally, I was hoping there might be a function
(perhaps like the string rfind function, but for
files) that would make this quick and simple. Any
suggestions?

Thanks,
Mark


__________________________________________________
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
http://sbc.yahoo.com



From robin at jessikat.fsnet.co.uk  Mon Sep 16 14:36:44 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Mon, 16 Sep 2002 19:36:44 +0100
Subject: pyro & parallel processing
References: <rLOb1LAK5fh9EwRi@jessikat.demon.co.uk> <uy9a1ritm.fsf@online.no>
 <am53og$hdl$1@news1.xs4all.nl>
Message-ID: <IWzDioA8Sih9Ewla@jessikat.demon.co.uk>

In article <am53og$hdl$1 at news1.xs4all.nl>, Irmen de Jong <irmen at NOSPAMRE
MOVETHISxs4all.nl> writes
>Syver Enstad wrote:
>> Robin Becker <robin at jessikat.fsnet.co.uk> writes:
>> 
>>>Has anyone used pyro to do symmetric parallel processing. The pyro
>>>framework is nice with master slave type applications, but the remote
>>>objects don't seem to be able to message back to their invoker.
>> 
>> They can, checkout the callback examples in the Pyro distro. Basically
>> what happens is that you pass a Pyro proxy as a parameter to the a
>> method on the Pyro object. 
>
>This is a good way of letting the 'other' Pyro object call methods on one
>of your 'own' (or 'local') objects. It is a more symmetric way but
>you have to pass callback object proxies as parameters.
>
>The other - fully symmetric - way, is to build some Pyro objects
>on both sides and register them all with the Name Server.
>This way everybody can discover the other's objects by themselves.
>You won't need to pass callback object proxies around.
>Note that this requires a Pyro server (daemon) on all nodes,
>but that is no different from the callback object solution.
>
>Irmen de Jong
>
>PS  the Pyro mailing list is usually a better place to ask questions
>about Pyro, see http://lists.sourceforge.net/lists/listinfo/pyro-core
>
well actually I think I need a mostly master-slave relation, but would
like the slaves to be able to pass information back to their masters for
redistribution to the other slaves. The kind of thing I'm thinking of is
parallel alpha-beta search with cutoffs. I think this is best done by
having a single master which has all the workers as its slaves. When the
worker finds a new global best it can pass this to the master using a
reasonably simple call back mechanism.

The master can presumably pass the new global cutoff to the slaves, but
I'm not sure yet how interruptible I can make them, so I guess my second
question is can we have asynchronous calls into the workers or do I need
to have multiple published classes with a common thread protected state?
-- 
Robin Becker


From krissepu at vip.fi  Sat Sep 28 14:54:50 2002
From: krissepu at vip.fi (Pekka Niiranen)
Date: Sat, 28 Sep 2002 21:54:50 +0300
Subject: What does Python fix?
References: <e15b6fa3.0209271418.14cd5642@posting.google.com> <flaapucbduj6g4ifrdf3oq99hdrepi1o3v@4ax.com>
Message-ID: <3d95fb09@news.vip.fi>

Sure, Python's syntax looks nice, but to use another language to get 
performance
(i.e speed) ? That sucks.

I have a feeling Python might renew peoples interest to various lisps.

After studying Python for 2 years I myself have
recently switched to DrScheme because I have no interest
relearning C to write extensions to Python.

What an eye opener it has been to study scheme...

-pekka-


Courageous wrote:

>>I'm curious to read a bit more about why Tim Peters (and presumably
>>others) think that Lisp-inspired languages (even non-prefix ones) are
>>doomed to eternal obscurity.  Would anyone care to comment and/or give
>>me some pointers to commentary on the subject? 
>>    
>>
>
>Shrug. Lisp trades one form of complexity for another. By reducing
>the solution to a universal grammar, they increase the symbolic load
>on the user, who is awash in a page full of very similar symbols.
>This is intimidating in particular to novices, for whatever reason,
>who are instantly alienated. Without sweeping in novices, you lack
>a grass roots movement, and without a grass roots, a language is dead.
>
>The parts of lisp that really matter to the community have since
>been adopted in part by many other languages. Lisp is dead. Long live
>Lisp.
>
>C//
>
>  
>




From gregory.p.green at boeing.com  Thu Sep 19 17:36:25 2002
From: gregory.p.green at boeing.com (Greg Green)
Date: Thu, 19 Sep 2002 21:36:25 GMT
Subject: Python COM and variable number of arguments
Message-ID: <ptv9y84m.fsf@boeing.com>

I am using python with a third-party closed source application. It has
Visual Basic for Applications embedded in it for use as an extension
language. My question is what to do about com calls that have a
variable number of arguments in Visual Basic. It seems that I have to
use all of the arguments when I call the method using python. But I
don't know what to use for a value when I don't want that argument.

For instance, In VB i can call the method .Add with 1, 2, or 3
arguments. The 2nd and 3rd arguments are optional. In this case the
2nd argument is called 'Before', and the 3rd is 'After'. So if I want
to add an item after the 3rd item. I do: 

        .Add(Item,,3) 

In python I don't know what to use as the second argument:

        Object.Add(Item=Item,Before=????,After=3)

In case you are wondering, I have tried 'None', or leaving off the
argument all together, but that doesn't work. :(

Thanks for any replies. This is causing me a lot of pain.

-- 

Greg Green



From phil at river-bank.demon.co.uk  Tue Sep 17 18:03:48 2002
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Tue, 17 Sep 2002 23:03:48 +0100
Subject: Python GUI app to impress the boss?
In-Reply-To: <96904b50.0209171305.6d65f574@posting.google.com>
References: <96904b50.0209171305.6d65f574@posting.google.com>
Message-ID: <200209172303.48222.phil@river-bank.demon.co.uk>

On Tuesday 17 September 2002 10:05 pm, Lee Gray wrote:
> I'm a fairly new programmer with experience in ASP/VBScript/DHTML.
> Other than small utilities, so far I've mostly written ASP apps to
> talk to an Oracle database on HP/UX, with Windows NT/2000 clients, and
> I've done maintenance work on VB and Oracle Forms apps.
>
> We're currently looking at .Net, and I'm not very impressed, for a
> variety of reasons. A little research makes me feel that either Java
> or Python would be a good alternative, and given my level of
> experience, I'd much prefer Python over Java.  I *think* I could
> actually get up to speed faster with Python than with VB and ASP.Net,
> in spite of my ASP/VBS experience.
>
> My boss is pretty open-minded, but also needs to see something working
> to be convinced Python is even a viable platform (he had never heard
> of it).  Otherwise, .Net is a given, whether it's any good or not,
> since corporate is MS all the way.
>
> Can someone point me to a good Python GUI app to demonstrate (other
> than IDLE)?
>
> It may very well be that .Net is best for me/us, but from what I've
> seen of Python, I *really* like the language and I'd at least like to
> see a good demonstration of a full-blown app before ruling it out.
> (Something along the lines of what I mentioned above would be ideal,
> but not a necessity.)

http://www.abo.fi/~iporres/Projects/fog0000000023.html

For SQL based applications PyQt contains a set of data aware widgets. PyQt 
includes an example SQL browser, query, reporter in 180 lines of Python.

Phil



From bkc at Murkworks.com  Thu Sep 26 17:32:30 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Thu, 26 Sep 2002 17:32:30 -0400
Subject: a python puzzle
References: <Pine.LNX.4.44.0209251736470.16806-100000@hydrogen.sabren.com> <mailman.1032992492.8970.python-list@python.org> <amtgmd$c00$1@usenet.Stanford.EDU> <mailman.1032998374.8738.python-list@python.org>
Message-ID: <3d937a44_4@goliath.newsgroups.com>

_
"Chad Netzer" <cnetzer at mail.arc.nasa.gov> wrote in message > at all, except
a rather

> routine piece of code.  If I has bothered to
> 'decrypt' it, I would be pissed.  (Reminds me of little Ralphie, in "A
> Christmas Story", waiting weeks to get his secret decoder wheel, only
> to find the secret message is 'Be sure to drink your Ovaltine')

Ha ha ha, love that movie.

"You'll shoot your eye out kid!"

(and I did decrypt it.. not pissed, it was only a 3 minute job)






-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From syver-en+usenet at online.no  Tue Sep 10 12:50:03 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Tue, 10 Sep 2002 16:50:03 GMT
Subject: How do i print to a printer using python?
References: <6afb9d2c.0209041044.43c5f5f1@posting.google.com> <al73gj$lbt$1@wanadoo.fr> <6afb9d2c.0209091641.1d9b652e@posting.google.com> <3d7d6fd5@news.sentex.net>
Message-ID: <uvg5d6b8b.fsf@online.no>

Peter Hansen <peter at engcorp.com> writes:

> Search for Mark Hammond's "win32all" package and install it, or
> install the ActiveState distribution of Python which has it already
> included.  win32ui is one of the modules included in that package.

Just to confuse further, win32all is not a python package. It's a
distribution of python modules and packages. What I mean is that
win32ui is a module that is not in a package, although it is included
in the win32all distribution. I think I'll quit now, before I make
things clearer. ;-)

-- 

Vennlig hilsen 

Syver Enstad


From Oschler at earthlink.net  Fri Sep 27 10:00:12 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Fri, 27 Sep 2002 14:00:12 GMT
Subject: Is Boost Python version 2 ready?
Message-ID: <MvZk9.12124$Lg2.3584877@news2.news.adelphia.net>

Is Boost Python version 2 ready for primetime yet?

thx





From fetus at AUTO  Thu Sep 26 17:16:28 2002
From: fetus at AUTO (A. Dimsdale and A. Moser)
Date: Thu, 26 Sep 2002 21:16:28 GMT
Subject: Searching for and sorting files by date and compiling stats in Windows 
Message-ID: <MOKk9.66578$U7.23791943@twister.socal.rr.com>

I (Aaron)'ve been called upon at work to compile statistics on MS Word temp-file
cruftup on the company's nationwide WAN, including the number of Word tempfiles
per year for the last three years before this one (by Date Modified), per
networked drive, and the amount of disk space used up by the aforementioned
cruft. (The purpose of this project being to give my boss something to use to
convince higher-up management to use a Python script that I am to write, which
would delete all Word cruft older than a specified date--which is how Alex and
I got into Python; I was the one who suggested we use it for the task, and got
the company to pay for a couple of Python reference books for me to teach myself
by.) How can I do the following in Python:
* Search through networked Windows drives for the string "~w*.tmp" (* meaning
wildcard of course) in the filename
* Calculate how much disk space is taken up by Word temp files by date modified
(i.e., files with "~w*.tmp" as their filenames with date modified in 2001 on
drive G: on SDspam01 use 75 MB total, files with "~w*.tmp" as their filenames with date modified in August 2002 on drive H: on DCspam04 use 60 MB total, etc.)
* Optimally, automagickally put this data into an MS Excel spreadsheet file
Sorry that's a lot of funk, but I've been combing MHammonds' win32all docs and
his "Programming Python on Windows 32" book, the other Python ref books the
company bought, and any online docs I can find and can't figure out how to do
all that jazz. Anybody who can help with this will have my eternal gratitude.
Thanks
-- 
-----------------------------------
Aaron L. Dimsdale and Alex S. Moser
"But I don't like SPAM!"
"You don't like SPAM???"


From mennosimons at gmx.net  Thu Sep 26 05:52:49 2002
From: mennosimons at gmx.net (Ulli Stein)
Date: Thu, 26 Sep 2002 11:52:49 +0200
Subject: unicode woes
References: <amuhrk$4cn$05$1@news.t-online.com> <3d92d120$0$76313$e4fe514c@dreader4.news.xs4all.nl>
Message-ID: <amuldd$bur$01$1@news.t-online.com>

Boudewijn Rempt wrote:

> Ulli Stein wrote:
> 
>> 
>> Some weeks ago the management decided to deploy unicode so that we can
>> handle every charset uniformly. The problem which showed at first: we
>> realized that we cannot change the encoding in site.py afterwards, i.e.
>> we have to specify the encoding all the way.
> 
> Use a sitecustomize.py file that contains:
> 
> import sys
> 
> sys.setappdefaultencoding=sys.setdefaultencoding
> 
> Then, in the application starter script use:
> 
> if hasattr(sys, 'setappdefaultencoding'):
>     sys.setappdefaultencoding('utf-8')
> elif sys.getdefaultencoding() != 'utf-8':
>     print 'Warning: encoding not set for unicode - see ReadMe file'
> 
> That should solve a lot of your problems. Then you need to carefully
> determine how the data that enters and leaves your application should
> be encoded. For instance, you might be getting data from files, databases,
> network or paste operations. If that data is many different encodings, you
> need to check the encoding at every entry point. And whenever you write
> data, you need to be aware of the encoding.
> 
> But at least _inside_ your application all data will be utf-8, which makes
> it easy to handle.
> 
> Don't be too sanguine about the unicode handling of Java, by the way. Lots
> of room for hard-to-find unicode-related bugs there.
> 
>> It got worse and worse because
>> the unicode encoding spread like a virus through all the source code.
>> Simple Exceptions which we wanted to print to the console throw another
>> Exception:
>> UnicodeError: ASCII decoding error: ordinal not in range(128)
> 
> Other people will be able to tell you why it was a good design decision to
> make it difficult to set a default application encoding different from
> ASCII. I'm still not convinced, but that might be just my denseness.
> 

But nevertheless you would have to use everywhere the Python unicode() 
function.

Or how would you do this (nonsense string):
str = str + "bla??" + str[:5] + "???"

Would you write then:
str = str + unicode("bla??") + str[:5] + unicode("???")?

And what about the "encode" variable in site.py: If another Python 
application running in parallel to our application changes the encode 
string, will it affect our app? Or has every application its own encode 
string?

U.


From barry at zope.com  Tue Sep 10 23:32:37 2002
From: barry at zope.com (Barry A. Warsaw)
Date: Tue, 10 Sep 2002 23:32:37 -0400
Subject: RELEASED email package version 2.3
Message-ID: <15742.47445.811488.515719@anthem.wooz.org>

I've just released the (standalone) email package version 2.3.  This
is identical to the version that is in the current cvs development
tree for Python 2.3.  It is made available as a distutils package for
installing in earlier versions of Python.  It is compatible with
Python 2.1.3 and Python 2.2.1.

The email package is a new package for parsing, handling, and
generating email messages and other RFC 2822 style documents.  It
conforms to most of the email related RFCs including 2045-2047 (the
MIME RFCs) and 2231.  It is intended to replace several older modules
in the standard distribution, such as rfc822, mimetools, multifile,
mimify, and MIMEWriter, and such non-standard modules as mimecntl.

email-2.3.tar.gz is available from the mimelib project on SourceForge:

    http://sf.net/projects/mimelib/

Enjoy,
-Barry



From dallasm at aiinet.com  Thu Sep 19 11:03:15 2002
From: dallasm at aiinet.com (Mahrt, Dallas)
Date: Thu, 19 Sep 2002 11:03:15 -0400
Subject: Python as shared library on Unix
Message-ID: <638AA0336D7ED411928700D0B7B0D75B02E1EFBC@aimail.aiinet.com>

If you the application is being done locally only (in other words, you have
control of all interpreters that will run your code) then the patch is very
simple to do against 2.2.1. I have successfully back ported it with minimal
changes. I did have to do it manually, but there aren't any major changes
outside of configure.in and Makefile.pre.in .

> -----Original Message-----
> From: martin at v.loewis.de [mailto:martin at v.loewis.de] 
> Sent: Wednesday, September 18, 2002 11:46 PM
> To: python-list at python.org
> Subject: Re: Python as shared library on Unix
> 
> 
> "Rolf Kalbermatter" <r.kalbermatter at hccnet.nl> writes:
> 
> > There is probably little chance that this will be officially 
> > backported to the 2.2.1 branch of Python?
> 
> No, there is no such chance.
> 
> Regards,
> Martin
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 



From aleax at aleax.it  Thu Sep 19 11:43:48 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 19 Sep 2002 15:43:48 GMT
Subject: strong/weak - dynamic/static [Was: Getting started]
References: <uojinmcfapjf45@corp.supernews.com> <tyf8z1y5cp5.fsf@pcitapi22.cern.ch> <Bqli9.36669$216.1428116@bin4.nnrp.aus1.giganews.com>
Message-ID: <Ugmi9.114712$ub2.2441284@news1.tin.it>

Terry Reedy wrote:
        ...
>>explaining or defining the concepts of ... dynamic and static typing.
> 
> Python objects are typed at creation.  Is that dynamic or static?  I'm

Judge for yourself:

>>> class X(object): pass
...
>>> class Y(object): pass
...
>>> x = X()
>>> type(x)
<class '__main__.X'>
>>> x.__class__ = Y
>>> type(x)
<class '__main__.Y'>
>>>

If you find a way to name this behavior "static", you'll be well
placed for the yearly Humpty Dumpty award.


Alex



From see_reply_address at something.invalid  Thu Sep 19 21:12:04 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Fri, 20 Sep 2002 13:12:04 +1200
Subject: Stackless Python, eventual merge?
References: <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <mailman.1032329366.6813.python-list@python.org> <j4ptvbaapm.fsf@informatik.hu-berlin.de> <3D890FB0.4030408@something.invalid> <m3admeo4u1.fsf@mira.informatik.hu-berlin.de> <ambso1$4g8vh$1@ID-59885.news.dfncis.de> <j4fzw6tofh.fsf@informatik.hu-berlin.de> <mailman.1032441028.15013.python-list@python.org> <j4elbqrrex.fsf@informatik.hu-berlin.de>
Message-ID: <3D8A75E4.4050206@something.invalid>

Martin v. L?wis wrote:

> Regardless, you are saying that the random module is not
> thread-safe.


While it may produce erroneous results, at least
that kind of thing won't crash the interpreter.

Stackless, on the other hand, takes thread-unsafety
to new heights...:-)

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From wlfraed at ix.netcom.com  Wed Sep 25 20:34:44 2002
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Wed, 25 Sep 2002 17:34:44 -0700
Subject: Newbie References Question -> Exemple
References: <3d91c72f$0$8558$79c14f64@nan-newsreader-02.noos.net> <3d91d2cf$0$1484$79c14f64@nan-newsreader-01.noos.net>
Message-ID: <4nktma.mg3.ln@ix.netcom.com>

Guy Rabiller fed this fish to the penguins on Wednesday 25 September 
2002 08:14 am:

> Ok,
> 
> thanks for your answers, I understand now that a variable name is
> 'just' a name in a dictionary, and not a container.
> 
> What I want to do is simple, and I suppose that this is my 'way of
> thinking' regarding Python that is wrong.
> 
> I have a list of points, and a list of polygons that should contain
> 'references' to points.
> 
> So for me:
> -> points list:
> p = []
> p.append([0,0])
> p.append([0,1])
> p.append([1,1])
> p.append([1,0])
> -> polygon:
> quad = []

        Created a /new/ list

> quad.append(p[0])

        appended the value of p[0] to the new list

> For me the quad list of point is not 'updated' with the change of one
> of its point.
>
        So just use

quad = p

which binds "quad" to the /same/ value that p is bound to.


        Strangely, I had the opposite problem two years ago, when trying to 
"save state" for small test set. I was using the simple assignment and 
finding that I wasn't saving the state but making a reference (your 
term) to it.

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <


From Oschler at earthlink.net  Thu Sep 12 19:17:29 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Thu, 12 Sep 2002 23:17:29 GMT
Subject: Piper Pickling?
Message-ID: <dg9g9.2373$Lg2.427869@news2.news.adelphia.net>

Forgive me :)

If Peter Piper pickles a peck of Python packages, can he pack the pickled
Python peckers in a proper packing place?

Translated: Anybody seen a module that can override the standard pickling
process and store and retrieve Python objects to a MySQL database?

thx

P.S. - Whell hell the language is named after the world's best comedy
troupe, what did you expect, green eggs and spam? :)




From pyth at devel.trillke.net  Mon Sep  2 06:03:44 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 2 Sep 2002 12:03:44 +0200
Subject: Why it is so dramatical?
In-Reply-To: <200209021107.25761.b.maryniuk@forbis.lt>; from b.maryniuk@forbis.lt on Mon, Sep 02, 2002 at 11:07:25AM +0200
References: <200209021107.25761.b.maryniuk@forbis.lt>
Message-ID: <20020902120344.I3893@prim.han.de>

Bo M. Maryniuck wrote:
> The task:
> 	Adding a text to a text.
> 
> The problem:
> 	It is tragically slow to add string to huge string.

That's because strings are immutable objects. When you issue

    somestring += otherstring

a new string object is allocated where the contents of
somestring and otherstring are copied in.  Then the
name 'somestring' is bound to this new string.

If you have something like this in a loop it's often
*much* better to work with a list and later do a 

    "".join(stringlist)

As you observed yourself this is around 100-1000 times faster.  

And consider: if strings were not immutable but modified in-place
(like lists) you couldn't use them as dictionary keys. 

if-it-hurts-don't-do-it-ly, yours Holger



From andrewm at object-craft.com.au  Thu Sep 12 00:35:22 2002
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Thu, 12 Sep 2002 14:35:22 +1000
Subject: Pickling slotted instances
Message-ID: <20020912043522.37B143C9B1@coffee.object-craft.com.au>

Instances using __slots__ can't be pickled - is there a fundemental reason
why this is so, or is it simply yet to be implemented?

At first glance, a __getstate__ for a slotted class could be as simple as:

    def __getstate__(self):
        attrs = [getattr(self, attr, None) for attr in self.__slots__]
        return self.__slots__, attrs

What am I missing?

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/



From tim.one at comcast.net  Mon Sep  9 13:06:01 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 09 Sep 2002 13:06:01 -0400
Subject: Can't exponentiate zero in NumPy
In-Reply-To: <alifc4$gsi$1@naig.caltech.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCCECMBDAB.tim.one@comcast.net>

[Rick Muller]
> I found out more about this problem, and I'm still hoping that someone
> can suggest a good fix.
>
> The problem comes when I take a power of a NumPy array that has very
> small values in it:
>
>  >>> from Numeric import *
>  >>> a = zeros(3,Float)
>  >>> a**2
> array([0., 0., 0.])
>  >>> a[1] = 1.e-310
>  >>> a**2
> Traceback (most recent call last):
>     File "<stdin>", line 1, in ?
> OverflowError: math range error
>
> This crashes on Linux, but not on Mac OS X.

Does this require wrapping in a Numeric array, or does it also happen in
*any* of these straight Python expressions?:

    1e-310 ** 2
    pow(1e-310, 2)
    import math
    math.pow(1e-310, 2)

    1e-310 ** 2.1
    pow(1e-310, 2.1)
    math.pow(1e-310, 2.1)

In addition, exactly which version of Python are you using (2.0, 2.1, 2.1.1,
2.1.2, 2.1.3, 2.2, 2.2.1, ...)?

> Is this a bug or a feature?

You first <wink>.  Python intends silent underflow to 0, but C99 changed the
rules for how errno works, and there was some class of related problem in
core Python right after glibc started exploiting the looser C99 rules.  As
far as we know that's been fixed.




From sabren at manifestation.com  Wed Sep 25 17:42:40 2002
From: sabren at manifestation.com (Michal Wallace)
Date: Wed, 25 Sep 2002 17:42:40 -0400 (EDT)
Subject: a python puzzle
Message-ID: <Pine.LNX.4.44.0209251736470.16806-100000@hydrogen.sabren.com>



Here's a python puzzle. Solve it and you win a prize. :)



"""
Ha'f ahoy ad ayfa wdqr srxhekdpyr! H phmm zhby
fhv odeajf dc cryy pys jdfahez ad ajy chrfa kyrfde
ad fdmby ajhf irwkadzrxo. Ixe wdq chzqry ha dqa?

Fheiy wdq jxby axgye ow ijxmmyezy xet tyirwkayt
ajy fyirya oyffxzy, H phmm ybye zhby wdq ajy
irwkadzrxo-oxghez fdqriy idty ad xoxuy wdqr
crhyetf!
"""
hokdra fwf, farhez, rxetdo
ihkjyr = mhfa(farhez.myaayrf[:26])
rxetdo.fjqccmy(ihkjyr)
ihkjyr = "".ldhe(ihkjyr)
ihkjyr += ihkjyr.qkkyr()
oxk = {}
cdr kmxhe, irwka he uhk(farhez.myaayrf, ihkjyr):
    oxk[kmxhe] = irwka
cdr mhey he dkye(fwf.xrzb[1]):
    krhea "".ldhe([oxk.zya(ij,ij) cdr ij he mhey])[:-1]



Cheers,

- Michal   http://www.sabren.net/   sabren at manifestation.com 
------------------------------------------------------------
Switch to Cornerhost!             http://www.cornerhost.com/
 Low Priced, Reliable Blog Hosting, With a Human Touch. :)
------------------------------------------------------------




From gerhard.haering at opus-gmbh.net  Wed Sep 25 10:24:25 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 25 Sep 2002 14:24:25 GMT
Subject: Python mixed with HTML like PHP
References: <o9h3pu4f24ffnupaen8ut8v8dgokacdna6@4ax.com>
Message-ID: <slrnap3hol.1uo.gerhard.haering@haering.opus-gmbh.net>

Johann wrote:
> I heard about project which can mix python cgi code with html content.
> I forgot its name. Who remember?

See http://www.python.org/cgi-bin/moinmoin/WebProgramming - "Templating
Systems".

or-write-your-own-like-everybody-else-ly yours,

-- Gerhard


From huxley1 at westnet.com.au  Fri Sep 20 03:18:32 2002
From: huxley1 at westnet.com.au (Jeremy Rew)
Date: Fri, 20 Sep 2002 15:18:32 +0800
Subject: How to access XML passed from an XForm
Message-ID: <5.1.1.6.0.20020920150837.00b8d848@localhost>

I dont know if this has been answered before now, but I couldnt find it 
after a fair bit of searching. (is there any way of running a search on the 
archives?)

I have an XForm that passes a serialised XML document to a URI ( a python 
script). Im not exactly sure of the internal representation of what it 
sends, but I am having trouble just finding out how to access the XML from 
within the python script.  I just need to know if theres a way to store the 
XML in a string or find out where I can access it from, as I already know 
how to process/manipulate the XML once I can get at it. Any ideas (Im 
rather new to python)?

Any help will be greatly appreciated.

Jeremy




From Tom_Good1 at excite.com  Mon Sep  9 20:07:00 2002
From: Tom_Good1 at excite.com (Tom Good)
Date: 9 Sep 2002 17:07:00 -0700
Subject: tree data structure
References: <FM0e9.2860$cP3.6165@news.iol.ie> <ala6c3$6i2$1@bob.news.rcn.net> <Ti1e9.2864$cP3.6124@news.iol.ie>
Message-ID: <ac677656.0209091607.59f08ccb@posting.google.com>

Padraig Brady <Padraig at Linux.ie> wrote in message news:<Ti1e9.2864$cP3.6124 at news.iol.ie>...

> > {'S2': ['S0', 'S1', 'S3'], 'S1': ['Q2'], 'S0': ['Q1'], 'L0': ['S2']}
> 
> Interesting I didn't know about setdefault.
> However how would you iterate over the above?

from __future__ import generators

def treeIterator(tree):
    for key in tree.keys():
        yield key
        for val in tree[key]:
            if not tree.has_key(val):
                yield val


if __name__ == '__main__':
    tree = {'S2': ['S0', 'S1', 'S3'], 'S1': ['Q2'], 'S0': ['Q1'], 'L0': ['S2']}
    for node in treeIterator(tree):
        print node


From eric.brunel at pragmadev.com  Wed Sep  4 04:19:34 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Wed, 4 Sep 2002 10:19:34 +0200
Subject: Privacy and Inheritance
References: <mailman.1031112136.4594.python-list@python.org>
Message-ID: <al4f8e$sk$1@wanadoo.fr>

Dennis B. wrote:

> Greetings;
> 
> Have enough confidence in Python to realize this is probably just my
> misunderstanding, but wondering why Python mangles private, double
> underscore attributes with the name of the base class even if they're
> inherited and the instance is that of the sub class.
> 
> Like the following doesn't even seem to work:
> 
> class StandardNormal:
> 
>     def letEventZScore(self, eventZScore):
>         self.__eventZScore = float(eventZScore)
> 
> class DerivativeNormal(StandardNormal):
> 
>     def getCentral(self):
>         if self.__eventZScore < 0:
> 
> Unless I change that last line to something like:
> 
>         if self._StandardNormal__eventZScore < 0:
> 
> Maybe just missing the point all together, but wondering if anyone'll
> point out anything a little more elegant and/or intuitive.

In languages that have actual "private", "protected" and "public" 
attributes, private attributes are not inherited. They are visible only in 
the class that define them, not outside it and not in its sub-classes. What 
you want is a protected attribute, that is visible in the defining class 
and in its sub-classes.

Unfortunately, Python doesn't know anything about protected attributes. 
AFAIK, there's no simple way to do what you want: in Python, attributes are 
either private or public; there's nothing in between...
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From jubafre at brturbo.com  Tue Sep 24 11:30:48 2002
From: jubafre at brturbo.com (jubafre at brturbo.com)
Date: Tue, 24 Sep 2002 12:30:48 -0300 (GMT-03:00)
Subject: hex numbers??
Message-ID: <574747556.1032881448716.JavaMail.nobody@webmail2.brturbo.com>

>>> y=hex(x)
>>> y
'0xa'
>>> type(y)
<type 'str'>

the hex() function put the hexadecimal number in a string type like this '0xa', but the real hexadecimal number of 10 is 'a', why python puts a '0x',  and how i can get just the real hexadedcimal number, for example i want to write in a file 10 in hexadecimal, but if i use hex(10) the result is '0xa' as a string, is not the same as 'a'.

'0xa' != 'a'

x = [12,13,14]
y = ['0xc', '0xd', '0xe']

hex() create a list of string,  how i can write in a file just the real value of hexadecimal number?

	
Juliano Freitas
www.gebrasil.hpg.com.br 

From zopestoller at thomas-guettler.de  Tue Sep 10 05:08:58 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Tue, 10 Sep 2002 11:08:58 +0200
Subject: run-mailcap in python?
References: <alkcfb$9ql$07$1@news.t-online.com>
Message-ID: <alkco4$j0k$06$1@news.t-online.com>

Thomas Guettler schrieb:
> Hi!
> 
> On Unix there is a script called run-mailcap, which
> opens the corresponding programm with the given file:
> 
> run-mailcap foo.html
> --> Default browser starts

[cut]

> run("foo.html")
>  --> Browser starts

I just discoverd os.startfile(), but it is unfortunately
only available on win32.

A Unix version would be nice.

  thomas



From marklists at mceahern.com  Mon Sep 23 13:18:36 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 23 Sep 2002 12:18:36 -0500
Subject: parsing tabs in a string
In-Reply-To: <006201c26324$4148ae90$025e10ac@calibredd.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOOELOEAAA.marklists@mceahern.com>

[Andrew Alzner]
> I'm trying to parse information extracted from a spreadsheet
> which is tab separated. Here is a simplified version of what I have:
>
> list=['"103-01a-17"\t2\t33\t256\t227\n']
> y=list[0]
> print y.find("\t")
> print y.find("\t",1)

Don't use list as a variable name, you'll stomp all over the builtin list.
Rather than finding each tab, why not split on tabs:

a = ['"103-01a-17"\t2\t33\t256\t227\n']
y = a[0]
for x in y.split("\t"):
  print x

// m




From pearu at cens.ioc.ee  Mon Sep 30 13:47:25 2002
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Mon, 30 Sep 2002 20:47:25 +0300 (EEST)
Subject: rank-0 arrays?
In-Reply-To: <U00m9.6260$_32.1554@fe04>
Message-ID: <Pine.LNX.4.21.0209302045110.10208-100000@cens.kybi>

On 30 Sep 2002, Steve Holden wrote:

> <brobbins333 at shaw.ca> wrote in message news:3d988160.79509684 at news...
> > In Numeric Python, what are rank-0 arrays?
> > Must not arrays have at least 1 dimension?
> >
> 
> Only if they must contain data. A xero-rank array will (sort of) be the None
> of Numeric arrays.

This is not correct.
In Numeric Python rank-0 arrays are like mutable scalars.

Pearu




From kilaruba at copper.net  Wed Sep 18 20:16:37 2002
From: kilaruba at copper.net (bert)
Date: Wed, 18 Sep 2002 17:16:37 -0700
Subject: multiple index inconsistency
References: <TP5i9.9293$Rf3.1003108@newsfeed.slurp.net>
Message-ID: <mI8i9.10037$Rf3.1101537@newsfeed.slurp.net>

Thanks, I see how things work. I'm now using, say, 
>>> N = 3
>>> a = []
>>> for j in range(N):
...     a.append(range(N))

to create a square matrix whose size is set by N (and I don't
need to type thousands of elements for a big matrix).


From sholden at holdenweb.com  Thu Sep 19 07:25:25 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 19 Sep 2002 07:25:25 -0400
Subject: Need advice on choosing the technology for intranet based application 
References: <3d88fd1e@news.ivm.net> <Gn7i9.1769$ik2.430@fe02> <3d89121f@news.ivm.net> <tFei9.18265$V7.4672511@twister.socal.rr.com>
Message-ID: <Utii9.11160$ik2.8281@fe02>

"Jeff Davis" <jdavis at empires.org> wrote ...
> The advice I was about to give you followed what Steve had to say, so I
> have included my comments here.
>
> > Whenever I read "CGI" I also read "a pain to maintain" some words
> > later... ? It's a project with a (modest) possibility to turn into a
> > product, so both portability and maintanability are issues.
>
> Well, CGI isn't considered the best development environment. It isn't
> horrible though (unless performance is an issue...).The main advantage to
> CGI is that it's portable, and you don't want to go rewrite your app. One
> thing you might do is write your program in python, and use libraries to
> handle everything. That way if you do need to put it on something
> unexpected, you can just rewrite a few lines.
>
> > On the other hand, Mark also meant CGI/Apache, writting Python/Apache,
> > right?
>
> Not necessarily. If you think you'll at least be able to install some
> software on the target webserver, you could use mod_python. Use the
> publisher handler, and read the simple docs on the publisher handler (a
> part of mod_python) and the cgi module (thats a python module with
> routines that are useful for any web programming, not just CGI).
>
> Or, you might want to stick with python CGIs. CGIs seem safer, because I
> don't see as many possible problems or stumbling blocks.If it get's turned
> into some other software later, it will be easy to change it from CGI to
> something else if you code defensively.

You might also want to look at WebWare, which goes pretty well in the Apache
environment and offers framework support for Python web applications.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Previous .sig file retired to                    www.homeforoldsigs.com
-----------------------------------------------------------------------





From whisper at oz.net  Mon Sep  9 19:35:19 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 9 Sep 2002 16:35:19 -0700
Subject: Diving into the Python pool....
In-Reply-To: <005c01c25855$dc5752d0$e680cf3f@leader1>
Message-ID: <GCEDKONBLEFPPADDJCOEMECFEOAA.whisper@oz.net>

Hi Jordan;

(Firstly, you're likely to get more responses if you post in plain text and
not html - many people just delete such messages.)

In my personal opinion (one shared by many others), Python is THE language
for almost any development task. My background is with C, C++ and scripting
languages and a smattering of Java (which I happen to dislike intensely).

There are an enormous quantity of application development frameworks for
developing web apps with Python. The preeminenant one is Zope
(www.zope.org). You can find a wealth of additional information about Python
at www.python.org. You'll also find a link to the python wiki there and it
has many pointers to python web app frameworks, python advocacy and other
tools.

HTH,

David LeBlanc
Seattle, WA USA
-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]On
Behalf Of Jordan Faris
Sent: Monday, September 09, 2002 16:09
To: python-list at python.org
Subject: Diving into the Python pool....


Hi...after strolling through these threads for some while now, and
comparing the generally benign tone here with the rather elitist edge of
other programming communities, I think I'd like to get my feet firmly onto
Terra Pythonia.
I'm a little reluctant to plow in headfirst, however, mainly because I had
my heart set on becoming a Java programmer, but I think I need to be more
realistic about programming fundamentals, since I am doing this completely
self-study.
Python is, if I understand correctly, an object-oriented language with a
much simpler syntax than Java, yet compatible in applicability, especially
web-related, is this correct?
My primary focus is to create streamlined, rapid-deployment web applications
(PHP and JSP portal sites, etc), but from what I've seen, Python can be used
or adapted to achieve comparable results; am I on the right track, or should
I continue to nurture the Java bean?
Any frank, level-headed input would be much appreciated.
Thanks......
Jordan




From op73418 at mail.telepac.pt  Tue Sep 17 07:34:36 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Tue, 17 Sep 2002 12:34:36 +0100
Subject: how to cast an instance id
References: <3D86E888.90AF569@millfilm.co.uk>
Message-ID: <hl4eou0q4e8ovjs9942m271r000jrvd1pm@4ax.com>

On Tue, 17 Sep 2002 09:32:08 +0100, Eric Texier <erict at millfilm.co.uk>
wrote:

>
>It there a way to create a new reference of  an object
>from the string representation of its id?
>Thanks for any help.
>
>class foo:
>    def info(self):
>        print self
>
>a = foo()
>a.info()
>ccc = "newa = 0x%x" % id(a)
>exec(ccc)
>
>

AFAIK there isn't - You probably have to go down the bowls of the C
implementation to do it. But why would you need such a contortionist's
act? If you explain your problem maybe the people here in the list can
figure out a Pythonic way to do it.

Best regards,
Gon?alo Rodrigues


From somebody at hotmail.com  Fri Sep  6 02:45:05 2002
From: somebody at hotmail.com (stephan)
Date: Fri, 06 Sep 2002 08:45:05 +0200
Subject: newsgroup for zope
References: <al4tru$o3j$03$1@news.t-online.com>
Message-ID: <al9ipd$moi$00$1@news.t-online.com>

stephan wrote:

> Hi perhaps a stupid question,
> 
> but does there exist a newsgrup for zope centric items?
> 
> bye
> 
>  Stephane

Thank you all very much for the infos :-)

Stephane


From Peer.Griebel at entire.de  Wed Sep 18 10:12:27 2002
From: Peer.Griebel at entire.de (Griebel, Peer)
Date: Wed, 18 Sep 2002 16:12:27 +0200
Subject: Copy with __slots__
Message-ID: <673E8D215B3BD311A9C90090279BBB8402829E7C@mail.ebssoftware.de>

> Griebel, Peer wrote [I'm only leaving the relevant parts of 
> the script]:
> 
> > class C1(object):
> >     __slots__ = "s1";
> > 
> > class C2(C1):
> >     __slots__ = "s2";
> 
> Funny use of semicolons here.  I'd suggest omitting them!

Yes sure... This is the result of constantly switching between Java and
Python. Nomally I don't use semicolons...
 
> > o2 = C2()
> > 
> > import copy
> > p2 = copy.copy(o2)
> > print p2.s2           # attribute s2 does not exist (has 
> not been copied)
> > 
> > The last print statement fails with the an AttributeError. Since
> 
> Traceback (most recent call last):
>   File "sloco.py", line 25, in ?
>     p2 = copy.copy(o2)
>   File "/usr/local/lib/python2.2/copy.py", line 84, in copy
>     y = _reconstruct(x, reductor(), 0)
>   File "/usr/local/lib/python2.2/copy_reg.py", line 68, in _reduce
>     dict = getstate()
> TypeError: a class that defines __slots__ without defining 
> __getstate__ 
> cannot be pickled

I'm using Python 2.2 (Cygwin). There I don't get the exception you
mentioned.
 
> This is with Python 2.2.1 -- as you see, it doesn't get anywhere
> as far as the print statement.  The error message is somewhat
> peculiar (who's trying to _pickle_ anything...?) -- and if you
> do try to define __getstate__ (and possibly __setstate__), you'll
> see they're no use for copy.copy (nor copy.deepcopy).  _reconstruct
> gets in the way.
> 
> If you only care about copy.copy, the simplest alternative is:
> 
> class C2(C1):
>     __slots__ = "s2"
>     def __copy__(self):
>         x = self.__class__()
>         x.s1 = self.s1
>         x.s2 = self.s2
>         return x

Yes I know this works. But this hat two drawbacks (at least in my eyes):

1. Formerly I used the approach of copy: I simply updated the dict of a
newly created empty object. This is fast and works well - but not with the
new classes. I think your approach runs much more slowly.

2. The approach is error prone. When simply updating a dict I'm not going to
miss a newly added attribute. But using your method I have to remember the
method whenever I add an attribute.

> The way to cover copy.copy, copy.deepcopy AND pickle is to have the
> class supply a special method __reduce__.  __reduce__ can return a
> pair where the first item is callable, the second item is the tuple
> to pass as arguments to the first.  Thus, for example:

Here once again I have the two disadvantages mentioned above.

> The nicest thing about it is that a class with slots will often
> already supply some nice function to build instances of the class
> given values to put in the slots -- you don't necessarily have
> to write factory functions such as _build just for the purpose
> of supporting copy, deepcopy and pickle... you often have them
> lying around!-).  For example, a more typical case might be:
> 
> class C2(C1):
>     __slots__ = 's2'
>     def __init__(self, s1=None, s2=None):
>         C1.__init__(self, s1)
>         self.s2 = s2
>     def __reduce__(self):
>         return self.__class__, (self.s1, self.s2)
> 
> Here I'm assuming C1 has a similar __init__, but that's a detail.
> The point is that __reduce__ can use class C2 itself as the
> first item of its result -- the class IS callable, after all.

Yes this would work. But I'm looking for speed (as you already abserved).
And my constructor is a little more heavy weight. So it is much faster to
use the old mechanism (update dict) than always calling the constructor.
 
> Alex

Thanks Alex!
I'm still in hope to find a fast and secure solution...

Peer



From ggittings at arraybiopharma.com  Mon Sep  9 14:42:17 2002
From: ggittings at arraybiopharma.com (Gabe Gittings)
Date: 9 Sep 2002 11:42:17 -0700
Subject: COM/Active X newbie question
References: <5523ec5.0209061446.1404cd10@posting.google.com> <ejee9.1387$_l1.3889@news-server.bigpond.net.au>
Message-ID: <5523ec5.0209091042.39c8be68@posting.google.com>

> If it is a true OCX, then it probably needs to be hosted in a GUI 
> environment, such as pythonwin.  See pywin\demos\ocx for examples.
> 
> If you can find any VBScript or JScript code that works with the object, 
> it should be usable the way you are trying.  If the only examples you 
> can find are hosted inside a form in a full-blown VB sample, then the 
> Pythonwin thang should work.

Thanks Mark first off your book is very helpful though it would be
nice if this was covered in the next edition ;). It is a true OCX
which seems odd to me because there is no visual UI pieces to the OCX
it is mostly events and getters. Anyway one last question. How to I
capture the events that it generates. In VB it is ridiculously easy
but I am not sure at all how to do it in python.

Thanks again for the help
Gabe


From senux at senux.com  Mon Sep 16 03:13:16 2002
From: senux at senux.com (Brian Lee)
Date: Mon, 16 Sep 2002 16:13:16 +0900
Subject: reading meta data in image files by PIL
Message-ID: <20020916071316.GA23449@mercury.senux.com>

Is it possible to read meta data in image files like
jpeg, png or gif by PIL(Python Image Library)? I read
handbook in PIL website but I am not sure. 

What I want to do finally is to make a script which
resize image, and insert a black border with titled
by meta data in each image file. Can I do this with PIL?

Any advice?

-- 
Brian Lee <senux at senux.com>



From hancock at anansispaceworks.com  Thu Sep 19 20:24:16 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Thu, 19 Sep 2002 17:24:16 -0700
Subject: How to use PyDoc?
Message-ID: <3D8A6AB0.95CFE28@anansispaceworks.com>

Hi all,

I'm trying to generate HTML documentation using PyDoc
for a Zope product. I'd like to make a collection of
HTML documents in a "doc" subdirectory -- pretty much
like what HappyDoc does, except I'm hoping that PyDoc
will collect some information that HappyDoc misses
(such as methods which are created by factory functions).

I can find a pydoc-generated documentation of the API,
but it doesn't make it very clear how you would actually
use it, and so far, I can't find a How-To on it.

Naturally, I need to do this from within the Python
interpreter (or in a script), not using the commandline
tool (otherwise the module I need to document won't
import -- since it requires Zope).

Does anyone have an example of doing this?

Thanks,
Terry

-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
Anansi Spaceworks                 
http://www.anansispaceworks.com 
------------------------------------------------------



From barry at zope.com  Thu Sep  5 14:06:59 2002
From: barry at zope.com (Barry A. Warsaw)
Date: Thu, 5 Sep 2002 14:06:59 -0400
Subject: New `spambayes' project on SourceForge
Message-ID: <15735.40259.117828.402419@anthem.wooz.org>

There's been a ton of press about applying Bayesian classifiers to
spam detection lately, spurred on by Paul Graham's recent paper "A
Plan for Spam"

    http://www.paulgraham.com/spam.html

Tim Peters has done an incredible amount of work on our Python
implementation of this idea.  Some of the reasons why I think Tim's
work is so cool is that he's brought along his deep knowledge of
speech recognition's related issues, and his obsessive devotion to
reducing the amount of spam I ultimately have to delete <wink>.

In order to encourage more participation from the wider open source
community, we've moved the code from a backwater of the Python cvs
tree to its own project on SourceForge.  The hope is that more people
will be able to contribute to ideas, testing, and integration of the
basic algorithms with other systems such as mail daemons, mailing list
managers, and mail clients.

The project is called "spambayes" (for lack of creativity on our part
:) and is hosted here:

    http://sf.net/projects/spambayes

If you're interested in becoming a developer on the project, let me
know.  Otherwise you can of course get anonymous checkouts of the code.

There are also two mailing lists related to the spambayes project.
The first is a general discussion list:

    http://mail.python.org/mailman-21/listinfo/spambayes

and the other is a list for cvs checkin message notices:

    http://mail.python.org/mailman-21/listinfo/spambayes-checkins

Feel free to join those lists (and help be a guinea pig for Mailman
2.1 :).

Enjoy,
-Barry

PS to Python-devers: the code has been removed from
nondist/sandbox/spambayes, so you won't be able to hack on it there.
Also, please move discussion about this from python-dev at python.org to
spambayes at python.org



From tim.one at comcast.net  Thu Sep 26 11:51:57 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 26 Sep 2002 11:51:57 -0400
Subject: SET_LINENO
In-Reply-To: <j47kh9b6kp.fsf@informatik.hu-berlin.de>
Message-ID: <BIEJKCLHCIOIHAGOKOLHCEBODLAA.tim.one@comcast.net>

[Bruce Dawson]
> However one of the posts on this topic (can't find it now) suggested
> that the hoped for improvements had not actually materialized.

[Martin v. Lowis]
> I think this suggestion is incorrect; the improvements can be
> demonstrated.

It was platform-dependent, and was a real loss on Windows.  I'm not worried
about that, though, because it's just "ceval disease":  the performance of
the eval loop goes up or down approximately at random.  MSVC6 has, for
years, done better than gcc on it, and seemingly by accident.  I figured
that MSVC had been knocked out of a local minimum by accident here, and
that, over time, other people making other random changes to ceval would
eventually nudge it pack in to the same-- or an even better --local minimum.

As of today, on a rather noisy box,

C:\Code\python\PCbuild>\python22\python -O ../lib/test/pystone.py
Pystone(1.1) time for 50000 passes = 3.81114
This machine benchmarks at 13119.4 pystones/second

C:\Code\python\PCbuild>\python22\python -O ../lib/test/pystone.py
Pystone(1.1) time for 50000 passes = 3.78183
This machine benchmarks at 13221.1 pystones/second

C:\Code\python\PCbuild>\python22\python -O ../lib/test/pystone.py
Pystone(1.1) time for 50000 passes = 3.79681
This machine benchmarks at 13169 pystones/second


C:\Code\python\PCbuild>python ../lib/test/pystone.py
Pystone(1.1) time for 50000 passes = 3.41775
This machine benchmarks at 14629.5 pystones/second

C:\Code\python\PCbuild>python ../lib/test/pystone.py
Pystone(1.1) time for 50000 passes = 3.41381
This machine benchmarks at 14646.4 pystones/second

C:\Code\python\PCbuild>python ../lib/test/pystone.py
Pystone(1.1) time for 50000 passes = 3.40596
This machine benchmarks at 14680.1 pystones/second

C:\Code\python\PCbuild>

Now lots of things have changed, but performance is so much better under 2.3
now on Windows without -O than it used to be with -O in 2.2 that "pinning
the blame" isn't an interesting game anymore.

Still, for peak speed over time, it would obviously have been better to
remove SET_LINENO via a method that *didn't* add new instructions to the
eval loop's critical path.  That was much harder to do, though.  I'm quite
happy with the compromise Michael crafted.




From thoa0025 at mail.usyd.edu.au  Mon Sep 16 04:22:03 2002
From: thoa0025 at mail.usyd.edu.au (:B nerdy)
Date: Mon, 16 Sep 2002 08:22:03 GMT
Subject: raw post data
References: <Wt8h9.9583$Ee4.22928@news-server.bigpond.net.au> <w_fh9.113093$pX1.4054147@news2.tin.it>
Message-ID: <Lwgh9.10771$Ee4.29455@news-server.bigpond.net.au>

thanks alex

:B

"Alex Martelli" <aleax at aleax.it> wrote in message
news:w_fh9.113093$pX1.4054147 at news2.tin.it...
> :B nerdy wrote:
>
> > how would i retrieve the raw post data in cgi python scripts ?
>
> I think something like:
>
> import sys
> all_posted_lines = sys.stdin.readlines()
>
> should do it.  Raw POST'ed data should be no more and no less
> than the contents of the CGI script's standard-input.
>
>
> Alex
>




From spike at orangeware.net  Fri Sep  6 16:12:41 2002
From: spike at orangeware.net (Spike^ekipS)
Date: Sat, 07 Sep 2002 05:12:41 +0900
Subject: (no subject)
Message-ID: <alb0uf$4ti$1@news1.kornet.net>

fsdfasdfasd



From martin at v.loewis.de  Wed Sep 18 08:32:12 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 18 Sep 2002 14:32:12 +0200
Subject: Stackless Python, eventual merge?
References: <JHEOKEOOLIGLDHCMAHMOAEMJDOAA.marklists@mceahern.com> <mailman.1032107006.6411.python-list@python.org> <am2g4a$o6h$1@bob.news.rcn.net> <3D854B7D.1060209@one.net> <j4znuio0kr.fsf@informatik.hu-berlin.de> <mailman.1032204706.14737.python-list@python.org> <m3y9a1eh8v.fsf@mira.informatik.hu-berlin.de> <1032242013.2443.1329.camel@dsl-65-184-205-5.telocity.com> <m34rcpcfuk.fsf@mira.informatik.hu-berlin.de> <mailman.1032293422.22920.python-list@python.org> <3D87CE0A.9090609@something.invalid> <m31y7rj49r.fsf@mira.informatik.hu-berlin.de> <3D881B66.1080200@something.invalid> <j4wupjafun.fsf@informatik.hu-berlin.de> <3D8868C3.9090005@one.net>
Message-ID: <m3wupjmqb7.fsf@mira.informatik.hu-berlin.de>

Michael Schneider <michaels at one.net> writes:

>     "stackless python behaves the same as 'normal' python, until you
>     import the stackless module.  At that point, it becomes
>     stackless python.  This allows you to run 'stackless unsafe'
>     modules with the same interpretor as stackless modules, in a
>     predictable manner (as long as you don't import stackless,  and
>     an unsafe module)"
> 
> 
> Is this the correct understanding????

That is correct (atleast, it should be correct - I haven't actually
reviewed the changes in this respect).

Regards,
Martin


From cbrown at metservice.com  Sun Sep 22 21:41:21 2002
From: cbrown at metservice.com (Colin Brown)
Date: Mon, 23 Sep 2002 13:41:21 +1200
Subject: PyQt setup on Linux Mandrake 8.2
References: <3D8B4898.20808@nospam.attbi.com>
Message-ID: <3d8e7141$1@news.nz.asiaonline.net>

"dwelch" <dwelch91 at nospam.attbi.com> wrote in message
news:3D8B4898.20808 at nospam.attbi.com...
>
> I would like to setup PyQt 3.4 on my Mandrake 8.2 box.
> ----snip----
>

Hi Don

I few months ago I installed PyQt under Mandrake Linux 8.2 for i586
and ended up doing the following:
----------------------------------------------------------------------------

Put tmake_1.8.tar.gz into /usr/local
Put sip-3.3.2.tar.gz into /usr/local
Put PyQt_3.3.2.tar.gz into /usr/local

gunzip tmake_1.8.tar.gz
tar -xvf tmake_1.8.tar
gunzip sip-3.3.2.tar.gz
tar -xvf sip-3.3.2.tar.gz
gunzip PyQt_3.3.2.tar.gz
tar -xvf PyQt_3.3.2.tar.gz

TMAKEPATH=/usr/local/tmake_1.8/lib/linux-g++
export TMAKEPATH

cd /usr/local/sip-3.3.2
python build.py -q /usr/lib/qt2 -i /usr/lib/qt2/include -g
/usr/local/tmake_1.8/bin/tmake -l qt-mt
make
make install

cd /usr/local/PyQt_3.3.2
python build.py -q /usr/lib/qt2 -g /usr/local/tmake_1.8/bin/tmake -l qt-mt
make
make install

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

I hope you find this useful.

There was a compile error that have yet to follow up on to do with glu but
apart from that I ended up with a pyuic that works okay with output from
Qt_Designer except for the table and custom widgets. When I get time I
will follow up on these.

Cheers
Colin Brown
PyNZ






From imbosol at vt.edu  Fri Sep  6 18:04:23 2002
From: imbosol at vt.edu (Carl Banks)
Date: Fri, 6 Sep 2002 18:04:23 -0400
Subject: successor to htmllib
References: <mailman.1031345949.12743.python-list@python.org>
Message-ID: <alb8si$qpp$1@solaris.cc.vt.edu>

Erik Price wrote:
> I noticed that the htmllib module is really best suited to HTML 2.0 
> documents.  I was wondering if there was a newer (4+) HTML or even 
> XHTML parsing module in development right now.  (By XHTML parsing lib I 
> mean perhaps an XML parser that is specifically written for the XHTML 
> 1.0 DTD.  I realize that any XML parser -should- work.)


Probably a better idea is to deprecate it and sgmllib.

Having used sgmllib for an HTML preprocessor, I can say its design is
not versatile at all, and trying to make it do anything different than
it was originally intended tends to be a headache.

sgmllib's reason for existence is to support the very limited subset
(term used loosely) of SGML needed by htmllib.  htmllib's reason for
existence is to support the rendering of HTML by an event-based
program.  I suppose they do that job satisfactorily.  However, the
more you deviate from their narrow intended uses, the more useless
they become.  sgmllib appears to be fairly buggy as well, although I
am not an SGML expert so I can't say how much.

I'd say you're better off using the XML tools, even though I don't
know anything about them, I can't imagine they would be less
versatile.


-- 
CARL BANKS
http://www.aerojockey.com


From see_reply_address at something.invalid  Sun Sep 22 20:41:55 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Mon, 23 Sep 2002 12:41:55 +1200
Subject: Larry Wall's comment on python...
References: <wkr8g7rn4c.fsf@yahoo.com> <mailman.1031377930.4225.python-list@python.org> <i09louorcaro2mvmt0eris7aa0sbvs11fn@4ax.com>
Message-ID: <3D8E6353.7030303@something.invalid>

Courageous wrote:

> One should use spaces
> and not tabs in Python code. Otherwise, any python file where
> tabs and spaces are mixed is also a choice of which indentation
> level other viewers of the code MUST use in their editors.


If you use spaces only, you're also choosing how other
people will see it in their editors (unless they're
willing to use a tool to re-indent it).

Using *tabs* only is the only way to get truly
preference-independent indentation.

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From martin at v.loewis.de  Sat Sep 21 06:43:36 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 21 Sep 2002 12:43:36 +0200
Subject: Porting woes
References: <3a4b2539.0209200809.407d50d5@posting.google.com> <j4sn041tuz.fsf@informatik.hu-berlin.de> <3a4b2539.0209201928.10c577b9@posting.google.com>
Message-ID: <m3wupfk4h3.fsf@mira.informatik.hu-berlin.de>

jstavnitzky at atsautomation.com (Jay) writes:

> Thanks for the suggestion!! It appears as though this particular
> problem goes away if I extend the static initialization of the
> PyTypeObject PyBaseObject_Type to explicitly set the tp_bases value
> (and the other six values that are not explicitly assigned at the end
> of the structure to NULL).  If I don't explicitly zero these values,
> they pick up whatever happens to be left over in those memory
> locations, is it possible that this is a compiler bug?? 

If the compiler fails to zero-initialize fields that are defined in
the struct, but not provided in the initializer, then yes, that's a
compiler bug.

> I assume that this code compiles correctly on most other compilers,
> or is zeroing out memory locations before use an operating system
> function??

It depends on the system. Typically, the "system" zero-initializes all
objects which are completely uninitialized, by means of the .bss
section (although the "system" might be a user-space function here,
such ld.so.1). For partially-initialized data, the compiler puts them
into the .data section, which typically requires complete
initialization in the assembler file.

> I hope that there aren't other places in Python where structures are
> not fully initialized (and I may see a bug of this type again).

Dozens. Virtually every type object is incomplete, as are method
tables in many cases.

Regards,
Martin


From martin at v.loewis.de  Mon Sep 23 18:51:58 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 24 Sep 2002 00:51:58 +0200
Subject: Saving & Restoring Python objects in XML?
References: <3D8F7800.F757A72D@wyeth.gsfc.nasa.gov>
Message-ID: <m31y7kiak1.fsf@mira.informatik.hu-berlin.de>

Tom Bridgman <bridgman at wyeth.gsfc.nasa.gov> writes:

> Is there a way to write the parser handlers and/or other components (I'm
> using PyXML 0.7 and higher) such that they can assemble the required
> objects as they read in the file?

If you have mostly simple data types, xml.marshal.generic might be
appropriate. If you have classes that you want to marshal and
unmarshal, inheriting from that framework might help.

Regards,
Martin



From mwh at python.net  Mon Sep 30 05:56:52 2002
From: mwh at python.net (Michael Hudson)
Date: Mon, 30 Sep 2002 09:56:52 GMT
Subject: Python Compiling
References: <amfveg$5g64k$1@ID-53055.news.dfncis.de>
Message-ID: <lkit0nvm93.fsf@pc150.maths.bris.ac.uk>

"Terje Johan Abrahamsen" <spoermeg at voldelig.com> writes:

> But, I guess my question boils down to, why is not Python a compiled
> language?

The unhelpful answer: it is!  It compiles to a high-level bytecode.
See other replies for info about compiling to a lower-level
instruction set...

Cheers,
M.

-- 
  I sense much distrust in you.  Distrust leads to cynicism, cynicism
  leads to bitterness, bitterness leads to the Awareness Of True
  Reality which is referred to by those-who-lack-enlightenment as
  "paranoia".  I approve.    -- David P. Murphy, alt.sysadmin.recovery


From deblnononospammmmmyyy at theworld.com  Thu Sep 19 00:07:47 2002
From: deblnononospammmmmyyy at theworld.com (David Lees)
Date: Thu, 19 Sep 2002 04:07:47 GMT
Subject: How to build Python 2.2.1 on Windows?
References: <mailman.1032402712.14232.python-list@python.org>
Message-ID: <3D894D93.CAE3C658@theworld.com>

OK, I see the PCbuild directory and the readme.txt.

Thanks everyone for the pointers.

David Lees
(Sorry I accidentally posted this reply as a post instead of a reply)

David LeBlanc wrote:
> 
> I have built Python on Windows using both the distribution and CVS download.
> The necessary VC++ project files are included in both. You will have to also
> download some other components as described in the readme file, including a
> special version of BSDDB version 1.85, zlib, Tcl/Tk (Tk needs Tcl to build),
> expat and maybe some other things. These are for extensions, and are not
> required to build the core Python.
> 
> WRT BSDDB, there are projects out there to build Sleepycat BSDDB version 4.0
> and that's a better choice then the default version 1.85 used by Python (due
> to bugs). The downside is that Sleepycat BSDDB is free for personal use
> only.
> 
> See \dist\Python-2.2.1\PCbuild for complete details.
> 
> Oh yes, put your python source distro or cvs dump under a directory called
> 'dist'. I found the directions to do that unclear in the readme and it gave
> me a bit of trouble. The other packages you'll need also go in 'dist' as
> peers of the python source tree. Those should be built before Python.
> 
> David LeBlanc
> Seattle, WA USA
> 
> > -----Original Message-----
> > From: python-list-admin at python.org
> > [mailto:python-list-admin at python.org]On Behalf Of David Lees
> > Sent: Wednesday, September 18, 2002 18:43
> > To: python-list at python.org
> > Subject: How to build Python 2.2.1 on Windows?
> >
> >
> > I have been playing with Python for awhile and would like to know how to
> > try building Python on a Windows machine that has MS C++ version 6.  Is
> > there a standard download that is configured to do this?  If not could
> > someone point me in the direction of information on builds?  I
> > downloaded Python-2.2.1.tgz.tar on my Win98 box and WinZip 8 does not
> > even seem to know how to unzip it.  I just transfered the tar file over
> > to a linux box and it unzips without and problem using tar.  The README
> > seems to only talk about linux/unix type systems.
> >
> > David Lees
> > --
> > http://mail.python.org/mailman/listinfo/python-list


From uwe_becher at gmx.de  Tue Sep  3 14:28:22 2002
From: uwe_becher at gmx.de (Uwe Becher)
Date: Tue, 03 Sep 2002 20:28:22 +0200
Subject: trailing newline in string.splitfields()
References: <Gv3d9.287721$me6.36853@sccrnsc01>
Message-ID: <3D74FF46.3040301@gmx.de>

Did you try regular expressions?

list = re.split('\s{1,2}',stringtosplit)

HTH
Uwe

Micah Mayo wrote:
> I was splitting up some fields in a textfile using string.split, which 
> works fine, but that it usee ' ' as the delimiter. I've found that some 
> of the things I want to split include a space or two, so I've gone to 
> using string.splitfields, which also works fine, but for one thing. 
> Unlike string.split() it also grabs the newlne('\n') at the end of each 
> line.
> 
> The way I've worked around this is as such:
> 
> var1,var2,var3 = string.splitfields(stringtosplit,',')
> list.append(var1)
> list.append(var2)
> list.append(var3[:-1])
> 
> So, I slice of the character, easy enough. And I'm generally satisfied 
> with that - but I was wondering if there may be, for the cause of more 
> beautiful code, a way to tell string.splitfields() to do it automatically.
> 
> Thanks,
> Micah
> 



From see_reply_address at something.invalid  Tue Sep 24 22:25:07 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Wed, 25 Sep 2002 14:25:07 +1200
Subject: more on weak typing..
References: <uospj56t55o858@corp.supernews.com>
Message-ID: <3D911E83.5090603@something.invalid>

Lance wrote:
 >

>   dynamically typed language
>   A language in which types are discovered at execution time; the opposite
> of statically typed. VBScript and Python are dynamically typed, because they
> figure out what type a variable is when you first assign it a value.


That's not quite accurate. In most dynamically-typed languages,
variables don't have types *at all*; types are associated with
values, not variables. If you took the above statement literally,
you might think that once you had assigned something of a given
type to a variable, you couldn't assign something of a different
type to it later -- which is obviously not true of Python (or
Lisp, or Scheme, or Smalltalk, or any other d-t language I can
think of!)

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From peter at engcorp.com  Thu Sep 19 07:20:21 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 19 Sep 2002 07:20:21 -0400
Subject: Minimal debug/rep functionality
References: <mailman.1032406044.16132.python-list@python.org>
Message-ID: <3d89b2e9$1@news.sentex.net>

Delaney, Timothy wrote:
>>From: Peter Hansen [mailto:peter at engcorp.com]
>> >>> def trace(frame, event, arg):
>>...   if event == 'line':
>>...      print '%s: %s' % (frame.f_code.co_filename, frame.f_lineno)
>>...   return trace
>>...
>> >>> import sys
>> >>> sys.settrace(trace)
> 
> 1. frame.f_code.co_filename will return the file that the .pyc was compiled
> from. Particularly in the case of standard libraries, this will often not be
> the path to the file in *your* filesystem. There is a way to get the
> corresponding file in your filesystem (basically, determine the module the
> frame comes from, then get the module filename using inspect).

True... presumably either the OP is not concerned about standard
library modules anyway, or he would check the archives and add
a simple way of doing this.  YAGNI...

> 2. The above will be *slow*. Very very slow. For a short program, this may
> not matter. For a long program, it will (I'm talking order of magnitude
> slower).

Hmmm... I didn't see any implication that speed was even remotely
a concern for the OP.  After all, he did ask to *print* each line
as it executed.  What could possibly be slower than that?
(I doubt he would care to deal with the volume of information
generated when using this with a large program anyway... needs
a way to turn it on and off, at least.)

My standard comments about Python speed and Knuth's comments about
premature optimization apply...

I'm looking forward to seeing your tool though, Timothy. :)

-Peter



From spoermeg at voldelig.com  Tue Sep 17 19:35:05 2002
From: spoermeg at voldelig.com (Terje Johan Abrahamsen)
Date: Tue, 17 Sep 2002 18:35:05 -0500
Subject: Source code
References: <am3ms1$2e49b$1@ID-53055.news.dfncis.de> <mailman.1032152058.11043.python-list@python.org> <m3admizdzq.fsf@mira.informatik.hu-berlin.de> <37fh9.112911$pX1.4044383@news2.tin.it>
Message-ID: <am8eq0$3ml4d$3@ID-53055.news.dfncis.de>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:37fh9.112911$pX1.4044383 at news2.tin.it...
> Martin v. Loewis wrote:
>
> > Gerhard H?ring <gerhard.haering at gmx.de> writes:
> >
> >> You should perhaps reconsider if and why you need to hide your source
> >> code.
> >
> > You should also recognize that, if your computer can execute and
> > perform a certain algorithm, users can always observe simulate the
> > computer operations, thus breaking any protection scheme.
> >
> > You'll have to give your customers a sealed computer and no permission
> > to install additional software, or communicate with the internet, if
> > you want to reliably prevent them from analysing your program.
>
> Incidentally, the latter option can be feasible in certain situations:
some
> key parts of your program's functionality might reside on a computer you
> control, accessible to the rest of the application (installed more
> traditionally on users-controlled computers) only as a black box via the
> net.  XML-RPC, SOAP, and other distributed-computing approaches
> such as Corba, make implementing this particularly easy these days.
>
> Feasibility depends on how troublesome it is to demand that the
> users' computers be net-connected whenever the application may
> run, and how time-critical is the round-trip time whenever the
> code that runs on the users' computers must delegate some part
> of an operation to code running on the computer you control.  In
> some application areas these feasibility issues may be killers, in
> others they're no trouble at all.  When the latter conditions prevail,
> this approach does yield something like a protection scheme that
> you can, at least in theory, make "unbreakable" (by satisfying other
> conditions, such as statefulness and other ways to make the part
> running under your control arbitrarily hard to reverse-engineer).

I have been thinking about that. But, I assume a lot of people are not
having a connection that is always on. And, personally I am a little
sceptical to programs that tries to access the net, but doesn't have any
reason to.




From jdavis at empires.org  Tue Sep 24 16:25:47 2002
From: jdavis at empires.org (Jeff Davis)
Date: Tue, 24 Sep 2002 20:25:47 GMT
Subject: help
References: <amqclg$6uu$1@server.ltid.inpe.br>
Message-ID: <fT3k9.62348$U7.21804606@twister.socal.rr.com>

Juliana Braga wrote:

> I need help.
> For example:
> x = ?name1?
> How can I get the value of variable x (in this case the value of variable
> x is ?name1?). After , I?d like to do this:
> value of variable x = ?teste?.

You're not being entirely clear, but it looks like what you want to do is:

x = 'name1'
x = 'teste'

Values cannot change into other values. A value is a value, and a variable 
can change the value that it holds, but a value cannot change. So maybe 
you could clarify what you want to do?

Regards,
        Jeff


From cliechti at gmx.net  Sat Sep  7 09:14:44 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 7 Sep 2002 15:14:44 +0200
Subject: Embedded Python implementation
References: <MAfe9.266658$v53.14092531@news3.calgary.shaw.ca>
Message-ID: <Xns92829B12D2570cliechtigmxnet@62.2.16.82>

Evan <tagith1394 at hotmail.com> wrote in 
news:MAfe9.266658$v53.14092531 at news3.calgary.shaw.ca:

> Hello,
> 
> I am currently trying to embed Python in an application I am writing, 
> and all is going well, except for one inexplicable bug..
> 
> When I pass a variable (pointer to a string) to PyRun_String, it will 
> always error on line 2, character 5.. regardless of what is actually on 
> the line in that spot (for that matter, even if there is no line 2)..

that sounds as you woul not pass the pointer to a string but to something 
else...

> The script works perfectly when I place it in a file, and use 
> PyRun_File, and it also works perfectly when I place the string in 
> quotations, and pass it to PyRun_String directly inside the C code..
> 
> The C code in question is as follows:
> if ( PyRun_String( com_list, Py_file_input, pGlobals, pLocals ) == NULL )
>          PyErr_Print();

i think that sould work when com_list is a "char *", say:
char * com_list = "print hello\n";
 
> I am confident the problem is not with pGlobals or pLocals, as those 
> arguments are used with PyRun_File, as well as with the quoted string..

chris
-- 
Chris <cliechti at gmx.net>



From pyth at devel.trillke.net  Mon Sep 30 19:42:34 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 1 Oct 2002 01:42:34 +0200
Subject: Question about __getattribute__/__getattr__
In-Reply-To: <Pine.LNX.4.44.0209301746400.12980-100000@mercury.sabren.com>; from jpm@papercut.org on Mon, Sep 30, 2002 at 05:51:16PM -0400
References: <20020930234554.Y30315@prim.han.de> <Pine.LNX.4.44.0209301746400.12980-100000@mercury.sabren.com>
Message-ID: <20021001014234.Z30315@prim.han.de>

Joao Prado Maia wrote:
> On Mon, 30 Sep 2002, holger krekel wrote:
> 
> > The important lines from the above link are:
> > 
> > def __call__(self, *args):
> >         if not self.memo.has_key(args):
> >             self.memo[args] = self.fn(*args)
> >         return self.memo[args]
> > 
> > and 'args' is a tuple of arguments.  
> > 'args[0]' would give you the first argument, for example.  
> > 
> 
> Correct. However, the same is not valid for __getattr__.

For caching a function's results why do you need __getattr__?

It seems you really need another caching policy than the prescribed
memoize pattern.  For example, you may implement a dictish object that evicts
old cached values by means of counting accesses, oldest access times
or whatever.  I still see no use for expecting __getattr_ to know about
the arguments.  Not?

    holger



From robin at execulink.com  Sun Sep 15 14:40:14 2002
From: robin at execulink.com (robin at execulink.com)
Date: Sun, 15 Sep 2002 14:40:14 -0400
Subject: Good book choice?
References: <e6c4872c.0209150635.4816ef98@posting.google.com>
Message-ID: <05k9ou4o2m9e41puce9ni3953b5hbhn8c0@4ax.com>

eva_ag at eresmas.com (Eva) wrote:

>I am new to Python but an experienced programmer in Fortran95 (not too
>object
>oriented). I am looking for a suited Python book. I'd like it to have
>clear ideas about OO, as well as an straight pointing into the core
>language.
>
>It seems that 'Python Essential Reference (2nd Edition)' is a good
>choice. There is also the book 'Making use of Python' (by Rashi
>Gupta). Is there any user that bought both? Is it good the treatment
>of OO in the former? Is there a lack of applications point of view in
>the former? Could anybody compare them? pros/cons?

Can't help you with those, but have had luck with Learning Python
(Mark Lutz & David Ascher, pub. O'Reilly), a great introduction to the
language that doesn't pad out its length with stuff you can read in
the reference documentation. I think it's the single-best Python book
out there. So do the other developers I work with.

Then you'll want to get the Python Cookbook (edited by Alex Martelli &
David Ascher, pub. O'Reilly) for in-depth analysis of common tasks and
thoughtful solutions. This is more advanced than your typical
Cookbook. It's going to be by my side a long time!

Start here: http://python.oreilly.com

-- robin


From malan at club-internet.fr  Mon Sep 30 15:29:52 2002
From: malan at club-internet.fr (malan)
Date: Mon, 30 Sep 2002 21:29:52 +0200
Subject: About data structure
Message-ID: <3d98a5d8$0$496$7a628cd7@news.club-internet.fr>

Hi all,

I need to build a data structure like this :

size	4 bytes ( example 0x0000000A )
flag	1 byte  ( example 0x87 )
trans   8 bytes ( example 'URTVMI1' )
F5	3 bytes ( example 0x00125C )

How can I define that structure, format it and send it via a socket ?
My problem is to code / decode the message to and from the server.

malan



From DennisR at dair.com  Thu Sep 19 00:22:07 2002
From: DennisR at dair.com (Dennis Reinhardt)
Date: Thu, 19 Sep 2002 04:22:07 GMT
Subject: Python comment stripper
References: <mailman.1032398005.31709.python-list@python.org>
Message-ID: <Phci9.755$5%6.50095251@newssvr13.news.prodigy.com>

> Um ... if you are including Python22.dll in the download, surely that is
> dwarfing your code size?
> Therefore your package sizes would go from 368 KB -> 366 KB.

Compressing a single .py file is lost in noise, agreed.

> Even assuming you have 100 times as much (compressed) code, and the
comment
> savings are identical:
> 663 KB -> 499 KB
> Is the savings worth the hassle?

The entire lib directory is 177 files comprising 1.92 Mb.  The corresponding
savings are (approximately)
910 -> 610 Kb.  One packaging possibility here is that a Python EXE consists
of Python22.DLL and the entire lib directory at a cost of approximately 610
Kb.  There is no dependency checking.  Wham -- the entire library is
included.  Indeed, building this could be done on a web page by dropping
your main Python code into a web edit box (see
www.http://www.autorunning.com/dgram5.cgi for an online EXE builder I
operate)  That sounds pretty hassle-free to me.  It is also time-consuming
to unpack dynamically (not an obvious tradeoff).  At 910 Kb, we are getting
closer to doing a full install in terms of download size.

There are other packaging possibilities.  I cannot promise I will go the way
hinted above.  Right now, a comment stripper is not perfected and that is a
major holdback.   My own thinking is I will not do comment stripping in
first release.  There are other follow on ideas I am mulling over if this
app gains a following.

The savings vs. hassle tradeoff has some favorable and some unfavorable
domains.  I don't think there is a single yes/no answer.
--

Dennis Reinhardt

DennisR at dair.com
http://www.dair.com




From koenig at v-i-t.de  Sun Sep 29 05:38:14 2002
From: koenig at v-i-t.de (Mirko Koenig)
Date: Sun, 29 Sep 2002 11:38:14 +0200
Subject: tkinter app won't start on wmaker startup
Message-ID: <an6i35$imh$1@news.tu-darmstadt.de>

Hi

When i save session with hwindowmaker all applications that are open will
be startet at the net restart of windowmaker again.
Only my tkinter app is not recognized by windowmaker and so not started
at startup

What code have i to write that a X windowmaker can remember the app?

someone told me to write:

root = Tk(className="appname")
root.title("appname")

But this didn't work

Thanx for help

Mirko Koenig


From edcjones at erols.com  Wed Sep 18 18:23:16 2002
From: edcjones at erols.com (Edward C. Jones)
Date: Wed, 18 Sep 2002 18:23:16 -0400
Subject: linux python ideas
References: <3D88E578.6010001@diespammerdieuselesspython.com>
Message-ID: <amau0j$rkn$1@bob.news.rcn.net>

Rob Andrews wrote:
> I'll be giving a presentation on Python to my local LUG 
> (http://lugoj.org) in a few weeks, and would like to point out at least 
> a few things that would be of particular interest to linux users.
> 
> The idea is not "how to program in Python", which would take more than 
> one quick presentation, but a one-off demonstration of ways in which 
> Python is particularly useful for such a group.
> 
> Any suggestions would be appreciated, since my notes are pretty generic 
> so far.

I use Python instead of shell scripts for anything that is not 
completely trivial. Here is a small program for finding all the links in 
a page of html:

#! /usr/bin/end python

import sys

if len(sys.argv) != 2:
     raise Exception, 'program must have exactly one argument.'
infile = sys.argv[1]
page = open(infile, 'r').read()
lowerpage = page.lower()

start = 0
while 1:
	start = lowerpage.find('href="', start)
     if start == -1:
	break
     start = start + len('href="')
	end = lowerpage.index('"', start)
     print page[start : end]



From fromclay at maxitec.co.za  Wed Sep  4 03:31:45 2002
From: fromclay at maxitec.co.za (xulit77)
Date: Wed, 4 Sep 2002 09:31:45 +0200
Subject: Is there light at the end
Message-ID: <al4bor$sv1$1@ctb-nnrp2.saix.net>


 Could someone please help me with this?

> Im really new to any kind of programming and the people at python.org
> return my messages in some other code :not python(lol)The people there are
> cool I just dont seem to understand there answers.If posible would you be
> able to answer in detail.
>
> is it wrong of me to send the ppp.log file to just anyone?
>
> HERE IS WHAT I HAVE:
>
> import re
> import string
>
> secondsMatcher=re.compile("Connect time: (\d+) secs")
>
> def getStuff(file):
>     closeline = 0
>     errorcount = 0
>     secondscount = 0
>     connectcount = 0
>     data_outcount = 0
     data_incount = 0
     ppplist = file.readlines()

     for line in ppplist:

       if string.find(line,"LayerUp") >= 0:
           print "found connect"
           connectline = string.split(line)
           connectcount = connectcount + 1

       elif string.find(line,"ERROR") >= 0:
           print "found error"
           errorcount = errorcount + 1

        elif string.find(line,"Connect time") >= 0:
           print "found close"
connectSecs=secondsMatcher.search(line).group(1)
           print "Connected for",connectSecs
          closeline = string.split(line)
           secondscount = secondscount + string.atoi(connectSecs)

 def main():
    f=open("C:/my documents/ppp.log","r")
     getStuff(f)
     return None

 if __name__=="__main__":
     main()

 THIS IS WHAT I NEED THE PROGRAM TO DO:

 I want the program to read the attached ppp.log file and give me the
 following type of report:


 PPP Log file summary for 5 Jul 2002
 -----------------------------------
 Jul  5 18:15:01 Connected to internet (IP 155.239.150.146)
 Jul  5 18:18:08 Closed connection. ( Connect time: 197 secs: 5091 octets
in,
 1864 octets out )
 -----------------------------------
 Successful connections: 5
 Failed connections: 2
 Total time online: 456 seconds Total data in: 66576 octets
 Total Data out: 66349 octets

HERE IS part of the olg file seems noone wants to touch the Attachment(lol)

> Jul  5 18:00:00 maxigate newsyslog[87300]: logfile turned over
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: Connected to local client.
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Command: /var/run/internet: quit
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: /var/run/internet: Client
connection dropped.
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: Connected to local client.
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Command: /var/run/internet: set
timeout 180
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Command: /var/run/internet: dial
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: bundle: Establish
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: deflink: closed -> opening
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: deflink: Connected!
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: deflink: opening -> dial
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Chat: Phone: 0860007249
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Chat: deflink: Dial attempt 1 of
1
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Phase: /var/run/internet: Client
connection closed.
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Chat: Send: AT^M
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Chat: Expect(5): OK
Jul  5 18:15:00 maxigate ppp[34607]: tun0: Chat: Received: AT^M^M
Jul  5 18:15:01 maxigate ppp[34607]: tun0: Chat: Received: OK^M
Jul  5 18:15:01 maxigate ppp[34607]: tun0: Chat: Send: ATE1Q0^M
Jul  5 18:15:01 maxigate ppp[34607]: tun0: Chat: Expect(5): OK
Jul  5 18:15:01 maxigate ppp[34607]: tun0: Chat: Received: ATE1Q0^M^M
Jul  5 18:15:01 maxigate ppp[34607]: tun0: Chat: Received: OK^M
Jul  5 18:15:01 maxigate ppp[34607]: tun0: Chat: Send: ATx1DT0860007249^M
Jul  5 18:15:03 maxigate ppp[34607]: tun0: Chat: Expect(40): CONNECT
Jul  5 18:15:22 maxigate ppp[34607]: tun0: Chat: Received:
ATx1DT0860007249^M^M
Jul  5 18:15:22 maxigate ppp[34607]: tun0: Chat: Received: CONNECT
33600/ARQ/V34/LAPM/V42BIS^M
Jul  5 18:15:22 maxigate ppp[34607]: tun0: Phase: deflink: dial -> carrier
Jul  5 18:15:23 maxigate ppp[34607]: tun0: Phase: deflink: /dev/cuaa0: CD
detected
Jul  5 18:15:23 maxigate ppp[34607]: tun0: Phase: deflink: carrier -> login
Jul  5 18:15:23 maxigate ppp[34607]: tun0: Phase: deflink: login -> lcp
Jul  5 18:15:23 maxigate ppp[34607]: tun0: LCP: FSM: Using "deflink" as a
transport
Jul  5 18:15:23 maxigate ppp[34607]: tun0: LCP: deflink: State change
Initial --> Closed
Jul  5 18:15:23 maxigate ppp[34607]: tun0: LCP: deflink: State change
Closed --> Stopped
Jul  5 18:15:24 maxigate ppp[34607]: tun0: LCP: deflink: LayerStart
Jul  5 18:15:24 maxigate ppp[34607]: tun0: LCP: deflink: SendConfigReq(179)
state = Stopped
Jul  5 18:15:24 maxigate ppp[34607]: tun0: LCP:  ACFCOMP[2]
Jul  5 18:15:24 maxigate ppp[34607]: tun0: LCP:  PROTOCOMP[2]
Jul  5 18:15:24 maxigate ppp[34607]: tun0: LCP:  ACCMAP[6] 0x00000000
Jul  5 18:15:24 maxigate ppp[34607]: tun0: LCP:  MRU[4] 1500
Jul  5 18:15:2





From gerson.kurz at t-online.de  Sun Sep  1 14:31:29 2002
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Sun, 01 Sep 2002 18:31:29 GMT
Subject: Making regex suck less
Message-ID: <3d725881.345921@news.t-online.de>

I wrote a small wrapper module to "struct" today that will allow me to
write C declarations and have 

...
print p.declare("""

struct test1
{
   int a,b;
   float this,and,that;
};

struct test2
{
    int count;
    struct test1 data[80]
}; 

""")
...
and have it create instances I can easily assign data to
...
t2 = p.createInstance("test2")
t2.data[0].a = 42
...
and pack for C extensions by utilizing the struct module
...
data = p.pack(t2)
...

because of being fed up with that pack("qh34s>id",...) stuff. (I'll
post that to my website once its in a state I can let somebody else
see ;) 

Anyway, that got me thinking on why do we have to deal with regular
expressions like r"((?:a|b)*)", when in most cases the code will look
something like this:

r = re.compile("<some cryptic re-string here>")
...
r.match(this) or r.find(that)

which means the real time is not spent in the compile() function, but
in the match or find function. So basically, couldn't one come up with
a *human readable* syntax for re, and compile that instead? Python
prides itself on its clean syntax, and human readability, an bang -
import re, get perl-ish code instantly! 

Also, I think it would already be an improvement if the syntax
provided for clear and easy-to-understand special cases, like

re.compile("anything that starts with 'abc'")

and if you cannot find something in the special cases for you, you can
always go back to 

re.compile("<some cryptinc re-string here>")

After all, *everyone* starting with re thinks the syntax is cryptic
and mind-boggling, and only if you get yourself into the "re mindset",
you understand things like r"\s*\w+\s*=\s*['\"].*?['\"]" instantly. If
we had an easier syntax, more people would be using re ;) 

Is the idea utterly foolish? 



From KBASS1 at nc.rr.com  Mon Sep  2 05:30:08 2002
From: KBASS1 at nc.rr.com (Kevin Bass)
Date: Mon, 02 Sep 2002 09:30:08 GMT
Subject: Newby Question: os.rename
Message-ID: <AcGc9.68214$Xa.3523883@twister.southeast.rr.com>

I am attempting to create an archive function within my program. This
function should create a directory with a timestamp name (such as
/home/oracle/data/20020831) on the fly, list the contents of a current
directory then place the contain from the previous directory into the newly
created directory.

The newly created directory with the timestamp name is being created and the
list of files from the current directory has been retrieved but my problem
comes when moving the files to the newly created directory. I am trying to
achieve this tasks with os.rename but find it difficult to get the command
to recognize the new directory and place the files into it.

I have attempted to get os.rename to recognize the 'archloc' variable that
contains the name of the newly created directory but that did not work.  I
also tried to join the newly created directory with the file names then send
it through os.rename and that did not work either.
What am I doing wrong? Thanks!! (My code is listed below)


def archive_data(archive):

   the_date = strftime("%Y%m%d")
   archloc = ('/home/oracle/data/' + the_date)
   os.mkdir(archloc)

   if archive == 'yes':
      for file in os.listdir('/home/oracle/data'):
         file2 = os.path.join(archloc, file)
         #print file2
   os.rename(r'/home/oracle/data/' + file, file2)

Kevin






From hst at empolis.co.uk  Wed Sep 25 10:52:48 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Wed, 25 Sep 2002 15:52:48 +0100
Subject: Newbie References Question
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB04F41E@hendrix.empolisuk.com>

Guy Rabiller wrote:
> 
> Hi,
> 
> let say I have:
> i1 = 1
> i2 = 2
> and
> p = [i1,i2]
> 
> How can I have:
> p=[*i1,*i2]
> rather than
> p=[**i1,**i2] as it is curently ?
> 
> ( Sorry for this nasty hybrid syntax )
> 
> Textualy, how if I want that p[0] refere to the 'i1' 
> container, and not to
> the refererence it contains ?
> 
> What I want is that if now I set:
> i1 = 4
> that automaticaly:
> p -> [4,2]
> and not keeping [1,2]
> 
> Thanks in advance.
> 
> --
> guy rabiller
> 3d animator / td
> grabiller at 3dvf.net
> http://grabiller.3dvf.net
> 
Well it is just about possible, but its an ugly hack.

Basically, all the members of the list p will need to be mutable - in practice each will need to be a list.

So:

>>> i1 = [1]
>>> i2 = [2]
>>> p = [i1, i2]
>>> i1[0] = 3
>>> p
[[3], [2]]
>>> i2[0] = 4
>>> p
[[3], [4]]

But I think it would be better if you told us what your real problem was so we could se if we could formulate a solution in a different way.

Harvey

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.



From gerhard.haering at opus-gmbh.net  Mon Sep 23 04:35:17 2002
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 23 Sep 2002 08:35:17 GMT
Subject: mySQL and Python 2.2
References: <mailman.1032633449.19570.python-list@python.org> <Elzj9.128797$ub2.2802895@news1.tin.it>
Message-ID: <slrnaotkj5.ic.gerhard.haering@haering.opus-gmbh.net>

In article <Elzj9.128797$ub2.2802895 at news1.tin.it>, Alex Martelli wrote:
> Greg & Janet LINDSTROM wrote:
> 
>> Hello-
>> I am trying to convert from gadfly to mySQL for my database needs on
>> my Windows 98 box. I have installed mySQL and it works (I have
>> created my tables and can manipulate them manually), and have
>> downloaded mySQL-python-0.9.1, but when I python setup.py build, I
>> get an error about cl.exe not existing (and, sure enough, it
>> doesn't :-).
>
> If you don't have a C compiler, you can't build C-coded Python
> extensions from sources:

C compilers for Windows are available relatively cheaply at
http://sourceforge.net/projects/mingw ;-)

> you need to find and install pre-built
> binary versions of those extensions for the Python version you
> are using.  You can find many such versions listed at (LONG
> URL, mail sw might break it -- if so, then rejoin it...):
> 
> http://sourceforge.net/project/showfiles.php?group_id=22307&release_id=102893

Yep. I've just sent 0.9.2 binaries to the maintainer and I think they'll show
up there soon. Btw. these binaries are the same I have on my homepage.

> [...] I think you can also find pre-built binaries of MySQLdb 0.9.2, for
> example at Gerhard H?ring's page, but I'm not sure about the details --
> Gerhard mentions said binaries at http://www.cs.fhm.edu/~ifw00065/news.shtml
> but gives no link and I can't find them.

You're not a true hacker :-P Else you'd have looked at the HTML code and seen
that the download link was only commented out. The reason was that they
depended on a libz.dll which normally isn't available on Windows systems:

I've been trying to get away from MSVC++ for building Free Software like MySQL.
Unfortunately, MySQL cannot be built out of the box using mingw on Windows, so
I needed to use a third-party patch against MySQL to get at least the MySQL
client libraries built.

Building MySQLdb (the Python adapter) itself using mingw is no problem
afterwards. But unfortunately, the MySQL client library I built was for some
reason dynamically linked against a libz.dll, which caused problems for Windows
users. Which is why I (and Andy) pulled back the previous 0.9.2 binaries.

After a very loud aaaaaaaaaaaaargh and a month later I've now given in and
built statically linked MySQLdb 0.9.2 binaries with MS Visual C++. And removed
the HTML comment on my homepage ;-)

-- Gerhard


From jtk at yahoo.com  Fri Sep  6 23:14:36 2002
From: jtk at yahoo.com (Jeff Kowalczyk)
Date: Fri, 6 Sep 2002 23:14:36 -0400
Subject: Available data types in Gadfly 1.0?
Message-ID: <albquv$qp4$1@main.gmane.org>

I'm experimenting with gadfly, and I'm looking for some documentation on
what data types are available for columns in Gadfly 1.0? so far, I only know
integer, float, varchar to work. Is there a date type? Boolean?







From loewis at informatik.hu-berlin.de  Thu Sep 26 02:36:38 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 26 Sep 2002 08:36:38 +0200
Subject: SET_LINENO
References: <3D92A46E.2030706@cygnus-software.com>
Message-ID: <j47kh9b6kp.fsf@informatik.hu-berlin.de>

Bruce Dawson <bruce_deletethis_dawson at cygnus-software.com> writes:

> Executive summary: Has the SET_LINENO removal been tested with the HAP
> debugger, and wouldn't a breakpoint instruction give much larger
> benefits?

To my knowledge: no, and no.

> It *sounds* like it has been done in such way to not cause debugger
> problems, but I'm curious if anyone has tested HAP with 2.3.

It was definitely designed to not cause problems for debuggers. If you
want to find out, you probably can easily test it yourself.

> Also, it sounds like the reason for removing SET_LINENO was to improve
> the performance of Python scripts. 

Correct.

> However one of the posts on this topic (can't find it now) suggested
> that the hoped for improvements had not actually materialized.

I think this suggestion is incorrect; the improvements can be
demonstrated.

> It seems to me (from the sidelines/peanut gallery/wherever) that the
> biggest performance problems in the SET_LINENO area are when
> debugging, not when running free, and that therefore this optimization
> is targeting the lesser problem.

No. Performing the SET_LINENO instruction if there is no debugging
going on contributed significantly to execution speed, which is the
reason for not emitting them under -O.

> So, wouldn't a break-point opcode be a more substantial improvement
> than removing SET_LINENO? Just bringing up that old topic again to see
> if it gets any more traction :-)

No. Removal of SET_LINENO does improve performance in a "free-running"
application, which adding a breakpoint instruction would not.

That said: it might be still worthwhile to add a breakpoint
instruction. There are a number of issues to be considered, such as
getting the breakpoint instruction into the byte code, and getting the
original instruction back in. So if you'ld like to see one, you
probably have to write a PEP.

> P.P.S. A breakpoint instruction would also be a boon for asserts - I'd
> love to be able to stop on the assert when running under a debugger,
> to see what is wrong before the stack frame unwinds.

I can't see how this helps. Wouldn't you have to set a breakpoint on
all asserts for that? And wouldn't that require finding all asserts in
the first place?

Regards,
Martin


From tundra at tundraware.com  Sat Sep  7 01:25:07 2002
From: tundra at tundraware.com (Tim Daneliuk)
Date: 07 Sep 2002 05:25:07 GMT
Subject: Larry Wall's comment on python...
References: <mailman.1031339711.632.python-list@python.org> <pc5bla.off.ln@boundary.tundraware.com> <roy-780F0D.21015006092002@reader1.panix.com>
Message-ID: <vb2cla.6di.ln@boundary.tundraware.com>

Roy Smith wrote:
<SNIP>

  Writing for a machine which only has 256 bytes of RAM gives you a
> somewhat different outlook on programming :-)

The Microchip PIC 16F84A chip I have been using as *68* bytes of RAM -
well RISC register file, really, but its the same thing -  32 bytes of
EEPROM, and a whopping 1K of flash memory for storing the program itself.
And yes, it does give you a whole different perpective on programming...

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com



From jjl at pobox.com  Tue Sep  3 12:56:57 2002
From: jjl at pobox.com (John J. Lee)
Date: Tue, 3 Sep 2002 17:56:57 +0100
Subject: Peculiar ImportError traceback
Message-ID: <Pine.SOL.4.44.0209031755160.7655-100000@mimosa.csv.warwick.ac.uk>

>>> from mx import DateTime
*** You don't have the (right) mxDateTime binaries installed !
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "mx/DateTime/__init__.py", line 8, in ?
    from DateTime import *
  File "mx/DateTime/DateTime.py", line 9, in ?
    from mxDateTime import *
  File "/usr/local/lib/python2.2/site-packages/reportlab/__init__.py", line 13, in ?
    myDir = os.path.split(__file__)[0]
ImportError: No module named mxDateTime
>>> from mx.DateTime import mxDateTime
>>>


Two obvious problems:

1. Bizarre last line of traceback involving reportlab (whose
__init__.py I list in full below).  Reportlab and mxDateTime, AFAIK,
are entirely unrelated!

2. mxDateTime's __init__.py (mx/DateTime/mxDateTime/__init__.py) does
an import mxDateTime, which then fails, causing the error message
about mxDateTime's binaries (ie. mxDateTime.so) to be printed.
Obviously, Python isn't finding the library, but I can't remember how
extension modules are located, nor could I find anything in the Python
docs on this -- could anyone point it out for me?

Even better, explain why on earth I get that line about reportlab!  Up
until now reportlab has sat in /usr/local/lib/python2.2/site-packages
quietly minding its own business.  The egenix-mx-base `mx' directory
(2.0.3) is also in the same site-packages directory.  Grepping through
reportlab, I see that one of its demos making use of mxTextTools, but
I don't think that's relevant.


John

#copyright ReportLab Inc. 2000-2001
#see license.txt for license details
#history http://cvs.sourceforge.net/cgi-bin/cvsweb.cgi/reportlab/__init__.py?cvsroot=reportlab
#$Header: /cvsroot/reportlab/reportlab/__init__.py,v 1.16 2001/08/13 12:17:12 rgbecker Exp $
__version__=''' $Id: __init__.py,v 1.16 2001/08/13 12:17:12 rgbecker Exp $ '''
__doc__="""The Reportlab PDF generation library."""
Version = "1.09"

def getStory(context):
    if context.target == 'UserGuide':
        # parse some local file
        import os
        myDir = os.path.split(__file__)[0]
        import yaml
        return yaml.parseFile(myDir + os.sep + 'mydocs.yaml')
    else:
        # this signals that it should revert to default processing
        return None



From michaels at one.net  Sat Sep  7 23:30:53 2002
From: michaels at one.net (Michael Schneider)
Date: Sat, 07 Sep 2002 23:30:53 -0400
Subject: Larry Wall's comment on python...
References: <mailman.1031377930.4225.python-list@python.org> <1549197.0PZNOebDTH@cartman>
Message-ID: <3D7AC46D.5010101@one.net>

I use xemacs for most of my editing, but I have been using
Komoda  from Activestate,  They place a highlight around the
blocks.

Congrats to Activestate on such a great tool!!!!!!!!!!

Mike

PS.  I am not affiliated with Active state, just a happy customer

yka wrote:

>Kow Kuroda wrote:
>
>
>>Well, Python code may look awkward --- if not awk-like ---
>>if you don't know how to customize the tab width in your
>>editor of choice. For example, vi and emacs indent by 8
>>characters by default, which I suspect is too wide for Python.
>>Some code with deep nesting wouldn't look very nice with
>>this default.
>>
>>Kow
>>
>
>Well at vim :set ts=4 makes more fitting
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020907/78311a97/attachment.html>

From imbosol at vt.edu  Fri Sep 13 14:13:38 2002
From: imbosol at vt.edu (Carl Banks)
Date: Fri, 13 Sep 2002 14:13:38 -0400
Subject: question about dictionary type..
References: <alt22f$pf8$1@newsreader.mailgate.org>
Message-ID: <altah0$2gj$1@solaris.cc.vt.edu>

eugene kim wrote:
> hi..
> 
> i'd like to have a dictionary which looks like
> table = { category1 : { category2 : { category3 : 'garbage value' } } }
> 
> category1 has many category2
> category2 has many category3
> 
> i expected
> table[ 'computer' ][ 'language' ][ 'python' ] = 0
> to create {computer : { language : { python :0 }}}
> table[ 'computer' ][ 'language' ][ 'java' ] = 0
> to becomes { 'computer' : { 'language' : { 'python' :0 , 'java' :0 }}}
> 
> but python doesn't allow me to do this..
> can anyone help me?


Yes.  Python is not Perl.  It does not automatically create
dictionaries or lists; you have to create the dict or list and bind it
to something first.  In other words, you can't do this:

   a['b'] = c

so ou have to do this:

   a = {}
   a['b'] = c

========

However, I thought it might be useful initialize deeply-nested tables
without having to explicitly create each new dictionary, so I thought
I'd program something up that does that (requires Python 2.2, of
course):


    class autonestdict(dict):
        def __getitem__(self,attr):
            try:
                return super(autonestdict,self).__getitem__(attr)
            except KeyError:
                return _autonestclosure(self,attr)

    class _autonestclosure(object):
        def __init__(self,nest,attr):
            self.nest = nest
            self.attr = attr
        def __getitem__(self,attr):
            return _autonestclosure(self,attr)
        def __setitem__(self,attr,value):
            obj = autonestdict()
            obj[attr] = value
            self.nest[self.attr] = obj


Then, you can use it like this:

    table = autonestdict()
    table[ 'computer' ][ 'language' ][ 'python' ] = 0
    table[ 'computer' ][ 'language' ][ 'java' ] = 1


The big problem with this class is when you do something like this:

    result = table['computador']

You don't get a KeyError; instead result gets bound to a
_autonestclosure object.  Which makes detecting errors kind of hard.


-- 
CARL BANKS
http://www.aerojockey.com


From shalehperry at attbi.com  Wed Sep 11 15:46:03 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 11 Sep 2002 12:46:03 -0700
Subject: deleting the first and the last character of a string
In-Reply-To: <alo5df$hoa$03$1@news.t-online.com>
References: <20020911105818.GC2054@partagas.as.de.cw.net> <mailman.1031756234.22942.python-list@python.org> <alo5df$hoa$03$1@news.t-online.com>
Message-ID: <200209111246.03080.shalehperry@attbi.com>

On Wednesday 11 September 2002 12:25, Markus wrote:
> Sean 'Shaleh' Perry wrote:
> >> Is there anything which says delete the first or last
> >> character of a string? I checked the string module on the web, but I
> >> couldn't find anything.
> >>
> >>>> s = '|f|o|o|'
> >>>> s.strip('|')
> >>>
> >>> s[1:-1]
>
> 'f|o|o'
>
> -Markus

True, but I like that the strip() call handles the case where one of the 
delimiters is missing.  One of the bugs that always bit me in perl was using 
chop() to eat the newline character when there was not a newline there to 
eat.  I try to avoid putting myself in that situation now.



From marklists at mceahern.com  Fri Sep 20 09:50:15 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 20 Sep 2002 08:50:15 -0500
Subject: isFloat: Without Exception-Handling
In-Reply-To: <3D8B23E5.3040005@thinkware.se>
Message-ID: <JHEOKEOOLIGLDHCMAHMOEECIEAAA.marklists@mceahern.com>

[Thomas Guettler]
> Is there a way to write the following method without using exceptions?

Please help me understand your aversion to the original code you posted.
IMHO, it has several salient characteristics:

1.  It works.
2.  It works.
3.  It works.

Did I forget to mention:

It works.

Anyway, I'd modify your original function slightly to use True/False
(builtin as of 2.2.1) and to be explicit about which errors we're ignoring:

def isFloat(s):
    is_float = True
    try:
        float(s)
    except (ValueError, TypeError), e:
        is_float = False
    return is_float

assert isFloat(1.2)
assert isFloat(1)
assert not isFloat(4+0j)
assert not isFloat("banana")
assert not isFloat(isFloat)
assert not isFloat(__import__)
...etc...

<wink>

Cheers,

// m

-




From grabiller at 3dvf.net  Thu Sep 26 01:00:37 2002
From: grabiller at 3dvf.net (Guy Rabiller)
Date: Thu, 26 Sep 2002 07:00:37 +0200
Subject: Newbie References Question -> Exemple
References: <3d91c72f$0$8558$79c14f64@nan-newsreader-02.noos.net> <3d91d2cf$0$1484$79c14f64@nan-newsreader-01.noos.net> <4nktma.mg3.ln@ix.netcom.com>
Message-ID: <3d92946c$0$18957$79c14f64@nan-newsreader-03.noos.net>

>         So just use
>
> quad = p

That could be an option but what I wrote was just an 'exemple'.

In the 'real world' I have to deal with points order, different lists of
points, etc..

But thanks anyway.

--
guy rabiller
3d animator / td
grabiller at 3dvf.net
http://grabiller.3dvf.net

Dennis Lee Bieber wrote:
> Guy Rabiller fed this fish to the penguins on Wednesday 25 September
> 2002 08:14 am:
>
>> Ok,
>>
>> thanks for your answers, I understand now that a variable name is
>> 'just' a name in a dictionary, and not a container.
>>
>> What I want to do is simple, and I suppose that this is my 'way of
>> thinking' regarding Python that is wrong.
>>
>> I have a list of points, and a list of polygons that should contain
>> 'references' to points.
>>
>> So for me:
>> -> points list:
>> p = []
>> p.append([0,0])
>> p.append([0,1])
>> p.append([1,1])
>> p.append([1,0])
>> -> polygon:
>> quad = []
>
>         Created a /new/ list
>
>> quad.append(p[0])
>
>         appended the value of p[0] to the new list
>
>> For me the quad list of point is not 'updated' with the change of one
>> of its point.
>>
>         So just use
>
> quad = p
>
> which binds "quad" to the /same/ value that p is bound to.
>
>
>         Strangely, I had the opposite problem two years ago, when
> trying to "save state" for small test set. I was using the simple
> assignment and finding that I wasn't saving the state but making a
> reference (your term) to it.




From aleax at aleax.it  Mon Sep 23 02:56:13 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 23 Sep 2002 06:56:13 GMT
Subject: match or sub problem?
References: <mailman.1032756629.18422.python-list@python.org>
Message-ID: <hWyj9.128701$ub2.2801005@news1.tin.it>

jubafre at brturbo.com wrote:

> I have this list of strings:
> final=['20 0x9', '30 0xa', 'B4 0x8', '30 0xb', 'F0', ' 1', ' 3', ' 5']
> 
> and i want to erase the nodes with whitespaces in the begining,

>From the example you give, no, you don't: rather, you want to erase
any whitespace (or a single space? your code suggests the latter --
hard to guess! -- I'll guess the former) at each item's start.

space_at_start = re.compile(r'^\s+')
final2 = [space_at_start.sub('', x) for x in final]

> i use  "match" but doesn?t work and "sub", i think the problem is in the
> "sub", why?

I think your main mistake (out of several you made, such as needlessly
recompiling the RE each time, etc) was not anchoring the match at
start-of-string with the leading ^ (\A would also work here -- the
difference comes only for multiline strings).


Alex



From member at dbforums.com  Tue Sep 17 13:04:48 2002
From: member at dbforums.com (MPS)
Date: Tue, 17 Sep 2002 17:04:48 +0000
Subject: Modify AD key fields programmatically via ADSI or VB Script
References: <1824301.1032265587@dbforums.com> <mailman.1032274400.5379.python-list@python.org>
Message-ID: <1825562.1032282288@dbforums.com>

Skip (and all),

I put it under Python accidentally and was not allowed by the system to
delete it.

--
Posted via http://dbforums.com



From max at alcyone.com  Tue Sep 17 14:31:34 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 17 Sep 2002 11:31:34 -0700
Subject: CGI question
References: <slrnaoerk7.mgo.jboy@otaku.freeshell.org>
Message-ID: <3D877506.F7C18F47@alcyone.com>

"John D. Boy" wrote:

> Now, my problem is to find out what follows behind the question mark
> in
> my CGI script. Unless the key has a value
> (http://my.domain.name/?key=value), it is not in the cgi.FieldStorage
> dictionary. Is it simply impossible, or am I not using the right
> function calls?

Try os.environ['QUERY_STRING'].

Note that this region is reserved for form submissions via GET, so if
you have any intention of having the user interact, then this isn't a
great choice of places to put it.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Be able to be alone.  Lose not the advantage of solitude.
\__/ Sir Thomas Browne
    Computer science / http://www.alcyone.com/max/reference/compsci/
 A computer science reference.


From joec at mill.co.uk  Fri Sep 13 08:04:30 2002
From: joec at mill.co.uk (Joe Connellan)
Date: Fri, 13 Sep 2002 13:04:30 +0100
Subject: Pmw.ScrolledText programatically scrolling
Message-ID: <3D81D44E.C5FC4A8B@mill.co.uk>

I've got a Pmw.ScrolledText() that I'm using as a log and I've written
an addToLog() method like so:

 def addToLog(self, logString):
  self.logScr.configure(text_state = 'normal')
  self.logScr.insert(END, logString)
  self.logScr.configure(text_state = 'disabled')

I now want it to scroll when I add a line so the user doesn't have to
constantly scroll down to view the most recent lines.

I can't find any info on doing this with ScrolledText or its text
component - do I need to reposition the scrollbars for it to scroll?

Thanks

Joe




From syver-en+usenet at online.no  Tue Sep 10 13:07:41 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Tue, 10 Sep 2002 17:07:41 GMT
Subject: PySol: "global name 'win32_gethomedir' is not defined"
References: <all3u2$qgm$1@brokaw.wa.com>
Message-ID: <un0qp6aex.fsf@online.no>

"Alan Winston" <awinston at scn.org> writes:

> I've been trying to run the PySol 4.81 pyc files under Python 2.2.1
> and
> 
> WinXP and Win98, with no success.
> 
> The Python window closes much faster than I can read the error message
> (is
> 
> there any way to force the window to stay open?), 

Can't help with the cause, but can't you just start the program from
the command prompt instead of via explorer?

-- 

Vennlig hilsen 

Syver Enstad


From BPettersen at NAREX.com  Fri Sep 13 20:04:26 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 13 Sep 2002 18:04:26 -0600
Subject: file mode attributes...
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192012532EE@admin56.narex.com>

I was wondering if this was the best way of implmenting these functions
(windows)?

 import os, stat

 def isReadable(fname):
     os.stat(fname).st_mode & stat.S_IREAD

 def isWriteable(fname):
     os.stat(fname).st_mode & stat.S_IWRITE

 def makeReadable(fname):
     os.chmod(fname, os.stat(fname).st_mode | stat.S_IREAD)

 def makeWriteable(fname):
     os.chmod(fname, os.stat(fname).st_mode | stat.S_IWRITE)

It seems rather round-about, especially since the stat attributes I'm
using aren't even mentioned in the manual...

-- bjorn



From fb at frank-buss.de  Sat Sep 14 17:30:56 2002
From: fb at frank-buss.de (Frank Buss)
Date: Sat, 14 Sep 2002 21:30:56 +0000 (UTC)
Subject: question about dictionary type..
References: <alt22f$pf8$1@newsreader.mailgate.org> <alv4op$m67$1@newsreader2.netcologne.de> <61Jg9.16334$5r1.665293@bin5.nnrp.aus1.giganews.com> <am0390$cvf$1@newsreader.mailgate.org>
Message-ID: <am09qg$78d$1@newsreader2.netcologne.de>

eugene kim <eugene1977 at hotmail.com> wrote:

> 
> thank you.. i just checked replies..
> i haven't tried them..
> this is what i came up with..
> (making needed tables first.. as carl and padraig said)
> i'm having one problem(described below)

Your code doesn't look very Python-like, but I'm no expert :-)

Do you know regular expressions? If you use my add-function, you can write 
the following:

-----------------------------------------
from string import upper
from sys import stdin
import re

def add(tree, path, value):
  start=tree
  for key in path[:-1]:
    if start.has_key(key):
      start = start[key]
    else:
      start[key] = start = {}
  start[path[-1]] = value

tree={}
records = re.compile('\((.*?)\)', re.DOTALL).findall(stdin.read())
for record in records:
  fields = map(lambda w: upper(w), re.findall("'(.*)'", record))
  add(tree, (fields[0],fields[2],fields[3]),0)
  add(tree, (fields[1],fields[2],fields[3]),0)

print tree
-----------------------------------------

With your testdata the output will be:

{'BAYLOR': {'FOOTBALL': {'M': 0, 'F': 0}}, 'CALIFORNIA': {'FOOTBALL': 
{'M': 0}}, 'NEW MEXICO': {'FOOTBALL': {'M': 0}}, 'SAMFORD': {'FOOTBALL': 
{'F': 0}}}

-- 
Frank Bu?, fb at frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de


From tdelaney at avaya.com  Mon Sep 16 01:14:58 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 16 Sep 2002 15:14:58 +1000
Subject: Source code
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A59D@natasha.auslabs.avaya.com>

> From: Terje Johan Abrahamsen [mailto:spoermeg at voldelig.com]
> 
> If I would like to write some code that would not be possible 
> for others to
> read, after I distribute the program, would Python be 

Unreadable source code can be created in any language. However, to create
truly unreadable code in Python takes true discipline.

May I suggest that Perl or C++ would be a better choice? If you specialise
in a little-used subset of C++ no one will *ever* understand it. Even Bjarne
Stroustrup (well, maybe ...).

Tim Delaney



From peter.bittner at gmx.net  Mon Sep 23 02:33:04 2002
From: peter.bittner at gmx.net (Peter Bittner)
Date: 22 Sep 2002 23:33:04 -0700
Subject: Universally unpacking tuples
Message-ID: <158d3913.0209222233.123a9675@posting.google.com>

Hi there!

I have problems universally unpacking tuples. I would like to take the
first element of a tuple and remove it (thus creating a new tuple with
one element less). Something like in the following example:

arbitraryTuple = ('abc', '123', 'qwert', ...)
while len(arbitraryTuple) > 0:
  elem, arbitraryTuple = arbitraryTuple
  print elem


Well, I know this doesn't work since unpacking requires the correct
number of elements on the left side... :-(

The root of the problem is that I am forced to move from a
list-solution (iterating over the elements of a list) to a
tuple-solution (iterating over the elements of a tuple). So, what I am
trying to do with the above code is nothing more than migrating the
following list-oriented code to a tuple-thing:

mylist = ['abc', '123', 'qwert', ...]
for i in range(len(mylist)):
  elem = mylist[i]
  print elem


Does anyone know a simple solution to this problem?
Cheers, Peter


From jcc at ibdosnorte.com  Mon Sep 16 12:51:59 2002
From: jcc at ibdosnorte.com (=?ISO-8859-1?Q?Juan_Carlos_CORU=D1A?=)
Date: 16 Sep 2002 09:51:59 -0700
Subject: problems when parsing & with xml.dom.minidom.parseString
Message-ID: <a230569c.0209160851.7c6884f@posting.google.com>

Hello all,

I'm having the following problem:

>>> node = xml.dom.minidom.parseString('<record>one &amp;
two</record>')
>>> print node.firstChild.firstChild.nodeValue
'one '

Why does the print statement return only a part of the string? How can
I include the '&' (ampersand) in a string?

Anyone knows the solution?


From phr-n2002b at NOSPAMnightsong.com  Fri Sep 20 01:46:59 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 19 Sep 2002 22:46:59 -0700
Subject: newbie seeks advise.
References: <77c84db2.0209192137.51167ebb@posting.google.com>
Message-ID: <7xk7lh5i24.fsf@ruckus.brouhaha.com>

kodokuchef at yahoo.com (akirasugiura) writes:
> I've been learning python for about a month. I had no prior knowledge
> of programming.
> 
> My question is, I have a tuple which is, data = (69, 40, 74, 41, 56,
> 57, 38, 58, 26, 55). How I can make a function that picks up a largest
> number
>  from this "data" tuple. If "data" was a  'List' I would be able to
> use "sort" but since sort doesn't  work with tuple. I am stuck here.

The simplest way is just 

   m = apply(max, data)

or in newer versions of Python you can also say

   m = max(*data)

If x the list (1,2,3), then apply(f,x) is the same as saying f(1,2,3).
f(*x) is another way to say the same thing.

Also, if you want to convert a tuple to a list, you can say list(data)
where data is the tuple.


From nospam at horvath.com  Wed Sep 25 02:36:29 2002
From: nospam at horvath.com (Bob Horvath)
Date: Wed, 25 Sep 2002 06:36:29 GMT
Subject: combining mailbox and email modules
Message-ID: <NPck9.581867$UU1.104088@sccrnsc03>

I am having trouble figuring out how to get the mailbox and email 
modules to play together so that I can use email module functions on 
every mail in the mailbox.  Is there a way to feed a factory from 
the email module to mailbox module?